USP25 regulates atherosclerosis by restricting RIPK1-mediated inflammatory responses

Summary

Background

Atherosclerosis is a common vascular disease that poses a serious threat to global health. However, the mechanism underlying the pathogenesis and progression of atherosclerosis remains elusive.

Methods

We analysed the expression of deubiquitinating enzymes in human atherosclerotic lesions and found that USP25 was significantly downregulated. The role of USP25 in atherosclerosis was validated in mouse models with an ApoE^−/− background. The protein substrates of USP25 were identified by mass spectrometry. Various biochemical methods were adopted to study the role of USP25 in signal transduction.

Findings

USP25 was predominantly expressed in macrophages in atherosclerotic lesions, and ablation of macrophagic USP25 significantly exacerbated atherosclerosis in ApoE^−/− mice accompanied by increased lipid deposition, macrophage infiltration, and vascular inflammation. Upon stimulation with ox-LDL or TNF-α, USP25 inhibited inflammatory responses in macrophages by restricting the NF-κB pathway. Mass spectrometry analysis identified RIPK1 as a USP25 substrate. Mechanistically, USP25 physically interacted with RIPK1 and removed K63 ubiquitin chains from RIPK1 via the C178 active site, thereby attenuating RIPK1-mediated signal transduction.

Interpretation

This study elucidated the function and molecular mechanism of USP25 in atherosclerosis, identifying USP25 as a beneficial regulator for this disease.

Funding

This work was supported by the Natural Science Foundation of Zhejiang Province (LZ24H090003 to X.W. and LTGY23H090001 to W.W.), the National Natural Science Foundation of China (82150710557 and 82293642 to W.S.; 81971143 to X.W., and 82271347 to G.W.), and Wenzhou Municipal Science and Technology Bureau (Y2021094 to J.H.).

Research in context.

Evidence before this study

Atherosclerosis is a chronic inflammatory vascular disease, and its pathogenesis and development are tightly regulated by post-translational modifications (PTMs) of proteins. Ubiquitination is an important PTM that critically modulates inflammatory responses. However, the role of deubiquitinating enzymes in atherosclerosis remains largely unknown.

Added value of this study

This study found that the expression of USP25 was significantly reduced in macrophages in atherosclerotic plaques. Global or hematopoietic cell-specific deletion of USP25 aggravated atherosclerosis in mice, accompanied by increased macrophage aggregation, lipid deposition, and lesional inflammation. Consistently, USP25 overexpression in hematopoietic cells significantly ameliorated atherosclerosis. Moreover, we found that USP25 inhibited inflammatory responses by removing K63 polyubiquitinating chains from RIPK1.

Implications of all the available evidence

This study not only identifies USP25 as a key regulator of atherosclerosis but also sheds light on the molecular mechanism of USP25 in inflammatory signal transduction. Overexpression of USP25 in hematopoietic cells alleviates atherosclerosis in mice, highlighting the translational potential of targeting USP25 for the treatment of atherosclerosis.

Introduction

Atherosclerosis, characterised by excessive cholesterol deposition and plaque build-up in the walls of medium-to-large arteries, is a chronic inflammatory disease resulting from the complex interplay between dysregulated lipid metabolism and immune responses.^1,2 It is the primary underlying cause of acute thrombo-occlusive events in the cardiovascular and cerebrovascular system, such as myocardial infarction and ischemic stroke.³ During the development of atherosclerosis, circulating monocytes adhere to the dysfunctional endothelium and then transmigrate into the sub-endothelial space to differentiate into macrophages.⁴ In the arterial intima, macrophages engulf large amounts of oxidized LDL (ox-LDL) via scavenger receptors such as CD36 and the class A1 scavenger receptor (SR-A1), leading to the formation of foam cells and atherosclerotic plaques.⁵ Besides, macrophages produce pro-inflammatory cytokines and chemokines to increase endothelial activation, leukocyte recruitment, and local inflammation in plaques, thereby accelerating the progression of atherosclerosis.^4,6 Therefore, identification of proteins critically regulating macrophage function in atherosclerosis may inspire novel therapeutic strategies for this disease.

In recent years, accumulating studies have revealed the importance of ubiquitination, a post-translational modification, in the onset and perpetuation of atherosclerosis.⁷ During ubiquitination, a protein substrate is covalently conjugated with one or more ubiquitin molecules under the sequential catalyzation of E1, E2, and E3 enzymes.⁸ As a counter-regulating mechanism, ubiquitination is inhibited by deubiquitinating enzymes (DUBs), which have emerged as essential regulators in atherosclerosis.^7,8 For example, the DUB USP9X removes K63-linked polyubiquitin chains from the K27 site of SR-A1 and thereby inhibits SR-A1 internalization, oxLDL uptake, and foam cell formation, resulting in ameliorated atherosclerosis.⁹ Atherosclerosis is also attenuated by the DUB A20, which inhibits atherosclerotic lesion development by suppressing inflammatory responses.¹⁰

USP25 is a DUB encoded on chromosome 21, and it is closely associated with Down syndrome and Alzheimer's disease.^11,12 In addition, USP25 participates in various diseases, including multiple sclerosis,¹³ sepsis,¹⁴ diabetic nephropathy,¹⁵ and IgG4-related disease,¹⁶ by modulating inflammation, which is also a factor fostering the progression of atherosclerosis.⁴ In the circulation system, we have found that USP25 inhibits pathological cardiac hypertrophy¹⁷ and cerebral ischemic stroke injury.¹⁸ However, the role of USP25 in atherosclerosis remains unknown.

In this study, we found that USP25 was negatively associated with atherosclerosis in both humans and rodents. Genetic deletion of USP25 significantly exacerbated atherosclerosis in ApoE^−/− mice. USP25 was mainly expressed by macrophages in atherosclerotic lesions and inhibited macrophage-mediated inflammatory responses. In terms of the mechanism, USP25 removed K63-linked polyubiquitin chains from receptor-interacting protein kinase 1 (RIPK1) and thereby restricts inflammatory signalling induced by ox-LDL and TNF-α, preventing the overproduction of pro-inflammatory mediators in macrophages. Thus, our findings identify USP25 as an important protective regulator in atherosclerosis.

Methods

Animal experiments

All animal care and experimental procedures were approved by the Wenzhou Medical University Animal Policy and Welfare Committee (approval ID: wydw2023-0553). Experiments were performed according to the guidelines from the Directive 2010/63/EU of the European Parliament on animal protection. To rule out the effect of estrogen on the animal model, only male mice were used in this study. ApoE^−/− mice on a C57BL/6 background were purchased from Zhejiang Vital River Laboratory Animal Technology Co., Ltd. (Zhejiang, China). Usp25^−/− mice on a C57BL/6 background were kindly provided by Prof. Jian Yuan (Tongji University, China). Usp25^Tg mice on a C57BL/6 background were constructed by us as reported before.¹⁹ ApoE^−/− mice were crossed with Usp25^−/− mice to generate Usp25^+/+ ApoE^−/− and Usp25^−/− ApoE^−/− mice.

The mice were housed in a controlled environment at 22 ± 2.0 °C and 50% ± 5% humidity, with a 12-h light/dark cycle (light on at 7: 00 and off at 19: 00), and were fed a standard rodent chow and water.

For the induction of atherosclerosis, eight-week-old male mice were fed a ND containing 10 kcal% fat, 20 kcal% protein, and 70 kcal% carbohydrate (Cat#: H10010, HFK Bioscience, Beijing, China) or a HFD containing 40 kcal% fat, 20 kcal% protein, 40 kcal% carbohydrate, and 1.25% cholesterol (Cat#: H10540, HFK Bioscience) for 16 weeks. ND and HFD groups were assigned in a randomized fashion. All mice were euthanized under 2% isoflurane (Cat#: R510-22-10, RWD Life Science, Shenzhen, China) mixed with air until respiration came to a complete stop and reflexes in the animals’ paws could no longer be triggered. Then, blood and aorta samples were collected. After handling the mice, the investigators were blinded when assessing the outcome.

Reagents

Oxidized low-density lipoprotein (ox-LDL) was purchased from Peking Union-Biology (Beijing, China). TNF-α (Cat#: 10291-TA) was purchased from R&D Systems (Minnesota, USA). Antibodies against RIPK1 (Cat#: 3493, 1: 1000, RRID: AB_2305314), p-RIPK1 (Cat#: 31122, 1: 1000, RRID: AB_2799000), p-IKKα/β (Cat#: 2697, 1: 1000, RRID: AB_2079382), IKKα (Cat#: 2682, 1:1000, RRID: AB_331626), p-p65 (Cat#: 3033, 1: 1000 for Western bolt and 1: 200 for immunofluorescence, RRID: AB_331284), p65 (Cat#: 8242, 1: 1000, RRID: AB_10859369), p-p38 (Cat#: 4511, 1: 1000, RRID: AB_2139682), p38 (Cat#: 8690, 1: 1000, RRID: AB_10999090), p-JNK (Cat#: 4668S, 1:1000, RRID: AB_823588), JNK (Cat#: 9252S, 1:1000, RRID: AB_2250373), p-ERK1/2 (Cat#: 4370S, 1:1000, RRID: AB_2315112), ERK1/2 (Cat#: 4695S, 1:1000, RRID: AB_390779), GAPDH (Cat#: 5174, 1: 1000, RRID: AB_10622025), α-SMA (Cat#: 19245, 1: 200 for immunofluorescence, RRID: AB_2734735), and Ubiquitin (Cat#: 3936, 1: 1000, RRID: AB_331292) were purchased from Cell Signaling Technology (Danvers, MA, USA). Antibodies against FLAG (Cat#: 20543-1-AP, 1: 1000, RRID: AB_11232216), MYC (Cat#: 60003-2-Ig, 1: 1000, RRID: AB_2734122), HA (Cat#: 51064-2-AP, 1: 1000, RRID: AB_11042321), CD36 (Cat#: 18836-1-AP, 1:1000, RRID: AB_10597244), ASK1 (Cat#: 28201-1-AP, 1: 1000, RRID: AB_2782957), MLKL (Cat#: 66675-1-Ig, 1:1000, RRID: AB_2882029), and p-ASK1 (Cat#: 28846-1-AP, 1: 1000, RRID: AB_2881220) were purchased from Proteintech (Wuhan, China). The antibody against USP25 (Cat#: ab187156, 1: 1000) and p-MLKL (Cat#: ab196436, 1:1000, RRID: AB_2687465) were purchased from Abcam (Cambridge, UK). Antibodies against F4/80 (Cat#: sc-377009, 1: 50 for immunofluorescence, RRID: AB_2927461) and USP47 (Cat#: sc-100633, 1:500, RRID: AB_2241454) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies against RIPK3 (Cat#: ER1902-67, 1:1000, RRID: AB_3069451) and p-RIPK3 (Cat#: HA500330, 1:1000, RRID: AB_3071425) were purchased from HUABIO (Hangzhou, China). The quality of these commercially available antibodies was validated by the corresponding manufacturers and we did not further validate these antibodies in this study.

Cell culture

L-929 cells (Cat#: GNM28, RRID: CVCL_0462) were obtained from the National Collection of Authenticated Cell Culture (Shanghai, China), and cultured in MEM-α medium (Cat#: 109C12571500BT, Thermo Fisher Scientific) containing 10% fetal bovine serum (FBS; Cat#: F101-01, Vazyme Biotech) and 1% penicillin/streptomycin (Cat#: P1400, Solarbio, Beijing, China). HEK293 cells (Cat#: GNHu17, RRID: CVCL_0045) were obtained from Shanghai Institute of Biochemistry and Cell Biology (Shanghai, China), and cultured in high-glucose DMEM medium (Cat#: C11995500BT, Thermo Fisher Scientific) containing 10% FBS and 1% penicillin/streptomycin. All cell lines were analysed and validated by the suppliers, and were not re-validated by us in this study.

For the generation of bone marrow-derived macrophages (BMDMs), medullary cavities of tibia and femur were flushed with ice-cold RPMI-1640 medium (Cat#: 11875093, Thermo Fisher Scientific), and the bone marrow cells were collected by centrifugation. Cells were seeded in plates in RPMI-1640 medium containing 10% FBS and 1% penicillin/streptomycin for 2 h. Then, the medium was changed to remove non-adherent cells. The remaining cells were cultured in DMEM medium containing 20% L-929 medium and 10% FBS for 7 days. All cells were cultured in a humidified incubator with 5% CO₂ at 37 °C.

Quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA was extracted from cells or animal tissues using RNAiso Plus (Cat#: 9109, Takara, Kyoto, Japan), and then reverse-transcribed to cDNA using PrimeScript™ RT Reagent Kit with gDNA Eraser (Cat#: RR047A, Takara) according to the manufacturer's instructions. PCR was performed on a QuantStudio™ 5 Real-Time PCR System (Thermo Fisher Scientific) with TBGreen® Premix Ex Taq™ II (Cat#: RR820A, Takara) and specific primers (Supplementary Table S1). Target gene expression was normalized to the mRNA level of Actb.

Plasmid construction and transfection

FLAG-USP25, FLAG-USP25-ΔUBA, FLAG-USP25-ΔUBAΔUIM1, FLAG-USP25-ΔUBAΔUIM1ΔUIM2, FLAG-USP25-ΔUSP, FLAG-USP25-C178A, FLAG-USP25-H608A, and MYC-RIPK1 plasmids were constructed by Genechem (Shanghai, China). HA-tagged wild-type and mutant ubiquitin plasmids were purchased from Miaolingbio (Wuhan, China). All plasmids were verified by DNA sequencing. Lipofectamine 3000 (Cat#: L3000015, Thermo Fisher Scientific) was applied to transfect plasmids into cultured cells.

Western blot

Cells or tissues were homogenized in cold RIPA Lysis Buffer (Cat#: AR0105, Boster Bio, CA, USA) containing protease inhibitor cocktail (Cat#: HY-K0010, MedChemExpress, New Jersey, USA), Phosphatase Inhibitor Cocktail II (Cat#: HY-K0022, MedChemExpress), and PMSF (Cat#: P0100, Solarbio). Cell or tissue lysates were centrifuged at 12,000 rpm for 10 min at 4 °C. The supernatant was collected and quantified using the Quick Start™ Bradford Protein Assay Kit 3 (Cat#: 5000203, Bio-Rad). Thereafter, protein samples were denatured in SDS-PAGE Loading buffer (Cat#: FD006, FDbio, Hangzhou, China) and then separated by SDS-PAGE. Proteins were transferred to PVDF membranes (Cat#: 10600023, Cytiva, Shanghai, China), followed by blotting with specific antibodies. Images were captured on a Fusion FX.EDGE system (Vilber, France) and analysed by the Image J software (version 1.38e, NIH, Bethesda, USA).

Immunoprecipitation

Protein lysates were prepared as described in “Western blot”. The samples were incubated with Protein A + G Magnetic Beads (Cat#: P2108, Beyotime) for 2 h at 4 °C to reduce non-specific binding. The supernatant was collected and incubated with primary antibodies overnight at 4 °C under gentle rotation. Thereafter, Protein A + G Magnetic Beads were added and incubated for 2 h at 4 °C under gentle rotation. The beads were harvested and then washed 5 times with ice-cold lysis buffer prior to further analysis.

In vitro deubiquitination assay

FLAG-USP25 and MYC-RIPK1 + HA-K63 Ub plasmids were separately transfected into cells with Lipofectamine 3000. Twenty-four hours after transfection, USP25 and K63-ubiquitinated RIPK1 were harvested from transfected cells by immunoprecipitation. Subsequently, K63-ubiquitinated RIPK1 was incubated with or without USP25 in the deubiquitination buffer (50 mM Tris–HCl, 5 mM MgCl₂, 2 mM DTT, 2 mM ATP, 5% glycerol) for 2 h at 37 °C. After incubation, the samples were analysed by Western blot.

LC-MS/MS analysis

USP25 and interacting proteins were harvested from cells transfected with FLAG-USP25 plasmids by immunoprecipitation with the anti-FLAG antibody. Immunoprecipitation with IgG served as negative control. The LC-MS/MS analysis (Hangzhou Lingyu Bio-Technology, Hangzhou, China) was carried out to identify proteins in immunoprecipitation samples.

Atherosclerotic lesion analysis

After 4 months of HFD feeding, mice were fasted for 12 h before euthanasia. The entire aorta and aortic sinus were isolated and fixed with 4% paraformaldehyde. To analyse the en face lesion area, the entire aorta was opened longitudinally to expose the intimal surface of the aorta, followed by staining with Oil Red O (G1260, Solarbio). Images were captured using a high-resolution digital camera. To analyse lesions in the aortic root, the heart and proximal aorta were embedded in the OCT compound (Cat#: 4583, SAKURA, Osaka, Japan). Serial 8-μm-thick sections of the aortic sinus were stained with Oil Red O and hematoxylin. Images were captured on a Nikon ECLIPSE Ni-U 237 microscope (Nikon, Tokyo, Japan). Aortic sections were stained with BODIPY dye (Cat#: 790389, Sigma–Aldrich, USA) and anti-F4/80 antibody. Images were captured on a ZEISS LSM 980 with Airyscan 2 confocal microscope (Carl Zeiss AG, Oberkochen, Germany).

Hematoxylin-eosin staining

Paraffin-embedded tissues were cut into 8 μm-thick sections. Then, the sections were stained with hematoxylin and eosin (Cat#: G4520, Solarbio) according to the manufacturer's instructions. After mounting the sections with neutral balsam (Cat#: G8590, Solarbio), images were taken using a Nikon ECLIPSE Ni-U 237 microscope.

Immunofluorescence staining

Samples were permeabilized with 0.5% Triton X-100 (Cat#: T8200, Solarbio) for 10 min and then blocked with 5% BSA Blocking Buffer for 30 min. Thereafter, the samples were incubated with specific primary antibodies overnight at 4 °C, followed by incubation with secondary antibodies conjugated with Alexa Fluor 488 (Cat#: 33206ES60, Yeasen) or 594 (Cat#: 34212ES60, Yeasen) for 1 h. Nuclei were stained with DAPI (Cat#: S2110, Solarbio). A ZEISS LSM 980 with Airyscan 2 confocal microscope was used to capture images.

ox-LDL uptake assay

BMDMs were incubated with ox-LDL (50 μg/mL) for 24 h. Thereafter, BMDMs were fixed in 4% paraformaldehyde for 15 min. After washing with PBS, cells were stained with 0.5% Oil Red O working solution for 15 min. Images were taken on a Nikon ECLIPSE Ni-U microscope equipped with a digital camera.

Bone marrow transplantation

Male recipient mice (6–8 weeks old) were fed acidified sterile water supplemented with 10 mg/mL neomycin sulfate (Cat#: HY-B0470, MedChemExpress) for 1 week and then irradiated at a dose of 8.5 Gy. Bone marrow cells were isolated from Usp25^+/+, Usp25^−/−, and Usp25^Tg mice, and 5.0 × 10⁶ cells were injected through the tail vein into each recipient mouse within 6 h after irradiation. After 8 weeks of bone marrow reconstitution, the recipient mice were fed a HFD for 16 weeks to induce atherosclerosis.

AAV injection

AAV-Ripk1-shRNA (P25E0318) (Cat#: GIDV5020789) and AAV-Ctrl virus CON305 (type 9) (Cat#: AAV9CON305) were generated by Genechem (Shanghai, China). Mice were intravenously injected with AAV-Ripk1-shRNA or AAV-Ctrl through the tail vein at a dose of 5 × 10¹¹ vg/mice. Three weeks after injection, mice were used for subsequent experiments.

Clinical samples

Human carotid artery biopsy specimens were obtained from patients with atherosclerosis at the First Affiliated Hospital of Wenzhou Medical University. The study involving clinical samples was approved by the Ethics Committee in Clinical Research (ECCR) of the First Affiliated Hospital of Wenzhou Medical University (Approval No. KY2024-R032), and informed consent was obtained from patients. The study was carried out in accordance with the Declaration of Helsinki. The patient information is provided in Supplementary Table S2.

Transcriptome analysis

The GSE41571 dataset was analysed by the R package limma, with thresholds set at P < 0.05 and |log₂FC| > 0.25. In addition, the single-cell RNA sequencing dataset GSE159677 was analysed by the R package Seurat (version 4.3.0.1) (http://satijalab.org/seurat/), and Uniform Manifold Approximation and Projection (UMAP) was used for nonlinear dimensionality reduction and visualization. CellMarker²⁰ and Panglao DB²¹ were used to identify cell types with labelled genes. Cell communication was analysed using the R package CellChat.²²

Statistical analysis

All data are shown as mean ± SEM. Statistical analyses were carried out using GraphPad Prism 8 (GraphPad, CA, USA). Data between two groups were compared using two-tailed t test, and data among three or more groups were compared using analysis of variance (ANOVA) test. P < 0.05 were considered statistically significant (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001). The sample size is included in figures and figure legends, and “n” represents biological replicates.

Role of funders

The funding source provided financial support, but had no role in study design, data collection, data analysis, data interpretation, or writing of this report.

Results

USP25 is downregulated in advanced atherosclerotic lesions

To identify DUBs associated with atherosclerosis, we analysed the expression profiles of 53 DUBs of the ubiquitin-specific protease (USP) subfamily in human atherosclerotic lesions using open GEO transcriptome databases. As shown in the GSE100927 dataset, transcription of 6 DUBs was significantly downregulated in human atherosclerotic lesions compared to that in control arteries (Fig. 1A). In addition, the GSE41571 dataset showed that the expression of 10 DUBs was further downregulated in severe atherosclerotic plaques of humans compared to that in mild plaques (Fig. 1A). The expression of Usp47, Usp25, Usp16, and Usp9X was downregulated in both datasets, and we further found that the expression of Usp25, Usp47, and Usp9X was significantly downregulated in murine aortas from ApoE^−/− mice fed with high-fat diet (HFD) (Fig. 1A and B). Of note, USP9X has been shown to be downregulated in atherosclerotic lesions in a previous study.⁹ Next, we found that protein levels of USP25, but not USP47, were significantly reduced in severe atherosclerotic lesions in both mice and humans (Fig. 1C and D, Supplementary Fig. S1A–B). Therefore, USP25 was selected for further investigation. Immunofluorescence staining of the atherosclerotic plaque showed that USP25 co-localized with the macrophage marker F4/80 rather than the smooth muscle cell marker α-SMA and the endothelial cell marker CD31 (Fig. 1E–G), suggesting that USP25 is predominantly expressed in macrophages in the plaque. Moreover, USP25 expression was reduced in macrophages of atherosclerotic lesions in a time-dependent way (Fig. 1H). Consistently, immunofluorescence staining also revealed that USP25 expression in macrophages of severe plaques was lower than that in mild plaques of humans (Fig. 1I). Surprisingly, Usp25 expression was significantly upregulated in BMDMs stimulated with ox-LDL or TNF-α, two pro-atherosclerotic factors (Supplementary Fig. S1C), indicating that the downregulated USP25 expression in atherosclerotic lesions may be due to mRNA-degrading mechanisms. Overall, these results show that USP25 expression is inversely correlated with atherosclerosis in humans and rodents, suggesting that USP25 may serve as a regulatory molecule for atherosclerosis.

Fig. 1.

USP25 is downregulated in human and mouse atherosclerotic lesions. (A) Screening of USP family genes in human carotid artery plaques based on GSE100927 and GSE41571 datasets. In volcano plots, red and blue dots represent upregulated and downregulated USPs, respectively. (B) Transcriptional levels of selected USPs in aortas of ApoE^−/− mice fed with ND and HFD were determined by qRT-PCR. ∗P < 0.05, ∗∗P < 0.01, ns, not significant, unpaired two-tailed Student's t-test (n = 5). (C) Representative immunoblots (upper panel) and densitometric quantification (lower panel) of USP25 in aortas of ND and HFD-fed ApoE^−/− mice. ∗∗P < 0.01, unpaired two-tailed Student's t-test (n = 3). (D) Representative immunoblots (upper panel) and densitometric quantification (lower panel) of USP25 in mild and severe atherosclerotic lesions of human carotid arteries. ∗P < 0.05, unpaired two-tailed Student's t-test (n = 3). (E–G) Representative immunofluorescence staining of USP25 and F4/80 (E), α-SMA (F), as well as CD31 (G) in aortic roots. Scale bar: 15 μm. (H) Representative immunofluorescence staining of USP25 and F4/80 in aortas of ApoE^−/− mice fed with HFD for indicated time. Scale bar: 100 μm ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, two-way ANOVA with Bonferroni post hoc test (n = 5). (I) Representative immunofluorescence staining of USP25 and CD68 in carotid arteries from patients with mild and severe atherosclerosis. Scale bar: 100 μm.

USP25 ameliorates atherosclerosis in ApoE^−/− mice

To explore the role of USP25 in atherosclerosis, we crossed Usp25^−/− mice with ApoE^−/− mice to generate Usp25^−/− ApoE^−/− double-knockout mice and Usp25^+/+ ApoE^−/− control mice (Supplementary Fig. S2A–B). Usp25^−/− ApoE^−/− and Usp25^+/+ ApoE^−/− mice were fed a HFD for 16 weeks to induce atherosclerosis (Supplementary Fig. S2C). Although HFD-fed Usp25^−/− ApoE^−/− and Usp25^+/+ ApoE^−/− mice had comparable body weight and serum lipid profile (Supplementary Fig. S2D–E), USP25 deficiency significantly increased the atherosclerotic lesion area in the whole aorta, as revealed by Oil Red O staining (Fig. 2A and B). Moreover, H&E staining showed that USP25 deletion increased atherosclerotic lesion and necrotic core areas in the aortic root of HFD-fed ApoE^−/− mice (Fig. 2C and D). Further analysis of aortic root components found that lipid accumulation and macrophage infiltration were evidently increased and smooth muscle cell content was significantly decreased in Usp25^−/− ApoE^−/− mice as compared with Usp25^+/+ ApoE^−/− mice (Fig. 2E and F, Supplementary Fig. S3). Chronic low-grade inflammation is a hall mark of atherosclerosis and can accelerate plaque formation and destabilization.^23,24 We found that USP25 deletion increased the expression of pro-inflammatory cytokines and chemokines in aortas of HFD-fed ApoE^−/− mice (Fig. 2G and H). Taken together, these results demonstrate that the lack of USP25 exacerbates atherosclerosis in mice.

Fig. 2.

USP25 deficiency exacerbates atherosclerosis in HFD-fed ApoE^−/− mice. (A) Representative Oil Red O staining of aortas from HFD-fed Usp25^+/+ApoE^−/− and Usp25^−/−ApoE^−/− mice. Scale bar: 5 mm. (B) Data show the percentage of plaque area/total vessel area. AA, aortic arch; TA, thoracic aorta. ∗∗∗P < 0.001, unpaired 2-tailed Student's t test (n = 6). (C) Representative H&E staining of aortic root sections. Scale bar: 200 μm. (D) Quantification of lesion area (left) and percentages of necrotic core (right). ∗∗P < 0.01, unpaired 2-tailed Student's t test (n = 10). (E) Representative Oil Red O (top), anti-F4/80 immunofluorescence (middle), and anti-α-SMA immunofluorescence (bottom) staining of aortic root sections. Scale bar: 100 μm. (F) Percentages of Oil Red O (top), F4/80⁺ (middle), and α-SMA⁺ (bottom) area in aortic root sections. ∗∗P < 0.01, ∗∗∗P < 0.001, unpaired 2-tailed Student's t test (n = 5). (G) Protein levels of IL-6 and TNF-α in atherosclerotic lesions were analysed using ELISA. ∗∗P < 0.01, ∗∗∗P < 0.001, unpaired 2-tailed Student's t test (n = 6). (H) Transcriptional levels of indicated genes in aortas were determined by qRT-PCR. Target gene expression was normalized to the level of Actb mRNA. ∗P < 0.05, ∗∗P < 0.01, unpaired 2-tailed Student's t test (n = 5).

Hematopoietic cell-derived USP25 ameliorates atherosclerosis

Given that USP25 was mainly expressed in macrophages in atherosclerotic plaques (Fig. 1E and H) and that macrophages critically contribute to the onset and perpetuation of atherosclerosis,⁷ it is highly possible that USP25 affects atherosclerosis by regulating macrophages. To confirm this hypothesis, we reconstituted irradiated ApoE^−/− mice with bone marrow cells from Usp25^+/+ and Usp25^−/− mice to generate bone marrow chimeric mice with USP25-sufficient and -deficient macrophages, respectively (Supplementary Fig. S4A–B). In line with previous findings that USP25 deficiency had no influence on body weight and serum lipid profile in HFD-fed ApoE^−/− mice (Supplementary Fig. S2D–E), HFD-fed chimeric mice receiving USP25-sufficient and -deficient bone marrow had comparable body weight and serum lipid profile (Supplementary Fig. S4C–D). However, after 16 weeks of HFD feeding, ApoE^−/− mice transplanted with USP25-deficient bone marrow had significantly enlarged lesion area in the whole aorta (Fig. 3A and B). Moreover, transplantation with USP25-deficient bone marrow significantly increased atherosclerotic lesion area, necrotic core area, lipid accumulation, and macrophage infiltration in the aortic root of HFD-fed chimeric mice (Fig. 3C–F, Supplementary Fig. S5). Macrophages are a major source of pro-inflammatory cytokines and chemokines, which strongly promote the formation and development of atherosclerotic lesions. Consistent with results shown in Fig. 2G, hematopoietic cell-specific deletion of USP25 significantly increased the aortic expression of pro-inflammatory cytokines and chemokines (Fig. 3G and H). Together, these data demonstrate that targeted deletion of USP25 in hematopoietic cells increases aortic inflammation and exacerbates atherosclerosis, indicating that USP25 influences atherosclerosis by modulating inflammatory responses in macrophages.

Fig. 3.

Deficiency of USP25 in hematopoietic cells exacerbates atherosclerosis in HFD-fed ApoE^−/− mice. (A) Representative Oil Red O staining of aortas from Usp25^+/+→ApoE^−/− and Usp25^−/−→ApoE^−/− mice fed a HFD for 16 weeks. Scale bar: 5 mm. (B) Data show the percentage of plaque area/total vessel area. ∗∗∗P < 0.001, unpaired two-tailed Student's t test (n = 6). (C) H&E staining of aortic root sections. Scale bar: 200 μm. (D) Quantitative analysis of lesion area (left) and percentages of necrotic cores (right). ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 10). (E) Representative Oil Red O (top), anti-F4/80 immunofluorescence (middle), and anti-α-SMA immunofluorescence (bottom) staining of aortic root sections. Scale bar: 100 μm. (F) Percentages of Oil Red O (top), F4/80⁺ (middle), and α-SMA⁺ (bottom) area in aortic root sections. ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 5). (G) Protein levels of IL-6 and TNF-α in atherosclerotic lesions were analysed using ELISA. ∗∗P < 0.01, ∗∗∗P < 0.001, unpaired 2-tailed Student's t test (n = 6). (H) Aortas were isolated from Usp25^+/+→ApoE^−/− and Usp25^−/−→ApoE^−/− mice after 16 weeks of HFD feeding. Transcriptional levels of indicated genes in aortas were determined by qRT-PCR. ∗∗P < 0.01, ∗∗∗P < 0.001, unpaired two-tailed Student's t test (n = 6).

USP25 inhibits inflammatory responses in macrophages by limiting NF-κB activation

To verify the impact of USP25 on inflammatory responses in macrophages, we stimulated primary BMDMs from Usp25^+/+ and Usp25^−/− mice with ox-LDL, and found that USP25 deficiency significantly increased the expression of Il1b, Il6, Tnf, Cxcl1, and Cxcl10 in ox-LDL-stimulated BMDMs (Fig. 4A). Moreover, USP25 deletion significantly increased the uptake of ox-LDL by BMDMs (Fig. 4B and C). However, the expression of CD36 and SR-A1, two scavenger receptors responsible for ox-LDL uptake, was not altered by USP25 deletion (Supplementary Fig. S6A–B), indicating that the increased uptake of ox-LDL in USP25-deficient cells should be largely attributed to inflammation. Of note, multiple inflammatory signalling pathways, particularly the NF-κB pathway, were strongly activated in macrophages of human atherosclerotic lesions (Fig. 4D). Therefore, we analysed the effect of USP25 on ox-LDL-induced NF-κB activation. Western blot results revealed that USP25 deletion substantially enhanced ox-LDL-induced activation of NF-κB and MAPK signalling pathways in BMDMs (Fig. 4E). Besides, immunofluorescence staining showed that USP25 deletion significantly enhanced the nuclear translocation of p65 upon ox-LDL stimulation (Fig. 4F and G). In line with these in vitro findings, USP25 deficiency significantly increased the phosphorylation of p65 in macrophages at the aortic root of HFD-fed ApoE^−/− mice (Fig. 4H and I). Collectively, these data suggest that USP25 attenuates macrophage-mediated inflammatory responses by inhibiting NF-κB activation.

Fig. 4.

USP25 deficiency increases inflammatory responses in macrophages by enhancing NF-κB activation. (A) BMDMs isolated from Usp25^+/+ and Usp25^−/− mice were stimulated with or without ox-LDL (50 μg/mL) for 6 h. The transcriptional levels of the indicated genes were detected by qRT-PCR. ∗P < 0.05, ∗∗P < 0.01, two-way ANOVA with Bonferroni post hoc test (n = 3). (B) BMDMs isolated from Usp25^+/+ and Usp25^−/− mice were incubated with or without ox-LDL (50 μg/mL) for 24 h, followed by Oil Red O staining. Scale bar: 20 μm. (C) Quantitation of Oil Red O-positive areas in BMDMs. ∗∗∗P < 0.001, two-way ANOVA using Bonferroni's post hoc test (n = 5). (D) KEGG pathway enrichment analysis of signalling pathways increased in macrophages of human atherosclerotic lesions based on GSE159677. (E) BMDMs from Usp25^+/+ and Usp25^−/− mice were treated with or without ox-LDL (50 μg/mL) for 30 min, followed by Western blot analysis with indicated antibodies. (F–G) Representative immunofluorescence staining (F) and quantitation (G) of p65 nuclear translocation in BMDMs treated with or without ox-LDL (50 μg/mL). Scale bar: 50 μm ∗∗∗P < 0.001, two-way ANOVA with Bonferroni post hoc test (n = 3). (H–I) Representative immunofluorescence staining (H) and quantification (I) of p-p65 and F4/80 in aortic roots of HFD-fed Usp25^+/+ApoE^−/− and Usp25^−/−ApoE^−/− mice. Scale bar: 50 μm ∗∗∗P < 0.001, unpaired two-tailed Student's t test (n = 6).

USP25 physically interacts with RIPK1 through the USP domain

To explore the molecular mechanism by which USP25 regulates the NF-κB signalling, we screened potential substrate proteins of USP25 by mass spectrometry (Fig. 5A). Interestingly, RIPK1, which is not only a signalling molecule upstream of NF-κB and MAPK signalling but also a pro-atherosclerotic factor,^25,26 was identified in the USP25 interactome (Fig. 5B and C), indicating that USP25 may affect vascular inflammation in atherosclerosis by regulating RIPK1. The interaction between endogenous USP25 and RIPK1 in BMDMs was confirmed by immunoprecipitation and Western blot analysis (Fig. 5D and E). The direct interaction between the two proteins was further consolidated with exogenously expressed FLAG-tagged USP25 and MYC-tagged RIPK1 (Fig. 5F and G). In addition, ox-LDL stimulation strengthened the association of USP25 and RIPK1 (Fig. 5H and I). Structurally, USP25 has a UBA domain, 2 UIM domains, and a USP domain (Fig. 5J). To pinpoint the structural domain of USP25 that interacts with RIPK1, we constructed a series of USP25 mutants lacking certain structural domains (Fig. 5J). We found that only the FLAG-USP25ΔUSP mutant, which lacked the USP domain, could not bind RIPK1, indicating that the USP domain is required for USP25 to interact with RIPK1 (Fig. 5K). Taken together, these results show that USP25 interacts with RIPK1 through the USP domain.

Fig. 5.

USP25 interacts with RIPK1 through the USP domain. (A) Schematic diagram of the proteomic screening for USP25-interacting proteins. (B) Two-dimensional plot of USP25-binding proteins, with the Y axis showing protein intensity and the X axis showing protein molecular weight. (C) List of potential substrates with unique peptides ≥2 identified by the mass spectrometry analysis. (D–E) Whole cell lysates of BMDMs were immunoprecipitated with anti-USP25 (D) or anti-RIPK1 (E) antibodies, followed by Western blot analysis. Rabbit IgG served as negative control. (F–G) HEK293 cells were co-transfected with FLAG-USP25 and MYC-RIPK1 plasmids for 24 h. Whole-cell lysates were immunoprecipitated with anti-FLAG (F) or anti-MYC (G) antibodies, followed by Western blot analysis. (H) BMDMs were left untreated or treated with ox-LDL (50 μg/mL) for 30 min before lysis. Whole-cell lysates were immunoprecipitated with anti-USP25 antibody, followed by Western blot analysis. (I) Subcellular distribution of USP25 (red) and RIPK1 (green) in BMDMs treated with or without ox-LDL (50 μg/mL) for 30 min was detected by immunofluorescence. Scale bar: 5 μm. (J) Schematic diagram of structural domains and truncation mutants of USP25. (K) HEK293 cells were co-transfected with indicated plasmids for 24 h. Cell lysates were immunoprecipitated with anti-FLAG antibody and further analysed by Western blot with indicated antibodies.

USP25 reduces K63 ubiquitination of RIPK1 via its catalytic activity

Considering that ubiquitination is essential for RIPK1 function in signal transduction, we assessed the impact of USP25 on the ubiquitination of RIPK1 and found that USP25 ablation increased RIPK1 ubiquitination in ox-LDL-stimulated BMDMs (Fig. 6A). Similarly, USP25 deficiency increased RIPK1 ubiquitination in aortas of mice with atherosclerosis (Fig. 6B). Of note, K48-linked polyubiquitination marks RIPK1 for proteasomal degradation whereas K63-specific polyubiquitination promotes the NF-κB-inducing activity of RIPK1.27, 28, 29, 30 USP25 deficiency increased RIPK1 ubiquitination but did not alter RIPK1 protein abundance (Fig. 6A and B), implying that USP25 does not regulate K48 ubiquitination of RIPK1. Indeed, USP25 specifically reduced K63 polyubiquitination of RIPK1 but had no impact on K6, K11, K27, K29, K33, and K48 ubiquitination (Fig. 6C and D, Supplementary Fig. S7A–F), demonstrating that USP25 regulates RIPK1 activation by reducing its K63 ubiquitination. Furthermore, USP25 could directly remove K63-linked polyubiquitin chains that were already established on RIPK1 in an in vitro deubiquitination assay (Fig. 6E and F). The enzymatic activity of USP25 is mediated by cysteine 178 (C178) and histidine 608 (H608) residues (Fig. 6G). To determine the active site that is essential for USP25 to deubiquitinate RIPK1, we inactivated C178 and H608 active sites in C178S and H608A mutants, respectively. Although both mutants still interacted with RIPK1 (Fig. 6H), the C178S mutant failed to reduce K63 ubiquitination of RIPK1 (Fig. 6I), indicating that the C178 residue of USP25 is essential for the K63 deubiquitination of RIPK1. In summary, these findings show that USP25 can cleave K63-linked polyubiquitination chains on RIPK1 via the C178 active site.

Fig. 6.

USP25 cleaves K63 ubiquitinated chains on RIPK1. (A) BMDMs from USP25^+/+ and USP25^−/− mice were treated with ox-LDL (50 μg/mL) for 30 min before lysis. Whole-cell lysates were immunoprecipitated with anti-RIPK1 antibody, followed by Western blot analysis with indicated antibodies. (B) Aortas from HFD-fed Usp25^+/+→ApoE^−/− and Usp25^−/−→ApoE^−/− mice were lysed for protein extraction. The protein samples were immunoprecipitated with anti-RIPK1 antibody, followed by Western blot analysis with indicated antibodies. (C–D) HEK293 cells were co-transfected with indicated plasmids for 24 h, followed by treatment with ox-LDL (50 μg/mL) for 30 min. Cell lysates were immunoprecipitated with anti-MYC antibody and further analysed by Western blot with indicated antibodies. (E–F) Schematic diagram (E) and representative immunoblots (F) of the in vitro deubiquitination assay. (G) Schematic diagram showing USP25 active sites. (H) HEK293 cells were transfected with indicated plasmids for 24 h before lysis. Cell lysates were immunoprecipitated with anti-MYC antibody and then analysed by Western blot analysis with indicated antibodies. (I) HEK293 cells were transfected with indicated plasmids for 24 h before lysis. Cell lysates were immunoprecipitated with anti-MYC antibody, followed by Western blot analysis.

USP25 attenuates TNF-α-induced inflammatory responses in macrophages

RIPK1 is an essential upstream component of the signalling induced by TNF-α, a detrimental cytokine in atherosclerosis.^25,26 Interestingly, macrophages and monocytes were the major cell populations responsive to TNF-α in human atherosclerotic lesions (Fig. 7A). Besides, the TNF signalling pathway, ranked 2nd highest among all enriched pathways, was strongly activated in macrophages of human atherosclerotic lesions (Fig. 4D). Therefore, the TNF-α-induced RIPK1-dependent signalling is critical for macrophage-mediated inflammatory responses in atherosclerosis, and USP25 may also affect atherosclerosis by controlling TNF-α-induced macrophage activation. Indeed, phosphorylation of IKKα/β and p65 was strongly enhanced in USP25-deficient BMDMs upon TNF-α stimulation (Fig. 7B). However, phosphorylation of RIPK1 was not altered by USP25 deficiency (Fig. 7B), which is consistent with previous findings that ubiquitination, rather than phosphorylation, of RIPK1 is essential for TNF-α-induced NF-κB activation.^28,30 The unchanged phosphorylation of RIPK1 further confirmed that USP25 inhibited RIPK1 activation by regulating ubiquitination rather than other post-translational modifications. Consistent with Western blot results, immunofluorescence staining found that USP25 deficiency increased the nuclear translocation of p65 upon TNF-α stimulation (Fig. 7C and D). Furthermore, ablation of USP25 significantly increased TNF-α-induced production of cytokines and chemokines in macrophages (Fig. 7E). In addition to inflammation, TNF-α and RIPK1 can also induce necroptosis. However, USP25 deficiency had no impact on necroptosis induced by TNF-α + CHX + Z-VAD (Supplementary Fig. S8A–B). Collectively, these findings reveal that USP25 inhibits atherosclerosis by limiting RIPK1-dependent inflammatory responses in macrophages.

Fig. 7.

USP25 deletion enhances TNF-α-induced inflammatory responses in macrophages. (A) Plot of the TNF signalling network among 12 cell populations in human atherosclerotic lesions based on GSE159677. The right panel shows the intensity of intercellular signal communication. (B) BMDMs isolated from Usp25^+/+ and Usp25^−/− mice were treated with or without TNF-α (50 ng/mL) for 30 min. Cell lysates were analysed by Western blot analysis with indicated antibodies. (C–D) Representative immunofluorescence staining (C) and quantitation (D) of p65 nuclear translocation in BMDMs treated with or without TNF-α (50 μg/mL). Scale bar: 50 μm ∗∗∗P < 0.001, two-way ANOVA with Bonferroni post hoc test (n = 3). (E) BMDMs were stimulated with or without TNF-α (50 ng/mL) for 6 h. The mRNA levels of the indicated genes were detected by qRT-PCR. ∗P < 0.05, two-way ANOVA with Bonferroni post hoc test (n = 3).

USP25 ameliorates atherosclerosis by regulating RIPK1

To confirm the function of the USP25-RIPK1 regulatory axis in atherosclerosis, we applied adeno-associated virus (AAV) expressing Ripk1 shRNA (AAV-Ripk1-shRNA) to downregulate RIPK1 in ApoE^−/− bone marrow chimeric mice receiving Usp25^−/− bone marrow cells (Fig. 8A, Supplementary Fig. S9A). RIPK1 downregulation had no impact on body weight and plasma lipid profile after HFD feeding (Supplementary Fig. S9B–C). However, RIPK1 knockdown significantly reduced atherosclerotic lesions in the aorta (Fig. 8B and C). Consistently, AAV-Ripk1-shRNA treatment reduced lesion area, necrotic core, and lipid accumulation at the aortic root (Fig. 8D–G). Moreover, immunofluorescence analysis found that RIPK1 inhibition also mitigated macrophage infiltration in atherosclerotic lesions (Fig. 8H and I). Importantly, knockdown of RIPK1 significantly reduced the expression of pro-inflammatory cytokines and chemokines in the aorta of HFD-fed bone marrow chimeric mice (Fig. 8J and K). Together, these findings demonstrate that RIPK1 downregulation mitigates the exacerbated atherosclerosis caused by USP25 deficiency in hematopoietic cells, indicating that USP25 ameliorates atherosclerosis by regulating RIPK1.

Fig. 8.

USP25 affects atherosclerosis by regulating RIPK1. (A) Experimental flowchart for the establishment of atherosclerosis in AAV-infected bone marrow chimeric mice. (B) Representative Oil Red O staining of aortas from indicated mice fed a HFD for 16 weeks. Scale bar: 5 mm. (C) Data show the percentage of plaque area/total vessel area. ∗∗∗P < 0.001, unpaired two-tailed Student's t test (n = 6). (D) H&E staining of aortic root sections. Scale bar: 200 μm. (E) Quantification of lesion area (left) and percentages of necrotic core (right). ∗∗∗P < 0.001, unpaired 2-tailed Student's t test (n = 10). (F) Representative Oil Red O staining of aortic root sections. Scale bar: 100 μm. (G) Percentages of Oil Red O area in aortic root sections. ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 5). (H–I) Representative immunofluorescence staining (H) and quantification (I) of F4/80 in aortic root sections. Scale bar: 100 μm ∗∗∗P < 0.001, unpaired two-tailed Student's t test (n = 5). (J) Protein levels of TNF-α and IL-6 in atherosclerotic lesions were analysed using ELISA. ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 6). (K) Transcriptional levels of indicated genes in aortas were determined by qRT-PCR. ∗P < 0.05, ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 5).

USP25 overexpression alleviates atherosclerosis in mice

To investigate whether USP25 overexpression could ameliorate atherosclerosis, we transplanted bone marrow cells from USP25 overexpression (Usp25^Tg) mice or control (Usp25^+/+) mice into irradiated ApoE^−/− recipient mice, and then fed these bone marrow chimeric mice with HFD for 16 weeks to induce atherosclerosis (Supplementary Fig. S10A–B). USP25 overexpression in hematopoietic cells significantly reduced atherosclerotic lesions in the aorta of HFD-fed ApoE^−/− mice but had no discernible effect on body weight and plasma lipid profile (Fig. 9A and B, Supplementary Fig. S10C–D). Moreover, USP25 overexpression reduced RIPK1 ubiquitination in the aorta of HFD-fed ApoE^−/− mice (Fig. 9C). Furthermore, overexpression of USP25 reduced lesion area, necrotic core, lipid accumulation, and macrophage infiltration in the aortic root of HFD-fed ApoE^−/− mice (Fig. 9D–I). In line with our previous findings that USP25 deficiency increased the expression of pro-inflammatory cytokines and chemokines, USP25 overexpression significantly reduced the levels of these pro-inflammatory mediators in the aorta of HFD-fed ApoE^−/− mice (Fig. 9J and K). Taken together, these data show that overexpression of USP25 in hematopoietic cells significantly alleviates atherosclerosis in mice.

Fig. 9.

USP25 overexpression alleviates atherosclerosis in HFD-fed ApoE^−/−mice. (A) Representative Oil Red O staining of aortas from Usp25^+/+→ApoE^−/− and Usp25^Tg→ApoE^−/− mice fed a HFD for 16 weeks. Scale bar: 5 mm. (B) Data show the percentage of plaque area/total vessel area. ∗∗∗P < 0.001, unpaired two-tailed Student's t test (n = 6). (C) Aortas from HFD-fed Usp25^+/+→ApoE^−/− and Usp25^Tg→ApoE^−/− mice were lysed for protein isolation. The lysates were immunoprecipitated with anti-RIPK1 antibody, followed by Western blot analysis with indicated antibodies. (D) H&E staining of aortic root sections. Scale bar: 200 μm. (E) Quantification of lesion area (left) and percentages of necrotic core (right). ∗∗P < 0.01, unpaired 2-tailed Student's t test (n = 10). (F) Representative Oil Red O staining of aortic root sections. Scale bar: 100 μm. (G) Percentages of Oil Red O area in aortic root sections. ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 5). (H–I) Representative immunofluorescence staining (H) and quantification (I) of F4/80 in aortic root sections. Scale bar: 100 μm ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 5). (J) Protein levels of TNF-α and IL-6 in atherosclerotic lesions were analysed using ELISA, ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 6). (K) Transcriptional levels of indicated genes in aortas were determined by qRT-PCR. ∗P < 0.05, ∗∗P < 0.01, unpaired two-tailed Student's t test (n = 5).

Discussion

In this study, we discovered that USP25 expression was significantly downregulated in atherosclerotic plaques in both humans and mice. In mouse models, global or hematopoietic cell-specific ablation of USP25 exacerbated atherosclerotic lesions accompanied by increased lipid deposition, macrophage accumulation, and lesional inflammation. Mechanistically, USP25 exerted its anti-atherosclerotic effect by inhibiting inflammatory responses in macrophages through K63 deubiquitinating RIPK1 (Supplementary Fig. S11). Thus, our findings identified USP25 as a key regulator of atherosclerosis and provided insight into the functional role of USP25 in this disease.

Inflammation is a driving factor underlying the pathogenesis and development of atherosclerosis.^4,6,24 Besides, anti-inflammatory therapy has been shown to reduce atherosclerosis-associated cardiovascular diseases.^31,32 From this perspective, proteins that participate in inflammation may become potential druggable targets for the treatment of atherosclerosis and its complications. USP25 is a DUB that critically regulates inflammatory signalling and inhibits multiple inflammatory diseases.13, 14, 15, 16^,18 Here, we found that deletion of USP25 significantly exacerbated atherosclerosis by increasing macrophage-mediated inflammatory responses. In terms of its mechanism of action, USP25 is different from existing anti-atherosclerotic drugs, particularly statins, which primarily regulate lipid metabolism. Therefore, therapies targeting USP25 may provide additional benefits for patients with atherosclerosis who still have residual inflammatory risk despite receiving statin treatment.

As a member of metabolism-associated molecular patterns (MAMPs), ox-LDL activates scavenger receptors and Toll-like receptors (TLRs) to incite inflammatory signalling transduction and the production of inflammatory mediators.³³ In addition to ox-LDL, inflammatory responses in lesional macrophages are strongly induced by TNF-α, a detrimental cytokine in atherosclerosis.^34,35 Here, we found that USP25 deficiency enhanced the ability of macrophages to produce pro-inflammatory cytokines and chemokines upon stimulation with ox-LDL or TNF-α. Moreover, ox-LDL or TNF-α stimulation induced the transcription of Usp25 in BMDMs. In sharp contrast, both the mRNA and protein levels of USP25 were markedly downregulted in human and mouse atherosclerotic lesions. Notably, a study shows that HuR induces the downregulation of USP31 by degrading its mRNA, thereby promoting macrophage-mediated inflammatory responses and the development of atherosclerosis.³⁶ Similarly, the downregulated USP25 expression in atherosclerotic lesions may be due to increased mRNA degradation mechanisms. However, further studies are warranted to clarify this issue, which represents a limitation of this study. Of note, signalling pathways induced by ox-LDL and TNF-α converge on RIPK1, a pivotal driver of inflammation in atherosclerosis.²⁵ Upon receptor activation, RIPK1 is modified with K63 polyubiquitin chains, which serve as scaffolds for the recruitment and docking of signalling molecules such as TAK1 and IKKs, leading to the activation of downstream MAPK and NF-κB pathways.^28,37,38 In the current study, ox-LDL-induced K63 polyubiquitination of RIPK1 was found to be inhibited by USP25, resulting in diminished NF-κB activity. Previously, USP25 has been shown to inhibit NF-κB activation induced by Interleukin-17 or lipopolysaccharide through deubiquitinating TRAFs, including TRAF3/5/6.13, 14, 15^,39 The present study is consistent with these previous findings that USP25 inhibits NF-κB activation and shows a key inhibitory effect of USP25 on RIPK1, another upstream hub of inflammatory signal transduction, expanding the regulatory mechanism of USP25 in the NF-κB signalling.

USP25 is a versatile DUB that can deubiquitinate various substrates and cleave multiple ubiquitin linkages. For example, USP25 inhibits the activity of TRAF5/6 by reducing K63 polyubiquitination.¹³ In contrast, USP25 stabilizes TRAF3^14,39 and SERCA2a¹⁷ by removing K48 polyubiquitin chains. Interestingly, USP25 can inhibit the activity or degradation of TRAF6 under different stimulation scenarios.^13,39 Here, we found that USP25 had no impact on RIPK1 protein abundance but reduced its activity by K63 deubiquitination. Furthermore, we discovered that the enzymatic activity of USP25 toward RIPK1 was mediated by the C178 active site, which is consistent with previous reports that the C178 residue is required for USP25 to deubiquitinate SHLD2,⁴⁰ SERCA2a,¹⁷ KEAP1,⁴¹ and TRAF3/5/6.^13,14,39

RIPK1 has been shown to promote atherogenesis by regulating inflammatory responses in endothelial cells and macrophages.²⁵ We found that USP25 was mainly expressed in macrophages in the atherosclerotic plaque and that USP25 affected atherosclerosis by regulating hematopoietic cells as evidenced by the bone marrow transplantation experiment, indicating that USP25 inhibits atherosclerosis predominantly by regulating macrophages. However, experiments with macrophage-specific USP25 knockout mice can provide more direct evidence for this contention, which is a limitation of this study that should be acknowledged.

Under different pathological conditions, USP25 exerts both detrimental and beneficial functions. For example, USP25 has been shown to promote Alzheimer's disease^11,12 and colon cancer.⁴² Consistently, pharmacological inhibition of USP25 with the small-molecule inhibitor AZ1 ameliorated these diseases in animal models.^12,42 In contrast, USP25 can attenuate cardiac hypertrophy induced by angiotensin II and transverse aortic constriction through stabilizing the anti-hypertrophy protein SERCA2a.¹⁷ In addition, USP25 ameliorates cerebral ischemic stroke injury by mitigating microglia-mediated neuroinflammation.¹⁸ These reports and the current study demonstrate that USP25 generally acts as a beneficial factor in the circulation system, highlighting USP25 as a potentially new therapeutic target for treating cardiovascular and cerebrovascular diseases.

Contributors

X.W. conceived the study. X.S., B.Z., and Y.X. performed the majority of experiments. X.D., J.C., M.Y., Y.Z., Z.Z., and Z.L. contributed to experiments and provided methodology. X.M. generated mouse strains. D.W. contributed to data analyses. X.W. verified the data. X.S. and X.W. analysed data and wrote the original manuscript. X.S., F.C., W.W., G.W., J.H., W.S., and X.W. edited the manuscript. W.W., G.W., J.H., W.S., and X.W. acquired funding. J.H., W.S., and X.W. supervised the project. All authors read and approved the final version of the manuscript.

Data sharing statement

Data generated in this paper will be shared upon reasonable request. Requests should be sent to the corresponding author Xu Wang, sunrim@163.com.

Declaration of interests

The authors declare no conflict of interest.