Deubiquitinating Enzyme USP19 Regulates Ferroptosis and Mitochondrial Damage in SH-SY5Y Cells by Targeting the NOX4 Protein

INTRODUCTION

Alzheimer’s disease (AD) is a neurodegenerative disease commonly seen in the elderly, characterized by irreversible memory loss and cognitive decline. The pathology of AD is characterized by the deposition of brain amyloid-β (Aβ) and the formation of neurofibrillary tangles, which ultimately leads to neuronal loss and cognitive decline [1]. Studies have shown that mitochondrial dysfunction is involved in the onset and development of AD [2]. Mitochondria are the powerhouse of the cell and provide energy to the cell through oxidative phosphorylation. When mitochondrial dysfunction occurs, the mitochondrial membrane potential decreases, ATP is depleted in large amounts, and reactive oxygen species (ROS) accumulate, promoting the deposition of Aβ. At the same time, the cytoplasmic splitting protein Drp1 will be recruited to the outer membrane of the mitochondria, triggering the splitting of the mitochondria and damage. And mitochondrial damage has been shown to be associated with AD lipid peroxidation and ferroptosis [3].

Ferroptosis is a type of cell death catalyzed by iron ions and ROS-induced lipid peroxidation, which is characterized by accumulation of iron ions and lipid peroxidation [4]. NADPH oxidases (NOX) are a class of transmembrane oxidoreductases that produce large amounts of ROS [5]. Excessive production of ROS can then lead to ferroptosis [6]. There are seven NADPH oxidases in humans: NOX1, NOX2, NOX3, NOX4, NOX5, DUOX1, and DUOX2 [7]. Among the NOX isoforms, NOX4 induces neuronal oxidative stress and iron overload promoting neuronal ferroptosis after cerebral hemorrhage [8]. However, the mechanism by which NOX4 regulates cellular ferroptosis through oxidative stress to promote AD progression is unknown.

Ubiquitination and deubiquitination are important physiological processes related to protein-specific degradation in cells. Deubiquitinating enzymes (DUB) affect the fate of proteins by removing ubiquitin chains from protein substrates [9]. The ubiquitin specific protease (USP) family is the largest subfamily of ubiquitin decarboxylase. In neurodegenerative diseases such as AD and Parkinson’s disease (PD), the ubiquitination and deubiquitination systems are unable to regulate nerve fibers, disrupt mitochondrial homeostasis, and promote ubiquitination of pathogenic proteins/damaged mitochondria [9, 10]. Rong et al. reported that USP11 promotes autophagy activation by stabilizing Beclin 1, thereby leading to ferroptosis [11]. Thus, inhibition of the activity of the relevant DUBs may have a therapeutic effect on neurological disorders. In the present study, we aimed to investigate the role of USP19 in stabilizing NOX4 and gained insight into the mechanism of ferroptosis by USP19 in an AD cell model.

MATERIALS AND METHODS

RESULTS

The USP19 maintains NOX4 protein stability

Figure 1A showed that USP19 regulated the stability of NOX4 protein by detecting the intensity of the fluorescence released from the luciferase-catalyzed substrate; and NOX4 protein level decreased after knockdown of USP19 in 293T cells transfected with shUSP19 (Fig. 1B).

OPEN IN VIEWER

Fig. 1

Screening of deubiquitinating enzyme USP family genes affecting NOX4 protein stability: A) biochemiluminescence instrument to detect the intensity of fluorescence released from the substrate catalyzed by this luciferase; B) western blot assay to detect the protein level of NOX4-Dluc.

Regulation of ferroptosis and mitochondrial damage in SH-SY5Y cells by USP19

Construction and characterization of AD cell models

To investigate the effect of USP19 on AD model cells, cell modeling using Aβ_1-40 was performed in this study. SH-SY5Y cells were cultured in vitro and treated with 5 μmol/L Aβ_1-40 added to the culture medium for 0 h, 24 h, and 48 h, respectively. The results showed that the cell viability gradually decreased with the prolongation of time, dropping to 84% at 24 h, and to 66% at 48 h (Fig. 2A). The results of Annexin V-FITC flow assay showed that the number of apoptotic SH-SY5Y cells were gradually increased with the prolongation of the drug intervention time (Fig. 2B, C). the MDA content and the content of Fe²⁺ in the cells also increased accordingly (Fig. 2D, E), indicating that the cells were subjected to enhanced oxidative stress and ferroptosis after the intervention of Aβ_1-40, suggesting that the construction of the AD cell model was successful.

OPEN IN VIEWER

Fig. 2

Construction and characterization of AD cell model: A) Cell model construction treatment, Aβ_1-40 0 h, 24 h, 48 h cell viability; B, C) Annexin V-FITC flow assay for apoptosis; D) Changes in the MDA content of cells after Aβ_1-40 treatment; E) Changes in intracellular Fe²⁺ content after cell Aβ_1-40 treatment; ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

Knockdown of USP19 reduces ferroptosis and mitochondrial damage in SH-SY5Y cells

To investigate the regulatory effects of USP19 on ferroptosis and mitochondrial damage in AD model cells, USP19 vector was designed, constructed and then functionally verified in this study (Fig. 3A). AD model was constructed by treating SH-SY5Y cells with 5 of μmol/L Aβ_1-40 for 24 h. The shUSP19 vector and control were transfected respectively, and the experimental groups were as follows: Mock (negative control), Aβ_1-40, Aβ_1-40  + shNC, Aβ_1-40  + shUSP19 groups. The results of qPCR and western blot experiments showed that the mRNA and protein expression of NOX4 was significantly higher in the Aβ_1-40 and Aβ_1-40  + shNC groups compared with the Mock group, and the mRNA expression of NOX4 was higher and the protein expression was decreased after knockdown of USP19 (USP19 ubiquitylation affects the post-transcriptional level and does not affect the gene) (Fig. 3B–D). Our results further showed that the mitochondrial fusion Mfn1 and Mfn2 protein expression levels were decreased in the Aβ_1-40 and Aβ_1-40  + shNC groups, and the protein expression of the splitting protein Drp1 was elevated, and the expression of the above mentioned proteins was closer to the normal level after knocking down USP19 (Fig. 3D); CCK8 results showed a decrease in proliferation after the addition of Aβ_1-40, and a significant increase in proliferation levels after knockdown of USP19 (Fig. 3E); AnnexinV-FITC flow-through assay results showed an increase in apoptosis after the addition of Aβ_1-40, and a significant decrease in apoptosis after knockdown of USP19 (Fig. 3F, G).

OPEN IN VIEWER

Fig. 3

Regulatory effects of USP19 on AD cells. A) Vector construction and functional validation; B) PCR detection of cellular NOX4 and USP19 gene expression; C) cellular Aβ_1-40 treatment changes in MDA content; D) western blot detection of protein expression, ns indicates no statistical significance compared to Mock group; E) cellular Aβ_1-40 treatment changes in cellular viability; F) flow chart; G) Apoptosis rate analysis. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

The relative ratio of red and green fluorescence is commonly used to measure the changes in mitochondrial membrane potential. The results in Fig. 4A and B showed that the green/red fluorescence ratio increased rapidly in the Aβ_1-40, Aβ_1-40  + shNC group compared with the Mock group (p < 0.01), and the ratio decreased rapidly after knockdown of USP19. The fluorescent probe DCF-H dye assay for ROS showed that the fluorescence intensity of the Aβ_1-40, Aβ_1-40  + shNC group increased rapidly (p < 0.01) compared with that of the Mock group, and the intensity decreased rapidly after knockdown of USP19. These also suggest that knockdown of USP19 reduces mitochondrial damage (Fig. 4C, D).

OPEN IN VIEWER

Fig. 4

Modulation of mitochondrial damage by USP19 in AD cells. A, B) JC-1 mitochondrial membrane potential assay; C, D) Fluorescent probe DCF-H dye assay; ^***p < 0.001, ^****p < 0.0001.

The results of the assay at the biochemical level showed (Fig. 5): compared with the MOCK group, the MDA and intracellular Fe²⁺ content increased rapidly in the Aβ_1-40 and Aβ_1-40  + shNC groups (p < 0.01), and the MDA and intracellular Fe²⁺ content decreased rapidly after knockdown of USP19; whereas the GSH/GSSG ratio was the opposite. These suggested that knockdown of USP19 could attenuate the damage of Aβ_1-40 to SH-SY5Y cells and reduce ferroptosis.

OPEN IN VIEWER

Fig. 5

Exploration of USP19 on lipid peroxidation and intracellular Fe²⁺ content in SH-SY5Y cells. A) Changes in cellular MDA content; B) changes in GSH/GSSG ratio; C) changes in intracellular Fe²⁺ content; ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

Mechanism of action of USP19 on the target protein NOX4 in neuronal cells

To investigate the mechanism of action of USP19 on NOX4, we performed immunoprecipitation experiments. The results showed that NOX4 protein and USP19 protein had a reciprocal relationship (Fig. 6B). And the results of ubiquitination assay showed (Fig. 6C): the degree of ubiquitination increased after knocking down USP19, suggesting that USP19 protein has the role of stabilizing NOX4 protein. SH-SY5Y cells were transfected with shUSP19 knockdown vector, FLAG-NOX4 overexpression vector, and exogenously applied with 100 μg/mL of actidione CHX treatment for 0, 6, 9, and 18 h. The results of western blot showed that NOX4 protein levels gradually decreased with time extension (Fig. 6E).

OPEN IN VIEWER

Fig. 6

Mechanism of USP19 action on target protein NOX4 in neuronal cells. A) Vector construction and functional validation; B) USP19 interaction assay with NOX4; C) ubiquitination assay; D, E) half-life assay. ^***p < 0.001. OE, observed to expected; TCL, total cell lysis.

USP19 targeting of NOX4 stability affects the regulation of ferroptosis and mitochondrial damage in SH-SY5Y cells

After transfection with shUSP19, the western blot results (Fig. 7) showed that the protein expression of NOX4, Drp1 decreased (p < 0.01), while Mfn1 and Mfn2 protein expression increased rapidly; on the contrary, the protein expression of Drp1 increased rapidly (p < 0.01), and Mfn1 and Mfn2 protein expression decreased rapidly in the NOX4 over-expression vector group. The same was true for apoptosis and mitochondrial membrane potential changes measured by flow cytometry. These all suggest that USP19 regulates ferroptosis and mitochondrial damage in the AD models by targeting NOX4.

OPEN IN VIEWER

Fig. 7

Regulation of AD cells by USP19 targeting NOX4. A, B) western blot assay for protein expression; C) cellular viability assay; D, F) apoptosis assay; E, G) cellular mitochondrial membrane potential assay. ns, no statistical significance, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001.

Biochemical analysis (Fig. 8): ROS, MDA, and intracellular Fe²⁺ content increased rapidly in the NOX4 overexpression vector group compared with the shUSP19 group (p < 0.01). In contrast, the GSH/GSSG ratio decreased rapidly (p < 0.01). These suggest that USP19 regulates ferroptosis and mitochondrial damage in vitro AD models expressing Aβ_1-40 by targeting NOX4 stability.

OPEN IN VIEWER

Fig. 8

Exploration of USP19-targeted NOX4 stability regulating lipid peroxidation and intracellular Fe²⁺ content in AD cells. A, B) ROS detection; C) changes in cellular MDA content; D) changes in intracellular Fe²⁺ content; E) changes in GSH/GSSG ratio, ns indicates no statistical significance, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

DISCUSSION

In this study, we screened USP19, a deubiquitinating enzyme coding gene that maintains the stability of NOX4 protein. we found that knockdown of this gene attenuated mitochondrial damage and ferroptosis in SH-SY5Y cells caused by Aβ_1-40, and moreover, we demonstrated that USP19 interacts with the NOX4 protein. USP19 can stabilize the NOX4 protein by targeting it that reduces ferroptosis.

Ubiquitination/deubiquitination is a dynamic regulatory system of bi-directional protein modification, which is handled in vivo by the ubiquitin ligase (E1-E2-E3) and deubiquitinating enzyme (DUB) families, respectively. Bidirectional modification of ubiquitination determines the development of many physiological events and is considered to be a highly promising new therapeutic target for neurological and other diseases [12]. USP19 is a 150 kDa tail-anchored ubiquitin-specific protease. Studies have shown that USP19 is associated with the development of various diseases, such as muscle diseases, obesity and cancer [13–15]. However, whether USP19 regulates neuronal mitochondrial damage remains unclear. Our study results suggest that knockdown of USP19 affects NOX4 protein expression, reduces mitochondrial damage and promotes cell proliferation.

Ferroptosis, a non-apoptotic programmed cell death caused by catastrophic accumulation of ROS, has received increasing attention in recent years, with a growing number of reports suggesting a role for aberrant iron homeostasis in AD pathophysiology [16]. Bao et al. reported that Ferroportin1, the only mammalian non-heme iron export identified, is downregulated in the brains of APPswe/PS1dE9 mice (an AD mouse model) and AD patients; and that administration of ferroptosis-specific inhibitor effectively reduces Aβ aggregation-induced neuronal death and memory impairment [17]. In this study, we found that intracellular iron levels were elevated after Aβ_1-40 intervention in SH-SY5Y cells, suggesting that Aβ_1-40 intervention caused abnormally elevated intracellular iron homeostasis in SH-SY5Y cells; whereas intracellular Fe²⁺-containing levels were decreased by knockdown of USP19.

Mitochondria are an important source of ROS in most mammalian cells. Excessive ROS production will lead to mitochondrial dysfunction and an imbalance in a group of proteins that contain structural domains of GTPases, such as Drp1 (which facilitates mitochondrial division) and Mfn1, Mfn2, etc. (which facilitate mitochondrial fusion) [18]. When mitochondrial division is excessive, mitochondrial fragmentation leads to a decrease or even loss of mitochondrial membrane potential and an increase in mitochondrial outer membrane permeability. The role of mitochondrial dysfunction in ferroptosis is controversial, but there is also growing evidence that mitochondria play an important role in ferroptosis. In this study, it was found that after Aβ_1-40 intervention in SH-SY5Y cells, the expression levels of mitochondrial fusion proteins, Mfn1 and Mfn2 were decreased, the protein expression of the splitting protein Drp1 was elevated, while the mitochondrial membrane potential was decreased, suggesting that both mitochondrial dysfunction and iron-mediated cell death occurred after Aβ_1-40 intervention in SH-SY5Y cells; whereas, the above-mentioned protein expression and mitochondrial membrane potential were adjusted to normal after knockingdown USP19.

ROS are products of normal aerobic metabolism carried out by the organism and are important markers of cellular oxidative damage. Overproduction of ROS and activation of lipid peroxidation are associated with cellular ferroptosis [19]. Whereas excessive oxidative stress (ROS production, mitochondrial dysfunction, lipid peroxidation) promotes the production of Aβ, the overproduction of Aβ leads to lipid peroxidation, creating a vicious cycle [20]. NOX is a family of ROS-producing enzymes [21]. NOX4 is a member of the NOX family and is widely expressed in the brain [22]. NOX4 knockout mice were found to have attenuated brain damage and reduced oxidative markers [23]. Whole organism knockout of the NOX4 gene and neuronal knockdown in mice reduced the accumulation of pathological tau [24]. NOX4 protein expression was significantly higher in APP/PS1 mice than in wild type mice [25].

In this study, NOX4 protein expression was increased after Aβ_1-40 intervention in SH-SY5Y cells, whereas knockdown of USP19 led to a decrease in NOX4 protein expression, increased cell viability, decreased apoptosis, and decreased intracellular Fe²⁺ content. Meanwhile, experiments also confirmed that USP19 and NOX4 interacted, and these suggested that knockdown of USP19 attenuated neuronal mitochondrial damage and iron death by targeting NOX4.

It is necessary to acknowledge the limitations of this work. We do not have a positive control for the ferroptosis to show that NOX4 is specifically activated in USP19 over-expressing conditions, which may have affected the integrity of this work. Additionally, as this study aims to explore USP19 regulates ferroptosis and mitochondrial damage in SH-SY5Y cells by targeting the NOX4 protein, we have not yet performed animal experiments as well as clinical work, which may affect the generalizability of this work. In our future work, we will refine the experimental design and we intend to investigate the interaction between USP19 and NOX4 protein is the same for AD animal models or AD patients.

AUTHOR CONTRIBUTIONS

Wenzhen Yu (Methodology; Project administration; Writing – original draft); Shuting Zhuang (Data curation; Methodology; Writing – review & editing); Mengxiong Zhan (Resources; Software; Supervision); Yong Chen (Formal analysis; Methodology); Jieping Zhang (Supervision; Visualization); Ling Chen (Project administration; Validation); Chunxiang Tu (Methodology; Project administration); Linfei Zheng (Formal analysis; Investigation); Shi Chen (Conceptualization; Funding acquisition).