Binding of α-synuclein to ACO2 promotes progressive mitochondrial dysfunction in Parkinson's disease models

Abstract

The accumulation of α-synuclein (α-syn), a key protein in Parkinson's disease (PD), contributes to progressive neuronal damage associated with mitochondrial dysfunction and interactions with various proteins. However, the precise mechanism by which α-syn affects energy metabolism remains unclear. In our study, we used human α-syn (hα-syn) transgenic mice, which exhibit progressive neuronal decline. Through an immunoprecipitation assay specific to hα-syn, we identified an enzyme in the mitochondrial tricarboxylic acid (TCA) cycle as a binding partner—mitochondrial aconitase 2 (ACO2), which converts citrate to isocitrate. Hα-syn increasingly interacted with ACO2 in mitochondria as mice aged, correlating with a progressive decrease in ACO2 activity. The overexpression of ACO2 and the addition of isocitrate, a downstream metabolite of ACO2, were observed to alleviate hα-syn-induced mitochondrial dysfunction and cytotoxicity. Furthermore, we designed an interfering peptide to block the interaction between ACO2 and hα-syn, which showed therapeutic effects in reducing hα-syn toxicity in vitro and in vivo. Our research establishes a direct link between α-syn and the TCA cycle and identifies ACO2 as a promising therapeutic target for improving mitochondrial function and reducing α-syn neurotoxicity in PD.

1. Introduction

Parkinson's disease (PD) is characterized as a progressive neurodegenerative disease for which α-synuclein (α-syn) accumulation is a well-known pathologic biomarker [1,2]. α-Syn is implicated in the progression of PD, including causing mitochondrial dysfunction [[3], [4], [5], [6], [7]]. The following mechanisms have been proposed: 1) Disrupted mitochondrial dynamics: α-syn aggregates can disrupt the equilibrium between mitochondrial fission and fusion, resulting in a decrease in mitochondrial mass and impairment of oxidative phosphorylation [8,9]. 2) Oxidative stress: α-syn can enhance the production of reactive oxygen species (ROS), leading to oxidative damage to mitochondrial structures and impairing their function [[10], [11], [12]]. 3) Impaired mitochondrial protein import: α-syn aggregates can disrupt the function of the translocase of the outer membrane (TOM) complex, leading to reduced mitochondrial respiration and hindered protein import [13]. 4) Altered calcium homeostasis: α-syn interacts with mitochondrial porin, known as voltage-dependent anion-selective channel (VDAC), which is situated in the outer mitochondrial membrane and maintain the metabolic crosstalk between mitochondria and the rest of the cell. α-Syn blocks VDAC and changes ion selectivity, leading to higher Ca²⁺ flux through the channel [14]. Despite the evidence, the molecular mechanisms underlying α-syn-induced disturbances in mitochondrial energy metabolism have yet to be elucidated. It is likely that as PD progresses and α-syn continues to accumulate, new mechanisms of neurotoxicity related to high α-syn levels may develop.

The importance of mitochondrial energy metabolism in the pathogenesis of PD is increasingly recognized, with mounting evidence linking impaired energy metabolism to neurodegeneration [15,16]. The mitochondrial tricarboxylic acid (TCA) cycle serves as a central component of energy metabolism, with mitochondrial aconitase 2 (ACO2) playing a crucial role in converting citrate to isocitrate [17,18]. ACO2 also maintains mitochondrial function and integrity [19]. Reduced ACO2 activity is associated with worsening clinical symptoms in PD patients due to impaired energy production and increased oxidative stress, which exacerbates neuronal damage [20]. ACO2 is sensitive to ROS due to its [4Fe–4S]²⁺ group, which is prone to inactivation [21].

In our study, since transgenic (TG) mice expressing human α-syn (hα-syn) exhibit progressive neurodegeneration, we performed experiments to search for hα-syn binding proteins during disease progression. It has been shown that hα-syn increasingly interacts with ACO2 as hα-syn accumulates in the mitochondria. This interaction, coupled with elevated ROS levels, likely contributes to diminished ACO2 activity. Therapeutic interventions to rescue ACO2 deficits include overexpression of ACO2, providing a downstream product of ACO2 (isocitrate) and using interference peptides to block the hα-syn/ACO2 interaction. These methods alleviate hα-syn cytotoxicity both in vitro and in vivo.

2. Results

2.1. Hα-syn interacts with ACO2 in hα-syn transgenic mice

To investigate the progressive toxicity of α-syn, we used TG mice overexpressing hα-syn in neurons (Thy1-SNCA). These mice exhibited motor dysfunction at 12 and 15 months of age, but not at 9 months (Fig. 1A–D and S1A-C). Consistent with the observed behavioral changes, western blot analysis revealed that the levels of hα-syn increased with age in the striatum and midbrain of TG mice (Fig. 1E). To elucidate the mechanisms underlying the progressive toxicity of hα-syn, we searched for binding proteins of hα-syn in various brain tissue samples collected from wild-type (WT) and TG mice at 6 and 12 months of age. Using immunoprecipitation with a hα-syn antibody and mass spectrometry (MS), we identified an enzyme in the TCA cycle, ACO2, particularly enriched in 12-month-old TG mice (Fig. 1F). In the brain slices from the substantia nigra of 12-month-old WT and TG mice, we performed immunostaining with antibodies against hα-syn, ACO2, and tyrosine hydroxylase (TH), a marker protein for dopaminergic neurons. The results showed that hα-syn co-localized with ACO2 in TG mice (Fig. S1D). The interaction between hα-syn and ACO2 was confirmed through co-immunoprecipitation (Co-IP) in vitro (Fig. 1G). As a toxic form of hα-syn [22], the levels of phosphorylated α-syn (pα-syn) also increased with age in TG mice and interacted with ACO2 in both in vitro and in vivo experiments (Figs. S1E–G). However, the interaction of pα-syn with ACO2 was increased in TG mice compared to endogenous pα-syn in WT mice. These findings imply that the interaction between hα-syn and ACO2 may contribute to the progressive toxicity observed in hα-syn overexpressing TG mice.

Fig. 1.

Hα-syn interacts with ACO2 in hα-syn transgenic mice. (A) Mouse experimental design. Motor behavior was measured in WT and TG mice of different ages starting at 9 months. (B–D) Pole test (B), rotarod test (C), and open field test (D) were used to assess motor behavior of 9-, 12-, and 15-month-old WT and TG mice. Right: The representative trajectory diagrams of mice in the open field test. n = 8. (E) Levels of hα-syn expression in the striatum (left) and midbrain (right) of 9-, 12-, and 15-month-old WT and TG mice were determined by western blot. (F) SDS-PAGE and Coomassie blue staining were used to detect the enrichment of hα-syn and interacting proteins. The brain samples were collected from 6-month-old WT and 6- and 12-month-old TG mice. IgG served as a control. (G) Co-IP experiments showing the interaction between hα-syn and ACO2 using hα-syn antibody (above) and ACO2 antibody (below) in brain tissues of 12-month-old WT and TG mice. Homologous IgG was used as the experimental control. Hα-syn and pα-syn pure proteins (100 ng) were used as positive protein controls. 6 M, 9 M, 12 M, 15 M = 6, 9, 12, and 15 months, respectively. Data are expressed as the mean ± SEM. P values were determined by unpaired t-test (B–D). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs. WT. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

2.2. The increasing interaction between hα-syn and ACO2 in mitochondria affects ACO2 activity

Given the observed interaction between hα-syn and ACO2 in older TG mice, as well as ACO2's mitochondrial localization, we aimed to determine whether this interaction is age-related and specific to mitochondria. We collected mitochondrial and cytosolic fractions from the striatum and midbrain tissues of WT and TG mice at different ages. Western blot analysis showed an increase in hα-syn levels in the mitochondrial protein fractions during aging in TG mice (Fig. 2A and S2A). While the ACO2 protein level remained unchanged, the interaction between ACO2 and hα-syn was enhanced (Fig. 2B). This increased hα-syn/ACO2 interaction may inhibit ACO2's enzymatic activity. The measured ACO2 activity showed a significant decrease in the striatum and midbrain of 12- and 15-month-old TG mice, with no such reduction observed in 9-month-old TG mice (Fig. 2C–D). This decline in ACO2 activity correlated with a similar decrease in mitochondrial complex I activity (Fig. 2E–F). These results indicate that the increased interaction between hα-syn and ACO2 may inhibit ACO2 activity and downregulate mitochondrial function.

Fig. 2.

The increasing interaction between hα-syn and ACO2 in mitochondria affects ACO2 activity. (A) Western blots showing the expression of hα-syn and ACO2 in cytosolic (Cyto, left) and mitochondrial (Mito, right) lysates isolated from the striatum of 9-, 12-, and 15-month-old WT and TG mice. VDAC1 served as a mitochondrial marker and β-actin as a loading control. (B) Co-IP experiments showing the interaction between hα-syn and ACO2 using the ACO2 antibody in Mito lysates isolated from the striatum (left) and midbrain (right) of 9-, 12-, and 15-month-old WT and TG mice. The hα-syn/ACO2 interaction increased in the mitochondria of TG mice with age. Homologous IgG was used as the experimental control. (C–F) ACO2 activity (C–D) and mitochondrial complex I activity (E–F) in the striatum and midbrain of 9-, 12-, and 15-month-old WT and TG mice. n = 8. (G–J) Levels of MDA (G), GSH (H), GSSG (I), and Fe²⁺ (J) in the striatum of 9-, 12-, and 15-month-old WT and TG mice. n = 8. MDA: Malondialdehyde; GSH: glutathione; GSSG: oxidized glutathione. (K) With increased hα-syn protein levels and the same amount of GST-ACO2 protein in the system, GST pulldown experiments showing the interaction between GST-ACO2 protein and different levels of hα-syn protein. Right: Quantitation of hα-syn protein-to-GST-ACO2 protein level ratio after GST-pulldown. n = 3. (L) ACO2 protein activity assay. The system was supplemented with ACO2 protein and different amounts of hα-syn protein (0.25, 0.5, 1, and 2 μg) or 2 μg Pgk1 protein. n = 4. (I) ACO2 protein activity assay. The system was supplemented with ACO2 protein, 2 μg of hα-syn protein, and different concentrations of citrate (5, 10, and 20 mM). n = 4. Data are expressed as the mean ± SEM. P values were determined by unpaired t-test (C–J) and one-way ANOVA (K–M). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 vs. WT (C–J), 0.25 μg of hα-syn (K), or ACO2 (L–M), ###P < 0.001 vs. ACO2 + hα-syn (I).

Mitochondrial dysfunction can exacerbate oxidative stress, which in turn inhibits ACO2 activity. Malondialdehyde (MDA), glutathione (GSH), and oxidized glutathione (GSSG) are critical indicators of oxidative stress [23]. We found elevated levels of MDA, GSH, and GSSG in the striatum of TG mice at 12 and 15 months (Fig. 2G–I). This oxidative stress appears to impair ACO2 by promoting the loss of Fe²⁺ and increasing intracellular Fe²⁺ levels (Fig. 2J), suggesting that oxidative stress may also regulate ACO2 activity.

To investigate the direct impact of hα-syn on ACO2's enzymatic activity, we purified hα-syn and ACO2 proteins in vitro. In GST-pulldown experiments, we loaded the same amount of GST-ACO2 protein onto a GST column and added increasing amounts of hα-syn protein. The results showed that more hα-syn protein could interact with ACO2 when more hα-syn was available (Fig. 2K). This increase in hα-syn concentration directly reduces ACO2 activity (Fig. 2L). Control experiments using phosphoglycerate kinase 1 (Pgk1) did not affect ACO2 activity, indicating that the hα-syn/ACO2 interaction specifically inhibits ACO2. Moreover, the interaction of pα-syn with ACO2 or a disease-related mutant form of ACO2 (ACO2-A252T) [20] followed a pattern like hα-syn (Figs. S2B–E).

One proposed mechanism for hα-syn's inhibition of ACO2 is substrate binding interference. To test this, we measured the enzymatic activity of purified ACO2. Normally, ACO2 activity was enhanced by citrate in a dose-dependent manner (Fig. 2M). In the presence of hα-syn, ACO2 activity decreased, but adding citrate partially restored its activity (Fig. 2M). This suggests that hα-syn binding may obstruct citrate binding.

2.3. Hα-syn reduces cellular ACO2 activity in a dose-dependent manner

To further elucidate whether the accumulation of hα-syn inhibits cellular ACO2 activity in a dose-dependent manner, SH-SY5Y cells were transfected with 1, 2, or 3 μg of human SNCA plasmids (encoding hα-syn) for 24 and 48 h. This experiment resulted in gradient levels of hα-syn and pα-syn, without altering ACO2 levels (Fig. 3A). However, ACO2 activity exhibited a gradient decrease (Fig. 3B). Correlation analysis revealed an inverse relationship between ACO2 activity and levels of hα-syn and pα-syn, showing a linear correlation (Fig. 3C–D). The decline in ACO2 activity was accompanied by reductions in intracellular ATP levels (Fig. 3D) and cell viability (Fig. S3A–B). These results suggest that increased hα-syn levels reduce ACO2 activity in a dose-dependent manner in SH-SY5Y cells.

Fig. 3.

Hα-syn reduces cellular ACO2 activity in a dose-dependent manner. SH-SY5Y cells were transfected with vector, 1, 2, or 3 μg of human SNCA plasmids for 24 and 48 h. (A) Representative immunoblot of hα-syn, pα-syn, and ACO2 of lysates from SH-SY5Y cells. Below: Quantitative analysis of hα-syn, pα-syn, and ACO2 levels. β-actin was used as a loading control. n = 3. (B) ACO2 activity was determined in SH-SY5Y cells. n = 6. (C) Correlation analysis between ACO2 activity and the hα-syn:β-actin (left) or pα-syn:β-actin ratio (right). n = 26. (D) ATP levels were determined in SH-SY5Y cells. n = 6. (E) ACO2 activity was measured in SH-SY5Y cells transfected with either vector or SNCA plasmids for 24, 48, and 72 h, or treated with NAC for the same durations. n = 6. Data are expressed as the mean ± SEM. P values were determined by one-way ANOVA. ∗P < 0.05, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, vs. Vector (A-B, D) or SNCA + NAC (E).

Overexpression of hα-syn also caused an elevation in ROS levels (Fig. S3C). To mitigate the effects of oxidative stress on ACO2 activity, we employed N-acetylcysteine (NAC), a ROS scavenger, which effectively reduced ROS levels (Fig. S3D). NAC treatment partially alleviated the decline in ACO2 activity after prolonged overexpression of hα-syn (Fig. 3E). These findings indicate that oxidative stress also contributes to the inhibition of ACO2 activity; however, it does not appear to be the primary cause.

2.4. ACO2 overexpression rescues mitochondrial dysfunction and improves motor behavior and TH levels in TG mice

To investigate whether ACO2 overexpression can alleviate hα-syn-induced cytotoxicity, we used lentiviral expression of ACO2 and hα-syn in primary neuronal cultures (Fig. 4A–B and S4A–B). In primary neurons infected with SNCA lentivirus, overexpression of ACO2 showed an increase in ACO2 activity, mitochondrial complex I activity, ATP levels, and cell viability (Fig. 4C–D and S4C–D), as well as reductions in intracellular ROS levels (Fig. 4E). Notably, in WT neurons, ACO2 overexpression did not enhance mitochondrial complex I activity or ATP levels, likely due to saturation of the TCA cycle.

Fig. 4.

ACO2 overexpression rescues mitochondrial dysfunction and improves motor behavior and TH levels in TG mice. (A–B) Western blot (A) and immunofluorescence images (B) showing the expression of ACO2 in primary neurons infected with vector and ACO2 lentivirus. β-tubulin Ⅲ (green) was used as a neuron marker. Scale bar = 75 μm. (C–D) Mitochondrial complex I activity (C) and ATP levels (D) were determined in primary neurons infected with vector1 (Ubi-MCS-3FLAG-SV40-puromycin), SNCA (correspond to vector1), vector2 (Ubi-MCS-3FLAG-SV40-Cherry-IRES-puromycin), ACO2 (correspond to vector2), SNCA + vector2, SNCA + ACO2 lentivirus (co-infection of SNCA with vector2, or ACO2 lentivirus). n = 6. (E) ROS levels were assessed by H2DCFDA staining in primary neurons. Right: quantitative analysis of average fluorescence intensity. Scale bar = 100 μm. n = 6. (F) Schematic representation of AAV virus injection. AAV9-EGFP (Vector), AAV9-ACO2-EGFP (ACO2), and AAV9-ACO2-A252T-EGFP (ACO2-A252T) viruses were administered via bilateral striatal injection in 12-month-old mice and tested 21 d after injection. (G–I) Pole test (G), rotarod test (H), and open field test (I) were used to assess motor behavior of mice in each group. Right: The representative trajectory diagrams of mice in open field test. n = 8. (J) Immunofluorescence imaging of hα-syn (red), ACO2 (magenta), and EGFP in the striatum of mice in each group. Scale bar = 20 μm. The images on the right are an enlargement of the area indicated by the white box. Scale bar = 10 μm. (K) Western blot analysis of hα-syn, TH, and ACO2 levels in the striatum of mice in each group. β-actin was used as a loading control. Quantitative analysis of hα-syn, TH, and ACO2 levels are shown below. n = 8. Data are expressed as the mean ± SEM. P values were determined by one-way ANOVA. Ns: no significant differences, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To confirm that the enzymatic activity of ACO2 is crucial for the rescue effect, SH-SY5Y cells were transfected with a mutant variant of ACO2 (ACO2-A252T-3xFlag) (Fig. S4E), known to exhibit a 63.3 % reduction in enzymatic activity [20]. Despite displaying a similar subcellular distribution as WT ACO2 (Fig. S4F), the mutant failed to restore ACO2 activity, mitochondrial function, ATP levels, or cell viability (Figs. S4G–K). These results indicate that a loss-of-function ACO2 mutant does not provide a protective effect against hα-syn toxicity.

We then assessed the effects of ACO2 overexpression in vivo by bilaterally injecting AAV9 viruses encoding either EGFP (Vector), ACO2-EGFP (ACO2), or ACO2-A252T-EGFP (ACO2-A252T) into the striatum of 12-month-old WT and TG mice (Fig. 4F). The ACO2 and EGFP in the ACO2-EGFP and ACO2-A252T-EGFP vectors are expressed by separate promoters. Motor function assessments conducted 21 d post-injection demonstrated that ACO2 overexpression improved motor performance in TG mice, while the ACO2-A252T variant did not (Fig. 4G–I). Expression of EGFP confirmed successful transfection in the striatum and substantia nigra (Fig. 4J and S5A). Western blot analysis showed increased ACO2 levels in the striatum and midbrain of TG mice in the ACO2 and ACO2-A252T groups, with no significant differences in hα-syn levels (Fig. 4K and S5B). Moreover, ACO2 overexpression prevented the hα-syn-induced decline in TH levels in the striatum and midbrain, whereas ACO2-A252T overexpression did not have the same effect (Fig. 4K and S5B). Immunostaining with TH antibody in brain slices showed a similar rescue effect (Fig. S5C). ACO2 overexpression also increased both ACO2 activity and mitochondrial complex I activity in the striatum and midbrain, whereas ACO2-A252T did not (Figs. S5D–G). Control experiments in WT mice revealed that neither ACO2 nor ACO2-A252T injections affected motor behavior (Figs. S5H–J), although ACO2 overexpression did enhance striatal ACO2 activity without impacting mitochondrial complex I activity or TH levels (Figs. S5K–M). Collectively, these data demonstrate that ACO2 overexpression can rescue motor behavioral deficits in TG mice and elevate TH levels in the brain, emphasizing its potential therapeutic role in mitigating hα-syn toxicity.

2.5. Reduced ACO2 activity impairs the TCA cycle and isocitrate supplementation rescues mitochondrial function

Given the pivotal role of ACO2 in the TCA cycle, reduced ACO2 activity may cause impairment of the TCA cycle. We analyzed TCA cycle metabolites in 12-month-old WT and TG mice using high-performance liquid chromatography/mass spectrometry (LC-MS/MS). The results showed that citrate levels increased while isocitrate levels decreased in TG mice, consistent with the alterations caused by the inhibition of ACO2 [19]. Additionally, the levels of succinate and malate were also decreased (Fig. 5A).

Fig. 5.

Reduced ACO2 activity impairs the TCA cycle and isocitrate supplementation rescues mitochondrial activity. (A) Concentration of the TCA cycle metabolites in the midbrain of 12-month-old WT and TG mice by LC-MS/MS. These intermediates, the levels of which were significantly increased or decreased, are shown in red and blue, respectively. Black: no change, grey: not determined. n = 6. (B) Flow chart of SNCA lentivirus infection and isocitrate addition in primary neurons. The primary neurons were infected with SNCA lentivirus on day 3 and treated with 2.5 or 5 mM isocitrate for 24 or 48 h on days 5 or 6 before being detected on day 7. (C) Detection of intracellular isocitrate concentration. n = 3. (D–E) ATP levels (D) and mitochondrial complex I activity (E) in primary neurons supplemented with 2.5 or 5 mM isocitrate for 48 h following infection with SNCA lentivirus. n = 6. (F) ROS levels were assessed by H2DCFDA staining. Right: quantitative analysis of average fluorescence intensity. Scale bar = 100 μm. n = 6. (G–H) ATP levels (G) and mitochondrial complex I activity (H) in primary neurons supplemented with 5 mM succinate or malate for 48 h following infection with SNCA lentivirus. n = 6. Data are expressed as the mean ± SEM. P values were determined using an unpaired t-test (A) and one-way ANOVA (C–H). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ###P < 0.001 vs. SNCA (C). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To restore mitochondrial dysfunction induced by hα-syn expression, we supplemented primary neurons with isocitrate. The addition of 2.5 and 5 mM isocitrate increased intracellular isocitrate concentrations without affecting cell viability, whereas a 10 mM concentration exhibited cytotoxic effects (Figs. S6A–C). Subsequently, primary neurons infected with SNCA lentivirus were treated with isocitrate at concentrations of 2.5 and 5 mM for 24 and 48 h (Fig. 5B). This treatment increased intracellular isocitrate levels (Fig. 5C) and enhanced cell viability, with the 48-h treatment showing superior efficacy (Figs. S6D–E). In terms of mitochondrial function, the addition of 5 mM isocitrate for 48 h in primary neurons overexpression hα-syn resulted in increased ATP levels (Fig. 5D), mitochondrial complex I activity (Fig. 5E), decreased ROS levels (Fig. 5F), restoration of MMP (Fig. S6F), and prevention of mitochondrial membrane permeability transition pore (mPTP) opening (Fig. S6G). We also attempted to restore the mitochondrial function by supplementing with succinate and malate, but no rescue effect was demonstrated (Fig. 5G–H). These results indicate that hα-syn-induced mitochondrial damage can be mitigated by the exogenous addition of isocitrate in primary neurons.

2.6. Interfering peptides block the interaction between hα-syn and ACO2, reducing hα-syn-mediated cytotoxicity in vitro and in vivo

To evaluate the significance of the hα-syn/ACO2 interaction, we used protein-protein docking prediction tools (ClusPro and Peptiderive) and identified potential interaction sites, including the amino acid sequences 84–94 of hα-syn (GAGSIAAATGF) and 565–575 of ACO2 (TTDHISAAGPW) (Fig. 6A–B). To assess these predicted sites, we engineered a mutant form of hα-syn, termed hα-syn-del, by deleting the 11 amino acids at positions 84–94 and adding a myc tag. This mutant plasmid (SNCA-del) was transfected into SH-SY5Y cells (Fig. S7A). Following truncation, hα-syn-del was successfully expressed, and the pα-syn-del protein was detectable (Fig. S7B). Co-immunoprecipitation (Co-IP) analysis demonstrated that hα-syn-del and pα-syn-del could no longer interact with the ACO2-3xFlag protein co-expressed in the cells (Figs. S7C–D).

Fig. 6.

Interfering peptides block the interaction between hα-syn and ACO2, reducing hα-syn-mediated cytotoxicity. (A) Predicted protein regions and amino acids implicated in the interaction between ACO2 (PDB ID: 7ACN) and hα-syn (PDB ID: 1XQ8). Right: Enlarged view of the interaction area in the left circle. (B) Representation of the hα-syn-ACO2 complex with the peptide contributing most to the interaction (the ‘hot segment’) highlighted. The peptide includes amino acid sequences 84–94 of hα-syn. (C) Confocal images showing the distribution of CPs and IPs within SH-SY5Y cells. Scale bar = 20 μm. (D) Biotin-labeled pulldown assay showing the binding of IPs to hα-syn and pα-syn but not the hα-syn-del and pα-syn-del protein in SH-SY5Y cells lysates. (E–F) Co-IP experiments showing the interaction between ACO2 and hα-syn in SH-SY5Y cells treated with different doses of CPs or IPs (10, 20, and 40 μM) for 24 h (E) and with 40 μM CPs or IPs for 24, 48, 72 h (F). (G–H) ACO2 activity was measured in SH-SY5Y cells treated with different doses of CPs or IPs (10, 20, and 40 μM) for 24 h (G) and with 40 μM CPs or IPs for 24, 48, 72 h (H). (I) ACO2 activity was measured in SH-SY5Y cells transfected with either vector or SNCA plasmids for 24, 48, and 72 h, or treated with CPs, IPs, or NAC for the same durations. Right: ACO2 activity at 72 h for each group. n = 6. (J) ACO2 protein activity assay. The system was supplemented with ACO2 protein, 2 μg of hα-syn protein, CPs, IPs, or H₂O₂. n = 3. (K) The oxygen consumption rate (OCR) measured in SH-SY5Y cells treated with 40 μM CPs or IPs for 72 h n = 4. (L–M) ATP levels (L), mitochondrial complex I activity (M), and cell viability (N) were determined in SH-SY5Y cells treated with 40 μM CPs or IPs for 72 h n = 6. Data are expressed as the mean ± SEM. P values were determined by one-way ANOVA or two-way ANOVA (H). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ###P < 0.001 vs. CPs (G).

We then designed an interfering peptide (IP) derived from the 11 amino acids of ACO2. For cell penetration, a Tat sequence (YGRKKRRQRRR) was added to the N-terminus, along with N-terminal biotin and C-terminal FITC (biotin-Tat-ACO2-11aa-FITC). For the control peptide (CP), the 11 amino acids of ACO2 were randomized as IGASWPDHATT (biotin-Tat-ACO2-random-FITC). These peptides entered cells efficiently, as demonstrated by FITC fluorescence (Fig. 6C). There was no effect of different concentrations of control peptides (CPs) and interference peptides (IPs) on ACO2 activity, mitochondrial function, or cell viability (Figs. S7E–I). A biotin pulldown assay confirmed that the IPs could pull down both hα-syn and pα-syn, while failing to interact with hα-syn-del (Fig. 6D). Treatment with IPs disrupted the interaction between ACO2 and hα-syn or pα-syn in a dose-dependent and time-dependent manner with 40 μM IPs over 72 h markedly inhibiting the interaction (Fig. 6E–F and S7J–K). ACO2 activity demonstrated a temporal and dose-dependent increase, achieving up to 85 % recovery of normal levels (Fig. 6G–H). Furthermore, the combination of NAC and IPs restored ACO2 activity to 92 % (Fig. 6I). These findings suggest that both hα-syn and oxidative stress contribute to the inhibition of ACO2 activity, with the hα-syn/ACO2 interaction exerting a more substantial impact. Additionally, when ACO2 and hα-syn purified proteins were combined, ACO2 activity decreased, and the addition of H₂O₂ further inhibited this activity. In this context, IPs reactivated ACO2 activity (Fig. 6J). 40 μM IPs treated for 72 h also improved cellular oxygen consumption rate (OCR) level, ATP levels, mitochondrial complex I activity and cell viability against hα-syn toxicity (Fig. 6K–N).

To assess the role of IPs in vivo, the peptides were administered via stereotaxic injection into the bilateral striatum of 12-month-old mice. Seven days post-injection, peptides were detected in the striatum (Fig. 7A). IPs treatment improved the behavior of TG mice in the pole test and rotarod test compared to CPs treatment (Fig. 7B–C) and showed a trend toward improvement in the open field test (Fig. 7D). Immunostaining with FITC showed that CPs and IPs spread in the striatum and substantia nigra (Figs. S8A–B). Moreover, IPs disrupted the interaction between ACO2 and hα-syn or pα-syn in the striatum and midbrain (Fig. 7E and S8C–E), increasing ACO2 activity (Fig. 7F and S8F). Additionally, IPs restored TCA cycle metabolite levels and mitochondrial function, alleviated oxidative stress, and increased TH levels in the striatal tissues of TG mice (Fig. 7G–N and S8G–H). Importantly, these peptides had no effect on WT mice (Fig. S9). Collectively, these results suggest that IPs can block the interaction of ACO2 with hα-syn and ameliorate hα-syn-mediated cytotoxicity in vivo.

Fig. 7.

IPs block interactions and reduce hα-syn-mediated cytotoxicity in vivo. (A) Stereotactic injection of CPs and IPs into the bilateral striatum of 12-month-old mice. Immunofluorescence images of brain slices from WT mice injected with peptides over a period of 7 d. Scale bar = 50 μm. (B–D) The motor behavior of mice was detected by the pole test (B), rotarod test (C), and open field test (D). Right: Representative trajectory diagrams of mice in the open field test. n = 6–8. (E) Co-IP experiments showing the interaction between hα-syn and ACO2 using ACO2 (middle) and hα-syn antibody (right) in the striatum of mice in each group. The hα-syn and ACO2 levels in brain tissue lysates were shown by western blot (Input; left). Homologous IgG was used as the experimental control. (F–M) ACO2 activity (F), isocitrate concentration (G), citrate concentration (H), Fe²⁺ levels (I), mitochondrial complex I activity (J), MDA levels (K), GSH levels (L), GSSG levels (M) in the striatum of mice in each group. n = 5. (N) Western blot analysis of hα-syn, pα-syn, and TH levels in the striatum of mice in each group. β-actin was used as a loading control. Quantitative analysis of TH levels is shown on the right. n = 5. Data are expressed as the mean ± SEM. P values were determined by one-way ANOVA. ∗P < 0.05, ∗∗∗P < 0.001.

3. Discussion

The present study showed that TG mice overexpressing hα-syn exhibit progressive neurodegeneration associated with hα-syn accumulation in mitochondria. The increasing interaction between hα-syn and ACO2 in mitochondria caused a progressive decrease in ACO2 activity, which leads to mitochondrial dysfunction and oxidative stress. This increased oxidative stress further contributes to the suppression. By overexpressing ACO2 or inhibiting this interaction with IPs, we were able to alleviate mitochondrial function in vivo and in vitro.

The pathogenesis of PD involves α-syn infiltration of mitochondria, which leads to mitochondrial dysfunction [24]. Although this study indicated that hα-syn reduces ACO2 activity and that mitochondrial dysfunction can be alleviated by the addition of isocitrate and up-regulation of ACO2 expression, it is important to acknowledge that hα-syn also impacts mitochondria via other mechanisms. Our data showed that mitochondrial function was reduced approximately 40 % by hα-syn. The addition of isocitrate or upregulation of ACO2 expression rescued nearly 20 % of mitochondrial function, as assessed by mitochondrial complex I activity and the ATP level. It is likely that the remaining 20 % of mitochondrial function that was not rescued was affected by hα-syn through ACO2-independent pathways. For example, α-syn can interact with proteins on the outer mitochondrial membrane, such as TOM20 and TOM40, impeding the import of proteins into the mitochondria [13]. Additionally, α-syn can interact with VDAC and disrupt mitochondrial calcium homeostasis by inhibiting VDAC conductance [25].

Pα-syn has been identified as the predominant pathological form of α-syn [26,27]. In autopsies of PD patients, pα-syn comprises 90 % of the α-syn found in Lewy bodies [28]. Although a recent study showed that phosphorylation of α-syn is a trigger for protein-protein interactions [29], our data showed that pα-syn behaves similarly to hα-syn regarding the ACO2 interaction and inhibition. As for the precise role of pα-syn in inducing mitochondrial damage, it is likely that the combined actions of hα-syn and pα-syn contribute to the deleterious effects on mitochondria and cellular function.

Declined ACO2 activity is crucial in neurodegeneration, influenced by several factors [[30], [31], [32], [33]]. 1) ROS and reactive nitrogen species (RNS) can cause oxidative modifications of the iron-sulfur cluster within ACO2, leading to reduced enzyme activity [[34], [35], [36]]. 2) Phosphorylation, acetylation, and ubiquitination can alter ACO2 activity by impacting enzyme stability, localization, and interaction with other proteins [37]. 3) Mutations in the ACO2 gene can produce dysfunctional enzyme variants that exhibit reduced activity or stability [[38], [39], [40]]. In this study, we identified a new mechanism for reduced ACO2 activity due to its interaction with hα-syn during PD progression. Inhibiting this interaction could restore ACO2 activity even in the presence of ROS and hα-syn, indicating the important role of this interaction in regulating ACO2 activity. It is likely that hα-syn, ROS, and ACO2 form a vicious cycle: the interaction between hα-syn and ACO2 suppresses ACO2 activity, impairing mitochondrial function, which in turn increases ROS production that further inhibits ACO2. Moreover, hα-syn may enhance oxidative stress through additional pathways. Our findings suggest that restoring ACO2 activity could be a viable strategy to break this cycle.

Regarding the mechanism of hα-syn's inhibition of ACO2, it is probable that α-syn obstructs the substrate (citrate) necessary for ACO2, as adding citrate can rescue ACO2 activity. However, other possibilities, such as alterations in ACO2 conformation or kinetics, cannot be ruled out. Furthermore, the interaction between hα-syn and ACO2 may affect ACO2's interactions with other proteins, such as ATP1B1 (an ATPase for Na+/K+ transport), citrate synthase, vesicle-associated membrane protein-associated protein A (VAPA), and isocitrate dehydrogenase 2 (IDH2) [41,42]. Further investigation is warranted to fully elucidate these interactions.

In addition to influencing the TCA cycle, ACO2 plays a critical role in controlling mitochondrial iron levels; iron-sulfur clusters are particularly vulnerable to oxidative damage and subsequent Fenton reaction-mediated iron overload within mitochondria [43,44]. Additionally, decreased ACO2 activity may be implicated in regulating mitochondrial autophagy through suppression of autophagy gene transcription [45].

In summary, our study has identified ACO2 as a target of α-syn-induced mitochondrial damage and progressive toxicity. The findings suggest that strategies aimed at enhancing ACO2 activity and inhibiting the interactions between ACO2 and α-syn may hold promise as potential therapeutic interventions to mitigate the progression of PD.

4. Materials and methods

4.1. Mice

Thy1-SNCA transgenic mice overexpressing hα-syn (017682) were obtained from the Jackson Laboratory (Bar Harbor, ME, USA) and maintained on a C57BL/6 N background. Male C57BL mice were procured from Vital River Laboratories (Beijing, China). The mice were housed in the Laboratory Animal Center of Capital Medical University (Beijing, China) under standard conditions of room temperature (22–25 °C) and a 12-12h light/dark cycle. All animal experiments were conducted in accordance with the Animal Care and Use Guidelines of the National Institutes of Health (Bethesda, MD, USA). The Institutional Animal Care and Use Committee (AEEI-2020-017) reviewed and approved the animal study.

4.2. Pole test

The pole test was utilized to assess the degree of bradykinesia, following previously established protocols [46]. Specifically, a mouse was positioned atop a pole measuring 50 cm in length with a 4-mm radius and the duration of time for the mouse to descend to the base was recorded in 3 trials and averaged.

4.3. Rotarod test

The rotarod test was used to evaluate motor function [47]. Mice underwent training sessions twice daily for 3 d, during which the rotation speed progressively increased from 4 to 40 rpm over a 5-min period. The duration of time that the mouse was able to remain on the rotating rod during subsequent assessments over the ensuing 3 d was recorded and averaged.

4.4. Open field test

The open field test was used to assess the spontaneous motor ability of mice. Prior to testing, the mice were acclimated to the experimental room for a period of 30 min to reduce stress. Subsequently, the mice were positioned in the central area of the open field, where activities were automatically recorded by animal behavior analysis software for a duration of 10 min. To prevent contamination between tests with different animals, the test apparatus was meticulously cleaned with 75 % ethanol.

4.5. Adhesive removal test

The adhesive removal test was utilized to evaluate paw sensitivity and dexterity in mice and was performed in accordance with previously published methods [48]. Two distinct colored adhesive tapes (3 mm × 4 mm) were affixed to the left and right front paws of each mouse and the time taken to contact and remove the tape was measured with a maximum time limit of 120 s. The mean time from three trials was calculated for analysis.

4.6. Grip strength test

The grip strength test was used to assess muscle strength in the forelimbs and hindlimbs. Initially, the forelimbs were permitted to grasp the grip net. Then, the experimenter pulled the tail of the mouse in a straight line until the grip was released, thereby eliciting the forelimb grip response. Subsequently, the subjects proceeded to pull the tails of the mice in a similar manner until the hindlimbs disengaged from the grip net first, thereby indicating the hindlimb grip. This process was repeated five times. The data were categorized based on animal weight and quantified accordingly.

4.7. Sample preparation and western blot

The following primary antibodies were used for western blot analysis: anti-pα-syn (D1R1R; Cell Signaling Technology, Danvers, MA, USA); anti-hα-syn (ab138501; Abcam, Cambridge, UK); anti-ACO2 (A4524; ABclonal, Wuhan, China); anti-TOM20 (D8T4N; Cell Signaling Technology); and anti-β-actin (66009-1-Ig; Proteintech, Chicago, IL, USA). A Bicinchoninic Acid Protein Assay kit (23225; Pierce Biotechnology, Rockford, IL, USA) was utilized to quantify protein concentration in recombinant protein samples, cells, and brain tissue lysates. The proteins were resolved via 12 % sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto PVDF membranes (Millipore, Bedford, MA, USA). After blocking with 5 % skim milk for 1 h at room temperature, the membranes were exposed to primary antibodies overnight at 4 °C. The membranes were incubated with secondary antibodies for 1 h after 3 washes with TBST. The membranes were then analyzed using an Odyssey imaging system and quantified utilizing ImageJ software (NIH, Bethesda, MD, USA).

4.8. ACO2 and mitochondrial complex I activity

ACO2 activity was quantified using an Aconitase Assay kit (MAK051; Sigma, St. Louis, MO, USA) in accordance with the manufacturer's recommended protocol. Tissue samples (20–40 mg) or cells (1×10⁶) can be homogenized in 100 μL of ice-cold Assay Buffer. Subsequently, centrifuge the samples at 800 g for 10 min at 4 °C to remove insoluble material. The resulting supernatant can be utilized for the c-aconitase assay. For the m-aconitase assay, further centrifuge the supernatant at 20,000 g for 15 min at 4 °C, collect the pellet, and dissolve it in 0.1 mL of cold Assay Buffer. Sonicate the pellet for 20 s. One unit of aconitase is defined as the amount of enzyme that catalyzes the isomerization of 1.0 mol of citrate to isocitrate per minute at pH 7.4 and 25 °C. The manufacturer's guidelines were followed for assessment of mitochondrial complex I activity using the Mitochondrial Complex I Activity Assay kit (GMS50007; GENMED, Shanghai, China).

4.9. Cell cultures and transfection

SH-SY5Y cells were grown in DMEM (Gibco, Carlsbad, CA, USA) supplemented with 10 % fetal bovine serum (FBS; Gibco). Cells were incubated at 37 °C in 5 % CO₂. Transfection was performed with plasmids (pcDNA3.1, pcDNA3.1-SNCA, pcDNA3.1-ACO2-3xflag, pcDNA3.1-ACO2-A252T-3xflag, pCMV-myc, pCMV-myc-SNCA, and pCMV-myc-SNCA-del) using polyethylenimine [PEI] (23966, linear, MW 25,000; Polysciences, Inc., Warrington, PA, USA) according to the manufacturer's instructions. After transfection, fresh medium was replaced after 6 h.

4.10. Primary cortical neuronal cultures and infection lentivirus

Lentivirus gene transfer vector encoding human wild-type SNCA (LV-SNCA), ACO2 (LV-ACO2), and ACO2-A252T (LV-ACO2-A252T) were purchased from Genechem (Shanghai, China). Cortical primary neurons were obtained from WT and TG embryonic mice at gestational days 14.5–15.5. The dissociated neurons were then cultured in 24- or 96-well plates on cover slips that had been coated with 0.1 mg/ml poly-l-lysine in Neurobasal medium (21103-049; Gibco) supplemented with l-glutamine (0.5 mM) and 50 × B27 supplement (17504-044; Gibco) to achieve a final concentration of 1 × . Half of the medium was changed every 3 d. Lentivirus infection was carried out after 4 d of culture.

4.11. Cell viability and LDH assay

Cell viability was determined using the CCK8 assay with WST-8 and PMS as the primary constituents. A total of 1 × 10⁴ cells were seeded in the wells of a 96-well microplate. The CCK8 solvent was diluted by a factor of 10 and 100 μL of the resulting solution was dispensed into each well. The microplate was then maintained at 37 °C and absorbance was measured at 450 nm using a microplate reader (PerkinElmer, Inc., Waltham, MA, USA). The absorbance value at the second hour was documented for subsequent analysis in this investigation. Cytotoxicity was evaluated utilizing the LDH assay method, in which 1 × 10⁴ cells were plated in individual wells of a 96-well microplate. Subsequently, 100 μL of cell supernatant from each well was transferred to a new 96-well microplate, followed by the addition of 100 μL of LDH detection solution, thoroughly mixed, and incubated for 30 min at room temperature (25 °C) in the absence of light. After the incubation, 50 μL of stop solution was introduced to each well and agitated for 10 min. Absorbance values were then determined using a microplate reader at a wavelength of 490 nm.

4.12. Immunofluorescence and confocal microscopy

The following primary antibodies were used for immunofluorescence: anti-pα-syn (pSyn#64; Wako, Osaka, Japan); anti-hα-syn (ab138501; Abcam); anti-TH (ab76442; Abcam); anti-ACO2 (D6D9; Cell Signaling Technology); anti-TOM20 (D8T4N; Cell Signaling Technology); anti-β-actin (66009-1-Ig; Proteintech); anti-β-tubulin Ⅲ (ab78078; Abcam); and anti-flag (C1305; Applygen Technologies Inc., Beijing, China). Cells were initially seeded on poly-l-lysine-coated coverslips within confocal plates. The cells were then fixed with 4 % paraformaldehyde for a duration of 30 min, followed by 3 washes with 0.01 M PBS and permeabilization with 0.3 % TritonX-100 in 0.01 M PBS for 10 min at room temperature. Following a blocking step with 5 % normal goat serum (5424; Cell Signaling Technology) for 1 h, the cells were incubated overnight with primary antibodies at 4 °C, then a 1-h incubation with secondary antibodies at room temperature. Visualization of cell nuclei was achieved through counterstaining with 4′,6-diamidino-2-phenylindole (D9542; Sigma) and imaging was performed using a confocal microscope (TCS SP8; Leica, Solms, Germany). The brain slices were immersed in PBST (0.3 % Triton X-100 in 0.01 M PBS) for a duration of 10 min, followed by incubation in 5 % goat serum for a period of 1 h. The sections were then treated with primary antibodies at a temperature of 4 °C for 24 h. After washing three times in PBST, the sections were exposed to secondary antibodies at room temperature for 1 h and stained with 4′,6-diamidino-2-phenylindole for 15 min. The sections were analyzed using a confocal microscope.

4.13. Co-immunoprecipitation

Brain tissues (2000 μg) or cell extracts (1000 μg) were incubated at 4 °C on a rotating platform overnight with anti-hα-syn (ab138501; Abcam), -pα-syn (D1R1R; Cell Signaling Technology), -ACO2 (D6D9; Cell Signaling Technology), or -flag (C1305; Applygen Technologies Inc.) antibody. Magnetic beads (40 μL; Med Chem Express, San Rafael, CA, USA) were washed with 500 μL of 150 mM IP buffer (10 mM Tris-HCl [pH 7.5], 150 mM NaCl, 2 mM EDTA, and 0.5 % Triton X-100). Subsequently, the protein/antibody mixture was introduced and allowed to incubate at 4 °C for 6 h. Following 3 washes, the target antigen-antibody complex was eluted with 20 μL of loading buffer and heated at 95 °C for 10 min prior to western blot analysis.

4.14. JC-1 staining

The mitochondrial membrane potential (MMP) was assessed using the JC-1 probe (T4069; Sigma-Aldrich). The JC-1 probe transitions from a monomeric state that emits green fluorescence under low MMP conditions to an aggregated state that emits red fluorescence under high MMP conditions. Following 3 washes with 0.01 M PBS, cells were seeded in 96-well plates and treated with JC-1 at 37 °C for 30 min. Cellomics high content screening (HCS; Thermo Scientific, Rockford, IL, USA) was used to monitor changes in fluorescence at wavelength of 488/530 nm (green) and 549/595 nm (red), ultimately determining the green-to-red fluorescence intensity ratio.

4.15. ATP level measurement

The ATP levels in cultured cells were quantified using an ATP assay kit (G9241, Promega, Madison, WI, USA). A total of 1 × 10⁴ cells were seeded in the wells of an opaque 96-well black plate. The cells are treated in accordance with the requirements of the experiment. The plate was allowed to equilibrate to room temperature for approximately 30 min prior to the measurement. Subsequently, 100 μL of ATP detection working buffer was added to each well, followed by a 2-min mixing step to induce cell lysis. The plate was then incubated at room temperature for 10 min to stabilize the luminous signal. The luminescence values were measured using a microplate reader.

4.16. Cloning, expression and protein purification

The cDNAs of human full-length α-syn were cloned into the pGEX-4T-1 vector, while the cDNAs of human full-length ACO2 and ACO2-A252T were cloned into the pGEX-6P-1 vector. These plasmids were subsequently transformed into BL21 (DE3) Escherichia coli cells (Tiangen, Beijing, China). Protein expression was induced by the addition of 0.5 mM isopropyl-β-d-1-thiogalactopyranoside (IPTG). Recombinant GST, GST-α-syn, GST-ACO2, and GST-ACO2-A252T were purified from the bacterial BL21 lysate using glutathione SepharoseTM 4B (GE Health Life Sciences, Pittsburgh, PA, USA). The hα-syn protein was incubated with human thrombin at room temperature for 8 h to eliminate GST labels, while the ACO2 and ACO2-A252T proteins were incubated with preScission protease for the same purpose. Following elution, the buffer was exchanged to a working solution containing 10 μM MgCl2, 20 μM HEPES, and 20 μM dithiothreitol using an ultrafiltration centrifugation tube (Millipore). Protein purity was assessed through Coomassie brilliant blue staining and western blot analysis.

4.17. Preparation of pα-syn

Pα-syn preparation was in accordance with established protocols [49]. Hα-syn protein was phosphorylated at serine-129 utilizing Polo-like kinase 3 (PLK3; PV3812, Thermo Scientific). The reaction mixture was comprised of 1.2 μL of PLK3, 50 μL of hα-syn protein (2 mg/ml), and 0.5 μL of ATP (100 mM; A26209, Sigma-Aldrich), which was subsequently incubated in a water bath at 30 °C for a duration of 3 h.

4.18. Glutathione S-transferase (GST) pulldown

Glutathione Sepharose™ 4B beads (10 μL) were placed in 1.5 μL EP tubes and thrice-washed with 0.1 M PBS for 5 min each. Then, purified GST, GST-ACO2, and GST-ACO2-A252T proteins were added to the EP tubes and incubated at 4 °C on a rotating platform for 1 h. Subsequently, pure hα-syn or pα-syn proteins were added to the system and incubated at 4 °C on a rotating platform for 4 h. The mixture was then centrifuged at 4 °C 1000 g for 5 min to collect the precipitatant. The complex was washed 5 times with 0.1 M PBS foe 5 min each. The complex was then eluted with 20 μL of loading buffer and heated at 95 °C for 10 min prior to western blot analysis.

4.19. Biotin-labeled peptide pulldown assay

Biotin-labeled peptides were used to pull down the hα-syn or hα-syn-del protein in order to determine the binding affinities. SH-SY5Y cells were transfected with SNCA or SNCA-del plasmids for 48 h. Cellular proteins were extracted using 1 × RIPA lysis buffer (C1053; APPLYGEN) and subsequently incubated with biotin-labeled peptide at 4 °C overnight. The mixture was then incubated with streptavidin magnetic beads (88817; Thermo Scientific) at 4 °C for 6 h. The beads were thoroughly washed and the proteins that were pulled down were utilized for SDS-PAGE and western blot analysis.

4.20. Protein docking, design, and synthesis of peptides

The ACO2 (PDB ID: 7ACN) and SNCA (PDB ID: 1XQ8) 3D structures were obtained from the Protein Data Bank (PDB [www.rcsb.org]). Protein-protein docking was performed using the ClusPro online docking server (https://cluspro.bu.edu/queue.php) [50]. IPs designed to disrupt the interaction between ACO2 and hα-syn were generated by the Peptiderive server (http://rosie.rosettacommons.org/peptiderive) [51]. An 11-amino acid peptide (YGRKKRRQRRR) derived from the Tat protein transduction domain was utilized as a cell-penetrating peptide. Control peptides (CPs [IGASWPDHATT]) and interfering peptides (Ips [TTDHISAAGPW]) were chemically synthesized by attaching a biotin-labeled cell-penetrating peptide at the N-terminus and conjugating with fluorescein isothiocyanate (FITC) at the C-terminus. The purity of the peptides, determined to be >95 %, was confirmed using a reversed phase-high performance liquid chromatography assay.

4.21. UHPLC-MS/MS analysis

The tissue sample was frozen and treated with 80 % methanol in a TissueLyser (JX-24, Jingxin, Shanghai, China) for 4 min at 40 Hz. After centrifugation, the supernatant was dried, then redissolved in 50 % acetonitrile for UHPLC-MS/MS analysis. A quality control sample was prepared by pooling all samples. The UHPLC-MS/MS analysis was performed using an Agilent 1290 Infinity II UHPLC system with a 6470A Triple Quadrupole mass spectrometer. Samples were injected onto a Waters ACQUITY UPLC BEH Amide column with a flow rate of 0.2 mL/min and a mobile phase of 10 mM ammonium acetate and 90 % acetonitrile. The gradient elution program was as follows: 0–1 min, 90 % B; 4 min, 85 % B; 12 min, 70 % B; 14–16 min, 50 % B; and 16.1–20 min, 90 % B. The raw data were processed using MassHunter Workstation Software (version B.08.00, Agilent, Waldbronn, Germany) to determine the quantitative accuracies of each compound. Peak areas of each compound in all samples were integrated and calibration curves were constructed. Concentrations of sugar metabolites in prepared samples were quantified automatically and output for quantitative calculations were saved in Excel. Content (μg/mg sample) = 0.05∗40∗C/23, where C is the concentration quantified in the prepared sample (μg/mL).

4.22. Stereotactic injection

The bilateral striatum of 12-month-old TG mice was injected with 1x10¹¹ viral particles (1 μL [1x10¹³ V G/ml]) of AAV9-EGFP (Vector), AAV9-ACO2-EGFP (ACO2), AAV9-ACO2-A252T-EGFP (ACO2-A252T), which were prepared by Genechem Technology (Shanghai, China). Additionally, 12-month-old TG mice were injected with CPs or IPs (1 μL [0.5 mM]) in the bilateral striatum. The stereotactic injection was performed at a rate of 0.1 μL/min with the following coordinates: anteroposterior (AP) = ±1.0 mm; mediolateral (ML) = ±1.5 mm; and dorsoventral (DV) = ±3.5 mm from the bregma. Following the injection, the needle was left in place for an additional 10 min to ensure full absorption of the solution. Subsequent to the surgical procedure, animals were closely monitored and received appropriate postsurgical care.

4.23. Statistical analysis

All data were statistically analyzed using Prism 8.0.1 software (GraphPad, La Jolla, CA, USA). Each experiment was independently performed at least three times. All data was tested for normality prior to selection of the statistical test. For two-group comparisons, the unpaired t-test was used. For multiple comparisons, one-way ANOVA was performed. Data are expressed as the mean ± standard error of the mean (SEM). A P < 0.05 was considered statistically significant in all cases.

CRediT authorship contribution statement

Jie Jiao: Data curation, Formal analysis, Investigation, Methodology, Resources, Validation, Visualization, Writing – original draft, Writing – review & editing. Ge Gao: Data curation, Investigation, Methodology, Project administration, Resources, Validation, Writing – review & editing. Junge Zhu: Data curation, Formal analysis, Investigation, Methodology, Resources, Validation, Writing – review & editing. Chaodong Wang: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Lei Liu: Conceptualization, Funding acquisition, Supervision, Visualization, Writing – review & editing. Hui Yang: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Institutional review board statement

The animal study protocol was approved by the Institutional Animal Care and Use Committee of Capital Medical University (approval no. AEEI-2020-017).

Declaration of competing interest

The authors declare no competing interests.

Appendix A. Supplementary data

The following is the supplementary data to this article: