Lipofectamine 2000 was purchased from Thermo Fisher (CA, USA). A mitochondria isolation kit was procured from Transgen Biotech (Beijing, China). Complete protease inhibitor cocktail tablets, Cell Counting Kit-8 (CCK-) 8, and anti-Cyt C antibody were purchased from Bimake (TX, USA). Anti-myc, Bak, β-actin, and Cox I antibodies, were purchased from Santa Cruz Biotechnology (TX, USA). An anti-PARP antibody was purchased from Cell Signaling Technology (MA, USA). An anti-Bax antibody was purchased from R&D Systems (MN, USA). Both anti-mouse IgG and anti-rabbit IgG secondary antibodies were purchased from Bioss (Beijing, China). pcDNA3.1/LacZ-Myc-Flag, pcDNA3.1/SOD1-Myc-Flag, and pcDNA3.1/SOD1^G93A-Myc-Flag plasmids were amplified as described previously (Zeng et al., 2012).

NSC-34 cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Hyclone) supplemented with fetal bovine serum (10%), streptomycin (100 μg/mL), and penicillin (100 units/mL) at 37°C and in the presence of 5% CO₂. The cells were transfected with pcDNA3-LacZ, wtSOD1, or SOD1^G93A constructs using Lipofectamine 2000 following the manufacturer’s protocols.

The cell survival activity was measured by the CCK-8 method. Briefly, NSC34 cells were seeded into 24-well plates at a density of 105 cells/mL. The cells were transfected with LacZ, wtSOD1, or SOD1^G93A plasmids for 24 h. After 24 h-transfection, the CCK-8 reagent was added and cultured for 0.5–5 h at 37°C. The absorbance was finally measured at 450 nm following the manufacturer’s protocols.

A caspase-3 activity kit was purchased from Bioworld (Nanjing, China). NSC34 cells were seeded into six-well plates at a density of 5 × 10⁵ cells/mL. The cells were transfected with LacZ, wtSOD1, or SOD1^G93A plasmids for 24 h. Activated caspase-3 cleaved Ac-DEVD-pNA to generate pNA. The absorbance was measured at 405 nm to indicate caspase-3 activity.

siRNA duplexes to PSAP, Bax, and Bak, and scrambled siRNA and RNAi-mate reagent were purchased from Gene Pharma (Shanghai, China). The cells were treated with each siRNA for 48 h and transfected with LacZ, wtSOD1, or SOD1^G93A plasmids using Lipofectamine 2000 reagent. The cells were harvested and lysed for further analysis 24 h after transfection.

The cytosolic and mitochondrial fractions were extracted using the mitochondria isolation kit following the manufacturer’s protocols to explore Cyt C release and Bax translocation.

Cell apoptosis was determined using an Annexin V/propidium iodide (PI) kit and analyzed by flow cytometry. The cells were harvested after 24 h transfection, washed twice with an ice-cold phosphate-buffered solution (PBS), and stained with Annexin V/PI solution.

IP and immunoblotting were performed as previously described (Zeng et al., 2012).

For Western blot analysis, cells were harvested and lysed by sonication for 25 s on ice in lysis buffer. Denatured protein (20 μg) ofeach sample was separated by SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane. The PVDF membrane was blocked with 5% skim milk powder in Tris-buffered saline tween (TBST) for 45 min. The protein blot was then incubated with appropriate primary antibodies and HRP-conjugated secondary antibody separately. For immunoprecipitation (IP), cells were harvested and lysed by sonication for 20 s on ice in IP lysis buffer, and samples were centrifuged at 14,000 × g for 5 min at 4°C. Then the supernatants were incubated with anti-myc, anti-PSAP, or anti-Bax together overnight at 4°C. The samples were added with protein A-sepharose (Bimake, TX, USA) for 2 h. After washing five times with cold PBS, the immunoprecipitates were separated by SDA-PAGE and incubated with the primary antibodies.

The data were analyzed by using SPSS 22.0 statistical analysis software, and groups of data are presented as mean ± standard deviation (SD).Statistical significance was performed on the data by analysis of variance for significance test. **P < 0.01 was used as the criterion for determining statistical difference.

NSC-34 cells were transfected with flag-tagged LacZ or wtSOD1, SOD1^G93A to investigate whether SOD1^G93A promoted apoptosis. Fig. 1A shows the percentage of viable cells after the transfection of LacZ, wtSOD1, and SOD1^G93A. The results indicated that the overexpression of SOD1^G93A decreased cell survival by ~45% compared with lacZ- or wtSOD1-transfected cells (P < 0.01).

As shown in Figs. 1B and 1C, SOD1^G93A triggered PARP cleavage after 24 h-transfection. In contrast, no apoptotic events were observed on the transfection of LacZ or wtSOD1 constructs. Annexin V/PI double staining assay was performed to detect cells in the necrotic or apoptotic stage to further confirm the apoptotic activity of SOD1^G93A. As shown in Fig. 1D, the number of apoptotic and necrotic cells was ~40% (B2 and B4) in SOD1^G93A-transfected cells, while 5% of cells were B4 (Annexin V⁺/PI⁻) or B2 (Annexin V⁺/PI⁺) in LacZ- or wtSOD1-transfected cells, suggesting low levels of cell death. Taken together, these results strongly suggested that the overexpression of SOD1^G93A induced apoptosis in NSC-34 cells.

Since mitochondrial damage promotes motor neuron cell death, the study next examined whether SOD1^G93A induced the release of mitochondria-related molecules into the cytoplasm. As shown in Fig. 2, increased levels of Cyt C were detected in cytosolic extracts, while a concomitant decrease in the levels of these molecules was observed in the mitochondrial fractions on SOD1^G93A overexpression (P < 0.01). In contrast, Bax expression reduced in the cytosolic fractions and was higher in the mitochondrial fraction on exogenous SOD1^G93A expression (P < 0.01).

SOD1^G93A triggered Cyt C release and Bax translocation regulated by the Bcl-2 family of proteins, which are known regulators of the mitochondrial apoptotic pathway. The study therefore determined the effects of Bax and Bak on SOD1^G93A-induced apoptosis. Prior to SOD1^G93A overexpression, the cells were silenced with Bax, Bak-specific siRNAs, or scrambled controls. The cell viability was then assessed following the transfection of LacZ, wtSOD1, and SOD1^G93A. Fig. 3A shows that both Bax and Bak were involved in SOD1^G93A-induced apoptosis as their silencing enhanced NSC-34 cell viability compared with nonsilencing cells (P < 0.01). As shown in Fig. 3B, SOD1^G93A increased the activity of caspase-3 by ~40% compared with LacZ-transfected cells. However, SOD1^G93A-triggered caspase-3 activation was inhibited by Bax or Bak silencing.Furthermore, as shown in Figs. 3C, 3E, and 3F, SOD1^G93A-induced PARP cleavage was completely ablated following Bax or Bak silencing. The effect of Bax- or Bak-silenced cells on SOD1^G93A overexpression-induced cell death was confirmed by Annexin V/PI staining (Fig. 3G), indicating that the percentage of apoptotic cells was ~60% (B2 and B4) following SOD1^G93A overexpression, which decreased to ~10% in Bax- or Bak-silenced cells. These data indicated that both Bax and Bak were required for SOD1^G93A-induced apoptosis.

PSAP specifically localized to the outer membrane of the mitochondria and was associated with Bax or Bak. In addition, PSAP was required for DR6-induced apoptosis (Zhang et al., 2020). DR6 was highly expressed in samples from mice and patients with ALS and played a crucial role in ALS development. The effects of DR6 on PSAP-specific siRNAs were assessed to further determine the role of PSAP in SOD1^G93A-induced apoptosis. Similarly, the silencing of PSAP attenuated SOD1^G93A-triggered cell death (Fig. 3A), PARP cleavage (Figs. 3C and 3F), and caspase-3 activation (Fig. 3B). Annexin V/PI staining confirmed these effects. Taken together, these data revealed that PSAP was required for SOD1^G93A-induced apoptosis (Fig. 3G).

The possible interaction of Bax, Bak, and PSAP upon SOD1^G93A overexpression was explored to further investigate the mechanism by which Bax, Bak, and PSAP mediated SOD1^G93A-induced apoptosis. As shown in Fig. 4A, Bax co-immunoprecipitated with SOD1-myc using anti-myc antibodies (lanes 3 and 4, top panel) but not IgG (lane 5) on transfection with wtSOD1 or SOD1^G93A. The levels of this complex modestly decreased in SOD1^G93A-transfected cells (lane 4) compared with wtSOD1-transfected cells (lane 3). Similarly, Bak co-immunoprecipitated with anti-myc (SOD1), but the interaction did not increase in cells overexpressing SOD1^G93A. Furthermore, only a negligible change in the levels of PSAP–Bax or PSAP–Bak complex formation was observed on SOD1^G93A transfection compared with LacZ-transfected cells. On transfection with wtSOD1, Bax co-immunoprecipitated with PSAP, while PSAP–Bak complex formation decreased. In addition, Bak–Bax complexes were detected in SOD1^G93A-expressing cells (lanes 4 and 5). The effects on PSAP-specific siRNAs were assessed to further confirm the role of PSAP in SOD1^G93A-induced Bak–Bax interaction. Similarly, the silencing of PSAP (Figs. 4C and 4D) attenuated SOD1^G93A-triggered Bak–Bax complex formation (lane 7 compared with lane 4, Fig. 4B). These data indicated that PSAP regulated SOD1^G93A-induced apoptosis via Bax–Bak interaction.

This study demonstrated that a specific mitochondrial protein PSAP played an important role in SOD1^G93A-induced apoptosis through Bax-Bak interaction. These findings provided new insights into the role of PSAP in motor neuron death and axon degeneration in ALS, highlighting the potential of PSAP as a novel therapeutic target for ALS treatment and prevention.

ALS is a neurodegenerative disease characterized by motor neuron death and axon degeneration that rapidly progresses after onset (Rowland and Shneider, 2001). Statistically, the average survival time of patients with ALS is ∼18 months (Bonafede et al., 2016). Studies have shown that oxidative stress, excitotoxicity, inflammation, and apoptosis can induce ALS (Krishnaraj et al., 2016). The mechanisms of motor neuron death remain unclear, and hence current therapeutic strategies are limited. Riluzole and edaravone are approved by the Food and Drug Administration for treating ALS (Dash et al., 2018; Jaiswal, 2019). Although they delay the development of ALS and prolong the life of patients by months, the benefits are negligible. Huang et al. (2013) demonstrated that DR6 was involved in the etiology of motor neuron death and axon degeneration, and that blocking DR6 with antibodies promoted motor neuron survival, improved motor neuron function, and protected neuromuscular junction integrity. Further evidence indicated that DR6 played a pivotal role in neuronal cell death and induced apoptosis through its interaction with Bax via a unique mitochondrial pathway (Zeng et al., 2012).

PSAP, also known as mitochondrial carrier homolog 1 (MTCH1) due to its homology with a mitochondrial carrier protein, was identified as a specific mitochondrial resident protein and shown to induce apoptosis via Apaf-1 and Smac (Li et al., 2013). Another mitochondrial carrier homolog, mitochondrial carrier homolog 2 (MTCH2), also known as met-induced mitochondrial protein, has been reported to interact with tBid in the mitochondria, forming a protein complex containing Bax (Lamarca et al., 2008). PSAP is the key to enhancing the function of the permeability transition pore (MOMP), which is a target of mitochondrial pro-apoptotic proteins or other related proteins. However, the mechanism(s) by which PSAP regulates mitochondrial apoptosis remains elusive (Lamarca, et al., 2008). Previous data demonstrated that PSAP -Bax complex formation was detected on the overexpression of DR6 or following Tumor Necrosis Factor α (TNFα) (Zhang et al., 2020). More interestingly, PSAP was crucial for DR6-induced apoptosis. DR6 triggered PARP cleavage, caspase activation, Cyt C release, and Bax translocation following PSAP silencing in vitro. Transgenic mice carrying mutant human SOD1, particularly human SOD1^G93A, were similar to patients with ALS in the clinic and hence could be used as reliable fALS models to dissect the mechanisms of cell death (Xie et al., 2015). Thus, NSC-34 cells were transfected with SOD1^G93A to serve as an ALS cell model in vitro. PSAP specifically localized on the mitochondria, and Bcl-2 family members, including Bax and Puma, were associated with the pathology of ALS (Gould et al., 2006; Kieran et al., 2007). The study first examined the role of Bcl-2 family members in SOD1^G93A-induced apoptosis. CCK-8, PARP cleavage, and Annexin V/PI double staining showed that both Bax and Bak played important roles in SOD1-induced apoptosis, as SOD1^G93A-induced cell death was inhibited by Bax or Bak silencing. The study assessed cell activity via CCK-8 assay, PARP cleavage via Western blot analysis, and Annexin V/PI staining via flow cytometry using PSAP siRNA to determine the effects of PSAP on ALS. Similar to DR6, the silencing of PSAP attenuated SOD1^G93A-induced apoptosis.

Oxidative stress, excitotoxicity, inflammation, loss of growth factors, and apoptosis induced ALS. Interestingly, these processes were either directly or indirectly related to mitochondrial function. Mitochondria are crucial for ATP production through oxidative phosphorylation and are critical for neuronal function. Mitochondrial dysfunction results in the loss of ATP production and reactive oxygen species (ROS) generation, promoting cell death. Mitochondrial damage therefore leads to neurodegenerative diseases (Xie et al., 2015). Mutant SOD1 was first identified by Rosen et al. in 1993 and was shown to be involved in ALS. To date, ≥130 mut SODl cell and animal models have been established for clinical research and drug development. Despite the initial promise, SOD1 gene therapy for patients with ALS has proved ineffective. SOD1^G93A and other mutation models still hold an important research value because ALS is not a simple motor neuron disease, as changes in non-motor neuron cells also contribute to disease pathogenesis (Reyes et al., 2010). ALS-related genes, including wtSOD1 and mutSOD1, are associated with mitochondrial dysfunction (Cozzolino et al., 2013; Xie, et al., 2015). The key mitochondrial regulatory factors, the Bcl-2 family, also play an important role in SOD1^G93A-induced apoptosis, including Bax and Bak, which regulate the intrinsic mitochondrial apoptosis pathway. Bax and Bak shuttle between the cytosol and the mitochondria outer membrane (MOM) in healthy cells (Todt et al., 2015). In response to an apoptotic stimulus, Bax and Bak are activated and accumulate on the MOM, triggering MOMP, leading to Cyt C release, capsase activation, and apoptosis (Pena-Blanco and Garcia-Saez, 2018). In addition, the mitochondrial apoptosis -induced channel (MAC) was identified as an outer membrane channel associated with apoptosis, although the role of Bax and Bak in MAC regulation remains undefined (Kroemer et al., 2007).

IP was performed to determine the relationship between SOD1, PSAP, Bax, and Bak so as to further explore the mechanism of PSAP regulating SOD1^G93A-induced apoptosis. A higher number of SOD1 -Bax complexes were evident after the overexpression of wtSOD1, while SOD1 -Bax complex formation decreased in response to SOD1^G93A overexpression. Bax translocation from the cytosol into the mitochondria was also observed in SOD1^G93A but not wtSOD1-overexpressing cells. Generally, SOD1 localized to the cytosol, and SOD1 -Bax complexes were abundant following wtSOD1 transfection. Higher levels of Bax translocated into the mitochondria; SOD1^G93A transfection explained the decreased SOD1-Bax interaction observed in cells expressing this mutant. The overexpression of SOD1 could disrupt Bax-Bak, PSAP-Bax, and PSAP-Bak complex formation in wtSOD1-transfected cells. In contrast, only slight differences in PSAP-Bax or PSAP-Bak complex formation were observed on SOD1^G93A overexpression. However, SOD1^G93A dramatically increased the number of Bax-Bak complexes, thereby promoting apoptosis. Furthermore, the knockdown of PSAP inhibited SOD1^G93A-induced Bax-Bak complex formation. These data suggested that PSAP regulated the formation of the Bax-Bak complex, either directly or indirectly, and mediated its effects on apoptotic induction through MOMP.

In summary, the results indicated that the mitochondrial protein PSAP was required for SOD1^G93A-induced apoptosis mediated through mitochondrial apoptosis-induced channels. PSAP is pivotal for Bax-Bak interaction, leading to caspase activation and apoptosis. Inhibiting these activities of PSAP therefore represents a potential approach for treating ALS.