Brucine-Induced Neurotoxicity by Targeting Caspase 3: Involvement of PPARγ/NF-κB/Apoptosis Signaling Pathway

Brucine, a weak alkaline indole alkaloid, is one of the main bioactive and toxic constituents of Strychnos nux-vomica L., which exerts multiple pharmacological activities, such as anti-tumor, anti-inflammatory, and analgesic effect. However, its potential toxic effects limited its clinical application, especially central nervous system toxicity. The present study was designed to investigate the neurotoxicity and mechanism of brucine. Our results showed that brucine significantly induced Neuro-2a cells and primary astrocyte death, as evidenced by MTT assay and LDH release. Moreover, transcriptome analysis indicated that PPAR/NF-κB and apoptosis signaling pathways were involved in the brucine-induced cytotoxicity in Neuro-2a cells. Subsequently, in fact, brucine evidently inhibited PPARγ and promoted phosphorylation of NF-κB. Furthermore, PPARγ inhibitor aggravated the neurotoxicity, while NF-κB inhibitor substantially reversed brucine-induced neurotoxicity. Moreover, brucine also significantly induced neuronal apoptosis and triggered increase in ratio of Bax/Bcl-2 and level of cleaved caspase 3, as well as its activity as evidenced by TUNEL staining and Western blot. Furthermore, molecular docking analysis predicted that brucine directly bound to caspase 3. Intriguingly, a caspase 3 inhibitor (Z-DEVE-FMK) largely abolished the neurotoxicity of brucine. Our results reveal that brucine-induced neurotoxicity via activation of PPARγ/NF-κB/caspase 3-dependent apoptosis pathway. These findings will provide a novel strategy against brucine-induced neurotoxicity.


Introduction
Brucine is one of the major bioactive and toxic constituents of Strychnos nux-vomica L., which is a traditional Chinese medicine, and widely used for the treatment of rheumatic pain, cancers, and so on (Ma et al. 2018;Savalia and Chatterjee 2017), while clinical applications of brucine are limited due to its severe adverse drug effects, especially neurotoxicity Li et al. 2018). Brucine-induced neurotoxicity mainly involves seizure and amyotrophy (Liu et al. 2015;Seeman and Tantillo 2020). However, until now, the detailed mechanism of brucineinduced neurotoxicity remains a mystery. Emerging evidence indicates that neuroinflammation and apoptosis play crucial roles in neurotoxicity (Dang et al. 2021;Pei et al. 2021;Zheng et al. 2013). Upon stimuli, neuroinflammation is evoked and coupled with the increased secretion of pro-inflammatory cytokines, which exacerbates neurotoxicity and results in irreversible neuronal apoptosis (Hong et al. 2020;Seshadri 2021). Apoptosis, one type of programmed cell death, is regulated principally by interactions within the B-cell lymphoma-2 (Bcl-2) family of proteins (Ney et al. 2021;Wang et al. 2020); among them, Bcl-2 and Bcl-2-associated X protein (Bax) are the vital members, which mediate multiple effectors thereby activating caspases (Reyna et al. 2017;Yang et al. 2020). Especially, activation of caspase 3, an eventual protein in the apoptotic cascade, leads to neuronal apoptosis (Chung et al. 2018;Huang et al. 2021). Of note, previous study revealed that brucine induces cancer cell apoptosis, along with decrease of Bcl-2 and increase of Bax and caspase 3 (Seshadri 2021;Bahrami et al. 2017), whereas whether neuroinflammation and caspase 3-dependent apoptosis contribute to brucine-induced neurotoxicity remains unknown.
Thus, the present study was designed to unveil the potential mechanism of brucine-induced neurotoxicity. Transcriptomics, coupled with molecular docking analysis, indicated vital roles of neuroinflammation and apoptosis in brucineinduced neurotoxicity. Our results uncover that brucine induces neuronal apoptosis through mediating PPARγ/ NF-κB/apoptosis signaling pathway. The findings highlight inhibition of caspase 3 or activation of PPARγ to hinder apoptosis as a potential strategy against brucine-induced neurotoxicity.
Primary astrocytes were extracted from neonatal rat brains as our previous study . The primary astrocytes were fixed with 4% paraformaldehyde, then incubated with anti-GFAP antibody, and visualized by an Alexa Fluorconjugated secondary antibody. GFAP staining was used to identify the primary astrocytes up to more than 95% for the following experiment. Primary astrocytes were cultured with DMEM/F12 medium supplemented with 10% FBS and 1% penicillin/streptomycin. The cells were grown in a 5% CO 2 humidified atmosphere with 5% at 37 °C. Primary astrocytes were treated with different concentrations of brucine or 10 μM Z-DEVE-FMK (caspase 3 inhibitor) or 10 μM GW 9662 (PPARγ inhibitor) or gliotoxin (NF-κB inhibitor) 24 h.

Cell Viability Determination
Neuro-2a cells (1 × 10 4 ) or primary astrocytes (1 × 10 4 ) were treated as mentioned above. At the end of treatment, 20 μl MTT (5 mg/ml) was added into the medium and cultured for another 4 h. Thereafter, the medium was carefully removed and 150 μl DMSO was added into each well. The optical density value of the cells was measured by a microplate reader at a wavelength of 490 nm. Cell viability was expressed as the percentage relative to the absorbance of the untreated control cells.

Detection of LDH Level
The level of LDH of Neuro-2a or primary astrocytes after exposure of brucine was determined using a LDH assay kit. In brief, the Neuro-2a cells or primary astrocytes were treated as mentioned above. Briefly, supernatant was collected by centrifugation (3000 × g, 20 min), and the level LDH release was evaluated using a LDH assay kit according to the protocol's instructions. Thereafter, absorbance was detected at wavelength of 450 nm.

Cell Morphological Observation
For morphological observation, Neuro-2a cells or primary astrocytes were seeded in 6-well plates and treated with brucine as mentioned above. Then, the cells were observed by phase contrast microscopy (Olympus, IX53 + DP20, Japan).

Transcriptome Analysis
Total RNA was extracted from the Neuro-2a cells of control and brucine group using TRIzol reagent according the experimental protocol; then, the quality of RNA samples quantified was detected based on Nanodrop ND-2000 system (Thermo Scientific, USA) and Agilent Bioanalyzer 4150 system (Agilent Technologies, CA, USA). Finally, qualified RNA transcriptome sequencing was performed with an Illumina Novaseq 6000/MGISEQ-T7 instrument supported by Shanghai Applied Protein Technology. Principal component analysis (PCA), an unsupervised analysis that reduces the dimension of the data, was carried out to visualize the distribution and grouping of the samples, which was also commonly used to assess differences between groups and the quality of the biological replicates of samples within groups. Differential expression analysis was performed using the DESeq2 (http:// bioco nduct or. org/ packa ges/ relea se/ bioc/ html/ DESeq2. html); differential genes (DGEs) with |log2FC|> 1 and adjusted p < 0.05 were considered to be significantly different expressed genes. To directly observe the profile of potential DEGs, the up-and downregulation of DEGs were visualized by volcano plot. Differential gene clustering is used to judge the changes of differential genes between brucine and control group. KEGG (Kyoto Encyclopedia of Genes and Genomes, http:// www. kegg. jp/) is one of the databases commonly used in pathway research. We use cluster Profiler R software package for GO function enrichment and KEGG pathway enrichment analysis, and P < 0.05, it is considered that the GO or KEGG function is significantly enrichment.

TUNEL Staining
The apoptosis of Neuro-2a cells was detected using the TUNEL assay by the One Step TUNEL Apoptosis Assay Kit according to the manufacturer's instructions. Briefly, the Neuro-2a cells were treated as described above; then, the cells were fixed with 4% paraformaldehyde for 30 min at room temperature and then incubated with 0.3% Triton X-100 in PBS for 10 min. Thereafter, the Neuro-2a cells were incubated in TUNEL test solution at 37 °C for 1 h in the dark, while the cell nuclei were labeled using DAPI staining. And TUNEL-positive cells were observed using a fluorescence microscope (Olympus IX73, Olympus, Tokyo, Japan).

Detection of Apoptosis Using Flow Cytometry
The apoptosis of treated cells was examined using Annexin V-FITC/PI apoptosis reagents according to the manufacturer's protocol. Briefly, the Neuro-2a cells were treated as mentioned above. The cells were then labeled with 500 µl binding buffer containing 2 µl Annexin V and 5 µl PI in the dark for 15 min after washed with PSB. Subsequently, apoptosis was detected using a flow cytometer (Navios, Beckman Coulter, USA), and the Annexin V and PI values were set as the horizontal and vertical axes, respectively. Mechanically damaged, late apoptotic, dual negative/normal, and early apoptotic cells were located in the upper left, upper right, lower left, and lower right quadrants of the flow cytometric dot plot, respectively.

Measurement of Caspase 3 Activity
The Neuro-2a cells or primary astrocytes were treated as described above. Cells supernatant was collected by centrifugation (3000 × g, 20 min, 4 °C), and the caspase 3 assay was evaluated using caspase 3 assay kit according to the manufacturer's recommended protocol. Thereafter, absorbance was detected at wavelength of 450 nm, and all values of % caspase 3 activity were normalized to the untreated control group.

PPARγ Activity Assay
The primary astrocytes were treated as described above. The cells were dissolved in RIPI lysis buffer containing protease inhibitor and PMSF and kept on ice for 30 min. Then, cell supernatant was collected by centrifugation (1000 g, 15 min, 4 °C); protein concentration was determined by the BCA assay. The DNA-binding activity of PPARγ was evaluated by ELISA kit according to the manufacturer's recommended protocol. Thereafter, absorbance was detected at wavelength of 450 nm.

NF-κB DNA Binding Activity Assay
The primary astrocytes were treated with or without brucine as described above. Afterward, the NF-κB binding activity was following determined using the NF-κBp65 Transcription Factor Assay Kit according the manufacturer's instructions. Thereafter, absorbance was detected at wavelength of 450 nm.

Evaluation of Inflammatory Factors
The levels of inflammatory factors were detected using ELISA assay. Briefly, the Neuro-2a cells or primary astrocytes were treated as mentioned above. Then, the supernatants were collected with centrifugation (3000 × g for 10 min at 4 °C). Thereafter, the levels of IL-1β, IL-6, and TNF-α were detected according to the manufacturer's instructions.

Molecular Docking
Molecular docking analysis between brucine and caspase 3 was performed using Autodock 4.2 and Autodock Tools (ADT). In brief, the caspase 3 (Protein Data Bank ID: 3KJF) protein was obtained from the Protein Data Bank database. A three-dimensional structure of brucine was established using the ChemBio3D Ultra 14.0 (PerkinElmer Informatics, USA), which was further proceeded using ADT. Thereafter, the molecular docking of the brucine and caspase 3 proteins was performed using Autodock 4.2.

Statistical Analysis
All data were obtained from at least three independent experiments and expressed as mean ± SD and were analyzed using SPSS version 18.0 (SPSS, Inc., Chicago, USA). Comparisons of the different groups were performed with Student's t-test. P < 0.05 was considered the minimal level of significance.

Brucine-Induced Cytotoxicity in Neuro-2a Cells
In order to explore the effect of brucine on Neuro-2a cells, cell viability and cytotoxicity were evaluated by MTT method and LDH release, respectively. The results showed that brucine at the concentrations of 1-2 μM did not affect the cell viability, and 4-128 μM induced the Neuro-2a cell death. Since brucine induced Neuro-2a cell death up to 50% at the dose of 16 μM, we chose the concentrations of 1, 4, and 16 μM in the following experiments (Fig. 1B). In parallel, brucine (1, 4, 16 μM) markedly increased levels of LDH (Fig. 1C). Moreover, after exposure to brucine, a reduced number and morphological changes in Neuro-2a cells were observed than those of control group (Fig. 1D). These findings suggest that brucine exerted cytotoxicity on Neuro-2a cells.

Transcriptome Processing and Data Analysis
PCA was commonly used to evaluate the differences between groups and the quality of the biological repetition of samples within groups. The PCA score plot showed that the gene expression between brucine treatment and control groups was significant ( Fig. 2A). Moreover, the volcano plot analysis showed that 727 DEGs were involved; among them, 186 genes were upregulated and 541 genes were downregulated (Fig. 2B). The hierarchical clustering analysis of gene expression changes further revealed a clear separation between brucine and control group (Fig. 2C). Furthermore, the results revealed that the top 20 pathways enriched from KEGG database, including NF-κB, PPARs, and apoptosis signaling pathways (Fig. 2D). These findings indicate that PPARγ/NF-κB/apoptosis signaling pathway is involved in the brucine-induced cytotoxicity.

Brucine Elicited Inflammatory Cytokines through Regulating PPARγ/NF-κB Signaling Pathway
Based on the analysis of transcriptome, we further detected the levels of inflammatory cytokines and the protein expressions of PPARs and NF-κB by ELISA assay and Western blot, respectively. These results showed that the levels of IL-1β, IL-6, and TNF-α were remarkably increased after brucine treatment than those of control group (Fig. 3A-C). Moreover, brucine significantly increased the phosphorylation level of NF-κBp65 than that of control group (Fig. 3D, E). In addition, brucine evidently decreased the protein expression of PPARγ, but did not change the protein expressions of PPARα and PPARβ/δ than those of control group (Fig. 3F-I). These findings suggest that brucine promotes inflammatory cytokine release through activating PPARγ/NF-κB signaling pathway.

Brucine-Induced Cytotoxicity through Inducing Apoptosis
TUNEL staining was used to evaluate whether brucine induced apoptosis in Neuro-2a cells. The results showed that brucine significantly increased the number of TUNEL-positive cells than that of control group (Fig. 4A,  B). Furthermore, the effect of brucine-induced apoptosis in Neuro-2a cells was verified using Annexin V-FITC/ PI staining by flow cytometry. The results demonstrated that apoptotic cells accounted for 6.5%, 38.7%, and 54.3% after treatment with brucine (1, 4, and 16 µM) in a concentration-dependent manner (Fig. 4C, D). These findings suggest that brucine induced cytotoxicity through inducing apoptosis.

Brucine-Induced Apoptosis through Activation of Caspase 3 Pathway
Western blot was used to further explore the molecular mechanism of brucine-induced apoptosis. The results showed that brucine significantly increased Bax/Bcl-2 ratio and cleaved-caspase 3 level ( Fig. 5A-C), as well as caspase 3 activity than those of control group (Fig. 5D). These findings indicate that brucine induced cytotoxicity, at least partly, through caspase 3-dependent apoptosis pathway.  3). The data were presented as the mean ± SD. * P < 0.05, ** P < 0.01 vs control group

Brucine-Induced Neurotoxicity on Primary Astrocytes
Based on the findings mentioned above, we further investigate whether brucine induced neurotoxicity. Since astrocyte is an important kind of neurocyte, which is rich of PPARγ and response to inflammation upon neurotoxic stimuli, we used primary astrocytes to determine the neurotoxic effect and its possible mechanism of brucine. The results showed that brucine at the concentrations of 1-8 μM did not affect the cell viability, and 16-128 μM induced the primary astrocyte death. Since brucine induced primary astrocyte death up to 50% at the dose of 32 μM, we chose the concentrations of 8, 16, and 32 μM in the following experiments (Fig. 6A). In parallel, brucine (8, 16, 32 μM) markedly increased LDH level than that of control group in a dose-dependent manner (Fig. 6B). Moreover, brucine (8, 16, 32 μM) also decreased the number of cells and led to the astrocyte collapse than that . The data were presented as the mean ± SD. * P < 0.05, ** P< 0.01 vs control group of control group (Fig. 6C). These results suggest that brucine induced neurotoxicity.

Brucine Increased the Proinflammatory Cytokines in Primary Astrocytes through PPARγ/NF-κB Signaling Pathway
To confirm the mechanism of brucine-induced neurotoxicity, the levels of proinflammatory cytokines were measured by ELISA assay. The results showed that brucine significantly increased the levels of IL-1β, IL-6, and TNF-α in primary astrocytes than those of control group (Fig. 7A-C). Furthermore, the protein expressions of NF-κBp65 and PPARγ, as well as NF-κBp65 and PPARγ activities, were evaluated using Western blot and appropriate activity kits, respectively. The results showed that brucine markedly increased phosphorylation level of NF-κBp65 and its DNA-binding activity in primary astrocytes than those of control group (Fig. 7D-F). Additionally, brucine significantly decreased the protein expression of PPARγ and its DNA binging activity in primary astrocytes than those of control group (Fig. 7G-I). These findings revealed that brucine-induced proinflammatory cytokine release is also through mediating PPARγ/NF-κB signaling pathway. To further investigate the mechanism by which brucine-induced neurotoxicity. We detected Bax/Bcl-2 ratio and caspase 3 activity. The results showed that brucine significantly increased Bax/Bcl-2 ratio (Fig. 7J, K) and caspase 3 activity (Fig. 7L) in primary astrocytes than that of control group. These findings demonstrate that brucine induced neurotoxicity through caspase 3-dependent signaling pathway.

Brucine Directly Bound to Caspase 3
Molecular docking was utilized to further explore the interaction between brucine and caspase 3. The results showed that docking score of brucine with caspase 3 was − 5.21 kcal/ mol and the suppositive binding sites between brucine and caspase 3 including THR62, VAL-4, Gly122, and Cys163 (Fig. 9). These findings confirm that caspase 3 might be the potential target of brucine-induced neurotoxicity.

Discussion
The present study, for the first time, revealed that (1) brucine induced neurotoxicity, at least partly, through mediation of PPARγ/NF-κB/apoptosis signaling pathway. (2) Brucine induced apoptosis due to it directly bound with caspase 3 and inhibited its activity (Fig. 10).
Emerging evidence demonstrates that brucine exerts anti-tumor, anti-inflammatory, and analgesic effect in clinic. Albeit brucine has a striking pharmacological profile, severe neurotoxicity is the primary barrier to its clinical application (Lu et al. 2020;Zhou et al. 2019), and the toxicological mechanism of brucine was still unknown. The present study indicated that brucine significantly induced neurotoxicity as evidenced by determination of neuronal cell viability and LDH release, which is consistent with previous study Shi et al. 2018). Notably, PPARs, NF-κB, and apoptosis signaling pathway were predicted during brucine-induced cytotoxicity by transcriptome analysis. PPARs are the nuclear receptor superfamily and can be categorized as three isoforms of PPARα, PPARβ/δ, and PPARγ, which play vital roles in improving glucose and lipid homeostasis, inhibiting neuroinflammation and oxidative stress (Chistyakov et al. 2020;Zolezzi et al. 2017). Among them, PPARγ plays a crucial role in the mediation of inflammatory and immune reactions Zuo et al. 2019). The activation of PPARγ adjusts the release of pro-inflammatory factors Fig. 6 Brucine-induced neurotoxicity on primary astrocytes. Primary astrocytes were treated with brucine at different concentrations (8, 16, and 32 μM) for 24 h. A Cell viability was determined using MTT assay (n = 5). B LDH release from primary astrocytes was measured using commercial LDH assay kit (n = 5). C The morphology of primary astrocytes was observed after brucine treatment (200 ×, scale bar = 200 μm). The data were presented as the mean ± SD. ** P < 0.01 vs control group through mediating the transcription factor activity, such as NF-κB Wang et al. 2017). Our findings indicated that brucine apparently inactivated the PPARγ, but did not affect the PPARα and PPARβ/δ, then reduced pro-inflammatory factor release via promoting phosphorylation of NF-κB, in keeping with the predicted results of transcriptome analysis. Furthermore, we also utilized PPARγ inhibitor and NF-κB inhibitor to verify the role of PPARγ and NF-κB in brucine-induced neurotoxicity. As expected, PPARγ inhibitor aggravated the neurotoxicity, while NF-κB inhibitor substantially reversed brucineinduced neurotoxicity. Interestingly, recent studies report that PPARs directly associated with apoptosis (Cheng and Guo 2021;Thulasi Raman et al. 2020). The results showed that brucine significantly increased the pro-apoptosis protein Bax and decreased the anti-apoptosis protein Bcl-2, as well as elevated the level cleaved caspase 3 protein and its activity. Intriguingly, brucine directly bound to caspase 3, and caspase 3 inhibitor dramatically inhibited brucineinduced neurotoxicity. Our results showed that once PPARγ Fig. 7 Brucine promoted the release of inflammatory factors in primary astrocytes. Primary astrocytes were exposed to brucine (8, 16, 32 μM) for 24 h. A IL-1β level (n = 5). B IL-6 level (n = 5). C TNF-α level (n = 5). D Representative Western blot of phosphorylation level of NF-κBp65. E Quantitation of phosphorylation of NF-κBp65 protein level (n = 3). F NF-κBp65 binding activity (n = 5). G Represent-ative Western blot of PPARγ protein expression. H Quantitation of PPARγ protein expression (n = 3). I DNA binging activity of PPARγ (n = 5). J Representative Western blot of Bax and Bcl-2 protein expressions. K Quantitation of Bax/Bcl-2 ratio (n = 3). L Caspase 3 activity (n = 5). The data were presented as the mean ± SD. * P < 0.05, ** P < 0.01 vs control group was suppressed, Bax increased, and Bcl-2 decreased, which could evoke apoptosis, while brucine inhibited PPARγ, thereby increased the ratio of Bax/Bcl-2 and activation of caspase 3. These findings indicate that brucine induced neurotoxicity through mediating PPARγ/NF-κB/caspase 3 signaling pathway.
It should be noted that although low-level concentrations of brucine were found in the brain, it could induce metabolic acidosis and injure the brain due to its direct poisons or the influence on neurotransmitters in the central nervous system. Brucine has been proved to penetrate the blood-brain barrier (BBB) in animal model Ren et al. 2018); however, there is still lack of reference data about the detailed clinical pharmacokinetic studies. What is more, this study preliminary reveals the neurotoxicity of brucine in vitro; the exact mechanism or target of brucine remains still a mystery. Thus, the detailed pharmacokinetics including evidence that brucine crosses BBB and the levels of brucine in brain tissue or cerebrospinal fluid in patients and the mechanism of brucine-induced neurotoxicity will be further elucidated utilized clinic trial or the animal models in our next story. Fig. 8 Brucine-induced neurotoxicity via mediating NF-κB/PPARγ/ caspase 3 signaling pathway. Neuro-2a cells were treated with brucine at different concentrations (1, 4, or 16 μM) for 24 h and with or without DEVE-FMK, GW 9662, or gliotoxin. A Cell viability was determined using MTT assay. B LDH level. Primary astrocytes were treated with brucine at different concentrations (8, 16, or 32 μM) for 24 h and with or without DEVE-FMK, GW 9662, or gliotoxin. C Cell viability. D LDH release. The data were presented as the mean ± SD. * P < 0.05, ** P < 0.01 vs control group; ## P < 0.01, $$ P < 0.01, && P < 0.01 vs brucine 16 or brucine 32 group Fig. 9 Brucine directly bound with caspase 3. A Amino acid residues. B A visual of the binding sites between brucine and caspase 3. C The substrate binding pocket Fig. 10 Presentation of a proposed mechanism for brucine-induced neurotoxicity. Brucine not only directly binds to caspase 3 and increases its activity, but also mediates PPARγ/NF-κB/apoptosis signaling pathway, thereby promoting inflammatory cytokines