Upregulation of OPA1 by carnosic acid is mediated through induction of IKKγ ubiquitination by parkin and protects against neurotoxicity

Food and Chemical Toxicology xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Food and Chemical Toxicology journal homepage: www.elsevier.com/locate/foodchemtox

Upregulation of OPA1 by carnosic acid is mediated through induction of IKKγ ubiquitination by parkin and protects against neurotoxicity Chia-Yuan Lina, Wen-Jiun Chena, Ru-Huei Fub,c, Chia-Wen Tsaia,∗ a

Department of Nutrition, China Medical University, Taichung, Taiwan Graduate Institute of Biomedical Sciences, China Medical University, Taichung, Taiwan c Translational Medicine Research Center, China Medical University Hospital, Taichung, Taiwan b

A R T I C LE I N FO

A B S T R A C T

Keywords: Carnosic acid 6-hydroxydopamine Mitochondrial dynamics Optic atrophy 1 Parkin/IKKγ/p65 pathway

An imbalance in mitochondrial dynamics is strongly associated with Parkinson's disease. The fusion protein optic atrophy 1 (OPA1) is up-regulated through the activation of parkin-mediated IκB kinase γ (IKKγ)/p65 signaling. This study investigated whether the neuroprotection of carnosic acid (CA) from rosemary is involved in mitochondrial dynamics and OPA1 protein induction by parkin/IKKγ/p65 signaling. The neurotoxin 6-hydroxydopamine (6-OHDA) treated with SH-SY5Y cells decreased OPA1 and mitofusin 2 fusion proteins, but increased fission 1 and dynamin related protein 1 (DRP1) fission proteins. By immunofluorescence, 6-OHDA induced the fluorescence of green spots outside the mitochondria, indicating that cytochrome c was released to the cytoplasm. Except for the effects on DRP1 protein, CA pretreatment reversed these effects of 6-OHDA. Additionally, CA treatment increased the ubiquitination of IKKγ, nuclear p65 protein, OPA1-p65 DNA binding activity, and OPA1 protein. However, transfection of parkin small interfering RNA (siRNA) attenuated these effects of CA. Furthermore, transfection of OPA1 siRNA abolished the action of CA to reverse 6-OHDA-increased cytosolic cytochrome c protein, apoptotic-related protein cleavage, and cell death. In conclusion, the mechanism by which CA counteracts the toxicity of 6-OHDA is through modulation of mitochondrial dynamics and upregulation of OPA1 via activation of the parkin/IKKγ/p65 pathway.

1. Introduction Parkinson's disease (PD), a neurodegenerative disease, is characterized by the progression of dopaminergic neurons of the substantia nigra pars compacta (SNpc), leading to the depletion of dopamine and the impairment of movement (Santos and Cardoso, 2012). A disruption of mitochondrial dynamics is believed to play an important role in the pathogenic mechanism of PD. Mitochondrial dynamics is regulated by fission and fusion (Youle and van der Bliek, 2012), and the balance in dynamics is responsible for maintaining mitochondrial quality and morphology. Mitochondrial fission is controlled by the proteins dynamin-related protein 1 (DRP1) and fission 1 (Fis1) (Hall et al., 2014), whereas mitochondrial fusion is mediated by mitofusin 1 and 2 (MFN1 and MFN2) and optic atrophy 1 (OPA1). OPA1 is widely distributed among the inner membrane of mitochondrial cristae (Griparic et al., 2004). The physiological roles of OPA1 include controlling mitochondrial fusion (Hall et al., 2014), maintaining mitochondrial integrity (Elachouri et al., 2011), regulating mitochondrial cristae structure (Olichon et al., 2003), and inhibiting

∗

cytochrome c release, which is an activator of the mitochondrialmediated apoptotic cascade (Olichon et al., 2003). Inducible pluripotent stem cell line-derived neural cells from PD patients with an OPA1 mutation showed mitochondrial fragmentation, a defective respiratory complex I, and a reduction in ATP (Iannielli et al., 2018). Similarly, silencing of OPA1 in HeLa cells induces fragmentation of the mitochondrial network, collapse of mitochondrial membrane potential (MMP), and disorganization of cristae structure, leading to cytochrome c release and the induction of apoptosis (Griparic et al., 2004). Mitochondrial fragmentation is also detected in depolarized primary rat cerebellar granule neurons treated with extracellular potassium in which the function of OPA1 has been lost (Gray et al., 2013). However, OPA1 overexpression in SH-SY5Y cells inhibits the effects of 1-methyl4-phenylpyridinium on mitochondrial structural abnormalities, cytochrome c release, and cell death (Ramonet et al., 2013). Parkin is an E3 ubiquitin ligase that has recently been shown to be involved in neuroprotection (Lin and Tsai, 2016, 2017). One study showed that exposure of SH-SY5Y cells to PD-linked toxins, such as 6hydroxydopamine (6-OHDA), rotenone, and dopamine, inhibits parkin

Corresponding author. E-mail address: [email protected] (C.-W. Tsai).

https://doi.org/10.1016/j.fct.2019.110942 Received 10 September 2019; Received in revised form 29 October 2019; Accepted 1 November 2019 0278-6915/ © 2019 Elsevier Ltd. All rights reserved.

Please cite this article as: Chia-Yuan Lin, et al., Food and Chemical Toxicology, https://doi.org/10.1016/j.fct.2019.110942

Food and Chemical Toxicology xxx (xxxx) xxxx

C.-Y. Lin, et al.

Abbreviations used CA DMSO DRP1 Fis1 IKK EMSA MFN1 MFN2 MMP

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolim mide NEMO NF-κB essential modulator NF-κB nuclear factor-κB 6-OHDA 6-hydroxydopamine OPA1 optic atrophy 1 PARP poly ADP-ribose polymerase PD Parkinson's disease siRNA small interfering RNA SNpc substantia nigra pars compacta

carnosic acid dimethyl sulfoxide dynamin related protein 1 fission 1 IκB kinase electromobility gel shift assay mitofusin 1 mitofusin 2 mitochondrial membrane potential

bro-

Sigma-Aldrich, Co (St. Louis, MO). Penicillin/streptomycin, L-glutamine, non-essential amino acids, and trypsin were purchased from Gibco Laboratory (Gaithersburg, MD). Fetal bovine serum was purchased from Hyclone™ (Australia). Caspase 3, cleaved caspase 3, poly (ADP-ribose) polymerase (PARP), cleaved PARP, and parkin primary antibodies were purchased from Cell Signaling Technology, Inc. (Danvers, MA). OPA1, DRP1, Fis1, MFN2, β-tubulin, IKKγ, cytochrome c, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primary antibodies were purchased from Santa Cruz Biotechnology, Inc. (Dallas, Texas). Ubiquitin and p65 primary antibodies were obtained from EMD Millipore Corporation (Temecula, CA). Western Lightning Plus-ECL, enhanced chemiluminescence substrate, the horseradish peroxidaseconjugated goat anti-rabbit and goat anti-mouse IgG were purchased from PerkinElmer, Inc. (Waltham, MA). Rabbit anti-goat IgG was purchased from R&D Systems Inc. (Minneapolis, MN). Glycine, acrylamide, and tris were obtained from USBiological Life Sciences (Swampscott, MA).

protein and induces apoptosis, whereas overexpression of parkin protects against the pro-apoptotic effects of these toxins (Jiang et al., 2004). In 6-OHDA-lesioned rats, parkin overexpression inhibits 6OHDA-induced degeneration of dopaminergic neurons and motor impairment. Thus, parkin is essential for preventing PD. More importantly, the neuroprotective role of parkin is related to the upregulation of OPA1 by IκB kinase (IKK)/nuclear factor-κB (NF-κB) signaling (Henn et al., 2007; Winklhofer, 2014). Briefly, parkin can add ubiquitin chains to IKKγ/NF-κB essential modulator (NEMO) to activate IKKα/β, which in turn causes IκBα degradation. Then, the transcription factor NF-κB, a p50/p65 heterodimer, translocates into the nucleus and binds to its target gene OPA1 (Henn et al., 2007; Winklhofer, 2014). The rosemary constituent carnosic acid (CA) has multiple biological properties, such as antioxidant, antivirus, anticancer, and neuroprotective actions (Barni et al., 2012; Shin et al., 2013; Yoshida et al., 2014). Emerging lines of evidence recently indicate that the action of CA against toxins is involved in protecting mitochondria. For example, in young adult CF-1 mice administered CA before the isolation of cortical mitochondria, the reduction in mitochondrial complex I and II caused by treatment with 4-hydroxyl-2-nonenal is inhibited (Miller et al., 2013). In other research, pretreatment of SH-SY5Y cells with CA was shown to protect the cells against the cytotoxicity of paraquat. In that model, CA protects against the inhibition of activity of mitochondrial complex I and V, ATP synthesis, and MMP, as well as activation of the cytochrome c-elicited apoptotic pathway (de Oliveira et al., 2016). Moreover, CA pretreatment reverses hydrogen peroxide-suppressed tricarboxylic acid cycle enzymes, including aconitase, α-ketoglutarate dehydrogenase, and succinate dehydrogenase (de Oliveira et al., 2018). In our published reports, the protection mechanisms of CA in 6-OHDAtreated neurotoxicity are associated with the activation of ubiquitin proteasome system and autophagy by up-regulating the protein of parkin (Lin and Tsai, 2016, 2017). Additionally, CA also induces PTENinduced kinase 1/parkin-mediated mitophagy to remove the impaired mitochondria and reduce the 6-OHDA-elcited mitochondrial-mediated apoptotic cascades (Lin and Tsai, 2019). Although a role for CA has been established in protecting mitochondria, no research has shown evidence of its neuroprotective role in connection with the regulation of mitochondrial dynamics and the induction of OPA1 protein. Moreover, whether the parkin-mediated IKKγ ubiquitination pathway plays an important role in OPA1 level by CA has not been investigated. This study therefore aimed to explore whether CA protects against the apoptosis induced by 6-OHDA in SH-SY5Y cells by modulating mitochondrial fission and fusion proteins as well as upregulating OPA1 protein via the parkin/IKKγ pathway.

2.2. Cell culture and treatment The protocol for cell culture was performed as in our previous study (Lin et al., 2014). Human SH-SY5Y cells were cultured with DMEM at 37 °C under a humidified atmosphere of 95% air and 5% CO2. CA and 6OHDA were dissolved in DMSO. Cells were pretreated with 1 μM CA once the cells reached 80% confluence. After 24 h, cells were then treated with 100 μM 6-OHDA for an additional 12 or 18 h. Control cells were treated with 0.3% DMSO.

2.3. Cell viability assay 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolim bromide (MTT) was used to determine cell viability. The absorbance of purple formazan crystals was detected at 570 nm by use of an ELISA reader (Bio Rad, Japan). The value of control cells was considered to be 100%.

2.4. Nuclear protein extraction and electromobility gel shift assay The method of nuclear protein preparation was described in our previous study (Lin et al., 2014). The interaction of protein-nucleic acid was measured by electromobility gel shift assay (EMSA). We designed the synthetic biotin-labeled double-stranded human OPA1 p65 oligonucleotides, as follows: forward, 5′-TTCCTGGGTCATTCCTGGAC-3’; reverse, 5′-GTCCAGGAATGACCCAGGAA-3’. According to our previous study (Lin et al., 2014), each nuclear protein-DNA complex was separated by electrophoresis on a 6% tris-boric acid-EDTA-polyacrylamide gel, and was then electro-transferred to a Hybond-N+nylon membrane (GE Healthcare Life Sciences, Marlborough, MA). After DNA crosslinking for 10 min, the membrane was reacted with streptavidinhorseradish peroxidase for 0.5 h. The bands of nuclear protein-DNA were detected by using an enhanced chemiluminescence kit.

2. Materials and methods 2.1. Chemicals and reagents CA (purity 95%) was purchased from A.G. Scientific Inc. (San Diego, CA). 6-OHDA, sodium bicarbonate, sodium pyruvate, dimethyl sulfoxide (DMSO), triton X-100, and Tween 20 were purchased from 2

Food and Chemical Toxicology xxx (xxxx) xxxx

C.-Y. Lin, et al.

2.5. Western blot analysis

2.7. Immunofluorescence assay

Cells were lysed in lysis buffer and then lysates were obtained by centrifugation at 16,000 хg for 20 min at 4 °C. Protein concentrations of lysates were measured with a Pierce™ Coomassie Plus (Bradford) Assay Reagent (Thermo Fisher Scientific, Rockford, IL). Ten or fifteen micrograms of protein from each lysate was subjected to 7.5%, 10%, and 12.5% SDS-PAGE gels and was then electrophoretically transferred to polyvinylidene fluoride membranes (Merck Millipore Ltd., Co. Cork, IRL). Blots were blocked with 50 g/L nonfat dry milk and were then incubated with primary antibodies. Bands were detected by using an enhanced chemiluminescence kit under a luminescent image analyzer (LAS-4000, FUJIFILM, Japan).

The method was described in our previous study (Lin and Tsai, 2019). After fixation with 10% formaldehyde and 2% sucrose, cells were permeabilized with 0.3% Triton-X 100 and were blocked with 3% bovine serum albumin. Then, the primary cytochrome c antibody was added to the cells overnight at 37 °C followed by reaction with the goat anti-mouse IgG (H + L) secondary antibody-Alexa Fluor 488 conjugate (purchased from Thermo Fisher Scientific, Rockford, IL) at 25 °C for 2 h. Nuclei and mitochondria were stained with Hoechst 33258 dye (purchased from Sigma-Aldrich, Co, St. Louis, MO) and MitoTracker ® Red CMXRos (purchased from Molecular Probes, Eugene, OR), respectively. A fluorescence microscope was used to detect the fluorescence. 2.8. Transient transfection of small RNA interference

2.6. Immunoprecipitation assay

The small RNA interference (siRNA) system was done as in our earlier study (Lin and Tsai, 2016). We designed the sense sequences of human parkin (5′-UUCGCAGGUGACUUUCCUCUGGUCA-3′) and OPA1 (5′-AAGTTATCAGTCTGAGCCAGGTT-3). According to the manufacturer's instructions, cells were transfected with two siRNAs by using the DharmaFECT 1 siRNA transfection reagent (from Thermo Fisher Scientific, Lafayette, CO) for 24 h, and were then treated with 1 μM CA for 1 (ubiquitin), 3 (parkin), and 24 (OPA1) h, respectively. In the cell death experiment, cells were exposed to 1 μM CA for 24 h followed by

The protocol was measured according to our previous study (Lin and Tsai, 2017). The IKKγ primary antibody was mixed with each lysate at 4 °C overnight, and the protein A-sepharose beads (0.1 g/L) were added to each lysate with IKKγ primary antibody. The mixtures were centrifuged at 14,600 хg for 20 min at 4 °C. Each immunoprecipitationcomplex was boiled with sample buffer for 10 min at 95 °C, and was then measured by Western blotting.

Fig. 1. Effect of CA and 6-OHDA on the expression of fusion and fission proteins. (A) SH-SY5Y cells were cultured with 0.1% DMSO (control, C) or 100 μM 6-OHDA for the indicated time. (B) SH-SY5Y cells were pretreated with 0.3% DMSO (control, C) or 1 μM CA for 24 h, and were then exposed to 100 μM 6-OHDA for an additional 6 h (fission proteins) or 12 h (fusion proteins). Proteins were measured by Western blotting. GAPDH and β-Tubulin served as an internal control. The protein expression of control cells was regarded as 1. One representative immunoblot out of three individual experiments is shown. Values are means ± SD of three individual experiments. Groups not sharing a common letter differ significantly, p < 0.05. The results of CA pretreatment were analyzed by use of Student's t-test. *p < 0.05 compared with control cells. #p < 0.05 compared with 6-OHDA cells only.

3

Food and Chemical Toxicology xxx (xxxx) xxxx

C.-Y. Lin, et al.

treatment with 100 μM 6-OHDA for 12 (apoptotic-related proteins) and 18 (MTT assay) h.

c primary antibody was used to explore the action of CA on the distribution of cytochrome c upon 6-OHDA treatment. As shown in Fig. 2, the green fluorescence is cytochrome c, the red fluorescence is the mitochondria, and the blue fluorescence is the nucleus. In the merged images, control cells appeared yellow, indicating that cytochrome c was distributed over the mitochondria. Green dots were shown in 6-OHDAtreated cells, which suggested that 6-OHDA increased the existence of cytochrome c outside the mitochondria (white arrow). In the presence of CA, the appearance of green dots outside the mitochondria induced by 6-OHDA was reduced. Therefore, CA can improve mitochondrial cytochrome c release upon treatment of cells with 6-OHDA.

2.9. Statistical analysis Commercially available software (SAS Institute Inc., Cary, NC) was used to perform the statistical analysis. Statistical significance was determined by using one-way ANOVA followed by Tukey's test. In the results shown in Figs. 1B and 5, Student's t-test was used to compare two groups. In the results, p values < 0.05 were assumed to indicate statistically significant differences.

3.3. CA increases the protein expression of OPA1 and the interaction of IKKγ and ubiquitin protein

3. Results 3.1. CA reverses the effects of 6-OHDA on the expression of fusion and fission proteins

Culturing cells with CA induced protein expression of OPA1 at 18, 24, and 36 h (Fig. 3A). This elevation in OPA1 protein is related to activation of the parkin/IKKγ/p65 pathway (Muller-Rischart et al., 2013). Moreover, parkin can add ubiquitin chains to IKKγ/NF-κB essential modulator to activate the translocation of p65 to the nucleus and up-regulate the transcription of the OPA1 gene (Henn et al., 2007; Winklhofer, 2014). To evaluate how CA interacted in the relationship between IKKγ and ubiquitin, we performed an immunoprecipitation assay using primary IKKγ antibody. As shown in Fig. 3B, the expression of ubiquitinated protein with IKKγ was increased after exposure to CA for 1 h, which suggests that CA can promote the interaction of ubiquitin protein with IKKγ.

To investigate the roles of CA and 6-OHDA in mitochondrial dynamics, we examined the expression of fusion and fission proteins by Western blotting. As shown in Fig. 1A, exposure of cells to 6-OHDA decreased the expression of the fusion proteins OPA1 and MFN2 but increased the expression of the fission proteins Fis1 and DRP1. Pretreatment of cells with CA inhibited the effects of 6-OHDA on the expression of these proteins (Fig. 1B). Compared with 6-OHDA alone, CA pretreatment reversed the expression of OPA1 and MFN2 protein by 147.1% and 104.8%, respectively. Moreover, CA pretreatment decreased the 6-OHDA-induced induction of Fis1 by 56.6%. CA pretreatment had no significant effect on the expression of DRP1 protein.

3.4. CA induces activation and DNA binding activity of p65 3.2. CA attenuates mitochondrial cytochrome c release induced by 6-OHDA In our previous study, we showed that CA activates p65 in SH-SY5Y cells (Lin et al., 2014). Similarly, in the present study, p65 protein was activated in cells treated with CA for 3 h (Fig. 4A), which suggests that

The release of cytochrome c to the cytoplasm is a hallmark of impaired mitochondria. An immunofluorescence assay with a cytochrome

Fig. 2. Effect of CA on the release of cytochrome c after 6-OHDA exposure. SH-SY5Y cells were pretreated with 0.1% DMSO (control, C) or 1 μM CA for 18 h, and were then exposed to 100 μM 6-OHDA for an additional 12 h. The green, red, and blue fluorescence indicate cytochrome c, mitochondria, and nuclei, respectively. The merged images show the overlap of the three fluorescences. The white arrows in the merged image show the released cytochrome c outside the mitochondria. Representative images from three individual experiments are shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

4

Food and Chemical Toxicology xxx (xxxx) xxxx

C.-Y. Lin, et al.

Fig. 3. Effect of CA on the protein expression of OPA1 and the ubiquitination of IKKγ. (A) SH-SY5Y cells were treated with 0.1% DMSO (control, C) or 1 μM CA for the indicated time. OPA1 protein expression was measured by Western blotting. β-Tubulin served as an internal control. (B) Each lysate was mixed with the primary IKKγ antibody. After immunoprecipitation assay, immunoprecipitation-complexes were measured by Western blotting with ubiquitin and IKKγ primary antibodies. The protein expression of control cells was regarded as 1. One representative immunoblot out of three individual experiments is shown. Values are means ± SD of three individual experiments. Groups not sharing a common letter differ significantly, p < 0.05. IB: immunoblotting.

Fig. 4. Effect of CA on nuclear protein and OPA1-p65 DNA binding activity. (A) SH-SY5Y cells were treated with 0.1% DMSO (control, C) or 1 μM CA for 0.5, 1, and 3 h. Nuclear protein was measured by Western blotting. PARP served as an internal control. The protein expression of control cells was regarded as 1. Values are means ± SD of three individual experiments. Groups not sharing a common letter differ significantly, p < 0.05. (B) SH-SY5Y cells were treated with 0.1% DMSO (control, C) or 1 μM CA for 1, 3, and 6 h. The DNA binding activity of p65 was measured by EMSA. Unlabeled doublestranded oligonucleotide (cold) was used to confirm the specificity of DNAprotein binding for p65. One representative immunoblot out of three individual experiments is shown.

CA can induce the translocation of p65 from the cytosol to the nucleus. We further determined the action of CA on p65 DNA binding activity by EMSA. After treatment with CA for 3 h, the DNA binding activity of p65 was augmented. The unlabeled double-stranded oligonucleotide (cold) was considered a competitive assay to confirm the specificity of the DNA-protein interaction for p65 (Fig. 4B). Taken together, these data suggest that activation of p65 is involved in the upregulation of OPA1 by CA.

that parkin is required for CA-mediated activation of the IKKγ/p65/ OPA1 pathway. 3.6. OPA1 siRNA inhibits the anti-apoptotic effect of CA During the initiation of intrinsic apoptotic signaling, mitochondrial cytochrome c is released to the cytoplasm and then activates the downstream of executioner caspases, including caspase 3, which in turn increases the cleavage of PARP (McIlwain et al., 2013). In our previous study, we showed that CA protects SH-SY5Y cells against 6-OHDA-induced apoptosis (Fu et al., 2018). In the present study, OPA1 siRNA was transfected into cells to clarify the role of OPA1 in the anti-apoptotic effect of CA. Culturing cells with 6-OHDA decreased the protein expression of OPA1 and increased the release of cytochrome c and the ratio of apoptotic-related proteins, including cleaved caspase 3/caspase 3 and cleaved PARP/PARP. In CA-pretreated cells, however, the effects of 6-OHDA were alleviated. After transfection with OPA1 siRNA, CA could no longer reverse the inhibition of OPA1 and the induction of cytochrome c and apoptotic-related proteins by 6-OHDA (Fig. 6A). As shown by MTT assay, 6-OHDA treatment decreased cell viability, whereas cell viability was reversed in CA-pretreated cells. Moreover, OPA1 siRNA blocked the ability of CA to reverse 6-OHDA-reduced cell viability (Fig. 6B). These results revealed that OPA1 plays an important role in CA's anti-apoptotic effect.

3.5. Parkin siRNA attenuates ubiquitination of IKKγ and protein expressions of nuclear p65 and OPA1 Our previous study indicated that parkin protein is increased in SHSY5Y cells treated with CA (Lin and Tsai, 2016). We thus used parkin knockdown to further explore whether parkin plays a role in the upregulation of OPA1 by CA through activation of the IKKγ/p65 pathway. In the nontargeting control siRNA (si-control) group, the expression of parkin, nuclear p65, and OPA1 protein in CA-treated cells was induced by 144.1%, 215.6%, and 157.4%, respectively. In the parkin siRNA group, the protein expressions of parkin, nuclear p65, and OPA1 were attenuated (Fig. 5A–C). Moreover, by immunoprecipitation assay, CA treatment increased the ubiquitination of IKKγ by about 62.0% compared with the si-control group. Treatment with parkin siRNA caused a 31.3% decrease in IKKγ ubiquitination (Fig. 5D). These results suggest 5

Food and Chemical Toxicology xxx (xxxx) xxxx

C.-Y. Lin, et al.

Fig. 5. Effect of parkin siRNA on CA-induced OPA1 protein via parkin/IKKγ/p65 pathway. Nontargeting control siRNA (si-control) and human parkin siRNA were transfected into SHSY5Y cells. After 24 h, cells were cultured with 0.1% DMSO (control, C) or 1 μM CA for 1 (D), 3 (A and B), and 24 (C) h. (A) Parkin, (B) nuclear p65, and (C) OPA1 protein expression were measured by Western blotting. (D) Lysates were subjected to immunoprecipitation assayIP with the primary antibody IKKγ. The ubiquitin protein expression of IKKγ was measured by immunoblotting. β-Tubulin, IKKγ, and PARP served as internal controls, respectively. The protein expression of control cells was regarded as 1. One representative immunoblot out of three individual experiments is shown. Values are means ± SD of three individual experiments. Results were analyzed by use of Student's t-test. *p < 0.05 compared with control cells of si-control. #p < 0.05 compared with si-control plus CA.

4. Discussion

OHDA-treated SN4741 cells (Xi et al., 2018). There is also evidence that CA is a promising agent for protecting the mitochondria of neurons (de Oliveira, 2018). We recently showed that protection of mitochondria by CA is related to a reduction of MMP collapse and the elimination of impaired mitochondria (Lin and Tsai, 2019). In the present study, we showed that CA can reverse the changes in mitochondrial dynamic proteins induced by 6-OHDA. We also found that CA enhances the expression of the fusion protein OPA1 to prevent released cytochrome c, apoptotic-related protein activation, and reduced cell viability upon 6OHDA treatment. These results indicate that upregulation of OPA1 by CA is mediated by the parkin/p65 pathway. The augmentation of mitochondrial fusion proteins is a critical role in the maintenance of mitochondrial function (Xi et al., 2018). Both MFN and OPA1 can regulate mitochondrial fusion. In general, MFN1 and MFN2 are responsible for outer mitochondrial membrane fusion; however, OPA1 is linked to the fusion of the inner mitochondrial membrane (Chen and Chan, 2009). Importantly, OPA1 also takes part

Mitochondrial dynamics is controlled by fission and fusion processes and maintains the morphology, quality, and function of the mitochondria. Dysregulation of this homeostasis is associated with the loss of dopaminergic neurons, which in turn causes PD (Lim et al., 2012). 6OHDA, a PD-related neurotoxin, promotes mitochondrial fission by activating DRP1 protein, leading to induction of mitochondrial fragmentation in SH-SY5Y cells (Gomez-Lazaro et al., 2008). Notably, bioactive compounds in the diet regulate mitochondrial dynamic proteins to reduce abnormal mitochondria after 6-OHDA exposure (Liu et al., 2015; Xi et al., 2018). For example, treatment of PC12 cells with allicin from garlic ameliorates the reduction in MMP and mitochondrial DNA content induced by 6-OHDA by inhibiting DRP1 and Fis 1 proteins and inducing OPA1 protein (Liu et al., 2015). Another study showed that MitoQ increases the mRNA and protein expression of MFN2 to promote mitochondrial fusion and inhibit MMP depolarization in 66

Food and Chemical Toxicology xxx (xxxx) xxxx

C.-Y. Lin, et al.

Fig. 6. Effect of OPA1 siRNA on inhibition of apoptosis by CA. After OPA1 siRNA transfection for 24 h, SH-SY5Y cells were pretreated with 0.1% DMSO (control, C) or 1 μM CA for 24 h, and were then exposed to 100 μM 6-OHDA for an additional 12 (A) and 18 (B) h. (A) The proteins of OPA1, cytosol cytochrome c, cleaved caspase 3, caspase 3, cleaved PARP, and PARP were measured by Western blotting. β-Tubulin served as an internal control. The protein expressions of OPA1 and cytosol cytochrome c in control cells were regarded as 1. The ratios of cleaved caspase 3/caspase 3 and cleaved PARP/ PARP were expressed as 1. One representative immunoblot out of three individual experiments is shown. (B) MTT assay was used to analyze cell viability. The viability of control cells was regarded as 100%. Values are means ± SD of three individual experiments. Groups not sharing a common letter differ significantly, p < 0.05.

(Muller-Rischart et al., 2013). That finding suggests that parkin and NEMO play critical roles in OPA1 induction (Muller-Rischart et al., 2013). In our previous study, we showed that CA activates the phosphorylation of IKKα/β and IκBα, as well as the translocation of p65 to the nucleus (Lin et al., 2014). In the present study, we showed that CA promotes the interaction of ubiquitinated protein with IKKγ, the binding activity of OPA1-p65 DNA, and the induction of OPA1 protein (Figs. 3 and 4). After parkin siRNA transfection, the decreased parkin level could not activate the IKKγ/p65/OPA1 pathway in CA-treated cells (Fig. 5). This finding is also supported by the study of MüllerRischart et al. which showed that treatment of cells with tumor necrosis factor-α increases NEMO ubiquitination, NF-κB activation, and OPA1 upregulation. In the absence of parkin, the stimulation of tumor necrosis factor-α is abolished (Muller-Rischart et al., 2013). Thereby, the increase in the OPA1 level with CA treatment is dependent on parkin/ IKKγ/NF-κB signaling. OPA1 has an important role not only in the maintenance of mitochondrial dynamics, but also in the survival of neurons (Jahani-Asl et al., 2011; Olichon et al., 2003). Several lines of evidence have shown that loss of OPA1 disturbs the morphology and function of mitochondria and triggers the release of cytochrome c and apoptosis of cells (Bertholet et al., 2013; Olichon et al., 2003). Knockdown of OPA1 in rodent cortical primary neurons leads to an increase in the fragmentation of mitochondria (Bertholet et al., 2013). In a cellular study, down-regulation of OPA1 protein by use of OPA1 siRNA induced a reduction in MMP, cytochrome c release, and an induction of cleaved PARP protein, leading to cell death (Olichon et al., 2003). On the contrary, OPA1 overexpression inhibits the generation of fragmented mitochondria, cessation of mitochondrial fusion, and the death of

in inhibiting cytochrome c (Frezza et al., 2006). In our previous study by Lin et al., we showed that CA can inhibit cytosolic cytochrome c protein after exposure to 6-OHDA (Lin and Tsai, 2019). As shown by immunofluorescence assay in the present study, CA decreased the release of cytochrome c to the cytoplasm (Fig. 2). We assumed that the inhibition of released cytochrome c in the cytoplasm by CA was closely associated with the action of OPA1. It has been shown that OPA1 can reduce mitochondrial cytochrome c release by regulating mitochondrial cristae remodeling (Frezza et al., 2006). In healthy mitochondria, the oligomerization of OPA1 structure maintains cristae junctions and then retains the existence of cytochrome c within the cristae (Frezza et al., 2006). In response to damaged mitochondria, the desoligomerization of OPA1 increases the widening of cristae junctions and leads to cytochrome c redistribution from cristae to the mitochondrial intermembrane space. Then, cytochrome c permeabilizes the outer mitochondrial membrane and is released to the cytoplasm, resulting in initiating the downstream apoptosis cascade (Frezza et al., 2006). In primary neurons, the excitotoxicity by N-methyl-D-aspartic acid results in loose cristae. However, OPA1 overexpression attenuates this effect by N-methyl-D-aspartic acid (Jahani-Asl et al., 2011). Silencing of OPA1 promotes drastic disorganization of the cristae, resulting in cytochrome c release (Olichon et al., 2003). Our investigation revealed that transfection of cells with OPA1 siRNA inhibited the reversal by CA of 6-OHDA-induced enhancement of cytosolic cytochrome c protein (Fig. 6A). This finding supports an important role of OPA1 in the reduction of cytosolic cytochrome c by CA. Parkin can increase OPA1 expression by mediating the NF-κB pathway (Lim et al., 2012). In parkin-deficient mice and NEMO-deficient mouse embryonic fibroblasts, OPA1 expression is down-regulated 7

Food and Chemical Toxicology xxx (xxxx) xxxx

C.-Y. Lin, et al.

neurons induced by the excitotoxin N-methyl-D-aspartic acid in primary neurons (Jahani-Asl et al., 2011). In SH-SY5Y cells overexpressing OPA1, the induction of mitochondrial structure abnormalities, mitochondrial complex I inhibition, and cytochrome c release caused by MPTP is inhibited (Ramonet et al., 2013). Moreover, overexpression of OPA1 in mice prevents MPTP-induced dopaminergic neurodegeneration (Ramonet et al., 2013). In the present study, transfection of OPA1 siRNA inhibited the ability of CA to reverse the 6-OHDA-induced effects on cytosolic cytochrome c protein, caspase 3, PARP cleavage, and cell survival (Fig. 6). It is likely that CA-inhibition cytochrome c release is related to the modulation of cristae remodeling by OPA1. Moreover, the induction of OPA1 by CA plays an important role in protecting cells from 6-OHDA-activated mitochondrial-mediated apoptotic pathway.

induced apoptosis. Mol. Neurobiol. 55, 1786–1794. Gomez-Lazaro, M., Bonekamp, N.A., Galindo, M.F., Jordan, J., Schrader, M., 2008. 6Hydroxydopamine (6-OHDA) induces Drp1-dependent mitochondrial fragmentation in SH-SY5Y cells. Free Radic. Biol. Med. 44, 1960–1969. Gray, J.J., Zommer, A.E., Bouchard, R.J., Duval, N., Blackstone, C., Linseman, D.A., 2013. N-terminal cleavage of the mitochondrial fusion GTPase OPA1 occurs via a caspaseindependent mechanism in cerebellar granule neurons exposed to oxidative or nitrosative stress. Brain Res. 1494, 28–43. Griparic, L., van der Wel, N.N., Orozco, I.J., Peters, P.J., van der Bliek, A.M., 2004. Loss of the intermembrane space protein Mgm1/OPA1 induces swelling and localized constrictions along the lengths of mitochondria. J. Biol. Chem. 279, 18792–18798. Hall, A.R., Burke, N., Dongworth, R.K., Hausenloy, D.J., 2014. Mitochondrial fusion and fission proteins: novel therapeutic targets for combating cardiovascular disease. Br. J. Pharmacol. 171, 1890–1906. Henn, I.H., Bouman, L., Schlehe, J.S., Schlierf, A., Schramm, J.E., Wegener, E., Nakaso, K., Culmsee, C., Berninger, B., Krappmann, D., Tatzelt, J., Winklhofer, K.F., 2007. Parkin mediates neuroprotection through activation of IkappaB kinase/nuclear factorkappaB signaling. J. Neurosci. 27, 1868–1878. Iannielli, A., Bido, S., Folladori, L., Segnali, A., Cancellieri, C., Maresca, A., Massimino, L., Rubio, A., Morabito, G., Caporali, L., Tagliavini, F., Musumeci, O., Gregato, G., Bezard, E., Carelli, V., Tiranti, V., Broccoli, V., 2018. Pharmacological inhibition of necroptosis protects from dopaminergic neuronal cell death in Parkinson's disease models. Cell Rep. 22, 2066–2079. Jahani-Asl, A., Pilon-Larose, K., Xu, W., MacLaurin, J.G., Park, D.S., McBride, H.M., Slack, R.S., 2011. The mitochondrial inner membrane GTPase, optic atrophy 1 (Opa1), restores mitochondrial morphology and promotes neuronal survival following excitotoxicity. J. Biol. Chem. 286, 4772–4782. Jiang, H., Ren, Y., Zhao, J., Feng, J., 2004. Parkin protects human dopaminergic neuroblastoma cells against dopamine-induced apoptosis. Hum. Mol. Genet. 13, 1745–1754. Lim, K.L., Ng, X.H., Grace, L.G., Yao, T.P., 2012. Mitochondrial dynamics and Parkinson's disease: focus on parkin. Antioxidants Redox Signal. 16, 935–949. Lin, C.Y., Chen, J.H., Fu, R.H., Tsai, C.W., 2014. Induction of Pi form of glutathione Stransferase by carnosic acid is mediated through PI3K/Akt/NF-kappaB pathway and protects against neurotoxicity. Chem. Res. Toxicol. 27, 1958–1966. Lin, C.Y., Tsai, C.W., 2016. Carnosic acid protects SH-SY5Y cells against 6-hydroxydopamine-induced cell death through upregulation of parkin pathway. Neuropharmacology 110, 109–117. Lin, C.Y., Tsai, C.W., 2017. Carnosic acid attenuates 6-hydroxydopamine-induced neurotoxicity in SH-SY5Y cells by inducing autophagy through an enhanced interaction of parkin and Beclin1. Mol. Neurobiol. 54, 2813–2822. Lin, C.Y., Tsai, C.W., 2019. PINK1/parkin-mediated mitophagy pathway is related to neuroprotection by carnosic acid in SH-SY5Y cells. Food Chem. Toxicol. 125, 430–437. Liu, H., Mao, P., Wang, J., Wang, T., Xie, C.H., 2015. Allicin protects PC12 cells against 6OHDA-induced oxidative stress and mitochondrial dysfunction via regulating mitochondrial dynamics. Cell. Physiol. Biochem. 36, 966–979. McIlwain, D.R., Berger, T., Mak, T.W., 2013. Caspase functions in cell death and disease. Cold Spring Harb. Perspect. Biol. 5, a008656. Miller, D.M., Singh, I.N., Wang, J.A., Hall, E.D., 2013. Administration of the Nrf2-ARE activators sulforaphane and carnosic acid attenuates 4-hydroxy-2-nonenal-induced mitochondrial dysfunction ex vivo. Free Radic. Biol. Med. 57, 1–9. Muller-Rischart, A.K., Pilsl, A., Beaudette, P., Patra, M., Hadian, K., Funke, M., Peis, R., Deinlein, A., Schweimer, C., Kuhn, P.H., Lichtenthaler, S.F., Motori, E., Hrelia, S., Wurst, W., Trumbach, D., Langer, T., Krappmann, D., Dittmar, G., Tatzelt, J., Winklhofer, K.F., 2013. The E3 ligase parkin maintains mitochondrial integrity by increasing linear ubiquitination of NEMO. Mol. Cell 49, 908–921. Olichon, A., Baricault, L., Gas, N., Guillou, E., Valette, A., Belenguer, P., Lenaers, G., 2003. Loss of OPA1 perturbates the mitochondrial inner membrane structure and integrity, leading to cytochrome c release and apoptosis. J. Biol. Chem. 278, 7743–7746. Ramonet, D., Perier, C., Recasens, A., Dehay, B., Bove, J., Costa, V., Scorrano, L., Vila, M., 2013. Optic atrophy 1 mediates mitochondria remodeling and dopaminergic neurodegeneration linked to complex I deficiency. Cell Death Differ. 20, 77–85. Santos, D., Cardoso, S.M., 2012. Mitochondrial dynamics and neuronal fate in Parkinson's disease. Mitochondrion 12, 428–437. Shin, H.B., Choi, M.S., Ryu, B., Lee, N.R., Kim, H.I., Choi, H.E., Chang, J., Lee, K.T., Jang, D.S., Inn, K.S., 2013. Antiviral activity of carnosic acid against respiratory syncytial virus. Virol. J. 10, 303. Winklhofer, K.F., 2014. Parkin and mitochondrial quality control: toward assembling the puzzle. Trends Cell Biol. 24, 332–341. Xi, Y., Feng, D., Tao, K., Wang, R., Shi, Y., Qin, H., Murphy, M.P., Yang, Q., Zhao, G., 2018. MitoQ protects dopaminergic neurons in a 6-OHDA induced PD model by enhancing Mfn2-dependent mitochondrial fusion via activation of PGC-1alpha. Biochim. Biophys. Acta (BBA) - Mol. Basis Dis. 1864, 2859–2870. Yoshida, H., Meng, P., Matsumiya, T., Tanji, K., Hayakari, R., Xing, F., Wang, L., Tsuruga, K., Tanaka, H., Mimura, J., Kosaka, K., Itoh, K., Takahashi, I., Imaizumi, T., 2014. Carnosic acid suppresses the production of amyloid-beta 1-42 and 1-43 by inducing an alpha-secretase TACE/ADAM17 in U373MG human astrocytoma cells. Neurosci. Res. 79, 83–93. Youle, R.J., van der Bliek, A.M., 2012. Mitochondrial fission, fusion, and stress. Science 337, 1062–1065.

5. Conclusion CA rescues the mitochondrial impairment in SH-SY5Y cells caused by the neurotoxicity of 6-OHDA and performs this action by inhibiting fission proteins and inducing fusion proteins. In the presence of CA, the protein expression of OPA1 is upregulated through activation of the IKKγ/NF-κB pathway by parkin. Therefore, the protective mechanism of CA is associated with modulation of mitochondrial dynamics and upregulation of OPA1 via the parkin/p65 pathway. CA may be a potential therapeutic candidate for the prevention of PD. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgment This study was supported by the Ministry of Science and Technology (NSC-101-2320-B-039-052-MY2 and MOST 106-2320-B-039-055-MY3). Transparency document Transparency document related to this article can be found online at https://doi.org/10.1016/j.fct.2019.110942. References Barni, M.V., Carlini, M.J., Cafferata, E.G., Puricelli, L., Moreno, S., 2012. Carnosic acid inhibits the proliferation and migration capacity of human colorectal cancer cells. Oncol. Rep. 27, 1041–1048. Bertholet, A.M., Millet, A.M., Guillermin, O., Daloyau, M., Davezac, N., Miquel, M.C., Belenguer, P., 2013. OPA1 loss of function affects in vitro neuronal maturation. Brain 136, 1518–1533. Chen, H., Chan, D.C., 2009. Mitochondrial dynamics–fusion, fission, movement, and mitophagy–in neurodegenerative diseases. Hum. Mol. Genet. 18, R169–R176. de Oliveira, M.R., 2018. Carnosic acid as a promising agent in protecting mitochondria of brain cells. Mol. Neurobiol. 55, 6687–6699. de Oliveira, M.R., da Costa Ferreira, G., Peres, A., Bosco, S.M.D., 2018. Carnosic acid suppresses the H2O2-induced mitochondria-related bioenergetics disturbances and redox impairment in SH-SY5Y cells: role for Nrf2. Mol. Neurobiol. 55, 968–979. de Oliveira, M.R., Ferreira, G.C., Schuck, P.F., 2016. Protective effect of carnosic acid against paraquat-induced redox impairment and mitochondrial dysfunction in SHSY5Y cells: role for PI3K/Akt/Nrf2 pathway. Toxicol. In Vitro 32, 41–54. Elachouri, G., Vidoni, S., Zanna, C., Pattyn, A., Boukhaddaoui, H., Gaget, K., Yu-Wai-Man, P., Gasparre, G., Sarzi, E., Delettre, C., Olichon, A., Loiseau, D., Reynier, P., Chinnery, P.F., Rotig, A., Carelli, V., Hamel, C.P., Rugolo, M., Lenaers, G., 2011. OPA1 links human mitochondrial genome maintenance to mtDNA replication and distribution. Genome Res. 21, 12–20. Frezza, C., Cipolat, S., Martins de Brito, O., Micaroni, M., Beznoussenko, G.V., Rudka, T., Bartoli, D., Polishuck, R.S., Danial, N.N., De Strooper, B., Scorrano, L., 2006. OPA1 controls apoptotic cristae remodeling independently from mitochondrial fusion. Cell 126, 177–189. Fu, R.H., Huang, L.C., Lin, C.Y., Tsai, C.W., 2018. Modulation of ARTS and XIAP by parkin is associated with carnosic acid protects SH-SY5Y cells against 6-hydroxydopamine-

8