Introduction

Mitochondria, the major organelles involved in ATP synthesis, take a central role in the redox biology of mammalian cells. These double-membrane organelles are both a source and a target of reactive species [1–3]. The respiratory chain (complexes I–IV) generates the electrochemical potential necessary to the production of ATP by the complex V/ATP synthase [4, 5]. Disruption in the electron transport and consequent electron leakage from the respiratory chain leads to the production of the radical anion superoxide (O2 −•) [6–11], which undergoes dismutation after reacting with the mitochondrial form of superoxide dismutase (Mn-SOD) generating hydrogen peroxide (H2O2) [12–15]. H2O2 is a membrane-diffusible molecule, meaning that it may spread the pro-oxidant from one mitochondrion to another, as well as to other cellular compartments [16]. Indeed, H2O2 is utilized by mammalian cells as a signaling agent [17–19]. H2O2, which is not a free radical, is converted in water by catalase (CAT) or glutathione peroxidase (GPx) [20]. There is consumption of reduced glutathione (GSH) when GPx consumes H2O2 [21]. GSH is the major nonenzymatic antioxidant within mammalian cells and may be found in both cytosol and mitochondria [22]. In this regard, investigating strategies that lead to improvement of the mitochondria-located antioxidant defenses is of pharmacological interest due to the role of this intrinsic antioxidant in rescuing mammalian cells in cases of redox impairment [15, 23–29]. Actually, redox disturbances have been seen in several human diseases, such as neurodegeneration, cardiovascular dysfunction, diabetes mellitus, and inborn errors of metabolism, among others [30–34]. Furthermore, redox impairment may lead to disruption of the bioenergetics-related reactions in mammalian mitochondria, compromising cellular function and dynamics due to decreased access to ATP [35–37].

Carnosic acid (CA; C20H28O4; MW 332.43392 g/mol) is a diterpene found in rosemary (Rosmarinus officinalis L.) and exhibits antioxidant, anti-inflammatory, and anti-apoptotic capacities, as has been studied in several experimental models [38–44]. Moreover, CA is an anti-tumor agent [45–47]. We have recently demonstrated that CA protects mitochondria of SH-SY5Y cells challenged with different chemical stressors by a mechanism associated with the activation of the master regulator of the redox biology in mammalian cells, nuclear factor erythroid 2-related factor 2 (Nrf2) [48–50]. Nrf2 upregulation induced by CA caused an increase in the expression of antioxidant and phase II detoxification enzymes in SH-SY5Y cells, such as heme oxygenase-1 (HO-1), glutathione reductase (GR), GPx, and Mn-SOD, decreasing the susceptibility of mitochondria to pro-oxidant stressors. Other research groups also demonstrated that CA exhibited a protective role against different stimuli in neuronal cells in both in vitro and in vivo experimental designs [51–55]. Nonetheless, this is the first work reporting the ability of CA in preventing the H2O2-triggered mitochondria-related redox disruption in SH-SY5Y cells.

In the present work, we investigated whether CA would be effective in preventing mitochondrial impairment in SH-SY5Y cells exposed to the pro-oxidant agent H2O2.

Materials and Methods

Materials

Plastic materials utilized in cell culture were obtained from Corning, Inc. (NY, USA) and Beckton Dickson (NJ, USA). Culture analytical grade reagents were acquired from Sigma (MO, USA). Other chemicals and assay kits were obtained as described here.

Cell Culture and Chemical Treatment

Human dopaminergic neuroblastoma SH-SY5Y cells (obtained from the American Type Culture Collection (Manassas, VA, USA)) were cultured in Dulbecco’s modified Eagle’s medium (DMEM)/F-12 HAM nutrient medium (1:1 mixture; supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine, penicillin (1000 units/mL), streptomycin (1000 μg/mL), and amphotericin B (2.5 μg/mL)) in a 5% CO2 humidified incubator at 37 °C. SH-SY5Y cells were cultured until a confluence of 80–90% was achieved and then trypsinized.

H2O2 was utilized at 300 μM for different periods of incubation according to each specific assay. A pretreatment with CA (dissolved in DMSO) at 1 μM for 12 h was performed in order to test the ability of this diterpene in preventing the deleterious effects triggered by H2O2 in SH-SY5Y cells. More detailed information (specific concentrations and periods of treatment) may be achieved in the figure legends. The data are exhibited as the mean ± S.E.M. of three or five independent experiments each done in triplicate.

Analysis of Cell Viability

The viability of SH-SY5Y cells was analyzed by using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, as previously described [56].

Quantification of Mitochondria-Related Apoptotic Factors and Cell Death-Associated Parameters

The immunocontents of Bcl-2, Bax, cytochrome c (mitochondrial and cytosolic), and cleaved PARP were quantified by using ELISA assay kits according to the instructions of the manufacturer (Abcam, MA, USA). The activities of the apoptotic enzymes caspase-9 and caspase-3 were assessed through the utilization of fluorimetric assay kits according to the instructions of the manufacturer (Abcam, MA, USA), as previously described [48–50, 57]. DNA fragmentation in cell lysates was measured by the utilization of an ELISA kit following the manufacturer’s instructions (Roche, Germany).

Measurement of Intracellular Reactive Oxygen Species Production

We utilized the nonpolar compound 2′-7′-dichlorodihydrofluorescein diacetate (DCFH-DA) to assess the intracellular production of ROS in this experimental model, as previously described [58]. The cells were loaded with 100 μM DCFH-DA dissolved in medium and were incubated for 30 min at 37 °C to allow cellular incorporation. After this incubation, the medium was changed to fresh medium and the oxidation of DCFH was registered during 30 min (1-min intervals) at 37 °C in a fluorescence (Em/Ex = 535/485 nm) plate reader (Molecular Devices, USA).

Quantification of MDA, Protein Carbonyl, and 8-Oxo-dG Levels

The levels of MDA, protein carbonyl, and 8-oxo-dG were determined by utilizing commercial kits (Abcam, MA, USA), as previously described [48–50, 57].

Analysis of 3-Nitrotyrosine Levels

The levels of 3-nitrotyrosine in mitochondrial membranes were examined by using a polyclonal antibody to 3-nitrotyrosine (Calbiochem, Germany) diluted to 1:2000 in phosphate-buffered saline (PBS) pH 7.4 with 5% albumin in an indirect ELISA assay, as previously reported [59–62].

Quantification of Total and Mitochondrial GSH Contents

The SH-SY5Y neuroblastoma cells were washed and collected, and GSH was measured according to the protocol of a commercial kit (Abcam, MA, USA), as described elsewhere [48–50, 57].

Isolation of Mitochondria

In order to obtain mitochondria from cultured cells, SH-SY5Y cells were washed and re-suspended in a buffer (250 mM sucrose, 10 mM KCl, 1 mM EGTA, 1 mM EDTA, 1 mM MgCl2, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 1 mM pepstatin A, 10 mg/mL leupeptin, 2 mg/mL aprotonin, and 20 mM HEPES—pH 7.4), and differential centrifugations were performed, as previously reported [63].

Extraction of Submitochondrial Particles

Submitochondrial particles (SMPs) were obtained from SH-SY5Y cells according to a protocol previously published [64, 65]. Briefly, after the end of the incubations, the cells were homogenized in a buffer (230 mM mannitol, 70 mM sucrose, 10 mM Tris-HCl, and 1 mM EDTA—pH 7.4). The mitochondrial solution was frozen and thawed (three times) originating superoxide dismutase-free SMP. The solution containing SMP was washed (twice) using a buffer (140 mM KCl, 20 mM Tris-HCl—pH 7.4) in order to ensure Mn-superoxide dismutase leakage from the organelle. This protocol was used to quantify the production of superoxide anion radical (O2 −•) and to examine the effects of H2O2 and/or CA on the levels of malondialdehyde (MDA), protein carbonyl, and 3-nitrotyrosine in mitochondrial membranes [59].

Enzyme Activities

The activities of aconitase, α-ketoglutarate dehydrogenase (α-KGDH), succinate dehydrogenase (SDH), complex I, and complex V enzymes were measured through the utilization of commercial kits following the instructions of the manufacturer (Abcam, MA, USA).

Quantification of ATP Levels

ATP levels were measured by using a commercial kit (Abcam, MA, USA). Briefly, the cells were re-suspended and homogenized in a specific ATP assay buffer. After centrifugation, the samples rendered supernatants, which were collected and transferred to another tube. The samples were then deproteinizated and centrifuged, and the supernatants were utilized to determine the ATP levels. The reaction was read in a fluorescence (Ex/Em = 535/590 nm) plate reader (Molecular Devices, USA).

Determination of Mitochondrial Membrane Potential

Mitochondrial membrane potential (MMP) was studied by utilizing a kit that applies tetraethylbenzimidazolylcarbocyanide iodine (JC-1) as a lipophilic cationic dye that penetrates and accumulates in mitochondria according to variations in the membrane potential, as previously described [48–50, 57].

Nrf2 Knockdown by siRNA Transfection

Briefly, Nrf2 knockdown was performed through transient transfection of SH-SY5Y neuroblastoma cells by utilizing Nrf2 siRNA duplex, following the instructions of the manufacturer (Santa Cruz, CA, USA) [66, 67].

Statistical Analyses

Statistical analyses were performed with GraphPad 5.0 software. Data are exhibited as the mean ± standard error of the mean (S.E.M.) of three or five independent experiments each done in triplicate; p values were considered significant when p < 0.05. Differences between the experimental groups were determined by one-way ANOVA followed by the post hoc Tukey’s test.

Results

CA Prevented Loss of Cell Viability and Inhibited Apoptosis in SH-SY5Y Cells Exposed to H2O2

We first examined the effect of various H2O2 concentrations on the viability of SH-SY5Y neuroblastoma cells. As depicted in Fig. 1a, H2O2 at 25–400 μM for 24 h decreased cell viability in a dose-dependent manner. We also investigated the CA concentrations that efficiently suppressed the H2O2-induced loss of cell viability. According to Fig. 1b, CA pretreatment for 12 h at 1–2 μM efficiently prevented loss of viability in H2O2-treated SH-SY5Y cells. We have previously demonstrated that CA at 1 μM exerted cytoprotective effects in SH-SY5Y cells exposed to different chemical stressors. Therefore, we decided to use pretreatment with CA at 1 μM for 12 h in further assays in the herein presented experimental model.

Fig. 1
figure 1

The effect of H2O2 at different concentrations for 24 h on the viability of SH-SY5Y cells. CA pretreatment at different concentrations (0.1–2 μM) for 12 h on the viability of SH-SY5Y cells exposed to 300 μM H2O2 for additional 24 h. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs control cells, #p < 0.05 vs CA-treated cells exposed to H2O2

Full size image

H2O2 at 300 μM for 24 h increased the levels of pro-apoptotic markers, as may be viewed in Fig. 2. Otherwise, CA (1 μM) pretreatment (12 h) suppressed apoptosis in H2O2-treated cells by blocking the H2O2-induced decrease in Bcl-2 (Fig. 2a) and the increase in Bax (Fig. 2b) levels. CA also abrogated the release of cytochrome c to the cytosol (Fig. 2c), consequently maintaining the mitochondrial content of cytochrome c near the values observed in the control group (Fig. 2d). In this context, CA prevented the increase observed in the activities of the pro-apoptotic enzymes caspase-9 (Fig. 2e) and caspase-3 (Fig. 2f) in the cells exposed to H2O2.

Fig. 2
figure 2

The effects of a pretreatment with CA at 1 μM for 12 h on Bcl-2 immunocontent (a), Bax immunocontent (b), cytosolic cytochrome c content (c), mitochondrial cytochrome c content (d), caspase-9 activity (e), and caspase-3 activity (f). H2O2 was utilized at 300 μM for further 24 h. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs control cells; #p < 0.05 vs H2O2-treated cells

Full size image

In order to confirm the anti-apoptotic effect of CA in H2O2-treated SH-SY5Y cells, we examined the levels of hallmarks of apoptosis (Fig. 3). CA pretreatment significantly prevented the H2O2-elicited increase in the cleavage of PARP (Fig. 3a) and in the fragmentation of DNA (Fig. 3b).

Fig. 3
figure 3

The effects of a pretreatment with CA at 1 μM for 12 h on the levels of cleaved PARP (a) and DNA fragmentation (b). H2O2 was utilized at 300 μM for further 24 h. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs control cells; #p < 0.05 vs H2O2-treated cells

Full size image

CA Exerted Antioxidant Effects in H2O2-Treated SH-SY5Y Cells

We next examined whether CA would prevent redox impairment in cells exposed to H2O2 (Fig. 4). The reactions performed by reactive species occur at very high rates [20, 68]; therefore, we decided to investigate the levels of markers of redox disturbances 3 h after administration of H2O2 to the cells. H2O2 increased the cellular production of reactive species (Fig. 4a), causing a decrease in the levels of cellular GSH (Fig. 4b) and oxidizing total DNA, as assessed through the quantification of 8-oxo-2′-deoxyguanosine (8-oxodG) (Fig. 4c). Pretreatment with CA prevented the redox disturbances caused by H2O2 also in mitochondria by reducing the levels of lipid peroxidation (Fig. 5a), protein carbonylation (5b), and protein nitration (Fig. 5c) in mitochondrial membranes.

Fig. 4
figure 4

The effects of a pretreatment with CA at 1 μM for 12 h on the cellular levels of reactive species generation (a), GSH (b), and 8-oxo-dG (c). H2O2 was utilized at 300 μM for further 3 h (DCF assay) or 24 h (GSH and 8-oxo-dG). The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs control cells; a p < 0.05 vs control cells; #p < 0.05 vs H2O2-treated cells

Full size image
Fig. 5
figure 5

The effects of a pretreatment with CA at 1 μM for 12 h on the levels of lipid peroxidation (a), protein carbonylation (b), and protein nitration (c) in mitochondrial membranes. H2O2 was utilized at 300 μM for further 24 h. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs control cells; #p < 0.05 vs H2O2-treated cells

Full size image

CA Affords Mitochondrial Protection in Cells Exposed to H2O2

According to Figs. 6 and 7, H2O2 impaired mitochondrial function. However, pretreatment with CA efficiently abrogated the effects of H2O2 on the activities of aconitase (Fig. 6a), α-KGDH (Fig. 6b), and SDH (Fig. 6c). CA suppressed the H2O2-induced inhibition of complex I (Fig. 7a) and complex V (Fig. 7b) enzymes and prevented the decrease in ATP levels elicited by H2O2 (Fig. 7c). CA also protected mitochondria by inhibiting the H2O2-induced loss of MMP (Fig. 7d). CA attenuated the effect of H2O2 on the production of O2 −• by mitochondria (Fig. 7e). The levels of GSH in mitochondria were also modulated by CA and H2O2, as may be seen in Fig. 7f.

Fig. 6
figure 6

The effects of a pretreatment with CA at 1 μM for 12 h on the activities of aconitase (a), α-ketoglutarate dehydrogenase (b), and succinate dehydrogenase (c) enzymes. H2O2 was utilized at 300 μM for further 24 h. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs control cells; #p < 0.05 vs H2O2-treated cells

Full size image
Fig. 7
figure 7

The effects of a pretreatment with CA at 1 μM for 12 h on the activities of the complexes I (a) and V (b), on the levels of ATP (c), on MMP (d), on the production of O2 −• by SMP (e), and on the levels of GSH in mitochondria (f). H2O2 was utilized at 300 μM for further 3 h (measurement of O2 −•) or 24 h (other analyses). The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs control cells; a p < 0.05 vs control cells; #p < 0.05 vs H2O2-treated cells

Full size image

CA Protected Mitochondrial Function by an Nrf2-Dependent Manner in SH-SY5Y Cells Challenged with H2O2

In order to investigate by which mechanism CA protected mitochondria in SH-SY5Y cells exposed to H2O2, we examined the consequences of silencing the Nrf2 transcription factor in this experimental model. CA failed to restore mitochondrial function in cells in which Nrf2 was knocked down, as may be observed in relation to the activities of complexes I (Fig. 8a) and V (Fig. 8b). CA did not enhance the levels of GSH efficiently in SH-SY5Y cells in which Nrf2 was silenced (Fig. 9a) and failed to suppress loss of MMP in cells exposed to H2O2 (Fig. 9b).

Fig. 8
figure 8

The effects of Nrf2 siRNA (48 h) on the activities of the complexes I (a) and V (b) of SH-SY5Y cells treated or not with CA and/or H2O2. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs the CA and H2O2-treated cells transfected with negative control (NC) siRNA

Full size image
Fig. 9
figure 9

The effects of Nrf2 siRNA (48 h) on the activities of the mitochondrial content of GSH (a) and MMP (b) of SH-SY5Y cells treated or not with CA and/or H2O2. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs the CA and H2O2-treated cells transfected with negative control (NC) siRNA

Full size image

CA Protected SH-SY5Y Cells Through an Nrf2-Dependent Manner

Finally, we analyzed whether Nrf2 would present a role in the CA-induced cytoprotective effects in H2O2-treated SH-SY5Y cells. According to Fig. 10, CA failed to abolish the cellular impairment in Nrf2-silenced SH-SY5Y cells.

Fig. 10
figure 10

The effects of Nrf2 siRNA (48 h) on the activities of the viability of SH-SY5Y cells treated or not with CA and/or H2O2. The results are presented as the mean ± S.E.M. of three or five independent experiments each done in triplicate. One-way ANOVA followed by the post hoc Tukey’s test, *p < 0.05 vs the carnosic acid and H2O2-treated cells transfected with negative control (NC) siRNA

Full size image

Discussion

Redox impairment causes mitochondrial dysfunction by inhibiting mitochondria-located enzymes that maintain bioenergetics-related reactions associated with ATP production [69, 70]. Disruption of the tricarboxylic acid cycle (TCA), for example, affects directly the flux of electrons in the respiratory chain, causing a decrease in the electrochemical gradient across the inner mitochondrial membrane [71–74]. This may be accompanied by electron leakage of the oxidative phosphorylation system, enhancing the production of reactive species, such as O2 −•, among others [10, 11, 75]. Therefore, this is a vicious cycle that gradually amplifies mitochondrial dysfunction, leading to the activation of the intrinsic apoptotic pathway and cell death [76]. Actually, disruption of the mitochondrial function, increased production of reactive species by these organelles, and enhanced rates of cell death have been observed in several human pathologies, including neurodegeneration and neurotoxicity [37, 77–81].

In the present work, we have found that a pretreatment with CA prevented the H2O2-induced mitochondrial dysfunction, loss of cell viability, and apoptosis in SH-SY5Y neuroblastoma cells. CA suppressed the H2O2-mediated inhibition of components of the TCA and of the oxidative phosphorylation system and decreased the production of O2 −• by mitochondria exposed to H2O2. CA also upregulated the levels of GSH in the mitochondria, probably causing mitochondrial protection, as previously demonstrated by our research group [82–84]. However, we have also demonstrated previously that CA increased the levels of Mn-SOD in SH-SY5Y cells [48, 49]. This enzyme is located in the mitochondrial matrix and takes a crucial role in converting O2 −• in H2O2 [20]. Therefore, the participation of enzymatic antioxidant defenses against exposure to H2O2 and the consequences this reactive species causes should not be discarded in this experimental model. It has been demonstrated that oxidants, including H2O2, inhibit mitochondrial enzymes, such as aconitase, α-KGDH, and SDH, in different experimental models [85, 86]. Additionally, impaired function and/or content of the TCA enzymes in the brain of patients with Parkinson’s disease [35, 87–89], Alzheimer’s disease [90, 91], or Huntington’s disease [92] has been published. In this regard, there is increasing interest in investigating natural compounds that may prevent mitochondrial dysfunction in case of neurologic disorders, among other diseases [15, 23, 24, 26–29, 93, 94].

We found that Nrf2 knockdown abrogated the beneficial effects elicited by CA in SH-SY5Y cells exposed to H2O2. Indeed, a close relation between mitochondrial function and Nrf2-dependent signaling has been reported [95]. Even though Nrf2 is a master regulator of the redox environment in mammalian cells, this transcription factor may be associated with other mitochondria-related parameters, such as oxidation of fatty acids [96] and synthesis of ATP [97]. Indeed, Nrf2 silencing abolished the CA-induced upregulation of mitochondrial enzymes of the oxidative phosphorylation system related to the maintenance of the energetic status in SH-SY5Y cells in this work. In spite of this, it remains to be determined exactly how Nrf2 knockdown impacted the effects mediated by CA, since the regulation of Nrf2 is not dependent only on redox reactions [98]. We [48, 49] and others [99] have previously shown that CA activates the PI3K/Akt signaling pathway, leading to an upregulation of Nrf2 and a consequent increase in the expression of antioxidant and phase II detoxification enzymes. The activation of the PI3K/Akt axis is not associated only with redox disturbances and takes a crucial role in mediating metabolic effects, for example, in mammalian cells [100]. Nonetheless, further research is necessary in order to investigate whether there is a role for this signaling pathway in modulating the CA-induced Nrf2-dependent metabolic effects in cultured cells.

Overall, CA prevented the H2O2-triggered mitochondrial redox impairment and dysfunction (i.e., TCA impairment) and cell death in SH-SY5Y cells by a mechanism involving the transcription factor Nrf2. Further research is necessary to investigate whether CA would be able to trigger mitochondrial biogenesis in mammalian cells, among other mitochondria-related effects.