Background
It has been well-recognized that high doses of methamphetamine (MA) result in impaired nigrostriatal dopaminergic systems in both rodents [
1‐
4] and primates [
5]. Although the pathogenesis on the MA-induced dopaminergic neurotoxicity remains to be further elucidated, this neurotoxicity may be, at least in part, related to oxidative stress [
3,
4,
6‐
10], inflammatory changes [
4,
6,
11,
12], and pro-apoptosis [
4,
9,
10,
13‐
16]. Thus, dopaminergic neurotoxicity induced by high doses of MA may be a possible Parkinson’s disease (PD) model [
17‐
21].
Furthermore, previous investigations have suggested that humans who abuse MA have an increased risk for PD later in life [
22‐
25]. Earlier postmortem studies reported reductions in dopamine levels, tyrosine hydroxylase (TH) expression, and dopamine transporter (DAT) binding in the striatum of MA abusers [
26], and these changes paralleled neurochemical changes in Parkinson’s disease (PD) patients [
27,
28].
Accumulated evidence indicates that MA can also cause oxidative stress by shifting the balance between reactive oxygen species (ROS) production and the capacity of antioxidant systems to scavenge ROS [
3,
4,
29‐
31]. Recently, we have proposed that MA-induced mitochondrial oxidative stress and mitochondrial dysfunction promotes dopaminergic degeneration [
4,
8]. Interestingly, NADPH oxidase (PHOX) activation was observed in response to mitochondrial ROS formation in human leukocytes [
32].
PHOX is a multiunit enzyme that catalyzes the reduction of molecular oxygen to form superoxide radicals and is composed of gp91phox, p22phox, p47phox, p67phox, p40phox, and small GTPase Rac (Rac1 or Rac2) subunits. Under basal conditions, p47phox, p67phox, and p40phox are present in the cytosol as a complex [
33], and Rac is bound to its inhibitory protein, RhoGDP-dissociation inhibitor (RhoGDI) [
34]. These subunits are separated from the transmembrane gp91phox and p22phox subunits [
33,
34]. Upon activation, the p47phox subunit gets phosphorylated and translocates to the membrane as a complex to assemble with gp91phox, p22phox, and membrane-translocated Rac to form an active PHOX capable of reducing oxygen to a superoxide radical to generate microglial [
35‐
38] and/or mitochondrial-derived ROS [
32] and possibly neuronal and astroglial ROS [
35,
39].
Microglia-mediated neuroinflammation has been linked to multiple neurodegenerative diseases, including PD [
35,
37‐
45]. One recent therapeutic strategy has been to deviate from conventional anti-inflammatory targets and inhibit upstream mediators, such as PHOX [
35]. Once activated, PHOX produces extracellular and intracellular reactive oxygen species, which are critical in initiating and maintaining neuroinflammatory responses, leading to progressive dopaminergic neurodegeneration [
42,
43,
46]. For example, activated microglia secrete a variety of toxic factors, such as tumor necrosis factor α, interleukin-1, and other pro-inflammatory cytokines, which work in concert to cause neuronal damage [
41]. Hong and colleagues have recognized PHOX as a key mediator in bridging neuroinflammation and progressive dopaminergic neurodegeneration [
42,
43,
47].
Importantly, a recent investigation demonstrated that treatment with apocynin, a non-specific inhibitor of PHOX [
48], results in a significant reduction in MA-induced dopamine-release from rat striatal slices [
49]. Furthermore, Park et al. [
50] found that MA (10 μM) induces an increase in phosphorylation of the p47phox subunit and subsequently enhanced PHOX activity in endothelial cells. However, the information of PHOX in the MA-induced neurotoxicity in vivo remains unknown. Thus, we investigated whether apocynin affects dopaminergic neurotoxicity induced by MA in mice, and whether apocynin modulates p47phox in our system, because p47phox acts as a connector between the components of the membrane and the cytoplasm [
36,
51,
52]. We suggested here for the first time that inhibition of the extracellular signal-regulated kinase (ERK)-dependent phosphorylation and membrane translocation of p47phox are critical for apocynin-mediated protective potentials against oxidative stress (mitochondria > cytosol), neuroinflammatory change, and pro-apoptotic pathway induced by MA and that these morbid events require pro-apoptotic scenarios induced by a toxic dose of MA.
Discussion
Previous studies have shown that activation of PHOX activity requires p47phox phosphorylation, a protein that plays an important role in the translocation of cytosolic components to cytochrome b558, as well as in the assembly and activation of PHOX [
36,
51,
52,
73]. Phosphorylation of p47phox constitutes one of the key intracellular events associated with PHOX activation, and Ser345 phosphorylation of p47phox by the MAPK protein ERK plays a critical role in the potentiation of PHOX activation by pro-inflammatory agents [
36,
51]. Therefore, we first examined the levels of MA-induced p47phox phosphorylation using the anti-phospho-Ser345-p47phox antibody.
Two prominent features of this protective role of apocynin or p47phox depletion were observed in this study: (1) apocynin-mediated inhibition of p47 translocation is mediated through the inhibition of PHOX subunit p47phox phosphorylation at Ser345 mainly via suppression of the ERK-signaling pathway, and (2) apocynin attenuates MA-mediated oxidative stress (mitochondrial > cytosolic fraction), mitochondrial dysfunction, microglial activation (towards M1 phenotype), pro-apoptosis, and dopaminergic loss mainly through the inhibition of ERK-dependent p47phox activation.
Participation of ERK1/2 in the activation of PHOX was also proven by a previous study using microglial cells/rat primary mesencephalic neuron-glia cultures stimulated with lipopolysaccharides (LPS) [
36]. The fact that apocynin significantly inhibits the formation of ROS (oxidative damage) 2 h after MA stimulation led us to examine this factor in greater detail by using p47phox-deficient mice. The findings that apocynin could significantly lessen the MA-induced dopaminergic loss in WT, but has no significant effect in response to p47phox knockout mice, suggest that the protective effect of apocynin is most likely mediated through the inhibition of p47phox activity.
Translocation of p47phox, p67phox, p40phox, and rac2 to the plasma membrane are required for the activation of PHOX [
73]. The phosphorylation of Ser345 of p47phox by pro-inflammatory agents enhances this translocation event [
36,
51,
52]. While investigating the mechanism by which apocynin inhibits PHOX activity, we found that apocynin significantly inhibits this MA-induced p47phox phosphorylation at Ser345, resulting in the inhibition of p47phox translocation. As Ser345 is located in the MAPK consensus sequence [
51], we examined whether apocynin inhibits components of the MAPK-signaling pathway, and our results indicate that apocynin shows a significant inhibitory effect on MA-induced ERK phosphorylation. However, MA-induced induction in p38- or JNK-phosphorylation was much less pronounced than in ERK phosphorylation. Furthermore, a specific ERK inhibitor U0126 exhibited strong inhibitory effects against MA-induced p47phox phosphorylation, p47phox translocation, oxidative damage, pro-apoptosis, and neurodegeneration, suggesting a central role of ERK in these effects. These findings, coupled with previous findings on the role of ERK in PHOX activation [
36], strongly indicate that it is ERK that regulates p47phox phosphorylation and constitutes the crucial target for apocynin-mediated inhibition of PHOX activation.
A previous study demonstrated that disturbed Ca
2+ homeostasis may mediate dopaminergic degeneration, such as PD [
74] and MA intoxication [
4,
8]. We examined here whether genetically inhibiting p47phox would affect MMP and intramitochondrial Ca
2+ accumulation in the striatum of mice. Accumulating evidence suggests that mitochondrial damage links inflammation to neuronal death [
4,
8,
75]. Moreover, it is recognized that the role of glial cells in MA-induced neurotoxicity is essential to identify factors contributing to, or mitigating, MA-induced damage to DA nerve terminals [
4,
21,
65‐
67,
76‐
78]. Importantly, it has been proposed that microglia participate in neurotoxicity associated with MA intoxication [
4,
21,
65‐
67,
78].
We were interested in whether or not apocynin would affect this apoptotic signaling pathway after MA exposure. An earlier report demonstrated that MA induces apoptotic cell death in striatal neurons [
13]. In the present study, we chose TUNEL staining (which labels the occurrence of DNA fragmentation, which occurs late in apoptosis). We previously failed to observe MA-induced TUNEL-positive cells in the striatum of C57BL/6 mice 12 h, 1 day, or 3 days after the final MA administration (i.e., four injections of 7 mg/kg MA, intraperitoneally at 2 h intervals or a single injection of MA 20–40 mg/kg), suggesting that the C57BL/6 background is not sensitive to TUNEL staining [
9,
77]. Thus, according to previous reports [
9,
15,
16], we used 10-week-old male Taconic ICR mice. Because apoptotic cell death was detectable at 20 mg/kg MA and reached a significant level at 35 mg/kg in our previous study, a 35-mg/kg dose of MA was chosen for the present study. We also sacrificed animals 1 day after MA administration [
4,
8,
9,
15,
16], and TUNEL-positive cells were induced maximally at this time point.
The relationship between mitochondrial damage, oxidative stress, and neuronal dysfunction has been recognized by the effects of excessive production of ROS within mitochondria, which leads to a reduction of mitochondrial antioxidant activity, in turn causing impairment of mitochondrial function [
4,
8,
79,
80]. Our results clearly indicate that MA-induced toxic damage is more pronounced in the mitochondrial fraction than in the cytosolic fraction in WT mice, and that apocynin or genetically inhibiting p47phox significantly attenuates this oxidative damage, mitochondrial dysfunction, pro-apoptotic changes, and dopaminergic impairment. We demonstrated that PKCδ is an oxidative stress-sensitive kinase, and its activation via caspase-3-dependent proteolysis induces apoptotic cell death in MA-induced dopaminergic toxicity [
4,
63]. Therefore, the protective effect of apocynin against MA-induced PKCδ activation and dopaminergic deficits might reflect an anti-peroxidative (mitochondrial > cytosolic) potential by targeting p47phox gene. Indeed, apocynin does not significantly alter neuroprotective activity mediated by p47phox gene knockout, suggesting that p47phox is a critical target for the neuroprotective activity of apocynin. To the best of our knowledge, the current study is the first to investigate the role of p47phox per se in apocynin-mediated neuroprotective potential with recovery of mitochondrial function.
We propose here that MA potentiates mitochondrial oxidative stress and also impairs the mitochondrial detoxification system, and MMP, possibly due to Ca
2+ accumulation. Increased intracellular Ca
2+ promotes the accumulation of Ca
2+ within the mitochondrial matrix when total Ca
2+ uptake exceeds total Ca
2+ efflux from mitochondria [
81]. Mitochondrial Ca
2+ overload may also lead to the uncoupling of mitochondrial electron transport and may potentiate oxidative stress. Decreases in MMP and increases in oxidative damage after MA treatment could be mediated by Ca
2+ entry.
Based on the importance of MnSOD in our experimental condition, we sought to determine whether or not this mitochondrial enzyme would provide neuroprotection against MA neurotoxicity. It has been acknowledged that MnSOD overexpression attenuates dopaminergic toxicity induced by MA [
82] or 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [
83] and protects cells from apoptosis [
84]. Maragos et al. [
82] demonstrated that the formation of protein carbonyls is less pronounced in MnSOD transgenic overexpressing (Tg) mice than that in non-Tg mice against MA toxicity. Furthermore, a previous report indicated that increased MnSOD expression without a change in Cu/Zn-SOD, catalase, or glutathione peroxidase activities [
85] conferred neuroprotection against dopamine loss in a model of neuronal damage, indicating a possible role for detoxification of MA-induced ROS by scavenging of superoxide radicals in the mitochondria. Interestingly, MnSOD overexpression failed to protect MA-induced reductions in 5-hydroxytryptamine (5-HT) and 5-hydroxyindoleacetic acid (5-HIAA) [
82], suggesting that the mitochondrial mechanism may not be involved in serotonergic toxicity. Our results indicate that higher levels of MnSOD might be associated with enhanced mitochondrial maintenance and could contribute to reducing apoptosis that has been induced by mitochondrial damage in the condition with dopaminergic impairments.
Microglial activation and oxidative stress induced by mitochondrial toxins (i.e., 3-nitropropionic acid) caused neuronal loss in the striatum [
64]. Mitochondria can be a target of free radical stress initiated by activated microglia. The combination of mitochondrial dysfunction, oxidative stress, and exacerbated activation of microglia generates a cycle that appears to lead to progressive dopaminergic neuronal cell death [
86]. We raise the possibility that apocynin might, at least in part, block Ca
2+ entry through the mitochondrial translocation of p47phox, given that apocynin primarily attenuated mitochondrial dysfunction and mitochondrial ROS.
In our study, a toxic dose (35 mg/kg, i.p. × 1) of MA induced the transformation of ramified/resting microglia into reactive hypertrophic microglia, as evidenced by increases in the number of branches and cell body size. Based on morphological characteristics, microglia can be classified into at least four stages of activation: (1) ramified/resting, (2) hypertrophic, (3) bushy, and (4) amoeboid microglia [
57,
87]. According to this classification, hypertrophic microglia have large cell bodies and long, thick, highly branched processes, whereas bushy or amoeboid microglia have fewer thick and rarely branched processes, even though cell bodies become larger than hypertrophic microglia. Raineri et al. [
88] showed that a multiple dose regimen of MA (i.e., 5 mg/kg × 4) induces amoeboid, as well as hyper-ramified microglia, in mouse striatum; however, amoeboid microglia were rarely observed in a toxic dose regimen of MA in our study. Thus, this issue requires further exploration.
Current results are in line with our previous reports that MA treatment significantly increased the mRNA expression of M1 phenotypic markers (CD16, CD32, and CD86), suggesting that microglia after MA treatment existed primarily in the classically activated state [
4,
21], which is pro-inflammatory. Thus, our results indicate that neuroprotection by apocynin or p47phox knockout is mediated by its anti-inflammatory properties.
We have reported that the oligomeric form of α-synuclein was obviously increased after MA [
12]. Interestingly, earlier studies have shown that aggregated α-synuclein released into the extracellular space from dying or dead DA neurons can directly induce microglia towards M1 phenotype with the activation of NADPH oxidase, increasing production of ROS and pro-inflammatory cytokines [
89‐
92]. Overexpression of mutant α-synuclein solely in microglia switches microglia into a more reactive M1 phenotype characterized by elevated levels of pro-inflammatory cytokines [
93]. Similarly, typical characteristics of M1 phenotype, including the activation of PHOX as well as the release of various pro-inflammatory mediators, were observed in the MPTP-intoxicated models [
94], indicating that this phenomenon, at least in part, parallels current results. However, although inhibition of PHOX or genetic inhibition of its functional p47phox subunit switches microglial activation from M1 to M2 in response to LPS challenge [
95], either inhibition did not significantly alter the mRNA expression of M2 phenotypic markers induced by MA in this study. Similar to the current study, multiple doses of MA did not significantly decrease M2 phenotype markers in our previous study [
4]. Thus, the interactive modulation between M1- and M2-activated populations remains to be determined [
68‐
70].
Author’s contributions
DKD, EJS, and YN took part in the pilot studies, animal treatment, Western blotting, and statistics. SR provided the information on the p47phox in the earlier period of this study. JHJ and CGJ did the dose-related pilot study, other biochemical study, and histology. TN and JSH provided helpful comments for the discussion and revised this manuscript. HCK arranged this manuscript via full communications with all co-authors. All authors read and approved the final manuscript.