PD is the second most common neurodegenerative disease in the western world and the single most common movement disorder. Over 1 million people in the United States are affected . Symptoms including rigidity, resting tremor, bradykinesia and postural instability are due to degeneration of the brain's nigrostriatal system with progressive loss of DNs in the substantia nigra pars compacta (SNpc), accompanied by depletion of the transmitter dopamine in the striatum. Current pharmacological therapy for PD ameliorates symptoms for a limited period of time, without retarding or reversing disease progression. Currently administered drugs work by increasing the concentration of functional dopamine in the striatum by one of a number of mechanisms: replacing dopamine itself (L-Dopa), inhibiting dopamine-degrading enzymes to prolong its half-life (Entacapone, Selegeline), or mimicking the effect of dopamine on its receptors with dopamine agonists (Bromocriptine, Pergolide, Pramipexole, etc). L-Dopa has remained the single most effective PD drug since its introduction decades ago [2, 3].
New treatment strategy aimed at slowing or halting DN death is desired. In the course of elucidating pathogenic events that eventually lead to PD, at least four major mechanisms have been identified: oxidative stress, protein aggregation, inflammation and excitotoxicity [4, 5]. It is assumed that these pathways constitute a complex network of events that eventually leads to DN death. Consequently, effective disease-modifying therapy would require addressing a combination of neurodegenerative mechanisms within the SN.
Even though the vast majority of PD cases are sporadic idiopathic forms, recent identification of a number of genes (PARK 1-11) responsible for rare familial cases has provided tremendous insight into the pathogenesis of the disease. The rationale behind studying rare genetic forms of a common sporadic disease is the assumption that they share key biochemical pathways. Of the ten genetic loci linked to familial PD, six gene products have been characterized so far: α-Synuclein, Parkin, UCH-L1, DJ-1, PINK1, and LRRK2 [5, 6]. DJ-1 is a relatively small, evolutionarily conserved protein belonging to the ThiJ/PfpI/DJ-1 family. Members of the ThiJ/PfpI/DJ-1 family include chaperones, proteases and transcriptional regulators , yet DJ-1's biochemical function relevant to PD remains to be defined. DJ-1 has been implicated in diverse cellular processes, including cellular transformation and tumorigenesis [8, 9], transcriptional regulation and RNA binding , androgen receptor signaling [11, 12], spermatogenesis , and oxidative stress response [14, 15]. In vitro studies showed that DJ-1 responds to oxidative stress induced by paraquat exposure, with a shift of its iso-electric point towards a more acidic form (from pI 6.2 to pI 5.8) . Postmortem analysis of PD brains detected higher concentrations of the acidic DJ-1 isoforms, as compared to healthy controls . DJ-1 is a hydrogen peroxide (H2O2)-responsive protein. H2O2 exposure oxidizes all its cysteine residues (Cys 46, 53, 106) to cysteine sulfonic acid , with Cys 106 being most sensitive. These studies demonstrate a direct modification of DJ-1 protein by reactive oxygen species (ROS), nourishing the notion that DJ-1 might act as a free radical scavenger or sensor. Cell culture studies of DJ-1 deficient neuronal cells revealed increased susceptibility to H2O2, MPP+, 6-hydroxydopamine, and rotenone [12, 17, 18], whereas DJ-1 overexpression dramatically reduced H2O2-induced neuronal death [17, 18]. Oxidative stress results in mitochondrial relocalization of DJ-1, which is mediated by oxidation of Cys 106 . Consistent with DJ-1 playing an important role in oxidative stress response, genetic studies in Drosophila and mice showed accumulation of ROS and increased sensitivity to oxidative stimuli, including H2O2 and MPTP, in DJ-1 mutants [18, 20, 22].
Minocycline is a member of the tetracycline group of antibiotics. Minocycline has been found to have additional anti-inflammatory and antioxidant properties independent of its antibacterial activity [23, 26], which may be useful for the treatment of neurodegenerative diseases including PD. Minocycline has also been shown to have neuroprotective effects in animal models of other pathological conditions, including ischemia and stroke, traumatic brain injury, amyotrophic lateral sclerosis (ALS), multiple sclerosis (MS), Huntington's disease (HD), and the MPTP model of PD in mouse [27, 30]. Potency appears to arise from the modulation of inflammatory cytokine release, microglia activation, nitric oxide production, matrix metalloprotease activation, and apoptotic cell death. Minocycline has an oral bioavailability of almost 100% and its absorption, unlike other tetracyclines, is not reduced by ingestion with food . High lipophilicity allows its easy diffusion into brain tissues. Minocycline also has a lower urinary excretion than other tetracyclines and is thus safer in elderly patients with impaired renal function .
Celastrol is a triterpene extracted from the root bark of an ivy-like, creeping plant called Triperygium wilfordii (TW) that is indigenous to Southern China. Extracts of the plant have had a long history of use in traditional Chinese medicine for treating fever, chills, edema and joint pain, conditions commonly associated with inflammation . Celastrol was found to suppress microglial cell activation, release of the inflammatory cytokines TNF-α and IL-1β by human macrophages and monocytes, and the production of nitric oxide (NO) by iNOS . Furthermore, celastrol was demonstrated to be a potent inhibitor of induced lipid peroxidation in rat liver mitochondria, exhibiting over 15-fold more antioxidant potency than α-tocopherol . Celastrol protected both the outer and inner mitochondrial membrane from peroxidation, possibly mediated by its radical-scavenging dienonephenol moiety, while the anionic carboxyl group protects the inner membrane from radical attacks by stabilizing its negative surface charge. In a rat model of Alzheimer's disease (AD), celastrol improved memory and learning in psychomotor-activity tests (PMA) . Low nanomolar concentrations of the drug showed efficacy in all the studies mentioned above.
CoQ10 (also known as ubiquinone) is composed of a quinone ring and a 10-isoprene unit tail. It is an obligatory cofactor in the mitochondrial respiratory chain. As a bioenergetic agent, it serves as an electron acceptor in complexes I and II/III of the electron transport chain. Mitochondrial dysfunction has been frequently observed in PD, and several lines of evidence support its causative role in disease pathogenesis. For example, a 30% to 40% reduction in complex I activity was observed in sporadic PD, and MPTP induces parkinsonism by inhibiting complex I activity . CoQ10 is also a potent antioxidant distributed in all membranes throughout the cell. It is able to work in concert with α-tocopherol and participates in the recovery of cells from oxidative stress [36, 37]. Levels of coQ10 measured in mitochondria from PD patients were significantly lower than in age-matched controls , while at the same time the percentage of oxidized coQ10 was relatively increased . In in vitro models, coQ10 could protect against MPP+ and rotenone induced toxicity [40, 42]. In animal models of ALS and HD, coQ10 treatment also showed beneficial effects [43, 44]. In the MPTP mouse model of PD, oral treatment of young mice with coQ10 and nicotiamide attenuated the effect of low dose MPTP administration, while coQ10 alone attenuated dopamine depletion in the striatum . CoQ10 treatment (200 mg/kg/day) in aged (1 year old) MPTP-treated mice, whose nervous system might already exhibit degenerative changes, showed that it significantly attenuated MPTP-induced loss of striatal dopamine and loss of TH-immunoreactive fibers in the striatum . CoQ10 administration can increase mitochondrial content of coQ10 in the cortex of 1-year-old rats . Similar promising results have subsequently been generated using a monkey MPTP model of PD . CoQ10 is extremely lipophilic, making it easy to cross the BBB. Its absorption is improved by the inclusion of lipid in the formulation and by taking it with food.
NBQX (2,3-dihydroxy-6-nitro-7-sulphamoylbenzo[f]-quinoxaline) is a potent competitive AMPA-receptor antagonist, belonging to the quinoxalinediones group . AMPA-selective glutamate receptor antagonists constitute potential neuroprotective agents by counteracting the excitotoxic effects of excess glutamate. In the MPP+ mouse model of PD, AMPA receptor antagonists were shown to have greater therapeutic potential than NMDA receptor antagonists . The quinoxalindione derivatives were discovered in 1988 and are still undergoing intensive study . NBQX exhibits improved AMPA receptor selectivity compared to earlier quinoxalindiones . It has systemic activity and was first shown to have therapeutic effects by protecting against cerebral ischemia after carotid artery occlusion in mice . NBQX has thus been used as the antagonist of choice in many "in vivo" and "in vitro" models. Besides its anti-stroke properties, NBQX also showed efficacy against PD , demyelinating disorders , and trauma .
Given that celastrol, minocycline, coQ10, and NBQX have different mechanisms of action and all shown efficacy in various neurological disease models, we decided to test their efficacy in a Drosophila DJ-1 model of PD. There are two DJ-1 homologues in Drosophila, DJ-1A and DJ-1B. Despite the fact that DJ-1A is more closely related to human DJ-1 at the sequence level, DJ-1B has been the main focus of research simply because of its ubiquitous and higher level of expression. It should be pointed out that DJ-1A, despite its low level of expression, is expressed in adult brain and is inducible under certain conditions . Results on DJ-1B function in stress response and DN survival using various genetic mutants have been divergent [21, 22, 55]. Studies on DJ-1A function in stress response and DN survival using two different DJ-1A RNAi lines by independent groups have been consistent [20, 56], although one study reported that a genomic deletion mutant of DJ-1A, in which the entire DJ-1A locus and some surrounding sequence were deleted, showed stress sensitivity but no DN loss . The lack of DN loss phenotype in the DJ-1A genomic deletion mutant could be caused by a genome-wide compensatory response to the deletion of DJ-1A; alternatively, the phenotypes induced by tissue-specific DJ-1A RNAi may require some non cell-autonomous function of DJ-1A. The efficiency of the DJ-1A RNAi transgene used in this study and evidence that the DJ-1A RNAi phenotype is unlikely due to "off target" effects have been presented in previous studies . Given that the DJ-1A RNAi model recapitulates two features of PD: loss of DN and reduction of brain dopamine levels [20, 56], we decided to test the drugs on this model. We expect that results from this study will help understand the role of DJ-1 dysfunction in PD pathogenesis and validate the usefulness of fly PD models in pharmacological studies.