4-amino-7-chloroquinoline derivatives for treating Parkinson's disease: implications for drug discovery

1. Introduction

Parkinson's disease (PD) is the most common movement disorder affecting approximately 0.3% of the general population and 1–2% of the population over age 65, with a mean age of onset at 55 years.[1] PD was first described by James Parkinson in 1817 and is characterized by the selective loss of A9 dopaminergic (DA) neurons in the substantia nigra (SN) and the presence of intraneuronal cytoplasmic inclusions, termed Lewy bodies. A9 DA neurons innervate the dorsal striatum and this nigrostriatal pathway controls voluntary movements, and their selective loss results in depletion of dopaminergic input in the striatum, leading to clinical manifestations of PD, typified by resting tremor, rigidity, and bradykinesia.

The cause and origin of PD are largely unknown. Like other complex human disorders, PD is likely affected by both environmental and genetic factors. The environmental hypothesis was particularly strengthened by the discovery that exposure to 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) rapidly induces parkinsonian symptoms.[2] In addition, exposure to certain other environmental toxins used as herbicides or pesticides, e.g. rotenone and paraquat, was found to increase the risk of developing PD.[3] These environmental toxin-based models of PD have significantly increased our understanding of the cellular events underlying PD pathogenesis. Another prevailing hypothesis is that protein misfolding and aggregation are directly related to PD pathogenesis. This idea originated from one of the hallmark pathophysiological features of PD, the formation of Lewy bodies, which are composed of proteinaceous aggregates. In general, the ubiquitin-proteasome system protects cells from misfolded proteins. The activity of this system gradually declines with age,[4] which is consistent with the observation that age is a major risk factor for developing PD.

Recent evidence suggests that neuroinflammation is closely associated with the pathogenesis of PD.[5–7] Indeed, the presence of activated microglia has been reported within the SN of PD postmortem tissues and increased levels of pro-inflammatory cytokines in the blood or cerebrospinal fluid of PD patients and in animal models.[8] Under normal physiological conditions, microglia exist as deactivated cells that produce anti-inflammatory and neurotrophic factors. Tissue damage or pathogen exposure activates microglia triggering an inflammatory response resolving upon tissue repair or pathogen removal. However, when tissue damage remains (such as long-lasting protein aggregates or Lewy bodies), inflammatory responses may become chronic and lead to the generation of neurotoxic factors, aggravating the disease process. Chronic microglial activation leads to production of pro-inflammatory cytokines (e.g. TNF-α, IL-1β, and IL-6) which are thought to accelerate the death of A9 DA neurons, partly because of their vulnerability due to the synthesis of DA and its related metabolites.[9] Notably, recent in vitro studies showed that extracellular α-synuclein, the major component of Lewy body, can be oxidized and nitrated and induce microglial activation, thus forming a vicious feed-forward loop leading to accelerated degeneration of DA neurons.[10] Taken together, although not necessarily the initial event in the neurodegeneration process, chronic neuroinflammation appears to significantly contribute to the pathophysiology of PD.[11] In support of this concept, several epidemiological studies showed that chronic use of nonsteroidal anti-inflammatory drugs significantly reduces the risk of PD.[12]

2. Identifying a druggable PD target

During the last two decades, our understanding of how key signaling molecules and transcription factors orchestrate the development of midbrain DA (mDA) neurons in the mouse brain has significantly progressed.[13] In particular, development of mDA neurons is dependent on two major signaling molecules, Sonic hedgehog (Shh) and Wnt1, and their downstream factors.[14] These two critical pathways (i.e. Shh-FoxA2 and Wnt1-Lmx1a) merge to control the expression of NR4A2 (Nurr1), suggesting that it is a key regulator of mDA neurons. Indeed, NR4A2^−/- embryos are devoid of mDA.[15] Furthermore, conditional NR4A2 ablation in fully differentiated adult neurons results in loss of mDA neuron-specific gene expression and neuron degeneration,[16] demonstrating that NR4A2 is essential for both development and survival of mDA neurons. As a master regulator for development, survival, and maintenance of mDA neurons, it is well known that NR4A2 induces expression of genes involved in DA phenotypes (e.g. tyrosine hydroxylase (TH), aromatic amino acid decarboxylase (AADC), dopamine transporter (DAT), and vesicular monoamine transporter (VMAT)), neurotrophic genes involved in survival of mDA neurons (e.g. GDNF receptor, c-Ret genes and brain-derived neurotrophic factor (BDNF)), [17–20] as well as mitochondrial genes involved in energy metabolism.[21] Moreover, NR4A2 appears to function as a transcriptional suppressor in different cellular contexts such as microglia and astrocytes by recruiting the corepressor complex CoREST to the NF-κB target genes (Figure 1).[22] Knocking down NR4A2 in extraneuronal cells dramatically increased inflammatory responses induced by lipopolysaccharide (LPS) or overexpression of a pathogenic mutant form of α-synuclein accelerating mDA neuronal loss, indicating that NR4A2 is critical for protection of mDA neurons from inflammation-induced death in part by suppressing neurotoxic mediators in microglia and astrocytes.[6,22] Thus, it is of great interest to address whether putative activators or agonists of NR4A2 influence either or both of these contrasting and opposite functions.

Figure 1.

NR4A2 appears to work as a transcriptional activator and a transcriptional repressor depending upon the cellular context. In midbrain dopamine (mDA) neurons, it critically activates many DA marker genes such as tryrosine hydroxylase (TH), dopamine transporter (DAT), and c-ret. In contrast, NR4A2 is also shown to work as a repressor and suppresses neurotoxic pro-inflammatory genes in microglia and astrocytes, and thus protects mDA neurons from inflammation-induced death. Interestingly, our recent results showed that AQ and CQ influence both of these contrasting functions of NR4A2.

3. Identification of antimalarial drugs, activating NR4A2

Both academic and pharmaceutical researchers have a keen interest in identifying potential agonists that can activate NR4A2's function (Table 1). 6-Mercaptopurine (6-MP), an antineoplastic agent, was identified as the first Nurr1 activator from high-throughput screening. 6-MP regulates NR4A2 through a region in the amino terminus, AF-1 domain.[23] Several benzimidazole-based compounds were identified as potent NR4A2 activators with an EC₅₀ value of 8–70 nM.[24] Interestingly, several other compounds such as Isoxazolo-pyridinone 7e(IP7e) and 1,1-bis(3′-indolyl)-1-(p-chlorophenyl) methane were identified as activators of the NR4A2 signaling pathway and were shown to have anti-inflammatory effects as well.[25-31] In addition, SA00025, developed by Sanofi, showed a partial neuroprotective effect in a PD animal model induced by injection of poly(I:C) and 6-OHDA.[32] While these small molecules represent promising candidate drugs, none of them were shown to influence NR4A2 activity by directly binding to its LBD. In order to identify putative agonists that directly bind NR4A2's LBD and activate it, our laboratory established high-throughput assay systems and screened a chemical library composed of 960 FDA-approved drugs (MicroSource Discovery Systems, Inc., Gaylordsville, CT). This screening identified three hit compounds (i.e. two antimalarial drugs, amodiaquine (AQ) and chloroquine (CQ), and a pain-relieving drug (glafenine)).[33] Strikingly, all three compounds share the 4-amino-7-chloroquinoline scaffold, strongly supporting a structure–activity relationship (SAR). Furthermore, direct binding of AQ/CQ to NR4A2's LBD was demonstrated using several analyses such as the Biacore S51 SPR sensor, fluorescence quenching analysis, a radioligand-binding assay using [³H]-CQ, and nuclear magnetic resonance (NMR). We found that AQ and CQ enhance expression of mDA-marker genes, are neuroprotective when cells are exposed to neurotoxins such as 6-OHDA as well as anti-inflammatory by suppressing pro-inflammatory gene expression in microglia. Furthermore, AQ and CQ significantly improved behavioral deficits in an animal model of PD, 6-OHDA-lesioned rats, without any sign of dyskinesia-like side effects.[33] Together, two antimalarial drugs, AQ and CQ, enhanced the contrasting dual functions of NR4A2, activation of midbrain DA neuron-specific function (e.g. TH expression) and repression of microglial activation and neurotoxic cytokine gene expression, leading to significant neuroprotective and/or neurorestorative effects in rodent models of PD (Figure 1). Furthermore, a previous study demonstrated that administration of CQ to monkeys protected them from MPTP-induced parkinsonian motor abnormalities.[34]

Table 1.

Putative NR4A2 agonists under development.

Name	Structure	Target	Application	References
6-Mercaptopurine		NR4A2 NR4A3	Anticancer	[23]
Benzimidazoles		NR4A2	N/A	[24]
Isoxazolo-pyridinone 7e		NR4A2	Multiple sclerosis PD	[25–27]
1,1-Bis(3′-Indolyl)-1-(p-chlorophenyl) methane		NR4A1 NR4A2	PD Anti-inflammation	[28–31]
SA00025		NR4A2	PD	[32]
Amodiaquine		NR4A2	PD	[33]
Chloroquine

4. Conclusion

PD is the most frequent movement disorder and its burden on our society is expected to escalate, as the population ages. Currently available treatments are symptomatic and, in many patients, eventually trigger severe side effects, such as dyskinesia. Thus, there is an enormous unmet demand for mechanism-based neuroprotective therapeutics. Based on developmental and molecular studies from numerous laboratories including our own, NR4A2 is considered to be a promising drug target and, indeed, numerous small molecules have been identified that activate its transcriptional activity, which are reviewed here. In particular, we discuss the identification of antimalarials containing the scaffold 4-amino-7-chloroquinoline as synthetic agonists of NR4A2 and a potential SAR, which may translate into development of novel neuroprotective therapeutics for PD.

5. Expert opinion

Multiple lines of evidence support that NR4A2 is a promising drug target for the development of novel therapeutics for PD. In particular, it is of great interest that it exhibits unique and contrasting dual functions as an activator and repressor in mDA neurons and glial cells, respectively (Figure 1). Remarkably, recent studies identified several small molecules developed as agonists/activators of NR4A2 that can enhance both its activator and repressor functions (Table 1). However, since its close homologs (i.e. NR4A1 and NR4A3) may have overlapping functions and expressions, further analyses are warranted to confirm that the functional effects induced by putative agonist(s) are indeed acting uniquely through NR4A2. In particular, since the expression and function of NR4A2 is not well characterized in glial cells, this issue is critical to determine target engagement in the brain. Furthermore, it is not clear whether these small molecules activate NR4A2 via direct binding to its LBD. Based on the ‘proof-of-principle’ studies that antimalarial drugs, AQ and CQ, containing an identical scaffold can enhance NR4A2's dual functions and can alleviate behavioral deficits in animal models of PD,[33] it is tempting to speculate that it is possible to identify more potent NR4A2 agonists with less side effects than AQ or CQ, using medicinal chemistry based on the putative SAR (4-amino-7-chloroquinoline). One may ask if other antimalarial drugs with similar structure can activate NR4A2. We investigated whether related chemical scaffold such as other quinoline compounds can activate NR4A2 function (because many NRs are known to have promiscuous activities). Thus, we analyzed the putative transactivation function of various quinoline compounds and compared them with AQ and CQ. None of these compounds including the antimalarial drugs mefloquine and primaquine showed any detectable transactivation function over the concentration range tested.[33] It is noteworthy that although these compounds have similar quinoline structures, none of them contain 4-amino-7-chloroquinoline. Thus, future medicinal chemistry and in vitro and in vivo studies are warranted to further validate this SAR-based potency optimization. In addition, it is noteworthy that AQ and CQ are well known to modulate autophagy although it is not known whether this activity is directly related to Nurr1. Since autophagy is critical for neurodegenerative disorders including PD,[35] it will be critical to address whether AQ/CQ influence autophagy in dopamine neurons in healthy and diseased brains. Taken together, our findings strongly suggest that it is possible to identify small molecules that can activate NR4A2 function by direct interaction with its LBD and that the structure of 4-amino-7-chloroquinoline is a critical SAR for the activation of Nurr1 and provides the framework for further optimization of NR4A2 agonists.