|Home | About | Journals | Submit | Contact Us | Français|
Previous studies have documented that orphan nuclear receptor Nurr1 (also known as NR4A2) plays important roles in the midbrain dopamine (DA) neuron development, differentiation and survival. Furthermore, it has been reported that the defects in Nurr1 are associated with Parkinson’s disease (PD). Thus, Nurr1 might be a potential therapeutic target for PD. Emerging evidence from in vitro and in vivo studies has recently demonstrated that Nurr1-activating compounds and Nurr1 gene therapy are able not only to enhance DA neurotransmission but also to protect DA neurons from cell injury induced by environmental toxin or microglia-mediated neuroinflammation. Moreover, modulators that interact with Nurr1 or regulate its function, such as retinoid X receptor, cyclic AMP-responsive element-binding protein, glial cell line-derived neurotrophic factor and Wnt/β-catenin pathway have the potential to enhance the effects of Nurr1-based therapies in PD. This review highlights the recent progress in preclinical studies of Nurr1-based therapies and discusses the outlook of this emerging therapy as a promising new generation of PD medication.
Parkinson’s disease (PD), affecting 1.5% of the global population aged over 65 years , is a common neurodegenerative disease characterized by the progressive loss of dopamine (DA) neurons in the substantia nigra (SN) pars compacta and the following striatal DA deficiency. The main clinical manifestations of PD, including bradykinesia, rigidity, resting tremor, postural instability and non-motor symptoms, seriously impair patients’ quality of life. Although the pathogenesis of PD is still far from being clearly understood, it has been accepted that the occurrence of PD might attribute to the interplay of genetic and environmental factors . Until now, at least 15 causal genes have been identified to be related to PD, such as α-synuclein, Parkin, LARRK2, PINK and DJ1 . In addition, exposure to certain environmental toxins, such as 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), rotenone and paraquat, has been found to increase the risk of developing PD. Pathologically, PD has also been characterized by the presence of Lewy bodies containing aggregated and misfolded α-synuclein. Moreover, neuroinflammation, reactive oxygen species, mitochondrial dysfunction and autophagy or proteasome system impairment are considered as major pathogenic contributors to PD [4–6].
Current available PD therapies are mainly to ameliorate motor symptoms associated with DA deficiency, including levodopa, DA agonists, catechol-O-methyltransferase inhibitors and monoamine oxidase-B inhibitors . Unfortunately, the current pharmaceutical treatments cannot prevent the degeneration of DA neurons, and usually lose their efficacy over time, while often accompany severe side effects such as dyskinesia, on-off fluctuation and peripheral reactions. Other medications such as deep brain stimulation, cell transplantation and gene therapy have limitations too, concerning the surgery risk and potential safety hazard. In addition, the symptomatic alleviating effects from alternative therapies lack of robust improvement, so that they usually just serve as adjunctive medications in clinic. In general, none of these treatments can either slow the disease progression or maintain the viability and functions of DA neurons . Therefore, mechanism-based and/or disease-modifying therapies are in urgent need for better management of PD progression. For example, several preclinical tests of molecular targeted therapies have been conducted and the results seem promising . It is highly speculated that these therapies may not only have symptomatic effects, but also provide neuroprotective and neurorepairable benefits in the treatment of PD . Among those, Nurr1-based therapies for PD have attracted great attention. Thus, in this review we will highlight the progress of several potential Nurr1-based strategies, including Nurr1 activating compounds and Nurr1 gene therapy, hoping that they may become future medication against PD.
Nurr1 (also known as NR4A2) is a member of NR4A subgroup of nuclear receptor superfamily. The other two members in NR4A subgroup are Nur77 (NR4A1) and Nor1 (NR4A3), which may possess quite different functions as compared with Nurr1 . There are many studies demonstrating that Nurr1 plays an essential role in the development, differentiation, maintenance and survival of midbrain DA neurons that are mostly affected in the brain of PD . Nurr1, as a transcriptional factor, seems to express earlier than numerous phenotypic markers of DA neurons, such as tyrosine hydroxylase (TH), DA transporter (DAT), aromatic amino acid carboxylase (AADC) and vesicular monoamine transporter (VMAT) [10,11]. There is an age-dependent reduction of Nurr1 expression in the midbrain DA neurons, which arouses much interest because it may possibly explain the high morbidity of PD in the elderly people [12, 13]. In addition, it has been reported that Nurr1’s immunoreactivity is significantly decreased in the SN of PD patients [14–16] and numerous polymorphisms and mutations in Nurr1 have been identified to be associated with both familial and sporadic PD . In consistence with those clinical observations, preclinical studies have also demonstrated the essential role of Nurr1 in the midbrain DA neurons development and functional maintenance [11,18–21]. Genetic deletion of Nurr1 in mice can cause the absence of DA neurons in the SN and ventral tegmental area, which may result in the death of the newborn mice . Several DA neuronal phenotypes such as TH and AADC that are important in the neurotransmission of nigro-striatal pathway are missing in Nurr1-deficient mice . Meanwhile, the heterozygous Nurr1-deficient mice show a higher susceptibility of DA neurons to neurotoxins and exert the similar age-dependent DA dysfunction even before the DA neurons loss [18–20]. Moreover, recent Nurr1 conditional ablation mice simulate the early features of PD, showing a series of pathological changes, including DA neuronal loss, striatal DA reduction, and motor deficiency, which could serve as suitable model to study PD .
It is clear that Nurr1 works as a transcriptional factor to regulate several genes involved in the DA neuronal phenotypes, ranging from the DA metabolism, neurotransmission, axonal growth, mitochondrial function and cell survival [17, 21–23]. More and more target genes have been found that are transcriptionally controlled by Nurr1 in DA neurons, including Dlk1, Ptpru, Klhl1, GTP and VIP [24–26]. Emerging evidence has suggested that there is a complicated network between Nurr1 and other essential transcriptional factors during the DA neuron development. Pix3 and Wnt/β-catenin pathways are the two major signaling molecules contributing to the midbrain DA neurogenesis via cooperating with the Nurr1 transcription complex [27–29]. In addition, the overexpression and accumulation of α-synuclein may interrupt the function of Nurr1 through the direct or indirect interference with the signaling of glial cell line-derived neurotrophic factor (GDNF), whose receptor proto-oncogene tyrosine-protein kinase receptor Ret is regulated by Nurr1 [30, 31], and the interaction with α-synuclein makes GDNF fail to exert neuroprotective effect in PD . In addition, mitochondrial dysfunction and the correlative oxidative phosphorylation impairment have been indicated as major risk factors of PD . In Nurr1-ablated DA neurons, more than 90% of genes encoding respiratory chains are down regulated . More importantly, Saijo et al.  have recently identified a new pathogenic mechanism linking Nurr1 and PD, i.e. the suppressive effect of Nurr1 in the production of inflammatory factors induced neuronal injury. Corepressor for Repressor Element 1 Silencing Transcription Factor (CoREST) repressor complex is identified as an essential modulator in Nurr1-mediated transrepression that inflammatory signals promote the recruitment of CoREST and Nurr1 to the p65 subunit of NF-κB, resulting in the reduction of gene transcription of inflammatory factors . Interestingly, microglia and astrocytes seem to be the primary targets for the neuroprotective effects of Nurr1, rather than DA neurons themselves . These findings provide us an insight understanding of the effect of Nurr1 in PD pathogenesis and open a door for future research for the therapeutic strategies based on Nurr1.
Basically, Nurr1 is an immediate early gene and its transcription can be rapidly induced by various stimuli such as cAMP, inflammatory signals, hormones, calcium and growth factors. These modulators can influence the Nurr1 expression by directly acting on the promoters or transcription regulatory element (cAMP-response element, CArG-like element, SP-1 element) [34, 35]. Nurr1, as a member of NR4A subfamily, shares similar structural features with Nur77 and Nor1, including (1) a modulator domain, referred to as activation function (AF)-1 in the N-terminus; (2) a conversed DNA binding domain (DBD); (3) ligand binding domain (LBD) and its transactivation-dependent AF-2 in the C-terminus . The high conserved DBD has two zinc-fingers, which is able to bind to nerve growth factor inducible-β-binding response element as monomer or homodimers, or Nur-response element as homodimers, which are involved in the transcription process to activate TH and DAT genes [36, 37]. In addition, Nurr1 and Nur77 can bind as monomers, homodimers and heterodimers with retinoid X receptor (RXR). These RXR heterodimers bind a motif called DR5 and can be efficiently activated by RXR ligands . Unlike other nuclear receptors, NR4A subfamily does not have LBD cavity due to the occupation of several bulky hydrophobic residues . Instead, its transcriptional activity appears to be dependent on the AF-1 domain . This feature of Nurr1 makes it difficult to discover compounds that can directly activate Nurr1 through LBD. However, several co-regulator interaction surfaces in Nurr1 LBD have been identified, such as residues 592, 593 and 577, especially the groove between helices 11 and 12, which provides the possibility in developing Nurr1-activating compounds based on their binding to LBD [41, 42]. Furthermore, several compounds have already been identified to activate Nurr1 or Nur77 via their LBDs [43–45]. In general, based on the activated functions of these regions of Nurr1 and numerous studies of Nurr1-activating compounds, it is possible to identify small molecules to activate Nurr1 through its different domains.
Great efforts have been taken to explore novel Nurr1-activating compounds in recent years (as summarized in Table 1). These compounds can activate Nurr1 functions, and further exert neuroprotective, neurogenic or anti-inflammatory effects.
Mercaptopurine (6-mercaptopurine or 6-MP), an anti-leukemia drug, is the first identified Nurr1/Nor1 agonist from high-throughput screening, which stimulates Nurr1/Nor1 through directly binding to the N-terminal AF-1 domain [46, 47]. In addition to 6-MP, a series of benzimidazole-based compounds with Nurr1-activating properties (EC50: 8–70nM) have been identified and synthesized . Given that no Nurr1 ligands have been found, a group of compounds with benzimidazole scaffold have been screened through a reporter-gene (luciferase) assay, whose expression is regulated by Nurr1-specific DNA-binding elements. However, only the Nurr1 biological functions, not the anti-parkinsonian effects of these compounds have been tested.
Previous reports have demonstrated that 1,1-bis (3’-indolyl)-1-(aromatic) methane (C-DIM) analogs can influence the expression of NR4A subfamily in several cancer cells [49–52]. Recently, the anti-parkinsonian effects of C-DIM analogs have been identified. Interestingly, it shows similar activating capacities with different p-substituted phenyl (OCH3, CI, CF3, Br, t-Bu, CN, I and OCF3), suggesting a structure activating relation (SAR) existing in these compounds that contain a bis (3’-indolyl) moiety. In addition, Li et al. have demonstrated that both N- and C- terminal domains are involved in the activation of Nurr1, but without evidence to confirm the direct binding between C-DIM analogs and Nurr1 . Among these C-DIM analogs, C-DIM5 and C-DIM8 have a higher affinity for Nur77 but with opposite effect [49, 51]. After a comparative analysis, C-DIM12 (higher affinity for Nurr1) has the most potent neuroprotective activity and anti-inflammatory effect in MPTP-lesioned rat model of PD . It could enhance the expressions of Nurr1 and Nurr1-regulated proteins such as TH and DAT, and Nurr1-mediated recruitment of CoREST and the NF-κB-mediated inflammatory genes expression are suppressed in SN [53–55].
Another Nurr1 activator, isoxazolo-pyridinone 7e (IP7e), has been reported to attenuate the inflammation and neurodegeneration via inhibiting NF-κB-dependent process . IP7e seems to be a promising candidate due to its high oral bioavailability of 95% plus rapid and extensive brain absorption and distribution . However, the Nurr1-activating effect of IP7e has only been tested in multiple sclerosis models, not in PD models. As for PD, the analogue of isoxazolo-pyrinone derivative, SH1, has been demonstrated to be effective to improve behavioral performance in lactacystin-lesioned PD mouse model, of which inhibition of microglia-mediated neuroinfammation and increasing the DA specific phenotypes may be the main mechanisms to explain its effect . In addition, SA00025, another novel Nurr1 agonist (EC50: 2.5nM), shows a partial neuroprotective effect in PD models induced by inflammatory stimulant poly (I:C) and 6-hydroxydopamine (6-OHDA). It can modulate several DA target genes and display anti-inflammatory activity by reducing the activation of microglia and astrocytes .
While all the above compounds have been identified as Nurr1-activating compounds, none of them are shown to activate Nurr1 by direct physical interaction with LBD. Strikingly, Kim et al. have recently identified antimalarial drugs amodiaquine (AQ) and chloroquine (CQ), together with a pain-releasing drug Glafenine as a novel group of Nurr1 activators and finally demonstrated that AQ/CQ (EC50: 20–50µM) can interact with Nurr1-LBD via direct physical binding . They used a series of analyses, including Biacore S51 SPR sensor, fluorescence quenching analysis, a radioligand-binding assay by [3H]-CQ and nuclear magnetic resonance to estimate this phenomenon. This finding advances our current understanding of the LBD binding properties and opens the door for further research and development of Nurr1 agonists. Interestingly, these compounds contain an identical 4-amino-7-chloroquinoline scaffold, which may predict a possible SAR. As for the underlying mechanisms, the inhibition of microglia-mediated neuroinflammation and increase of DA-specific genes expression are involved in the anti-parkinsonian effect of AQ/CQ. Moreover, the autophagy-regulating effects of AQ/CQ may also predict potential interactions between Nurr1 and autophagy for their anti-parkinsonian effects .
These Nurr1 activating compounds provide much promising possibilities for PD treatment and most of them exert the anti-parkinsonian effect through both the activating and suppressing functions of Nurr1. Although these compounds have the ability to activate Nurr1, it is still not clear the exact regions of Nurr1 targeted by these compounds. Among them, only 6-MP and AQ has been confirmed to directly bind to the AF-1 or LBD of Nurr1, so further studies are in urgent need to discover the specific binding regions on Nurr1. Furthermore, the pharmacodynamics and pharmacokinetic properties and the side effects of those compounds are also required to evaluate. Since the other two members of NR4A subfamily, Nur77 and Nor1, share high similar structure with Nurr1, and they are also involved in the regulation of neurologic functions and inflammation, it is of importance to test the specificity of the existing Nurr1 activating compounds [61, 62]. Wei et al. have recently indicated that Nurr1 and Nur77 might present a contra-directionally coupling interaction in memantine-mediated neuroprotection . This complicated network makes it very difficulty to predict the therapeutic outcome if both Nurr1 and its homologs are activated through the conserved regions. Moreover, SAR has been a research hotpot recently, especially after the presentation of the Nurr1-activating effects of C-DIM analogs and AQ/CQ. Meanwhile, we should notice the fact that subtle alternations of the functional structure might result in a great or even opposite effects [49,51]. Only after those concerns are well addressed, we might be able to determine if Nurr1-activating compounds can provide an alternative therapy for PD.
In addition to activate Nurr1 by recognition on the specific target region of Nurr1, some other compounds, such as DA agonists, memantine, retinoic acid-loaded polymeric nanoparticles (RA-NPs) and herbal extracts (as summarized in Table 2), have also been demonstrated to up-regulate Nurr1 expression in preclinical trials of PD treatment although there have not any report to confirm that they can activate Nurr1 by direct binding [63, 66, 70–73, 78].
The neuroprotective effect of DA agonists has been a controversial topic for decades . One recent clinical trial supports that DA agonists can exert neuroprotection by inducing Nurr1 expression in peripheral blood mononuclear cells . Additionally, in DA neuron cell lines, D2/D3 agonist pramipexole has a profound effect to increase Nurr1 expression, which precedes the up-regulation of DAT and VMAT2 expressions , suggesting that Nurr1 may serve as a key target for DA agonists-mediated neuroprotection.
Memantine, an N-methyl-D-aspartate receptor antagonist, has been demonstrated to restore PC12 cells survival from 6-OHDA-induced neurotoxicity through up-regulating Nurr1 and downregulating Nur77 and partially inhibit migration of Nur77 from nucleus to mitochondria . Different from Nurr1, Nur77 usually triggers apoptotic process when migrate to mitochondria and induce inflammation via NF-κB pathway [67,68]. Interestingly, administration of DA activators could decrease Nur77 expression , suggesting that Nurr1 together with the contra-directional coupling of Nurr1/Nur77 is involved in PD pathogenesis.
Besides those chemically synthesized compounds, several Nurr1-modulators from natural source, such as moracenin D, EGb 761, radicicol and Bushen Huoxue Decoction (BHD), are also effective in PD therapy [70–73]. Moracenin D, an extract from Mori Cortex radices, significantly up-regulates Nurr1 expression and down-regulates α-synuclein expression . EGb 761, extracted from Ginkgo biloba leaves, has been applied in clinical trials of treating several neuropsychiatric diseases [74,75]. It also exerts neuroprotective effect in MPTP-lesioned mice via increasing the expression of a series of DA-related genes, of which Nurr1 is increased by 148% in SN . Our previous study has shown that radicicol could induce HSP70 expression in SH-SY5Ycells against rotenone-induced apoptosis through the inhibition of P53 and induction of Nurr1 . It has been reported that the medical decoction BHD could also increase Nurr1 expression. The anti-parkinsonian effect of BHD has been tested in clinical trials [73,76].
Biomaterials as therapeutic tools have been developed to help Nurr1-based PD therapy recently. Retinoic acid (RA) can enhance the survival and maturation of neuronal cells and RA receptors are highly expressed in DA neurons . In contrast to conventional RA formulations, the recently developed novel nanoparticles coupled with RA could rapidly transport into cells to release RA, and exert neuroprotection against DA neuronal damage in MPTP-mouse model of PD. More importantly, RA-NPs administration increases the expression levels of Pitx3 and Nurr1, showing its supportive effect on the development and functional maintenance of DA neurons in PD .
Although the specific mechanisms for the Nurr1 up-regulating effects of the above mentioned compounds, natural products and biomaterials still remain poorly understood, those candidates have a common feature to promote Nurr1 expression or functions, and part of those candidates have been shown to exert potent anti-parkinsonian properties in clinic [66, 73]. Further exploration of those modulators should focus on identifying their direct binding regions to Nurr1 and the Nurr1 upstream molecules. All of these efforts may lead to a better understanding of their Nurr1 up-regulating mechanisms, and discovery of more potent Nurr1-modulators for the treatment of PD.
Recent animal trials of Nurr1 gene therapy have shown a gratifying progress. Two months after midbrain AAV-Nurr1/Foxa2 injection, the DA neurons density and neurotrophic factors expression are significantly increased in MPTP-lesioned mice model of PD together with a suppression of pro-inflammatory cytokines secretion . Remarkably, this Nurr1/Foxa2 mediated cytoprotective effect is observed even one year after injection, suggesting that transfection of these genes is an intriguing approach in PD therapy . Although the outcome of this trial seems exciting, the safety and feasibility of Nurr1 gene therapy still needs further verification. Firstly, the long-term effect of constitutive overexpression is not clear. Secondly, this therapeutic strategy has only been tested in toxin-induced animal models (such as MPTP), but not in transgenic PD models that may mimic the broader pathology of PD. Thirdly, in order to achieve a better therapeutic outcome, the time window of treatment and a controllable regulation of Nurr1 expression are yet to be further explored.
Interestingly, it has been reported that GDNF fails to restore DA neurons loss is due to the toxicity of α-synuclein . Considering the overexpression of Nurr1 may protect DA neurons against α-synuclein , a combinatory delivery of AAV-Nurr1 with GDNF or neurturin into the midbrain might improve the outcome of the existing preclinical or clinical trials, and might predict a very attractive alternative therapy for PD .
Nurr1 serves as a promising target for PD therapy by promoting the survival or genesis of DA neurons . Since Nurr1 is encoded by an immediate early gene, many modulators are implicated in the Nurr1 pathway. A wide range of physiological signals, such as prostaglandins, fatty acids, calcium, stress, growth factors, inflammatory cytokines, and even physical stimuli, such as membrane depolarization have been shown to induce Nurr1 expression [81–86]. Those factors usually influence the activity of the upstream regulatory molecules thus makes Nurr1 respond rapidly to the cellular environmental changes. Aging is a negative factor causing the decline of Nurr1 expression , which in turn may reduce the compensatory protective capability of Nurr1 and increase the susceptibility to develop PD. Restore the impaired Nurr1functions by Nurr1-based therapies could provide tentative therapeutic strategies for PD.
RXR, neurotropic factors, cAMP-responsive element-binding protein (CREB) and Wnt/β-catenin pathway seem to play positive roles in regulating Nurr1 expression [87–89]. Thus, we propose that it may be an alternative way to facilitate Nurr1 functions through activating or increasing those modulators. Inspiringly, RXR ligands, such as docosahexaenoic acid, bexarotene, LG100268 and XCT0139508 have been demonstrated to protect DA neurons through interaction with Nurr1/RXR heterodimers [87, 90, 91]. Although the biological functions of bexarotene still remains to be defined, one recent study has revealed that bexarotene might have the capacity to rescue the disrupted GDNF signaling and regulate oxidative phosphorylation and Nurr1-related genes . In addition, animal tests have indicated that transplantation of GDNF-pretreated NSCs could dramatically enhance the expression of Nurr1 and TH . Given that GDNF has a poor blood-brain-barrier penetrating feature, GDNF inducers such as telmisartan, carcitriol, rhus verniciflua stokes and cabergoline can stimulate GDNF expression [93–98]. Moreover, CREB as an upstream regulator of Nurr1 may play an important role in the process of most modulators-mediated Nurr1 up-regulation, such as prostaglandin E2, thromboxane A2, vascular endothelial growth factor and N-methyl-d-aspartate [81, 84, 99, 100]. Thus, CREB activators or other cAMP increasing agents may also have the capacity to restore Nurr1 functions. However, considering CREB activates many signal pathways, they in general may not be suitable to treat PD. Furthermore, several studies have indicated that activation of Wnt/β-catenin signaling pathway could induce the transcription of Nurr1 [101–103]. Given that Wnt/β-catenin pathway is involved in the PD pathogenesis and exogenous wnt1 could attenuate DA neuron loss in animal models of PD [104, 105], we propose that activating Wnt/β-catenin pathway, such as GSK-3β antagonists, may be effective in treating PD through the enhancement of Nurr1 expression . Furthermore, Nurr1 activity can be affected by several PD-related molecules, such as α-synuclein and microRNA-132 (miR-132) [106,107]. Therefore, these molecules may be potential therapeutic targets against PD. Indeed, it has been reported that oligomer selective α-synuclein antibodies and prolyl oligopeptidase inhibitors are effective in PD models through decelerating the accumulation of α-synuclein [108–110]. It is plausible that Nurr1 may be involved in the anti-parkinsonian effects of those compounds. Moreover, several studies have described a potential role of microRNAs in PD pathogenesis [111,112], and miR-132 can directly inhibit Nurr1 expression ; thus miR-132 antagonist may be a potential modulator for PD treatment .
Nurr1 based cell transplantation is another therapy currently under investigation. Several animal experiments have shown the therapeutic potential of Nurr1-based cell transplantation, not only ameliorating the behavioral abnormality, but also restoring the DA neuron deficits [114–119]. However, they have not been tested in PD patients concerning the biosafety and long-term efficacy of cell transplantation in general .
It has been noticed that Nurr1 modulators may have other functions in addition to regulating Nurr1. In addition, most of those compounds have not been evaluated in PD models yet, nor their side effects have been systematically determined. Furthermore, Nurr1also expresses in non nervous system . Therefore, the non-neurological effects of Nurr1 should not be neglected.
In summary, Nurr1-activating compounds, Nurr1 modulators, and Nurr1 based gene therapy have shown their therapeutic potentials for PD treatment (as summarized in Figure 1). The pharmacological effects of Nurr1-based therapies against PD are proposed as follows: (1) increase the expression of DA-related genes; (2) protect or repair DA neurons against neurotoxins; (3) inhibit the microglia activation and suppress neuroinflammation. These versatile functions make Nurr1 an attracting target for treating PD. Further studies should focus on identifying small molecules, which can effectively activate Nurr1 functions through either direct binding to the specific regions of Nurr1 or modulating the Nurr1-related signaling. Furthermore, the study of Nurr1 conditional knockout mouse models would be beneficial to better understand the critical roles of Nurr1 in PD pathogenesis, and help screen more specific Nurr1-activating or Nurr1-modulating compounds. Although it is still a long way to go for the clinical use in PD patients, the Nurr1-based therapies provide a very promising outlook as the next generation PD treatment.
The authors are supported in part by Chinese National Sciences Foundation (NO. 81370470 and 81430021), the Collaborative Innovation Center for Brain Science, and the Program for Liaoning Innovative Research Team in University (LT2015009), and the intramural research program of NIA (AG000928).
Conflict of interests
The authors confirm that this article content has no conflict of interest.