MANF is widely expressed in the nervous system (31, 32). In the brain of postnatal and adult rats, MANF expression is mainly localized in neurons (31, 32). Relatively high levels of MANF mRNA and protein were distributed in the areas of brain including anterior olfactory nucleus and mitral cell layer of olfactory bulb, II-VI layers of cerebral cortex, piriform cortex, CA1-CA3 regions of hippocampus, dentate gyrus, paraventricular nucleus, bed nucleus of stria terminalis, hypothalamus, cerebellar Purkinje cells, and also in the spinal cord. In the striatum, low MANF expression was detected. Relatively intermediate levels of MANF expression was present in the pars reticulata of substantia nigra (SN). Of note, some of the tyrosine hydroxylase (TH) positive dopaminergic neurons in the pars compacta expressed MANF (31). MANF expression was high on postnatal day 3 and 5, and declined gradually during brain maturation (32). MANF expression is also widely distributed in the non-neuronal tissues. High MANF expression was detected in testis, seminiferous tubules, salivary gland and pancreas (31, 33).

Patterns of MANF mRNA and protein expression in mouse embryonic development were revealed by in situ hybridization and immunohistochemistry. MANF mRNA expression was widely present in the brain from E12.5. mouse embryos. High level of MANF expression were detected in the neopallial cortex, choroid plexus of the lateral ventricles. Relatively low levels MANF expression were detected in the embryonic midbrain, striatum, trigeminal and dorsal root ganglia. A wide MANF expression was also detected in the cartilage primordium of head and vertebra, liver, umbilical vessels, submandibular gland, and pancreas. A low level of MANF expression was detected in lung, metanephros and gut (31).

In human tissues, MANF expression has not been detected extensively. MANF mRNA expression was revealed by RT-PCR in several brain regions and in peripheral tissues (31). Recently, a robust MANF expression was detected by Immunofluorescence and Western blot in retinal ganglion cells of human retina (34). A recent study revealed that MANF expression was also present in human blood serum (35).

It has been evidenced that MANF is important for dopamine system maintenance and survival in Drosophila. In larval Drosophila brain, MANF expression was present in astrocyte-like glia surrounding dopaminergic cells (36). In Drosophila adult brain, MANF expression was detected both in glia and dopaminergic neurons (37). The knockout of MANF gene in Drosophila could lead to a loss of axons of dopaminergic neurons. In addition, the lack of MANF in Drosophila could decrease dopamine levels and increase the transcripts of the dopamine producing enzymes such as tyrosine hydroxylase and DOPA decarboxylase (38). The knockout of MANF gene in Drosophila resulted in several genes related to processes altered in PD including oxidative phosphorylation, protein ubiquitination, mitochondrial function and dopamine metabolism (38). Overexpression of MANF could rescue dopaminergic neurons against oxidative stress (38).

In zebra fish, a wide expression of MANF mRNA was detected by qPCR and in situ hybridization during embryonic development and in adult organs (39). Highest levels of MANF mRNA were expressed in whole embryo two hours after fertilization and declined gradually during brain maturation. In adult brain of zebra fish, MANF mRNA expression was most prominently detected in in neurons of several regions of brain including the thalamic, hypothalamic, forebrain, basal ganglia, cerebellum, and optic tectum. However, only few MANF positive cells were found to co-express tyrosine hydroxylase. Notably, the knockdown of MANF expression could decrease dopamine levels and tyrosine hydroxylase gene transcripts. In addition, these defects could be rescued by injection of exogenous MANF mRNA (39). This evidence suggests that MANF is involved in the regulation of the development of dopaminergic system in zebra fish.

It has been revealed that purified MANF protein could selectively enhance the survival and sprouting of nigral dopaminergic neurons in vitro (16). MANF protein has also been reported to mediate the presynaptic enhancement of GABAergic inhibition to protect dopamine neurons (40). Since MANF expression was present in the striatum and MANF is considered to involve in maintenance of dopaminergic neurons (31). There is an increasing interest to probe the function of MANF in PD treatments. MANF was able to protect nigrostriatal dopaminergic nerves from 6-OHDA-induced degeneration when it was intrastriatally administrated 6 h before neurotoxin (41). More importantly, MANF was also able to restore the function of the nigrostriatal dopaminergic system when microinjected 4 weeks after 6-OHDA administration in the striatum (41). In addition, 125I-labeled MANF was distributed throughout the striatum more readily than GDNF (41). These results suggest that MANF has significant therapeutic potential for the treatment of PD. In addition, the infusion of MANF and gadolinium-DTPA with convection-enhanced delivery (CED) catheter system was tested in porcine brain. The finding showed that distribution of gadolinium-DTPA on magnetic resonance imaging correlated well with the distribution of MANF with immunohistochemistry, suggesting that the MANF with the CED system may be a potential strategy in PD treatments (42).

A comparative study related to the effects of MANF, CDNF and GDNF, was carried out in a severe unilateral 6-OHDA model (43). It has been shown that CDNF with a dose of 40μg was able to inhibit 6-OHDA-induced loss of TH positive cells of the pars compacta of the substantia nigra and TH positive fibers in the striatum and attenuate amphetamine-induced turning behavior. The results suggest that CDNF could inhibit degeneration of nigrostriatal dopamine nerve tract after 6-OHDA injection. However, MANF at similar doses failed to protect nigrostriatal dopamine nerve tract against 6-OHDA induced severe degeneration (43). Although after chronic infusion, 125I MANF protein exhibited a larger distribution volume than CDNF in rat brain tissue and was retrogradely transported from the striatum to the substantia nigra and frontal cortex.

MANF protein has also been studied in a cellular model of PD. Purified MANF protein improved the viability of SH-SY5Y cells against 6-OHDA toxicity, and protected the cells from 6-OHDA or α-synuclein induced apoptosis (44). Recent studies have reported that extracellular application of MANF attenuated 6-OHDA-induced neurotoxicity in SH-SY5Y cells (45, 46).

Since intracranial injection of neurotrophic factors is difficult to develop into a long-term therapy, gene delivery using viral vectors or non-viral vectors mediated neurotrophic factors into rats brain would be an attractive therapeutic method for treatment of PD.

Cordero-Llana et al indicated that intrastriatal lentiviral vector-mediated overexpression of MANF did not decrease amphetamine-induced rotational behavior and did not increase TH positive fibers in the striatum and TH positive neurons in the substantia nigra in parkinsonian rats (47). In addition, intranigral lentiviral vector-mediated overexpression of MANF also had no beneficial effects on amphetamine-induced rotations or TH striatal fiber density but resulted in a significant preservation of TH positive cells (47). However, these results are inconsistent to the results in our recent study. Our study reported that intrastriatal injection of adeno-associated virus serotype 9 (AAV9) mediated human MANF could reduce rotational asymmetry and promote the regeneration of TH positive fibers in the striatum and survival of TH positive neurons in the substantia nigra in the 6-OHDA model of PD (45). We also reported that intrastrital injection of AAV9 – human MANF could increase the concentration of striatal dopamine (DA) and dihydroxyphenylacetic acid (DOPAC, indicator of DA metabolism) and homovanillic acid (HVA, indicator of DA metabolism) and reduce the ratio of DOPAC + HVA /DA (indicator of DA turnover) in 6-OHDA-lesioned rats (45). There are several factors for this inconsistency. Firstly, 6-OHDA injection in Cordero-Llana’s study resulted in a severe lesion (75-90% TH positive cells loss in the substantia nigra), whereas 6-OHDA injection in our study led to a 29-38% loss of TH positive cells in the substantia nigra. In the severe 6-OHDA-lesioned model, nigrostriatal dopaminergic pathway is prone to be damaged severely. Thus dopaminergic neurons in substantia nigra cannot be rescued because of impaired axonal transport and/or neuron death. Secondly, lentivirus application with a low titer (8× 108 TU/ml) only led to a local transduction, and the low-level expression of MANF failed to exert its neuroprotection in the severe 6-OHDA-lesioned model.

MANF plays a significant role in the regulation of ER stress, which triggers unfolded protein response to restore ER homeostasis. It has been observed that MANF remains inside the cells and localizes in the ER lumen (25, 28, 30, 45, 48). Endogenous MANF protein expression is up-regulated by inducers of ER stress in vitro (20, 28, 45, 48-50) and in ER stress response induced by cerebral ischemia, epileptic insult and myocardial ischemia in vivo (31, 49, 50).

We previously showed that endogenous MANF was up-regulated with a similar pattern as chaperone Bip in SH-SY5Y cells after 6-OHDA treatment (45). In addition ER stress chaperone Bip, downstream molecules including p-eIF2α/eIF2α/ATF4/CHOP, ATF6α and XBP1s were also induced by 6-OHDA. We also observed that overexpressed MANF localized surrounding nucleus and co-localized with protein disulfide isomerase (PDI, an ER-resident marker protein) (45). Notably, intracellularly overexpressed MANF can significantly increase cell viability against 6-OHDA induced insult, and inhibited the activation of protein kinaselike ER kinase (PERK), inositol-requiring enzyme 1 (IRE1) and ATF6α pathway (ER stress-associated three pathways) (45). Our results demonstrate that intracellular overexpressed MANF can at least partially alleviate ER stress to protect cells against 6-OHDA neurotoxicity (Table 2).

The C-terminal domain of human MANF is structurally similar to the SAP domain of Ku70 protein (17). Ku 70 binds to the proapoptotic protein BAX via its C-terminal SAP domain to exert its anti-apoptotic function (22-24). Due to the highest homologous to SAP domain of Ku70, Hellman et al. speculated that MANF may prevented apoptosis via interaction with BAX. Accordingly, cellular studies confirmed that overexpression of the full length or C-terminal domain MANF could protect neurons intracellularly against BAX-mediated apoptosis and C-terminal domain of MANF is responsible for the intracellular protection against the BAX-dependent apoptosis (17). However, various data for protein-protein interaction did not confirm any evidence for interaction between MANF and BAX, suggesting that MANF may mediate its anti-apoptotic function via upstream molecules of BAX (25).

In addition, caspase-3 is also involved in the protection effect of MANF. Caspases play essential roles in apoptotic pathway. Caspase-3 is a central member in the execution phase of cell apoptosis. MANF was found to inhibit the cleavage of caspase-3 apoptosis induced by focal cerebral ischemia (51). Two recent studies have also reported that MANF protein repressed cleaved caspase-3 expression induced by 6-OHDA treatment or overexpressed α-synuclein in vitro (44, 52). MANF also prevented 6-OHDA-induced apoptosis by activation of PI3K/Akt/GSK3ß signaling pathway, activation and nuclear translocation of Nrf2 (Nuclear factor erythoid-2-related factor, Nrf2) (53).

Autophagy and apoptosis are two common phases in programmed cell death to maintain cellular homeostasis. It has been indicated that autophagy plays an important role in dopaminergic cells in vitro (54, 55). Moreover, autophagy system was unregulated in both PD patients and animal models of PD (56, 57). Zhang et al recently described that MANF could inhibit autophagy induced by 6-OHDA via AMPK/mTOR pathway in vitro (46) (Table 2).

MANF is a member of CDNF/MANF family and has a neurotropic effect to rescue apoptotic neurons (7, 17). It was therefore hypothesized that MANF would have a neuroprotection in animal models of PD. Numerous in vitro and in vivo studies have confirmed this hypothesis (41, 44-47, 53). But the neuroprotective mechanisms of MANF have not been fully understood. Our recent study showed that extracellular MANF could protect SH-SY5Y cells against 6-OHDA mediated neurotoxicity. However, the neuroprotective mechanisms of extracellular MANF may be different from intracellular MANF (45). Our data demonstrated that extracellular MANF protein did not inhibit the levels of p-eIF2α/eIF2α/ATF4/CHOP, ATF6α and XBP1s, which were induced by 6-OHDA. We found that extracellular MANF protein could increase the levels of p-AKT/AKT and p-mTOR/mTOR, which were decreased after 6-OHDA treatment. Hellman et al. showed that extracellular administration of MANF cannot enter the cells (17). Therefore, we concluded that the neuroprotection of MANF on dopaminergic cells seem to be via both intracellular and extracellular modes of action. On one hand, intracellular overexpression of MANF protects dopaminergic cells by inhibiting the ER stress, and on the other hand extracellular MANF protein protects cells by activation of PI3K/Akt/mTOR pathway. It has also been reported that PKC and NFκB pathways are also involved in cytoprotective effects of MANF in inducible spinocerebellar ataxia 17 mice and in inflammation diseases (58, 59) (Table 2). Whether these pro-survival pathways are involved in MANF-mediated neuroprotective effects are still needed to be further determined.