MANF Inhibits α-Synuclein Accumulation through Activation of Autophagic Pathways

Progressive accumulation of misfolded SNCA/α-synuclein is key to the pathology of Parkinson's disease (PD). Drugs aiming at degrading SNCA may be an efficient therapeutic strategy for PD. Our previous study showed that mesencephalic astrocyte-derived neurotrophic factor (MANF) facilitated the removal of misfolded SNCA and rescued dopaminergic (DA) neurons, but the underlying mechanisms remain unknown. In this study, we showed that AAV8-MANF relieved Parkinsonian behavior in rotenone-induced PD model and reduced SNCA accumulation in the substantia nigra. By establishing wildtype (WT) SNCA overexpression cellular model, we found that chaperone-mediated-autophagy (CMA) and macroautophagy were both participated in MANF-mediated degradation of SNCAWT. Nuclear factor erythroid 2-related factor (Nrf2) was activated to stimulating macroautophagy activity when CMA pathway was impaired. Using A53T mutant SNCA overexpression cellular model to mimic CMA dysfunction situation, we concluded that macroautophagy rather than CMA was responsible to the degradation of SNCAA53T, and this degradation was mediated by Nrf2 activation. Hence, our findings suggested that MANF has potential therapeutic value for PD. Nrf2 and its role in MANF-mediated degradation may provide new sights that target degradation pathways to counteract SNCA pathology in PD.


Introduction
Parkinson's disease (PD) is a neurodegenerative disease characterized by abnormal deposits of SNCA/α-synuclein aggregates and progressive loss of dopaminergic neurons in the substantia nigra [1,2]. The abnormal accumulation of SNCA could induce neurodegeneration through disrupting of axonal transport, as well as impairing mitochondrial, lysosomal, proteasomal, and endoplasmic reticulum (ER) functions [3]. SNCA exists in a dynamic equilibrium among different conformations and oligomers [4], and the propensity for its aggregation may be reversed by reduction in monomeric SNCA levels which results in disaggregation of soluble oligomers [5]. Considering accumulation and propagation of misfolded SNCA in the brain is integral to the disease pathogenesis, drugs, or herbs aiming at promoting SNCA degradation may work as an efficient therapeutic strategy for PD.
Mesencephalic astrocyte-derived neurotrophic factor (MANF) has been confirmed to possess a more specific neuroprotection for dopaminergic neurons compared to other neurotrophic factors such as glial cell line-derived neurotrophic factor (GDNF) and brain-derived neurotrophic factor (BDNF) [6]. Knocking out the homologous gene of  2 Oxidative Medicine and Cellular Longevity MANF in Drosophila leading to an abnormal development of dopaminergic (DA) neurons is suggesting a key role of MANF on the development and functional maintenance of DA system [7]. Our previous studies also demonstrated that MANF inhibited the loss of DA neurons in PD rats, mice, and transgenic nematode models and improve the motor function of model animals [8][9][10][11][12]. Considering the vital role of SNCA accumulation in the neurodegeneration of PD, we established an A53T mutant α-synuclein nematode model and verified that MANF could regulate the expression of autophagy-related genes, inhibited the accumulation of SNCA A53T , and subsequently promote the survival of DA neurons [11]. However, the specific mechanism that triggering SNCA clearance by MANF remains unknown. Autophagy lysosome pathway (ALP) system plays a vital role in the degradation of SNCA [13]. Inhibiting ALP activity exacerbates the abnormal aggregation of SNCA [14]. ALP includes chaperone-mediated autophagy (CMA), macroautophagy, and microautophagy. Among them, CMA and macroautophagy have been verified to be associated with the clearance of SNCA [13]. CMA is a selective protein degradation process in which cytosolic proteins bearing a KFERQ motif are recognized by Hsc70, then delivered to lysosomal associated membrane protein 2a ( Figure 1: The protective effect of AAV8-MANF on PD model and its role on the expression of lysosome-related genes in substantia nigra. (a) AAV8-MANF (with His tag) was injected into the substantia nigra of rat brain. The expression of MANF in DA neurons was detected by double immunofluorescence staining using anti-His tag antibody and anti-TH antibody. Scale bar is 200 μm. (b) The motor function of AAV8-MANF on rotenone-induced PD mouse models was evaluated by rotarod analysis. Ctrl group: n = 14; rotenone group: n = 14; rotenone + PBS group: n = 15; rotenone + AAV8 − MANF group: n = 13. (c, d) The effect of AAV8-MANF on rotenone-induced DA neuron degeneration was determined by TH staining. Scale bar is 500 μm.
(e, f) The level of SNCA in substantia nigra was detected using western blot analysis. Rotenone + PBS group and rotenone + AAV8 − MANF group: n = 5. (g, h) RNA-sequence analysis of gene expression in substantia nigra followed by AAV8-MANF injection. (e) Heatmap analysis. (f) KEGG pathway analysis. Data were expressed as mean ± SD from three independent experiments. # P < 0:05 and ## P < 0:01 vs. Ctrl group; * P < 0:05 vs. rotenone + PBS -treated group.   Oxidative Medicine and Cellular Longevity and are eventually transported into the lysosomal lumen for degradation [15]. In macroautophagy, portions of the cytoplasm, including protein aggregates, are sequestered inside a double membrane structure which, in turn, can fuse directly with the lysosome [16]. Previous study revealed that MANF gene mutation in Drosophila inhibited the gene expression of lysosome-associated membrane protein (LAMPs), decreased the expression of proton pump V-ATPase genes, and downregulated the lysosomal hydrolysis activity, which eventually disrupted protein degradation process [17]. Thus, we assumed that the degradation of MANF on SNCA may be mediated by CMA and/or macroautophagy pathway.
In the present study, we used the rotenone-induced PD mice model to investigate the effects of MANF on Parkinsonian behavior and SNCA pathology. Meanwhile, Lenti-X™ Tet-On® 3G Inducible Expression System and SH-SY5Y cells were used to realize doxycycline (Dox) induced expression of SNCA WT and SNCA A53T . We characterized the role of CMA and macroautophagy pathway in MANF-mediated degradation of SNCA and investigated the potential interaction between CMA and macroautophagy activation.

AAV8-MANF Relieves Parkinsonian Behavior and
Alleviated the Accumulation of SNCA in the Substantia Nigra. AAV8-MANF (with His tag) was established to achieve the continuous supply of MANF. As shown in Figure 1(a), AAV8-MANF injected into substantia nigra could effectively infect dopaminergic neurons and express MANF protein. AAV8-MANF was injected into the bilateral substantia nigra 1 week before rotenone (3 mg/kg) was subcutaneously injected for 5 weeks. Rotarod analysis showed that AAV8-MANF significantly ameliorated motor impairments of PD mouse (Figure 1(b)). TH staining indicated that AAV8-MANF inhibited rotenone-induced DA neuron degeneration (Figures 1(c) and 1(d)). Afterward, as indicated in Figures 1(e) and 1(f), the increased accumulation of SNCA in substantia nigra was decreased following AAV8-MANF treatment. RNA-seq sequencing revealed that AAV8-MANF could promote the expression of genes include membrane protein genes (Abcb9), hydrolase genes (Ppt2, Hexa, Arsa, Nagpa, and Pla2g15), and proton pump V-ATPase gene (Atp6v0c) related to the lysosomal pathway, indicating an involved of ALP in MANF-mediated SNCA degradation (Figures 1(g) and 1(h)).

MANF Inhibits the Accumulation of SNCA WT in PD Cellular
Model. Lenti-X™ Tet-On® 3G inducible expression system and SH-SY5Y cells were used to realize Doxinduced expression of SNCA WT ( Figure S1 A-C). MANF treatment alone did not show any obvious effect on the viability of SNCA WT SH-SY5Y cells for 48 h ( Figure S3). Whereas MANF treatment for 24 h significantly reduced the accumulation of SNCA WT in a concentration dependent manner (Figures 2(a)-2(d)). RNA-seq sequencing and subsequently BP analysis showed that the expression of genes associated with "Macroautophagy," "Chaperonemediated autophagy (CMA)," "Protein targeting to lysosome," and other genes involved in ALP process was changed after MANF treatment for 24 h (Figures 2(e)-2(g)).

Macroautophagy Was Involved in MANF-Mediated
Degradation of SNCA WT . Macroautophagy was reported to be activated to promote SNCA clearance when CMA system is impaired caused by SNCA overaccumulation [13]. Here, we tested whether MANF-mediated degradation of SNCA WT was modulated by macroautophagy. As shown in Figures 4(a) and 4(b), when cells were incubated with Dox and MANF for 24 h, the content of SNCA WT was significantly reduced, but the expression of macroautophagy markers such as light chain 3 (LC3), Beclin-1, and P62 were not changed, indicating that macroautophagy system was not activated in the early stagy. However, when cells were incubated with Dox and MANF for 48 h, the conversion of LC3-I/II and the expression of Beclin-1 were increased, and simultaneously, the expression of P62 was decreased by MANF treatment, suggesting macroautophagy was activated during SNCA accumulation (Figures 4(c) and 4(d)). By using Lenti-mCherry-eGFP-LC3B plasmid and macroautophagic inhibitor CQ to detect autophagic flux, we found that MANF could partly inhibit the decreased ratio of mCherry/eGFP fluorescence intensity caused by CQ, indicating an activation of macroautophagy (Figures 4(e)-4(h)). More importantly, the clearance of MANF on SNCA WT was partially abolished by the treatment of macroautophagic inhibitor CQ (Figures 4(i)-4(j)). These findings indicated that, besides CMA, macroautophagy system was also involved in MANF-mediated degradation on SNCA WT .

Nrf2 Was Involved in the Early Activation of Macroautophagy by MANF when CMA System Was
Blocked. To evaluate the crosstalk or compensatory interaction between CMA and macroautophagy, CMA activation was suppressed by RNAi-mediated LAMP-2A inhibition. As shown in Figures 5(a) and 5(b), the conversion of LC3-I/II and the expression of Beclin-1 were increased, while the expression of P62 was decreased when CMA system was blocked, indicating a beforehand activation of macroautophagy by MANF. Both CMA and macroautophagy pathway could be regulated by Nrf2 [18,19]. Upregulating Nrf2 reduced the abnormal accumulation of SNCA by promoting CMA and macroautophagy activity [20]. Besides, when macroautophagy was blocked, P62 could activate Nrf2 by binding to keap1 [21] that may trigger other degradation pathway, such as CMA. We reported that MANF increased Nrf2 expression and promoted its nuclear translocation to exerting antiapoptotic effects [9]. Hence, we speculated that Nrf2 may be involved in the crosstalk between CMA and macroautophagy on MANF-induced clearance of SNCA WT . As shown in Cuervo et.al reported that SNCA A53T could block CMA due to the strongly bound to LAMP-2A and disrupt the degradation of substrate proteins [22]. Hence, SNCA A53T overexpression cellular model was used to mimic CMA dysfunction situation, and Nrf2 mediated activation of macroautophagy was investigated. Lenti-X™ Tet-On® 3G inducible 9 Oxidative Medicine and Cellular Longevity expression system and SH-SY5Y cells were used to realize Dox induced expression of SNCA A53T (Figure S2A-C). We showed that MANF treatment significantly reduced the accumulation of SNC A53T in a concentration dependent manner (Figures 6(a)-6(d)). As expected, CMA was not activated in MANF-mediated clearance of SNCA A53T due to impaired CMA system (Figures 6(e)-6(h)), while macroautophagy was confirmed to be participated in the degradation of SNCA A53T by MANF (Figures 7(a)-7(h)). Noticeably, Nrf2 inhibitor ML385 could partially abrogate MANF-mediated activation of macroautophagy and the subsequent clearance of SNCA A53T (Figures 8(a)-8(f)).
Accordingly, these findings suggest that Nrf2 was involved in the early activation of macroautophagy by MANF when CMA system was impaired.

Discussion
SNCA plays a central role in the aetiology and pathophysiology of PD [23]. Drugs or herbs that own SNCA clearance properties may be potential candidates for PD treatment. MANF, as a newly identified neurotrophic factor, has been confirmed to possess neuroprotective effects on PD [24]. Our previous study further demonstrated that MANF could   activate autophagy-related genes and alleviate the aggregation of SNCA A53T11 , but the exact mechanisms remain unknown. In the current study, we firstly confirmed that CMA and macroautophagy, the two main subtypes of ALP, were both involved in MANF-induced degradation of SNCA WT . In addition, we found that MANF could activate macroautophagy in a Nrf2-dependent manner when CMA system was impaired. CMA is a selective degradative process for cytosolic proteins that contributes to the maintenance of proteostasis [25]. Reduced levels of CMA markers have been observed in postmortem brain samples from PD patients [26], indicating that CMA dysfunction was associated with the pathogenesis of PD. In addition, SNCA, the neuropathological hallmark, also contains KFERQ sequence, which was verified as a natural substrate of CMA [26]. Increasing evidence confirmed that CMA represents a major pathway for SNCA WT degradation [27]. CMA inhibition was accompanied by formation of detergent-insoluble or high molecular-weight (HMW) oligomeric SNCA conformations, leading to dopaminergic neurodegeneration and parkinsonian behavior [28]. Our observations revealed that MANF could upregulate the expression of HSC70 and LAMP-2A, the main components of CMA, and promote the combination between HSC70 and SNCA WT , indicating an activation of CMA system, which eventually contributed to the degradation of SNCA WT .
Only SNCA monomers and dimers, but not oligomers, could be degraded by CMA [29]. Excess levels and/or abnormal aggregated of SNCA impaired CMA, and then, macroautophagy system will be activated to participate in the clearance of oligomer SNCA [30]. Using a Dox-induced cellular model of SNCA WT , we also found that macroautophagy system would be activated by MANF due to the continuously accumulation of SNCA WT , which suggested that CMA could participate in the degradation of SNCA WT in the early stagy and macroautophagy system may act as a subsequent compensation mechanism. Interestingly, we revealed that when CMA was blocked by using a LAMP-2A targeting siRNA-based approach, or using Dox-induced SNCA A53T overexpression cell model to mimic CMA dysfunctional condition, macroautophagy could be early activated by MANF to eliminate the accumulation of SNCA WT . These results indicated that MANF could promote the degradation of SNCA WT from SNCA accumulation to aggregation. Meanwhile, even when CMA was blocked due to uncontrolled accumulation of SNCA, MANF still could opportunely activate macroautophagy to further degrade SNCA, which is important for inhibiting SNCAinduced neurotoxicity.
Nrf2, as a leucine zipper transcription factor, controls the basal and stress-inducible expression of over 250 genes [31]. A majority of these genes are involved in the different phases of the macroautophagy process, from cargo recognition to autolysosome clearance [32][33][34][35][36]. Nrf2 can also activate macroautophagy through the Nrf2-p62-keap1 feedback loop [19]. Besides, Nrf2 could recognize and bind to the ARE functional region in the LAMP-2A gene, thereby regulating the expression of LAMP-2A and eventually upregulating CMA activity [18]. Thus, the Nrf2 pathway could serve as an upstream signal to regulate the activities of both CMA and macroautophagy [18,20]. Genetic increase of astrocyte-specific Nrf2 could attenuate the functional deficiency of macroautophagy and CMA, which promoted the degradation of SNCA A53T in SNCA A53T mouse model [20]. Hence, Nrf2 might act as a regulatory node in the proteolytic network represented by macroautophagy and CMA in PD and may serve as a link whereby one reacts in a compensatory manner to the loss of activity in the other [19][20][21]. In this study, we found that MANF upregulated Nrf2 expression while stimulating CMA or macroautophagy system. Blockage of Nrf2 using ML385 (a specific inhibitor) could 14 Oxidative Medicine and Cellular Longevity almost completely counteract the early activation of macroautophagy induced by MANF. These results revealed that Nrf2 was involved in the functional compensatory between CMA and macroautophagy by mediating macroautophagy activation in CMA dysfunction situation.

Conclusions
Our results demonstrated that CMA and macroautophagy system are involved in MANF-induced degradation of SNCA WT . Meanwhile, Nrf2 was confirmed to be a regulator to activate macroautophagy in CMA dysfunctional situation. There might be some limitations in the present study. (I) Ubiquitin-proteasome system (ULP) is also involved in the regulation of SNCA. Monoubiquitinated SNCA is targeted for degradation by the proteasomal system [37]. Whether MANF-mediated degradation of SNCA is associated with ULP needs further exploration. (II) The exact mechanisms of Nrf2 on MANF-induced regulation of CMA and macroautophagy still need to be clarified. (III) Considering the multiple mechanisms involved in the activation of CMA and macroautophagy, besides Nrf2, other signaling pathways need to be further figured out. Despite some limitations, the findings from our investigation indicated MANF may be a candidate for the treatment of PD. Meanwhile, Nrf2 and its role in MANF-mediated degradation may provide new therapeutic strategies that target degradation pathways to counteract SNCA pathology in PD.

Animals and Treatment.
In previous studies, adult male rats and mice have been widely used in the study of PD [38][39][40][41]. In this study, adult male Sprague-Dawley rats (used for detecting the expression of AAV8-MANF and RNA-seq) and male C57/BL6 mice (used for exploring the protection of AAV8-MANF on Rotenone-induced PD models) were obtained from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China) and used following the National

Construction of AAV8-MANF-His and AAV8-NULL.
The coding regions of the MANF with His tagged to its Cterminal were synthesized by Genewiz bio (South Plainfield, NJ, USA). The synthesized genes were digested with the relevant restriction enzymes, and DNA fragments were ligated into the Stratagene CMV AAV expression vector yielding the vector pAAV-MANF-His was cloned into the Stratagene CMV AAV expression vector to generate the control virus. Constructs were verified by sequencing and packaged into AAV8 viral particles with the Rep2/Cap8 AAV encapsulation construct. AAV8-NULL was packaged as the negative control of AAV8-MANF-His. These two viral vectors were purified by a 2-step chromatographic step.    for 60 min at 37°C. After the slices or cells were washed three times in PBS, the fluorescence was observed using fluorescence microscopy.

Western Blot.
After treating under various conditions, the tissues or cells were harvested, the total proteins were extracted, and the total protein concentrations were detected by BCA Protein Assay Kit (Beyotime Biotechnology, Nantong, China). Equivalent amounts of protein of each sample were electrophoresed on SDS-polyacrylamide gels and transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA). After the membranes were blocked for 1 h at RT, primary antibodies (in TBST-5% BSA) against SNCA, LAMP-2A, Hsc70, LC3, Beclin-1, P62, and Nrf2 (anti-SNCA, anti-LAMP-2A, anti-Hsc70, and anti-Nrf2 antibodies are obtained from Abcam, MA, USA; anti-LC3, anti-Beclin-1, and anti-P62 antibodies are obtained from Cell Signaling Technology, Beverly, USA) were added and incubated overnight at 4°C. After being washed 3 times in TBST, the membranes were incubated with HRP-conjugated secondary antibodies (KPL, Gaithersburg, MD, USA) for 1 h at 37°C. An ECL kit (Millipore, Bedford, MA, USA) was used to visualize membrane immunoreactivity.
5.9. Coimmunoprecipitation. The cell samples were lysed in lysis buffer (Beyotime Biotechnology, Nantong, China) and centrifuged for 15 min at 16,000 g. The supernatant was incubated with anti-SNCA antibody and Protein A/G-agarose (Beyotime Biotechnology, Nantong, China) overnight at 4°C. After extensive washing with lysis buffer, the bound proteins were eluted from the beads by boiling in loading buffer and were subjected to western blot analyses.
5.11. Statistical Analysis. Data visualization and analysis were performed with GraphPad Prism 8 (GraphPad Software Inc., La Jolla, CA, USA). And a one-way analysis of variance (ANOVA) followed either student's t-test by was used to compare each experiment. For all statistical analyses, P < 0:05 was considered significant.

Data Availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethical Approval
All animal research was approved by the Ethics Committee of Shanghai Tongji Hospital, Tongji University School of Medicine. All efforts were made to minimize the number of animals used and to minimize animal suffering.

Conflicts of Interest
The authors have declared that no conflicts of interest.