SHP2-mediated mitophagy boosted by lovastatin in neuronal cells alleviates parkinsonism in mice

Dear Editor, The loss of PINK1/Parkin-dependent mitochondrial clearance causes loss of dopaminergic neurons in the substantia nigra and contributes to the pathogenesis of Parkinson’s disease (PD). Several kinases were reported to regulate the ubiquitin E3 ligase activity of Parkin through phosphorylation but the involvement of protein tyrosine phosphatase (PTPase) for Parkin activity remains elusive. Although the roles of Src homology 2 domain-containing tyrosine phosphatase-2 (SHP2) in development, hematopoiesis and cancer immunology have been intensively reported, knowledge of regulation and function of SHP2 in neuronal diseases remains scant. We previously showed that SHP2 maintains mitochondrial homeostasis through dephosphorylating ANT1 at Tyr-191 during NLRP3 inflammasome activation in macrophages. This previous study prompted us to investigate whether SHP2 regulates mitophagy and mitochondrial quality in neurons and, if so, whether targeting SHP2 could be a novel strategy for neuronal protection in PD. As shown in Fig. 1a, b and Supplementary Fig. 1a–b, CCCPinduced mitochondria ubiquitination, reduction of mitochondrial mass as well as TOM20 ubiquitination and degradation were attenuated by SHP2 knockdown. Mitophagic flux examined by mtKeima also suggests that SHP2 positively regulates mitophagy (Fig. 1c, d, Supplementary Fig. 1c–f). Next, the mitochondrial translocation of Parkin and TOM20 degradation was remarkably decreased after SHP2 knockdown (Fig. 1e, Supplementary Fig. 2a–b). Parkin ubiquitination induced by CCCP treatment was also significantly inhibited by SHP2 knockdown and augmented after SHP2 overexpression (Fig. 1f, Supplementary Fig. 2c). Coimmunoprecipitation assay showed that both endogenous SHP2 (Fig. 1g) and exogenously SHP2 interacted with Parkin (Supplementary Fig. 3a). SHP2 and Parkin colocalization in the mitochondria was validated by Structured Illumination Microscopy (SIM) and immunoblot of mitochondrial and cytosolic fractions (Supplementary Fig. 3b–d). The PTP domain of SHP2 was shown to interact with Parkin (Supplementary Fig. 3e). Moreover, the interaction of SHP2 and Parkin as well as its involvement in mitophagy was also confirmed in primary neuron cells (Supplementary Fig. 4). These findings demonstrate that SHP2-Parkin interaction is required for Parkin-mediated mitophagy. Since the major activity of SHP2 relied on its PTPase activity, a PTPase gain-of-function SHP2 mutant (SHP2-D61A) and a loss-offunction SHP2 mutant (SHP2-C459S) were overexpressed in HeLa cells together with EGFP-Parkin. Mitochondrial translocation of LC3B was remarkably suppressed by SHP2-C459S and promoted by SHP2-D61A (Supplementary Fig. 5a). TOM20 clearance was evidently decreased by the overexpression of SHP2-D61A compared to vector control or SHP2-C459S (Fig. 1h, Supplementary Fig. 5b). These results suggest that the PTPase activity of SHP2 is indispensable for mitophagy regulation. Since it functions as a PTPase, SHP2 might regulate Parkin activity through the Tyr dephosphorylation of Parkin. Consistent with a previous report, the Ser phosphorylation of Parkin was enhanced upon CCCP treatment. However, in contrast, the Tyr phosphorylation of Parkin was reduced (Supplementary Fig. 5c). Noticeably, SHP2 knockdown abolished these changes in Ser and Tyr phosphorylation (Supplementary Fig. 5d). By using the purified SHP2 and Parkin, we confirmed that SHP2 could directly bind and dephosphorylate tyrosine of Parkin (Supplementary Fig. 5e–h). Next, we wondered whether SHP2 could be a potential target to boost Parkin-mediated mitophagy. A SHP2 enzyme activity screen system was employed for identifying compounds that could promote the catalytic activity of SHP2 (Supplementary Fig. 6a). A lactone ring structure-containing compound library (Chengdu, Biopurify) was screened and lovastatin was found to be able to elevate SHP2 activity both in a cell-free system and in cells (Fig. 1i, Supplementary Fig. 6b–c). A surface plasmon resonance (SPR) assay confirmed their interaction (KD= 36 μM, Fig. 1j), which was further demonstrated by the increased thermal stabilization of SHP2 (Supplementary Fig. 6d–e). To examine whether lovastatin could protect mitochondria in neurons, rotenone was utilized to mimic the pathological conditions of mitochondria in PD. Lovastatin dose-dependently increased cell viability in rotenone-treated SH-SY5Y cells (Supplementary Fig. 7a). ROS generation and mitochondrial membrane potential collapse were also suppressed by lovastatin (Supplementary Fig. 7b–c). Elevated mitophagy levels in the cells were also evidenced by red fluorescence from mt-Keima in cells treated with lovastatin (Supplementary Fig. 7d). Finally, as shown in Fig. 1k for the transmission electron microscopy (TEM) images, rotenone treatment led to significant mitochondria swollen and cristae disruption. In the lovastatin-treated group, the damaged mitochondria localized near a lysosome in the autophagolysosome and were surrounded by a double membrane, suggesting that damaged mitochondrion was removed via mitophagy. To clarify the relationship between SHP2 enzyme activity and the ability of lovastatin to promote mitophagy, SHP2 localization after lovastatin treatment was examined. SHP2 as well as Parkin translocation to mitochondria was increased after lovastatin treatment (Supplementary Fig. 8a–c). Furthermore, lovastatin treatment triggered mitochondrial protein degradation as well as Parkin ubiquitination (Supplementary Fig. 9a–b). A significant interaction between SHP2 and Parkin was observed after lovastatin treatment (Fig. 1l, Supplementary Fig. 9c). However, its ability to promote mitochondrial protein clearance and the neuroprotective effect of lovastatin was reversed after SHP2 knockdown (Supplementary Fig. 9d–e). To assess the protective effect of lovastatin on PD, a 1-methyl-4phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced PD murine model (MPTP-PD) was employed. Lovastatin and levodopa were administered as depicted in Supplementary Fig. 10a-b, and a


Cell culture
Human SH-SY5Y, HeLa and HEK293T cells were obtained from Shanghai Institute of Cell Biology (Shanghai, China). SH-SY5Y cells were cultured in RPMI 1640 and HeLa and HEK293T cells were cultured in DMEM, supplemented with 10% fetal bovine serum.
The assay for catalytic activity of SHP2 The catalytic activity of SHP2 was monitored using the surrogate substrate DiFMUP in a prompt fluorescence assay format 3  SH-SY5Y cells were incubated with or without compound for 2 h, then the cells were collected and subjected to Cellular Thermal Shift Assay (CETSA) assay 4,5 . Briefly, incubated cells were equally divided into 10 parts, each parts got heated for 3 min under different temperature (43, 46,  49, 52, 55, 58, 61, 64, 67, 70°C), then the heated cells were put into -80°C for 12 h, put into room temperature for 5 min, then repeated one more time. After that, cell lysates were extracted by centrifuged at 20,000 g, 20 min. Level of SHP2 was detected by western blot.

Molecular docking
The X-ray crystal structure of SHP2 was downloaded from Protein Databank (PDB, http://www.rcsb.org/). Subsequently, based on the active binding pocket, molecular docking of candidate compounds was performed by Accelrys Discovery Studio (version 3.5, Accelrys, San Diego, CA, USA). Candidate compounds were constructed using Discovery studio with CHARMm force field parameters. Molecular dockings were performed by CDOCKER protocol.
The other parameters were fixed as default values.
Surface plasmon resonance (SPR) SPR assay was performed using the Biacore T200 as follows. Recombinant human SHP2 protein was immobilized on a Biacore CM5 sensor chip via the primary amine groups. The compounds were flowed at a rate of 30 μl/min for 60 s to allow for association, followed by 150 s for dissociation over immobilized protein in PBS/5% DMSO running buffer (1.05×PBS, 0.5% P20 surfactant, 5% DMSO, pH 7.4). Andrographolide was tested for binding at 1.5 μM to 80 μM.
Normalization of the data involved transformation of the y-axis such that the theoretical maximum amount of binding for a 1:1 interaction with the protein surface corresponded to a sensor response of 100 relative units (RU).
Immunofluorescence assay Adherent cells on coverslips were fixed in 4% paraformaldehyde in PBS for 20 min at room temperature. Cells then permeabilized with 0.5% Triton X-100 in PBS for 20 min. After blocking with 5% BSA for 1 h, cells were incubated with primary antibodies overnight at 4 °C followed by incubation with Alexa Fluor-conjugated secondary antibody for 2 h and 1 μg/ml DAPI for 5 min.
The slides were then mounted with ProLong Gold (Life Technologies) and imaged with a Leica TCS SP8 fluorescent confocal microscope.

Co-immunoprecipitation assay
Proteins from cells were incubated with 1 μg of appropriate antibody and precipitated with protein A/G-agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA). The immunoprecipitated proteins were separated by SDS-PAGE and immunoblot analysis was performed with the indicated antibodies as described previously.

Western blot
The protein lysates were separated by 10% SDS-PAGE and subsequently electrotransferred onto a polyvinylidene difluoride membrane (Millipore, Bedford, MA). The membrane was blocked with 5% nonfat milk for 1 h at room temperature. The blocked membrane was incubated with the indicated primary Abs, and then with a horseradish peroxidase-conjugated secondary Ab. Protein bands were visualized using Western blotting detection system according to the manufacturer's instructions (Cell Signaling Technology, MA).
(Nanjing, China). The TH expression neuron-specific knockout mice (SHP2 TH-/-) were generated by crossing SHP2 flox/flox mice with TH-Cre transgenic mice. The animals were maintained with free access to pellet food and water in plastic cages at 21 ± 2 °C and kept on a 12 h light-dark cycle.
All mice are in C57BL/6 background and are harbored in the specific pathogen-free facility in Nanjing University. Eight-week-old female cSHP2-KO mice and WT littermates were used.
Animal welfare and experimental procedures were carried out strictly in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, USA) and the related ethical regulations of our university. All efforts were made to minimize animals' suffering and to reduce the number of animals used.
Adult male C57BL/6 mice were randomly divided into five groups (10 mice per group). In brief, control group was treated with PBS (vehicle). Mice were intraperitoneally administered with MPTP (dissolved in PBS) in a final concentration of 25 mg/kg daily for five consecutive days. One hour after MPTP injection, mice were orally administrated 12.5 or 25 mg/kg lovastatin once a day for 10 days. Levodopa was orally administrated (75 mg/kg, once a day for 10 days) as a positive drug. After the final treatment, behavioural procedures including pole test, beam hang test, rotarod task, and forced swimming tests were conducted to assess behavioural alternations as described before 6 . All the evaluators were blinded to the groups.
Immunohistochemical analysis, TUNEL assay and Nissl staining Immunohistochemical analysis was performed on paraffin-embedded brain tissue sections (3 μm).
Briefly, the sections were deparaffinised, rehydrated, and washed in PBS, and then treated with 2% hydrogen peroxide, blocked with 3% goat serum, and incubated with anti-tyrosine hydroxylase TH (TH) (1:500) overnight at 4 °C. The slides were then processed with GTVisinTM™ antimouse/anti-rabbit immunohistochemical analysis KIT according to the manufacturer's instructions.
TUNEL and Nissl staining were performed to examine apoptosis and neuron damage, respectively, according to the manual.

Isolation and culture of the primary neuron cells from mice
Ventral mesencephalon was dissected from E12.5 C57BL/6 embryos in cold 0.01μM PBS on ice, and DMEM containing 0.25% trypsin was added and incubated for 10 min in a 37 °C water bath.
The tissue pieces were dissociated by a 1 ml pipette tip through pipetting five to seven times. The