α-Synuclein disrupts microglial autophagy through STAT1-dependent suppression of Ulk1 transcription

Abstract

Background

Autophagy dysfunction in glial cells is implicated in the pathogenesis of Parkinson’s disease (PD). The previous study reported that α-synuclein (α-Syn) disrupted autophagy in cultured microglia. However, the mechanism of microglial autophagy dysregulation is poorly understood.

Methods

Two α-Syn-based PD models were generated via AAV-mediated α-Syn delivery into the mouse substantia nigra and striatal α-Syn preformed fibril (PFF) injection. The levels of microglial UNC-51-like kinase 1 (Ulk1) and other autophagy-related genes in vitro and in PD mice, as well as in the peripheral blood mononuclear cells of PD patients and healthy controls, were determined via quantitative PCR, western blotting and immunostaining. The regulatory effect of signal transducer and activator of transcription 1 (STAT1) on Ulk1 transcription was determined via a luciferase reporter assay and other biochemical studies and was verified through Stat1 knockdown or overexpression. The effect of α-Syn on glial STAT1 activation was assessed by immunohistochemistry and western blotting. Changes in microglial status, proinflammatory molecule expression and dopaminergic neuron loss in the nigrostriatum of PD and control mice following microglial Stat1 conditional knockout (cKO) or treatment with the ULK1 activator BL-918 were evaluated by immunostaining and western blotting. Motor behaviors were determined via open field tests, rotarod tests and balance beam crossing.

Results

The transcription of microglial ULK1, a kinase that controls autophagy initiation, decreased in both in vitro and in vivo PD mouse models. STAT1 plays a critical role in suppressing Ulk1 transcription. Specifically, Stat1 overexpression downregulated Ulk1 transcription, while Stat1 knockdown increased ULK1 expression, along with an increase in LC3II and a decrease in the SQSTM1/p62 protein. α-Syn PFF caused toll-like receptor 4-dependent activation of STAT1 in microglia. Ablation of Stat1 alleviated the decrease in microglial ULK1 expression and disruption of autophagy caused by α-Syn PFF. Importantly, the ULK1 activator BL-918 and microglial Stat1 cKO attenuated neuroinflammation, dopaminergic neuronal damage and motor defects in PD models.

Conclusions

These findings reveal a novel mechanism by which α-Syn impairs microglial autophagy and indicate that targeting STAT1 or ULK1 may be a therapeutic strategy for PD.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12974-024-03268-4.

Introduction

Parkinson’s disease (PD) is one of the most common neurodegenerative disorders, affecting approximately 2% of the aged population [1]. It is a movement disorder characterized by resting tremor, muscle stiffness and bradykinesia. In fact, patients with PD often suffer from various nonmotor symptoms, such as depression, constipation and sleep dysfunction, during the prodromal phase. Both nonmotor and motor symptoms are caused mainly by dopamine (DA) depletion due to the loss of DAergic neurons in the substantia nigra (SN). Curative treatment for PD is still lacking, although patients benefit greatly from medicines such as levodopa and DA receptor agonists [2]. Therefore, it is imperative to gain deep insight into the cellular and molecular pathogenesis of PD.

Mounting evidence suggests a role for autophagy in PD [3–5]. Autophagy is a process that eliminates damaged organelles and misfolded proteins in cells. Neuronal autophagy deficiency leads to the buildup of protein aggregates, which are associated with neurodegenerative disorders, including PD [3, 6]. Lewy bodies, a pathological feature of PD, mainly consist of α-synuclein (α-Syn) aggregates, which trigger detrimental effects on synaptic function and neurodegeneration [7, 8]. The cell-to-cell propagation of α-Syn via a prion-like paradigm was identified in the past decade [9, 10]. Notably, brain-resident immune cells—microgliacan respond to α-Syn and induce neuroinflammation [11–13], which renders the surrounding microenvironment vulnerable to α-Syn spreading and neurotoxicity. For example, Choi et al. reported that microglia can engulf and degrade α-Syn via selective autophagy and demonstrated that disruption of microglial autophagy-related gene 7 (Atg7) promoted α-Syn accumulation and neurodegeneration [11]. Qin Y et al. reported that conditional knockout (cKO) of microglial Atg5 facilitated NLRP3 inflammasome activation and exacerbated neurotoxin-induced DAergic neuron loss in mice [14]. Tu et al.  previously reported that neuron-released α-Syn disrupted microglial autophagy, which contributed to neuroinflammation and neurodegeneration in an α-Syn-based PD mouse model [12]. These findings indicate a critical function of microglial autophagy in health and suggest that disruption of microglial autophagy may contribute to PD pathogenesis. However, the mechanism that underlies microglial autophagy dysregulation in PD remains to be defined.

In this study, we showed decreased transcription and expression of UNC-51-like kinase 1 (ULK1), a serine/threonine kinase that regulates autophagy initiation, in α-Syn fibril (PFF)-stimulated microglia and two α-Syn-based PD mouse models. We further demonstrated that microglial Ulk1 transcription is suppressed by the transcription factor signal transducer and activator of transcription 1 (STAT1) in α-Syn PFF-stimulated microglia. More importantly, we demonstrated that the ULK1 activator and microglial Stat1 ablation attenuated neuroinflammatory responses and mitigated DAergic neuron damage in PD model mice.

Materials and methods

Cell culture and treatment

Primary microglial cultures were prepared as previously reported [15]. In brief, cortical tissues were dissected from neonatal pups (P1–P3), digested and filtered through a 70-µm strainer. After centrifugation, the pellets were suspended in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 1% penicillin/streptomycin and 10% fetal bovine serum (FBS; Gibco, Gaithersburg, MD, USA). The cells were plated into T75 flasks, and the medium was changed every other day. Upon reaching confluency, microglia were harvested by shaking the flasks orbitally at 180 rpm for 2 h and seeded into culture dishes. After attachment for approximately 2 h, the microglia were gently rinsed with prewarmed phosphate-buffered saline (PBS) and cultured in fresh medium for further experiments. BV2 cells, MEFs, and HEK293 cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin in an incubator with 5% CO₂. For in vitro studies, microglia were treated with α-Syn PFF or lipopolysaccharide (L2880, Sigma, USA). The ULK1 activator BL-918 (provided by Prof. Liang Ou-Yang, Sichuan University) and the Toll-like receptor 4 (TLR4) inhibitor TAK-242 (HY-11109, MCE, USA) were added 30 min before α-syn PFF or LPS treatment.

Quantitative polymerase chain reaction (qPCR)

Total RNA was extracted and reverse transcribed via a cDNA synthesis kit (R323; Vazyme, China) according to the manufacturer’s instructions. qPCR was performed via SYBR Green PCR Master Mix (Q711, Vazyme, China) with the specific primers listed in Table 1 via the 7500 Real-Time PCR system (Applied Biosystems). The mRNA levels of the target genes were calculated via the 2^−△△Ct method and normalized to those of the housekeeping gene Gapdh.

Table 1.

Sequences of the primers used for the qPCR analysis of gene expression

Gene	Species	Primer	Sequence (5’-3’)
Sqstm1	mouse	Forward	GAGGCACCCCGAAACATGG
		Reverse	ACTTATAGCGAGTTCCCACCA
Tfeb	mouse	Forward	TGTGATTGTCTTTCTTCTGCCG
		Reverse	GACTCAGAAGCGAGAGCTAACA
Atg5	mouse	Forward	AGCCAGGTGATGATTCACGG
		Reverse	GGCTGGGGGACAATGCTAA
Atg7	mouse	Forward	GTTCGCCCCCTTTAATAGTGC
		Reverse	TGAACTCCAACGTCAAGCGG
Ulk1	mouse	Forward	AAGTTCGAGTTCTCTCGCAAG
		Reverse	ACCTCCAGGTCGTGCTTCT
ULK1	human	Forward	GGCAAGTTCGAGTTCTCCCG
		Reverse	CGACCTCCAAATCGTGCTTCT
Lamp1	mouse	Forward	CCATTCGCAGTCTCGTAGGTG
		Reverse	CAGCACTCTTTGAGGTGAAAAAC
Atg12	mouse	Forward	TGAATCAGTCCTTTGCCCCT
		Reverse	CATGCCTGGGATTTGCAGT
Atg16l	mouse	Forward	TTGTCCTTCTGCTGCATTTG
		Reverse	ACATGATGGTGCGTGGAAT
Becn 1	mouse	Forward	ATGGAGGGGTCTAAGGCGTC
		Reverse	TGGGCTGTGGTAAGTAATGGA
Il-1β	mouse	Forward	TGGAAAAGCGGTTTGTCTTC
		Reverse	TACCAGTTGGGGAACTCTGC
Gapdh	mouse	Forward	GAAGGTCGGTGTGAACGGAT
		Reverse	AATCTCCACTTTGCCACTGC
GAPDH	human	Forward	GGAGCGAGATCCCTCCAAAAT
		Reverse	GGCTGTTGTCATACTTCTCATGG
Atg13	mouse	Forward	CCAGGCTCGACTTGGAGAAAA
		Reverse	AGATTTCCACACACATAGATCGC
Rb1 cc1	mouse	Forward	GACACTGAGCTAACTGTGCAA
		Reverse	GCGCTGTAAGTACACACTCTTC
Atg101	mouse	Forward	ATGAACTGTCGATCAGAAGTGC
		Reverse	CCTATGGAGTACGTGCCCT

Western blot and antibodies

Cells and tissues were lysed in RIPA buffer supplemented with protease and phosphatase inhibitors as previously described [12]. Protein concentrations were determined with a BCA assay kit (20201ES90, Yeasen, China). Proteins were separated by SDS‒PAGE and transferred onto polyvinylidene difluoride membranes (ISEQ00010, Millipore, USA). The membranes were blocked with 5% (w/v) nonfat milk and incubated with primary antibodies against the proteins of interest at 4 °C overnight. Next, the membranes were washed and incubated with an appropriate secondary antibody (1:5000, 115-035-003/111-035-003, Jackson ImmunoResearch Lab, USA) at room temperature (RT) for 1 h. The protein bands were visualized via the ECL reagent (P10300, NCM Biotech, China) on a ChemiDoc XRS + System (Bio-Rad, USA). Densitometric analysis was performed via ImageJ software.

The primary antibodies used were as follows: anti-ULK1 (1:500, 8054 S, CST, USA), anti-LC3 (1:2000, NB100–2220, Novus, USA), anti-tyrosine hydroxylase (anti-TH, 1:1000, T1299, USA), anti-SQSTM1/p62 (1:1000, P0067, Sigma, USA), anti-p-STAT1^Y701 (1:500, AP0054, ABclonal, China), anti-STAT1 (1:500, 14994, CST, USA), anti-α-Syn (1:500, AB138501, Abcam, USA), anti-ACTB (1:5000, A3854, Sigma, USA) and anti-GAPDH (1:5000, 60004-1-IG, Proteintech, China).

α‑Syn PFF preparation

α-Syn PFFs were generated as previously described [16]. In brief, Escherichia coli BL21 (DE3) cells were transformed with the pET-28b (+) plasmid encoding human α-Syn, incubated in LB medium and treated with 1 mM isopropyl β-D-thiogalactoside to induce α-Syn expression. Bacteria were harvested by centrifugation at 5000 rpm for 5 min and resuspended in Tris buffer (150 mM Tris at pH 8.0, 10 mM EDTA, 150 mM NaCl), followed by sonication and centrifugation at 13,000 rpm for 15 min. The resulting supernatant was collected for α-Syn purification. Endotoxin was depleted via a ToxinEraser kit (L00338, GenScript, China). α-Syn monomers were diluted to 5 mg/ml and shaken at 1,000 rpm in an Eppendorf Thermomixer for 7 days to facilitate α-Syn aggregate formation. For the experiments, α-Syn aggregates were diluted to the desired concentration and sonicated for 30 s (0.5 s on/off) at 10% power immediately prior to use. α-Syn PFF was visualized via transmission electron microscopy. The seeding capability of α-yn PFFs was validated by testing α-syn phosphorylation at S129 (pS129-α-Syn) in primary neurons (Supplementary Fig. 1e).

Animals and stereotaxic surgery

C57BL/6 mice (6–8 weeks old, male) were purchased from Shanghai Laboratory Animal Center (Shanghai, China). To generate mice with microglial Stat1 ablation, Cx3cr1^CreERT2 mice (kindly gifted by Prof. Jian Cheng, Soochow University) were crossed with Stat1^fl/fl mice (GemPharmatech, Nanjing, China), followed by genotyping. All 6-week-old male Stat1^fl/fl; Cx3cr1^CreERT2/+ (described as Stat1^cKO hereafter) mice and their littermates (Stat1^fl/fl) were intraperitoneally injected with 75 mg/kg/day tamoxifen (Sigma, T5648) for 5 days to induce Cre expression and recombination. Tlr4^−/− mice were obtained from the Model Animal Research Center (MARC) of Nanjing University. The mice were maintained at the Laboratory Animal Center of Suzhou Medical College at Soochow University and housed in a specific pathogen-free facility under a 12 h light‒dark cycle at 21 ± 2 °C, with food and tap water ad libitum. All procedures involving mice were performed in accordance with the Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee of Soochow University (ECSU-201600014).

α-Syn-based PD models were established via two different protocols: recombinant adeno-associated virus (AAV)-mediated α-Syn overexpression (OE) and PFF injection into the striatum via a stereotaxic apparatus. For surgery, all the mice were anesthetized with 2.5% avertin. To induce α-Syn OE, 1.5 µl of AAV2/9-hα-Syn (1.26E + 12 v.g., Obio, Shanghai) or AAV2/9-Vector was bilaterally delivered into the SN, with coordinates at anteroposterior (AP) -3.0 mm, mediolateral (ML) ± 1.25 mm, and dorsoventral (DV) + 4.5 mm relative to the bregma as previously described [12]. The needle was left in place for 5 min before being retracted slowly. To model α-Syn spreading in PD, α-Syn PFF (5 µg) was bilaterally injected into the striatum at the following coordinates: AP + 0.2 mm, ML ± 2.0 mm, and DV + 3.2 mm, as previously described [17, 18]. For controls, an equal volume of PBS was injected.

Microglia isolation from adult mice

Microglia were acutely isolated from adult mice as previously reported [12]. Briefly, the mice were anesthetized and perfused with chilled PBS. The brains were dissected and homogenized in serum-free DMEM. The homogenates were filtered through a 70-µm strainer followed by centrifugation at 1000×g for 10 min at 4 °C. Microglia were then isolated via density gradient centrifugation. Specifically, the pellets were suspended in 22% Percoll and centrifuged to remove myelin fragments. The resulting pellets were resuspended in 37% Percoll solution. Next, a density gradient was added to a 15-ml centrifuge tube and centrifuged at 1000×g for 40 min with no brake. Microglia were carefully collected at 70–37% confluence and resuspended in 3 volumes of Hank’s balanced salt solution, followed by centrifugation at 800×g for 10 min at 4 °C. The purity of the isolated microglia (gated on CD11b^HighCD45^intermediate and then CX3CR1⁺) was assessed via flow cytometry, which revealed that the purity reached > 90% (Supplementary Fig. 2).

Immunostaining

The mice were anesthetized and sequentially perfused with saline and 4% paraformaldehyde (PFA) for fixation. The brains were dehydrated in 10-30% sucrose solution and cut into 18 μm thick sections via a cryostat (Leica, German). For immunofluorescence staining, the sections were blocked in 5% BSA with 0.25% Triton X-100 for 1 h at RT and incubated with primary antibodies overnight at 4 °C. After being rinsed three times in PBS, the sections were incubated with Alexa Fluor 488- or 555-conjugated secondary antibodies (1:500, Thermo Fisher) in the dark for 1 h. Next, the sections were mounted onto slides and stained with DAPI. Images were taken under a confocal microscope (LSM700; Carl Zeiss, Germany). For immunohistochemical staining, the sections were soaked in 3% hydrogen peroxide for 10 min to reduce endogenous peroxidase activity before blocking. The sections were finally incubated with a horseradish peroxidase-conjugated secondary antibody. Staining was detected via a DAB kit (GK500705, Gene Tech). The following primary antibodies were used: anti-ULK1 (1:200, 8054 S, CST, USA), anti-IBA1 (1:1000, 019–19741, Wako, Japan), anti-NLRP3 (1:200, 15101, CST, USA), anti-TH (1:1000, T1299, Sigma, USA), anti-p-STAT1^Y701 (1:200, AP0054, ABclonal, China), and anti-GFAP (1:1000, z033429-2, Dako, Denmark).

For cell staining, cells grown on coverslips were fixed with 4% PFA for 10 min, rinsed twice with PBS, and then permeabilized in 0.25% Triton X-100/PBS for 5 min. Next, the cells were blocked in PBS with 5% FBS for 10 min, incubated with anti-p-STAT1^Y701 and anti-IBA1 antibodies at 4 °C overnight, and then with Alexa Fluor 488- and 555-conjugated secondary antibodies for 1 h at RT. Coverslips were mounted with Fluoroshield medium containing DAPI. Images were acquired via confocal microscopy.

TH⁺ neurons in the SNpc were quantified by an experienced researcher who was blinded to the experimental groups. After adjusting the threshold and carefully marking the borders of the SN, TH⁺ neurons with only a visible nucleus were counted by ImageJ via the Cell Counter plugin. Every fifth section throughout the SN was collected, and at least four sections were incorporated into the counting system for each mouse. The total number of TH⁺ neurons in the SN of the hemisphere was estimated by multiplying the counted number by five since every fifth section was used for analysis. For the mean fluorescent intensity (MFI) analysis, a threshold was selected with image/adjustment tools to achieve the desired range of intensity values for each experiment. Once determined, this threshold was applied to all the images. After exclusion of the background, the selected area in the signal intensity range of the threshold was measured via the measurement option under the Analyze/Measure menu. The Manders’ overlap coefficient was used to measure the colocalization of microglia with ULK1, NLRP3 or p-STAT1^Y701. Microglia were counted with the Analyze Particles function. At least three sections per mouse were included for analysis.

ULK1 mRNA quantification in peripheral blood mononuclear cells (PBMCs)

To quantify the ULK1 mRNA level in PBMCs, 5 ml of blood was collected from each individual via the median cubital vein and stored in an EDTA anticoagulant tube (BD Bioscience, USA) at -80 °C within 1 h after collection. Total RNA was extracted via a PAX-gene Blood RNA kit and reverse transcribed into cDNA. qPCR was performed as described above, using the primers for the human ULK1 gene listed in Table 1.

The study was approved by the Ethics Committee of the Second Affiliated Hospital of Soochow University in China, and written informed consent was obtained from all participants (IK-2018-061-01). All participants underwent a comprehensive general and neurologic examination. Patients with PD were referred by senior neurologists, and disease severity was assessed via the United Parkinson disease rating scale (UPDRS). Healthy controls without any signs of neurological disease were recruited from routine physical examinations.

RNA interference and plasmid transfection

To knockdown the Ulk1 gene, BV2 cells were transfected with Ulk1 siRNA (sense: 5′-GGGUAGUAAUGACACCACCUCGGAA-3′; antisense: 5′-UUCCGAGGUGGUGUCAUUACUACCC-3′) or scrambled siRNA (GenePharma, Shanghai, China) as a control via Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA, USA). The plasmids used for Stat1 OE (Flag-Stat1) and knockdown (sh-Stat1) were a gift from Professor Hui Zheng at Soochow University (Suzhou, China). Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) was used for plasmid transfection. The genetic editing efficiency was verified via western blotting.

Open field test

Locomotor activity was evaluated via an open field test. The apparatus consisted of a white chamber (40 × 40 × 40 cm). The mice were allowed to adapt to the apparatus for one hour before the test. Each mouse was placed in the center and allowed to explore the arena for 10 min. The locomotor trajectories were automatically recorded with a video camera and analyzed with an ANY-Maze system (Stoelting, USA). The total distance traveled was measured to evaluate locomotor activity. The arena was cleaned with 70% ethanol after each trial.

Rotarod test

The rotarod test was used to assess motor coordination. Before testing, the mice were trained at a constant speed of 20 rpm for three sessions on three consecutive days. Each session included two separate trials. During the testing, the rotarod speed was increased from 0 to 40 rpm within 5 min. The latency to fall off the rod was recorded automatically for each mouse, and the average latency of three trials, with an interval of 15 min, was calculated for further analyses.

Balance beam test

The balance beam test was used to assess motor function. The apparatus consisted of a beam (1 cm wide and 100 cm long), which was placed 60 cm above the floor with a black shelter at one end. The mice were trained to cross a thicker beam (3 cm wide) three days before formal testing. On the testing day, each mouse was placed on one end. The time to cross the beam was recorded. Three consecutive trials with 30-min intervals were performed for each mouse to minimize individual variance. The mean time of three trials was calculated for analyses.

Generation of Stat1-KO MEFs

The plasmids for Streptococcus pyogenes Cas9 and sgRNA expression were a gift from Guanghui Wang (Soochow University, Suzhou, China). MEFs were cotransfected with the Cas9 expression plasmid and Stat1 sgRNA (5’-TACTACAGAGCAAACGCTGG-3’; 5’-GGAGAGTGTCGACCTGCTGG-3’) plasmid via Lipofectamine 2000. After transfection, MEFs were cultured under blasticidin (2 µg/ml) and puromycin (2 µg/ml) selection for two weeks. Single clones were picked for culture, and the KO efficiency was verified by western blotting.

Luciferase reporter gene assay

To assess Ulk1 promoter activity, we constructed a plasmid carrying a reporter (pGL3-Ulk1-luc) by cloning the Ulk1 promoter cDNA into a pGL3 enhancer vector with a luciferase reporter. The pGL3-Ulk1 mutant was obtained via site-directed mutagenesis via the wild-type pGL3-Ulk1 plasmid as a template with the following primer sequences: 5’-ccccgcagggaaacccaagatccgcggcac-3’; 5’-gtgccgcggatcttgggtttccctgcgggg-3’. The constructs were verified via DNA sequencing.

To evaluate the impact of STAT1 on Ulk1 promoter activity, HEK293T cells were cotransfected with pGL3-Ulk1-luc and a Flag-Stat1 or vector, along with a Renilla plasmid, to normalize the transfection efficiency. Twenty-four hours after transfection, dual-luciferase reporter assays were performed via a Vazyme kit (DL101-01; Nanjing, China). The Ulk1 promoter activity was expressed as the relative firefly luciferase activity over the Renilla luciferase activity.

Chromatin immunoprecipitation (ChIP)

The ChIP assay was carried out according to the protocol of the SimpleChIP^® Plus Sonication Chromatin IP Kit (56383, CST, USA). In brief, BV2 cells were fixed with 1% formaldehyde solution for 10 min to crosslink proteins to DNA. Glycine was added to quench excess formaldehyde. After nuclear isolation, the chromatin was sonicated with a Bioruptor Plus sonication device (Diagenode, Belgium). ChIP was performed using 2 µg of either anti-STAT1 (9172, CST) or rabbit IgG at 4°C overnight. The protein‒DNA complexes were precipitated with protein G magnetic beads, followed by elution. Next, the crosslinking was reversed, and the DNA was purified. The enrichment of the ULK1 promoter fragments was assessed via qPCR with the following primers: forward, 5’- CATCAAGTGCTCCAACATGC-3’ and reverse, 5’- GTTAGAGGGGAAGTTTGAGG-3’. The results are presented as the percentage of input.

Statistical analyses

All the results are presented as the means ± SEM of at least three independent experiments unless otherwise specified. The statistical significance was evaluated via GraphPad Prism 8 (San Diego, CA, USA). Student’s t tests were used to compare the differences between two groups. For multiple group comparisons, statistical significance was determined by one-way analysis of variance (ANOVA) followed by Tukey’s test, Dunnett’s post hoc test, or two-way ANOVA followed by Tukey’s post hoc analysis, as indicated in the figure legends. Differences were considered significant at P < 0.05.

Results

α-Syn PFF suppresses Ulk1 transcription and expression in microglia

Microglial autophagy dysregulation is associated with PD and other types of neurodegeneration [14, 19]. Tu et al. previously demonstrated that α-Syn disrupted autophagy activity in cultured microglia through TLR4-dependent p38 and Akt-mTOR signaling [12]. However, it remains unknown whether other mechanisms may be involved. To clarify this, we studied the mRNA levels of several autophagy-related genes via qPCR and revealed that the Ulk1, Atg5, Atg7 and Tfeb gene mRNA levels decreased whereas Sqstm1/p62 transcription increased in α-Syn PFF-treated microglia compared with those in PBS-treated controls (Fig. 1a). However, the mRNA levels of other tested genes, including Lamp1, Atg12, Atg16l and Becn1, as well as the ULK1 complex components Rb1 cc1 (FIP200 homolog in mouse), Atg101 and Atg13, remained unaltered. Given that ULK1 is essential for autophagy initiation, we focused on its expression and detected a time-dependent decrease in microglial Ulk1 mRNA levels after α-Syn PFF exposure (Fig. 1b). In line with these findings, the ULK1 protein level and the level of LC3II, a classic autophagy marker, were also decreased in α-Syn PFF-treated microglia (Fig. 1c). The dose-dependent effects of α-Syn PFF on ULK1 mRNA and protein levels were also tested. Exposure to 2.5 µg/ml α-Syn PFF resulted in a decrease in the ULK1 mRNA level but not its protein expression in primary microglia. Higher concentrations of PFF (5 µg/ml and 10 µg/ml) downregulated ULK1 at both the mRNA and protein levels (Supplemental Fig. 1f-g). Therefore, 5 µg/ml α-Syn PFF was used in subsequent studies for primary microglia treatment.

Fig. 1.

α-Syn inhibits ULK1 expression in microglia. (a) qPCR results of autophagy-related gene mRNA levels in α-Syn PFF (5 µg/ml, 6 h)-treated primary microglia (n = 4 independent experiments). (b) Time-dependent changes in microglial Ulk1 mRNA levels, n = 3. One-way ANOVA followed by Dunnett’s post hoc analysis. (c, d) Western blot and quantification of ULK1 and LC3 protein levels in PFF-treated microglia (c, n = 4) and SN lysates (d) from AAV-hα-Syn- or Vector-injected mice (n = 4 animals per group). (e) qPCR data of Ulk1 mRNA levels in microglia isolated from AAV-hα-Syn- or Vector-injected mice (n = 4 independent repeats). (f) Representative images of IBA1 (red) and ULK1 (green) doublestaining in the SN reticulum (SNr). Arrowheads indicate the colocalization of IBA1 and ULK1, and enlarged images of the inset are shown on the right. Scale bar, 20 μm. ULK1 intensity and Mander’s coefficient analysis are shown in g and h; n = 3 mice per group. (i) PBMC ULK1 mRNA levels in healthy controls (HCs) and patients with PD. The data are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. Unpaired Student’s t test except b

To verify these observations in vivo, virus-mediated α-Syn OE was established by injecting AAV-hα-Syn into the mouse SN as previously reported [12]. This modeling procedure resulted in remarkable damage to DAergic neurons and pS129-α-Syn accumulation, as well as motor deficits at 8 weeks postinjection, recapitulating the pathological features of PD. Therefore, we tested whether ULK1 protein expression was altered in this PD model. Western blot analysis revealed that the ULK1 and LC3II protein levels in the SN of AAV-hα-Syn-injected mice were lower than those in the AAV-vector group at 8 weeks post injection (Fig. 1d). Microglia were then isolated from PD mice and controls, and Ulk1 mRNA levels were studied via qPCR. Compared with those in the control group, the microglial Ulk1 mRNA levels in the AAV-hα-Syn-injected group were approximately 50% lower (Fig. 1e). Furthermore, we performed double immunostaining to study changes in the microglial ULK1 protein in the SN reticular region (SNr), given that this region has the highest density of microglia [20]. Compared with other regions, the SNr is particularly sensitive to proinflammatory stimulation, which may directly affect the DAergic neurons in the neighboring SN pars compacta (SNpc) [21, 22]. As shown in Fig. 1f-h, the intensity and colocalization of ULK1 with ionized calcium binding adapter molecule 1 (IBA1)-positive microglia in AAV-hα-Syn-injected mice were lower than those in control mice. This finding was consistent with the above qPCR data. We also validated whether ULK1 expression was altered in human PD patients by recruiting patients with PD (n = 26) and age-matched controls (n = 26). ULK1 mRNA levels in PBMCs were tested because of the lack of human brains. A lower ULK1 mRNA level was detected in PD patients than in healthy controls (Fig. 1i). These data indicate that decreased transcription of microglial ULK1 may be involved in PD.

ULK1 activator inhibits microglia-mediated neuroinflammation in hα-Syn-overexpressing mice

We further explored the consequences of a decrease in microglial ULK1 in PD. To knockdown the Ulk1 gene, the BV2 microglial cell line was used instead. We tested the mRNA levels of the proinflammatory cytokine interleukin-1β (IL-1β) in Ulk1- or scramble siRNA-transfected BV2 cells in the presence or absence of α-Syn PFF. The BV2 cell line appeared to be less sensitive than primary microglia. Therefore, 10 µg/ml α-Syn PFF was applied for BV2 treatment. Compared with scramble siRNA (si-NC), Ulk1 knockdown alone tended to increase Il-1β mRNA levels and markedly increased cytokine production in the presence of α-Syn PFF (Fig. 2a&b). Conversely, pretreatment with the ULK1 activator BL-918 at concentrations of 5 µM and 10 µM suppressed the increase in microglial Il-1β mRNA levels triggered by LPS, a classic inflammasome, whereas 2.5 µM BL-918 was less effective (Fig. 2c). Similar results were observed in α-Syn PFF-stimulated cells pretreated with BL-918 (Fig. 2d).

Fig. 2.

Modulation of ULK1 affects microglia-mediated neuroinflammation in vitro and in vivo. (a) Verification of Ulk1 knockdown in BV2 cells, n = 3. (b) qPCR data of Il-1β mRNA levels in scramble (si-NC)- or si-Ulk1-transfected BV2 cells treated with 10 µg/ml PFF for 6 h, n = 3. (c, d)Il-1β mRNA levels in PFF (c)- or LPS (500 ng/ml, d)-stimulated BV2 cells pretreated with various concentrations of BL-918; n = 3. (e)In vivo study design. The ULK1 activator BL-918 (40 mg/kg) was given orally every other day for seven weeks, starting one week after AAV-hα-Syn or vector injection. Behavioral and histological studies were performed at 8 weeks post injection. (f) Weekly changes in body weight postinjection (n = 8 animals per group). (g-i) Representative images (g) of IBA1 staining in the SNr. Quantification of IBA1 intensity (h) and soma diameter (i, n = 4 mice per group, at least 30 cells per mouse). (j-l) Representative images of IBA1 (green) and NLRP3 (red) double-staining. Scale bar, 20 μm. Arrowheads indicate IBA1 and NLRP3 colocalization. NLRP3 intensity quantification and Mander’s coefficient analysis are shown in k and l, n = 3 mice per group. The data are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. Unpaired Student’s t test for a. One-way ANOVA followed by Tukey’s post hoc analysis for other panels

Next, we performed an in vivo study to assess the effect of the ULK1 activator on PD mice. To achieve this goal, BL-918 was given orally at a dose of 40 mg/kg every other day for seven weeks, starting one week after AAV-hα-Syn injection (Fig. 2e). Motor performance and pathological changes were studied at the end of treatment. There was mild body weight loss in the AAV-hα-Syn-injected mice, which occurred at five weeks postsurgery but remained unchanged thereafter (Fig. 2f). No difference was observed between the BL-918-treated and untreated PD mice. Immunostaining revealed that BL-918 treatment inhibited microglial activation, as indicated by the reductions in IBA1 intensity and soma diameter, in the SN of AAV-hα-Syn-injected mice (Fig. 2g-i). The fluorescence intensity of NLRP3, an inflammasome protein that is predominantly localized to IBA1⁺ cells, also decreased in the BL-918-treated group (Fig. 2j-l). These data suggest that the ULK1 activator potently suppresses neuroinflammation and imply that the decrease in microglial ULK1 expression may be involved in PD-related neuroinflammation.

ULK1 activator alleviates motor dysfunction and DAergic neuron loss in hα-Syn-overexpressing mice

The effects of the ULK1 activator on DAergic neuron loss and motor behavior were also studied. The results of the open field and rotarod tests revealed impaired locomotor and motor coordination in AAV-hα-Syn-injected mice, which was alleviated by BL-918 treatment (Fig. 3a&b). Immunostaining with an antibody against TH, the rate-limiting enzyme for DA synthesis, revealed that the number of SNpc TH⁺ neurons was approximately 60% lower in the AAV-hα-Syn group than in the AAV-vector group, and this reduction was attenuated in the BL-918-treated group (Fig. 3c-e). The quantification of striatal TH intensities, which reflect dopaminergic terminals, revealed similar results. These findings were corroborated by western blotting (Fig. 3f-i). In addition, a significant decrease in the LC3II protein level was detected in the SN of AAV-hα-Syn-injected mice, and this decrease was attenuated by BL-918 administration. No significant difference in the SQSTM1/p62 protein level was detected among these groups. These data indicate a neuroprotective effect of the ULK1 activator on PD.

Fig. 3.

The ULK1 activator BL-918 alleviates DAergic neuron damage in hα-Syn-overexpressing mice. (a) Latency to fall off the rod during the rotarod test (n = 8 animals per group). (b) Total travel distance during the open field test (n = 7 animals for each group). (c-e) Representative images (c) and quantification of SNpc TH⁺ neurons (d) and striatal fiber densities (STRs, e), n = 3–4 mice per group, with at least four slices per mouse for analysis. (f-i) Western blot analysis of TH, LC3II and SQSTM1/p62 protein levels in SN lysates normalized to those of actin (g-i); n = 5 animals per group. The data are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, n.s., not significant. One-way ANOVA followed by Tukey’s post hoc analysis

STAT1 serves as a negative regulator of Ulk1 transcription

Several posttranslational modifications of ULK1 have been reported [23]. However, the regulation of Ulk1 transcription is poorly understood. Therefore, we explored the mechanisms of microglial ULK1 downregulation in a PD model. First, we screened the transcription factors for the Ulk1 gene via four transcription factor prediction websites, namely, hTFtarget, ENCODE, cor_GTEx and KnockTF. A common set of 11 transcription factors was found (Fig. 4a).

Fig. 4.

STAT1 is responsible for the suppression of Ulk1 transcription in α-Syn-treated microglia. (a) Venn diagram demonstrating the number of predicted transcription factors (TFs) of the Ulk1 gene via the hTFtarget, ENCODE, cor_GTEx and KnockTF datasets. A list of potential TFs that regulate Ulk1 transcription is shown in the box. (b, c) qPCR data of Ulk1 mRNA levels (b) and western blot analysis of various protein levels (c) in sh-Vec- or sh-Stat1 (Stat1 KD)-transfected BV2 cells, n = 4. (d) Western blot and quantification of ULK1 and STAT1 protein levels in Stat1 plasmid- or vector-transfected BV2 cells; n = 4. (e)Ulk1 mRNA levels in IFN-γ-treated microglia; n = 3 independent repeats. (f) The binding motif of STAT1 in the promoter of mouse Ulk1 was predicted via JASPAR. The WT and mutated STAT1 regulatory sequences are located 67 bp downstream of the transcription start site (TSS) (+ 1). (g) Dual luciferase assay for Ulk1 promoter activity. HEK293 cells were transfected with pGL3-WT-Ulk1-Luc or pGL3-Mut-Ulk1-Luc, along with pcDNA3.1 or pcDNA3.1-Stat1, n = 3 independent repeats. (h) ChIP assay for STAT1 enrichment in the Ulk1 promoter. Fold changes are presented, n = 3. (i)Ulk1 mRNA levels in PFF-treated sh-Vec- or sh-Stat1-transfected BV2 cells, n = 3. (j) Western blot analysis of STAT1 and p-STAT1^Y701 levels in PFF-treated microglia in the presence of TAK-242 (1 µM), n = 4. (k) p-STAT1^Y701 fluorescence and subcellular distribution in PFF-treated WT or Tlr4^−/− microglia. Scale bar, 10 μm. Nuclei were stained with DAPI. The data are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, n.s., not significant. One-way ANOVA followed by Tukey’s post hoc test was used for i and j, and Student’s t test was used for the other panels

STAT1 was selected as the transcription factor of interest given that it functions as a crucial regulator of inflammation [24]. To test whether STAT1 regulates Ulk1 transcription, we transfected short hairpin RNA (shRNA) targeting Stat1 into BV2 microglia and detected higher Ulk1 mRNA levels in Stat1-knockdown (KD) cells than in control cells (Fig. 4b). Western blot analysis confirmed the increase in ULK1 protein levels in Stat1-KD cells (Fig. 4c). In addition, a marked increase in LC3II and a decrease in SQSTM1/p62 protein levels were observed, indicating increased autophagic flux after Stat1 KD. Notably, increased Ulk1 mRNA and protein levels were also observed in a stable Stat1 knockout (KO) MEF line (Supplemental Fig. 3). Conversely, a decreased ULK1 protein level was detected in Stat1-OE cells (Fig. 4d). We also observed decreased Ulk1 mRNA levels in microglia challenged with interferon (IFN)-γ, a potent cytokine that activates STAT1 (Fig. 4e). Moreover, a putative binding sequence for STAT1 (CTTTTGTTTC) was identified in the mouse Ulk1 promoter (67 ~ 76 bp downstream of the transcription start site) according to predictions from the hTFtarget and JASPAR websites. We then constructed a luciferase reporter plasmid encoding the Ulk1 promoter and a corresponding mutant at the STAT1 binding motif (Fig. 4f). A dual-luciferase reporter assay revealed that Stat1 OE reduced the activity of the wild-type (WT) Ulk1 promoter but not the mutant promoter (Fig. 4g), suggesting that STAT1 represses Ulk1 transcription.

We also verified whether STAT1 was responsible for the ULK1 decrease in α-Syn PFF-exposed microglia. A ChIP study revealed greater enrichment of STAT1 at the Ulk1 promoter in α-Syn PFF-treated BV2 cells than in control cells (Fig. 4h). Moreover, the α-Syn PFF-induced decrease in Ulk1 mRNA levels was abolished in Stat1-KD cells (Fig. 4i), indicating that STAT1 plays an essential role in suppressing Ulk1 transcription in α-Syn PFF-exposed microglia. We also studied the role of TLR4 in STAT1 activation, given that neuron-released α-Syn disrupted microglial autophagic flux through TLR4 signaling [12]. Pretreatment with TAK-242, a TLR4 inhibitor, abolished the increase in STAT1 phosphorylation at tyrosine 701 (p-STAT1^Y701) in α-Syn PFF-challenged microglia (Fig. 4j). Genetic depletion of Tlr4 resulted in similar observations, which demonstrated that α-Syn PFF triggered an increase in p-STAT1^Y701 and its nuclear translocation in WT microglia but not in Tlr4^−/− cells (Fig. 4k), suggesting that TLR4 was required for STAT1 activation and thus Ulk1 suppression caused by α-Syn PFF in microglia.

STAT1 activation is responsible for microglial Ulk1 suppression in hα-Syn-overexpressing mice

The change in STAT1 activity was validated in an α-Syn-based PD model. Western blotting revealed increases in p-STAT1^Y701 and STAT1 levels, as well as the ratio of p-STAT1^Y701 to STAT1, in the SN at 8 weeks after AAV-hα-Syn injection compared with those in the AAV-vector group. In parallel, we observed decreased ULK1 protein levels, accompanied by decreased TH and increased α-Syn protein levels, in PD mice (Fig. 5a&b). Immunostaining revealed an increase in p-STAT1^Y701 intensity in the SNr of the AAV-hα-Syn-injected mice. Moreover, an obvious p-STAT1^Y701 reactivity and its colocalization with the nucleus were observed in IBA1-positive cells in the AAV-hα-Syn-injected mice (Fig. 5c-e). However, p-STAT1^Y701 intensity was rarely detected in the nuclei of GFAP (an astrocyte marker)- positive cells (Fig. 5f&g). We speculated that microglia may be more susceptible to α-Syn than astrocytes are. To test this assumption, in vitro studies were performed. In microglia, p-STAT1^Y701 tended to increase as early as 10 min and was robustly elevated at 30 and 60 min after α-Syn PFF exposure, whereas the total STAT1 level remained unaltered (Fig. 5h). Neither the p-STAT1^Y701 nor the total STAT1 protein level changed in astrocytes following α-Syn PFF exposure for 60 min (Fig. 5i). However, longer exposure elicited an increase in p-STAT1^Y701 and a decrease in the ULK1 protein level in astrocytes (Supplemental Fig. 4 a-c). These findings support the in vivo data that microglial STAT1 is more responsive to α-Syn PFF.

Fig. 5.

Microglial STAT1 is activated in α-Syn OE mice. (a, b) Western blots (a) and quantification (b) of p-STAT1^Y701, ULK1, α-Syn and TH protein levels in the SN of AAV-hα-Syn- or AAV-Vector-injected mice at 8 weeks post injection; n = 4 animals per group. (c-g) Representative images of p-STAT1^Y701 (green) with IBA1 (c) or GFAP (f) costaining in the SNr. The p-STAT1^Y701 intensity (d) and Mander’s overlap coefficient analysis for p-STAT1^Y701 colocalization with IBA1 (e)- or GFAP (g)-positive cells are presented; n = 3 mice per group, with at least three slices per mouse. Aarrowheads indicate p-STAT1^Y701 and IBA1 colocalization, with the enlarged inset on the right. DAPI was used to stain the nuclei. Scale bar, 20 μm. Unpaired Student’s t test. (h, i) Western blot analysis of STAT1 and p-STAT1^Y701 levels in α-Syn PFF-treated microglia (h) and astrocytes (i); n = 3–5 replicates. *P < 0.05, **P < 0.01, ***P < 0.001, n.s., not significant. One-way ANOVA followed by Dunnett’s test was used for h and i, and Student’s t test was used for the other panels

Stat1 deficiency enhances ULK1 expression and autophagy activity in microglia

To define the role of microglial STAT1 in PD, we generated microglial Stat1^cKO mice by crossbreeding Stat1^fl/fl mice with Cx3cr1^CreERT2 mice (Fig. 6a). No difference in gross appearance or body weight was observed between 4-month-old Stat1^cKO and Stat1^fl/fl littermate controls (Fig. 6b).

Fig. 6.

Increased ULK1 expression and autophagic flux in Stat1-deficient microglia. (a) Strategy for microglial Stat1 cKO generation in mice. Male Stat1^cKO mice and Stat1^fl/fl mice were administered 75 mg/kg tamoxifen for five days and used for studies two weeks later. (b) The appearance and body weight of Stat1^fl/fl and Stat1^cKO mice; n = 9 mice per group. (c) Schematic diagram of primary microglial culture from Stat1^fl/fl or Stat1^cKO pups. (d-f) Western blot analysis of various protein levels in Stat1^fl/fl and Stat1^cKO microglia subjected to 4-OHT treatment for two (D2) and four (D4) days (n = 4). (g) qPCR data of microglial Ulk1 mRNA levels on D2 after 4-OHT treatment; n = 5. The data are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. Student’s t test

We then cultured primary microglia from Stat1^cKO and Stat1^fl/fl pups and treated the cells with 4-hydroxytamoxifen (4-OHT), an active metabolite of tamoxifen, to induce Cre expression and Stat1 depletion in Stat1^cKO-derived microglia. The protein levels of STAT1 and ULK1, as well as those of LC3II and SQSTM1/P62, were studied via western blotting (Fig. 6c). Compared with that in Stat1^fl/fl microglia, the STAT1 protein level in Stat1^cKO microglia decreased by approximately 50% on day 2 (D2) after 4-OHT treatment and was almost undetectable on day 4 (D4) (Fig. 6d-f). Moreover, Stat1^cKO microglia presented higher ULK1 protein levels than Stat1^fl/fl microglia did at D2 and even higher levels at D4 after 4-OHT treatment. This finding was in line with the data shown in Fig. 4, confirming the inhibitory effect of STAT1 on Ulk1 transcription. Notably, the LC3II protein level in Stat1^cKO microglia increased at D2 but decreased at D4 compared with that in Stat1^fl/fl microglia, whereas the p62 protein level persistently decreased in Stat1^cKO microglia at D2 and D4. These findings indicate increased autophagic flux in Stat1-deficient microglia. Additionally, qPCR revealed higher Ulk1 mRNA levels in Stat1^cKO microglia than in Stat1^fl/fl microglia (Fig. 6g). Collectively, these data suggest that STAT1 suppresses microglial Ulk1 transcription and thus autophagy activity.

Conditional ablation of microglial Stat1 alleviates neuroinflammation and DAergic neurodegeneration in α-Syn-based PD mouse models

We further assessed the impact of microglial Stat1 depletion on PD pathology and motor behaviors. First, we studied the effect of Stat1 cKO on microglial responses in the AAV-mediated hα-Syn OE model. A robust change in microglial morphology was found in the SNr of PD mice. The IBA1 fluorescence intensity increased, and the soma of IBA1⁺ microglia enlarged in the AAV-hα-Syn-injected Stat1^fl/fl mice compared with those in the vector group. This effect was alleviated in AAV-hα-Syn-injected Stat1^cKO mice (Supplemental Fig. 5), suggesting that Stat1 cKO prevented microglial activation in this PD model.

AAV-mediated hα-Syn OE often results in α-Syn OE at supraphysiological levels [25]. These models reproduce nigral degeneration, α-Syn pathology and motor deficits, greatly advancing our understanding of PD, especially familial PD with α-Syn elevation. However, α-Syn OE is not associated with idiopathic PD. Therefore, we applied another PD model via the delivery of α-Syn PFFs into the striatum of 6-month-old mice, which has been commonly used in recent publications [17, 18]. α-Syn PFFs serve as “seeds” to trigger normal levels of endogenous α-Syn to form aggregates, which can spread throughout the brain, especially in vulnerable regions of PD. Within 3–6 months after α-Syn PFF injection, neuroinflammation and DA neuron degeneration can be progressively detected in the nigrostriatal regions. In this study, a robust neuroinflammatory response, indicated by increased IBA1 intensity and microglial soma volume, was observed in the SN at three months after a single injection of α-Syn PFF into the striatum of Stat1^fl/fl mice compared with those in the vehicle (PBS) group (Fig. 7a-c). These changes were attenuated in PFF-treated Stat1^cKO mice. Double staining revealed diminished ULK1 reactivity in IBA1-positive cells from α-Syn PFF-injected Stat1^fl/fl mice (Fig. 7d&e), similar to the observations in the AAV-mediated α-Syn OE model (Fig. 1f). However, the intensity of ULK1 increased, and ULK1 colocalized more with IBA1-labeled microglia in both α-Syn PFF-treated and untreated Stat1^cKO mice. We attempted but failed to detect autophagosomes in microglia by immunostaining with anti-LC3 and anti-IBA1 antibodies, probably because of the relatively lower autophagy activity in normal and PD mice (data not shown). We then performed an in vitro study and reported a decrease in the LC3II level, along with an increase in the SQSTM1/p62 level, in α-Syn PFF-treated normal microglia. Stat1^−/− microglia presented increased LC3II levels and decreased p62 protein levels following α-Syn PFF treatment (Fig. 7f).

Fig. 7.

Microglial Stat1 depletion alleviates DA neurodegeneration and motor deficits in α-Syn PFF-induced PD mice. (a-c) Representative images of IBA1 staining at various magnifications in the midbrains of Stat1^fl/fl and Stat1^cKO mice 3 months after α-Syn PFF or PBS injection into the striatum. Quantifications of IBA1 intensity and soma diameter are shown in b and c. n = 3 mice per group, with at least three slices included for analysis. (d, e) Imaging for IBA1 (red) and ULK1 (green) costaining. The white arrowheads indicate the colocalization of IBA1 and ULK1, as analyzed by Mander’s coefficient (e), n = 3 mice per group. Scale bar, 20 μm. (f) Western blot analysis of LC3 and p62 protein levels in PFF-treated WT and Stat1^−/− microglia from Stat1^fl/fl and Stat1^cKO mice, respectively. (g, h) Representative immunohistochemistry images of TH⁺ neurons and quantification (h) in the SNpc; n = 4 mice per group. Scale bar, 200 μm. (i) Western blot analysis of TH protein levels in SN lysates; n = 8 mice per group. (j-l) Behavioral performance of α-Syn PFF- or PBS-injected Stat1^fl/fl and Stat1^cKO mice during the open field (j), rotarod (k) and beam crossing tests (l). n = 6–8 mice per group. The data are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. ^##P < 0.01 compared with the PBS-injected Stat1^fl/fl control. Two-way ANOVA followed by Tukey’s post hoc analysis

We further evaluated SNpc DAergic neuron damage at six months after α-Syn PFF or PBS injection. Immunostaining and quantification revealed that, compared with the control, α-Syn PFF caused a 49% loss of TH⁺ neurons in Stat1^fl/fl mice and a 21% loss in Stat1^cKO mice (Fig. 7g&h), suggesting protection of microglial Stat1 ablation against DAergic neuron damage. This finding was verified by western blot analysis, which revealed that α-Syn PFF-injected Stat1^cKO mice had higher TH protein levels than Stat1^fl/fl control mice did (Fig. 7i). Finally, locomotor and motor coordination abilities were assessed. The results of the open field test revealed that α-Syn PFF injection diminished locomotor distance in Stat1^fl/fl mice, which was attenuated in Stat1^cKO mice (Fig. 7j). Rotarod and balance beam crossing tests demonstrated that α-Syn PFF also reduced the latency to fall off the rod but prolonged the time to cross the beam in Stat1^fl/fl mice, indicating impaired coordination ability. The α-Syn PFF-induced motor coordination defects were alleviated in Stat1^cKO mice (Fig. 7k&l). Collectively, these data suggest that microglial Stat1 ablation exerts neuroprotective effects on α-Syn PFF-induced PD-like mice.

Discussion

Accumulating evidence reveals a role for autophagy in regulating microglial function in the adult brain [11, 19]. Autophagy enables microglia to engage amyloid plaques and prevents microglial senescence [26]. Microglia also clear neuron-released α-Syn via selective autophagy and prevent neurodegeneration [11]. Autophagy deficiency disrupts microglial lipid homeostasis, modulates neuroinflammation and aggravates tau pathology [27]. Microglial autophagy may be dysregulated during neurodegeneration. Tu et al. previously reported that short exposure to α-Syn inhibited microglial autophagy via TLR4-dependent p38 MAPK and Akt-mTOR signaling [12]. However, it remains unclear whether other mechanisms exist when microglia are chronically exposed to neuron-released α-Syn. In this study, we revealed decreased transcription of microglial Ulk1 in both in vitro and in vivo α-Syn-based PD mouse models. We reported a crucial role of STAT1 in repressing Ulk1 transcription. More importantly, microglial Stat1 depletion alleviated neuroinflammation, DAergic neuron damage and motor defects in α-Syn PFF-induced PD mice. Along with the previous report [12], we propose that neuron-released α-Syn may disrupt microglial autophagy through multiple mechanisms, depending on the timing, concentration and species of α-Syn.

ULK1 is critical for autophagy initiation and autophagosome‒lysosome fusion [28, 29]. Changes in microglial ULK1 expression or activity modulate neuroinflammation. He et al. reported that LPS triggered ULK1 phosphorylation at Ser757 and suppressed microglial autophagy and that an ULK1 inhibitor was sufficient to promote an inflammatory response in the absence of overt stimulation [30]. This finding was consistent with our results that Ulk1 knockdown increased the IL-1β mRNA level in α-Syn PFF-treated and untreated microglia, whereas ULK1 activation had the opposite effect. A marked decrease in microglial ULK1 expression was consistently demonstrated in the hα-Syn OE and α-Syn PFF-induced PD models, indicating that ULK1 expression may change in PD. Similarly, a lower ULK1 mRNA level was detected in the PBMCs of PD patients in this study, which is in line with the findings of a previous report [31]. An earlier study revealed a slight decrease in ULK1 protein levels in the SN of three patients with PD [32]. Further studies are warranted to confirm the alterations in ULK1 levels in the brains of PD patients.

Canonically, ULK1 modulates autophagy initiation by forming a complex with the ATG13, FIP200, and ATG101 proteins. The noncanonical function of ULK1 has also been revealed [33, 34]. Changes in ULK1 expression or activity are associated with human diseases, including cancer and neurodegeneration [35]. Small molecules that target ULK1 have been developed. A potent activator of ULK1, BL-918, enhances autophagy and protects against neurotoxin-induced PD in vitro and in vivo [36]. Herein, a significant reduction in ULK1 expression was revealed by western blot in the SN eight weeks after AAV-hα-Syn injection. BL-918 treatment, which started one week after AAV-hα-Syn delivery and continued for seven weeks, alleviated neuroinflammation and exhibited neuroprotective effects in PD mice, indicating the potential of ULK1 activators for PD therapy. An increased LC3II level was observed in the BL-918-treated PD mice compared with the untreated PD mice, indicating that BL-918 increased autophagy activity. Notably, a nonautophagic effect of BL-918 cannot be completely excluded. This activator may affect not only microglia but also DAergic neurons, thus exerting neuroprotection when given at an earlier stage. Neuronal autophagy impairment has been well demonstrated in PD [4, 37, 38]. In the future, whether neuronal ULK1 expression also changes in PD should be explored.

A variety of posttranslational modifications regulate ULK1 activity or stability [35, 39]. The transcriptional regulation of ULK1 has emerged as a hotspot. Here, we demonstrated that STAT1 functions as a suppressor of Ulk1 transcription, bridging microglial autophagy dysregulation and neuroinflammation in PD. Stat1-deficient cells presented increased ULK1 expression and autophagy, excluding the possibility of off-target effects resulting from RNA interference. Typically, STAT1 acts as a positive regulator of many inflammatory genes. However, the opposite effects also exist. Several genes, such as c-myc, cyclin D and thyroid-specific genes, are negatively regulated by STAT1 [40]. The negative regulatory effect of STAT1 on Ulk1 transcription was also reported in human fibrosarcoma cells [41], which is consistent with our data. This finding indicates a conserved effect of STAT1 on Ulk1 transcription across species.

Our results revealed a role for STAT1 in α-Syn-triggered microglial activation. Notably, astrocytes seemed to be less responsive to α-Syn PFF. A longer exposure (> 3 h) was required for STAT1 activation and ULK1 downregulation in cultured astrocytes (Supplemental Fig. 4a-c). This finding was consistent with the in vivo observations in hα-Syn OE mice, in which p-STAT1^Y701 was predominantly elevated in microglia. Nevertheless, microglial and astrocytic STAT1 activation and a decrease in ULK1 may synergistically contribute to PD pathogenesis, particularly at the later stage.

The JAK/STAT pathway plays critical roles in growth factor and cytokine signaling, orchestrating the regulation of the immune system and cell proliferation [42, 43]. Recent studies revealed a role for STAT1 in neurological disorders. For example, genetic and pharmacologic inhibition of STAT1 alleviated inflammation and long-term neurologic deficits after traumatic brain injury [24]. Notably, STAT1 may not be limited to regulating inflammation in the brain. STAT1 activation is responsible for amyloid β generation triggered by γ-secretase-cleaved Tau [44]. The JAK1/2 inhibitor AZD480 prevented neurodegeneration in a rat model of PD [45]. However, the relevance of STAT1 dysregulation in PD remains elusive. Here, we demonstrated an increase in microglial p-STAT1^Y701 in two α-Syn-based PD models and revealed that STAT1 suppressed Ulk1 transcription. More importantly, microglial Stat1 cKO attenuated neuroinflammation, DAergic neuron loss, and motor deficits in PD mice, indicating that STAT1 is a potential target for PD therapy. It is yet to be determined whether microglial Stat1 deficiency affects α-Syn spreading.

Conclusions

In summary, our findings reveal an inhibitory effect of STAT1 on Ulk1 transcription and expression, which contributes to microglial autophagy impairment in response to α-Syn. Our study highlights microglial STAT1 and ULK1 as promising therapeutic targets for PD.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

PD

Parkinson’s disease

α-Syn

α-synuclein

ULK1

UNC-51-like kinase 1

STAT1

Signal transducer and activator of transcription 1

PFF

Preformed fibrils

DAergic

Dopaminergic

SN

Substantia nigra

Atg7

Autophagy related gene 7

cKO

Conditional knockout

IBA1

Ionized calcium binding adapter molecule 1

shRNA

Short hairpin RNA

IL-1β

Interleukin-1β

IFN-γ

Interferon-γ

4-OHT

4-hydroxytamoxifen

SNr

Substantia nigra reticular

SNpc

Substantia nigra pars compacta

TH

Tyrosine hydroxylase

Author contributions

CPS, XOH and ZYM conceptualized and designed the experiments and analyzed the data; HYT conducted partial experiments and analyzed the data; HCQ analyzed the data; KL and CFL were involved in critical revision of the manuscript for important intellectual content. YL and LOY synthesized BL-918; JYL collected clinical samples and advised the project; JYL and LFH designed the study; and LFH wrote the manuscript with input from all the authors.

Funding

This work was supported by grants from the National Natural Science Foundation of China (82171251, 82104190), the Key Project of the Natural Science Foundation of Jiangsu Provincial Higher Education Institutions (21KJA180003) and the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). This was also partly supported by the Discipline Construction Program of the Second Affiliated Hospital of Soochow University (XKTJTD202004).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study involving human blood was approved by the Ethics Committee of the Second Affiliated Hospital of Soochow University in China (IK-2018-061-01). All the animal studies complied with all the relevant ethical regulations for animal testing and research and were approved by the Institutional Animal Care and Use Committee of Soochow University (ECSU-201600014).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.