LRRK2 G2019S promotes astrocytic inflammation induced by oligomeric α-synuclein through NF-κB pathway

Summary Parkinson’s disease (PD) is characterized by the irreversible loss of dopaminergic neurons and the accumulation of α-synuclein in Lewy bodies. The oligomeric α-synuclein (O-αS) is the most toxic form of α-synuclein species, and it has been reported to be a robust inflammatory mediator. Mutations in Leucine-Rich Repeat Kinase 2 (LRRK2) are also genetically linked to PD and neuroinflammation. However, how O-αS and LRRK2 interact in glial cells remains unclear. Here, we reported that LRRK2 G2019S mutation, which is one of the most frequent causes of familial PD, enhanced the effects of O-αS on astrocytes both in vivo and in vitro. Meanwhile, inhibition of LRRK2 kinase activity could relieve the inflammatory effects of both LRRK2 G2019S and O-αS. We also demonstrated that nuclear factor κB (NF-κB) pathway might be involved in the neuroinflammatory responses. These findings revealed that inhibition of LRRK2 kinase activity may be a viable strategy for suppressing neuroinflammation in PD.


Highlights
LRRK2 G2019S and oligomeric a-synuclein activate astrocytic neuroinflammation LRRK2 G2019S accelerates dopaminergic cell loss induced by oligomeric a-synuclein LRRK2 kinase inhibitor reduces NLRP3 activation and the level of inflammatory factors LRRK2 G2019S participates in neuroinflammation via regulating the NF-kB pathway

INTRODUCTION
Parkinson's disease (PD) is the second most frequent neurodegenerative disorder and the fastest growing neurological disease. 1 PD is mainly characterized by the loss of dopaminergic neurons in the substantia nigra, and the aggregation of misfolded alpha-synuclein (a-syn) to form Lewy bodies. 2 In recent years, both in vivo and in vitro evidences have revealed that oligomeric a-syn (O-aS) is the most cytotoxic component leading to neurodegeneration in PD. 3 A clinical longitudinal study also showed that O-aS in cerebrospinal fluid (CSF) was correlated with the severity of PD motor symptoms. 4However, the precise molecular mechanisms of O-aS toxicity remain to be elucidated.Release of pathological O-aS from damaged neurons could accelerate neuronal cell death in part via astrocytes and microglia activation, revealing that neuroinflammation induced by O-aS is an inescapable part of PD pathology. 5,6eucine-Rich Repeat Kinase 2 (LRRK2) G2019S mutation is a common genetic variant in PD patients, accounting for 4% of the familial and 1% of the sporadic PD cases. 2 Clinical studies have shown that PD patients with LRRK2 G2019S mutation have an increase in inflammatory factors compared with patients with primary PD. 7,8Recent study has also revealed higher levels of O-aS and tumor necrosis factor alpha (TNF-a) in the CSF of both symptomatic and asymptomatic LRRK2 mutation carriers, 9 indicating that there might be a close interaction between LRRK2 dysfunction and O-aS toxicity in the neuroinflammation associated with PD.
Exploring the interaction of LRRK2 and O-aS could be helpful for developing effective targeted disease-modifying therapies.In the present study, we found that LRRK2 G2019S mutation aggravated the glial inflammatory response induced by O-aS, mainly affecting the morphology of astrocytes, and inhibition of LRRK2 kinase activity reduced the production of inflammatory factors induced by O-aS.We also demonstrated that LRRK2 G2019S mutation and O-aS might interplay in the nuclear factor kB (NF-kB) pathway, promoting neuroinflammation.

LRRK2 G2019S induced early parkinsonism-like phenotypes and aggravated the loss of dopaminergic neurons in O-aS-induced mouse model
We first expressed monomeric a-syn in BL21 (DE3) Escherichia coli and purified the protein.Protein purity was confirmed by Coomassiestained SDS-PAGE and western blotting (Figures S1A and S1B).1][12] In native-PAGE, O-aS showed a diffused band with higher molecular mass compared with the monomer (Figure 1A).We also utilized negative staining to further characterize the O-aS.As previously reported, O-aS species appeared to be approximately 5-20 nm in height (Figure 1B). 10 The soluble O-aS is attributed to specific structural characteristics that confer damaging properties, causing lipid peroxidation and rapidly inducing a large number of intracellular reactive oxygen species (ROS) (Figure S1C). 12The b-sheet structure in the prepared O-aS was confirmed using circular dichroism (CD) spectra (Figure S1D).
We stereotactically injected O-aS into the striatum of 6-month-old LRRK2 G2019S-Tg mice, and the injected protein was detectable in tyrosine hydroxylase (TH)-positive nigra dopaminergic neurons using anti-human a-syn antibody 7 days post-injection (Figure S1E).4][15] Gait performance tests revealed that only G2019S+O-aS group showed increased stance time of both forepaw and hindpaw in the treadmill (Figures 1C and 1D).In the motor learning tests, G2019S+O-aS mice fell off from the accelerating Rota-rod faster compared with G2019S mice after 3 days training (Figure 1E).Repeated measures ANOVA revealed a significant interaction between group and day (Figure 1E).Rota-rod data obtained from each date and trial were presented in Figure S2 and Table S1.To investigate whether LRRK2 G2019S mutation affected the survival of dopaminergic neurons after O-aS treatment, we used immunohistochemical staining to quantified TH-positive neurons in O-aS-treated control and LRRK2 G2019S-Tg mice.The number of TH-positive cells decreased significantly in the O-aS-treated LRRK2 G2019S-Tg mice compared with the O-aS-treated control mice (Figures 1F and 1G).We also quantified the density of TH terminals in striatum.The results were consistent with those of THpositive cell counting (Figure S3).

LRRK2 G2019S regulated the glial inflammatory response induced by O-aS
We then asked whether LRRK2 G2019S mutation promoted the effects of O-aS via regulating neuroinflammation.Immunostaining results showed that the number of astrocytes and microglia increased significantly after O-aS injection in both control and LRRK2-G2019S-Tg mice (Figure 2A).However, G2019S group showed more cell proliferation compared with the control group (Figures 2A-2C and S4).Morphological analysis showed that the cell ending radius and branch number of astrocytes significantly increased in the O-aS-treated G2019S group compared with the O-aS-treated control group (Figures 2D-2F), while the morphology of microglia did not change significantly (Figures 2D, 2G, and 2H).Our study revealed that astrocytic inflammation maybe pronounced at a certain stage during the development of PD.

LRRK2 G2019S enhanced O-aS-induced inflammatory levels in primary astrocytes, and inhibition of LRRK2 kinase activity could rescue the effects of both LRRK2 G2019S and O-aS
In vivo studies are difficult to explore the role of astrocyte-specific inflammatory responses in neuroinflammation, excluding the contribution of microglia.Both microglia and astrocytes were involved in the inflammatory responses.We also confirmed that NOD-like receptor thermal protein domain associated protein 3 (NLRP3) inflammasomes could co-localize with both microglia and astrocyte in O-aS-treated LRRK2 G2019S mouse (Figure S5).To resolve the difficulty, we used primary cultured astrocytes (labeling with GFAP) for further studies (Figure S6A).We also investigated whether expression of LRRK2 was time-dependent in primary astrocytes using western blotting.Expression of LRRK2 in primary astrocytes increased over time, and expression in the G2019S group was significantly higher than that in the control group after 21 days (Figures S6B and S6C).However, LRRK2 in primary microglia was barely detectable (Figure S6D).We cultured G2019S astrocytes for 21 days to study the effect of O-aS-induced neuroinflammation.Extracellular O-aS may interact with various cellular receptors, which can upregulate interleukin (IL)-1a, IL-1b, IL-6, and TNF-a transcription of proinflammatory factors. 16We treated cultured primary astrocytes with O-aS (0.5 mg/mL) and detected the inflammation-related factors 24 h later (Figure 3A).Western blotting results showed that O-aS induced cultured primary astrocytes to produce high levels of inflammation-associated NLRP3, and LRRK2 G2019S mutation significantly enhanced the effects of O-aS (Figures 3B and 3C).We collected the cell supernatant from each group and detected the inflammatory factors IL-1b and TNF-a and anti-inflammatory factor IL-10 using ELISA.Both IL-1b and TNF-a increased significantly in the G2019S+O-aS group (Figures 3D and 3E), while the level of IL-10 was significantly reduced in the G2019S+O-aS group (Figure 3F), compared with the CTR+O-aS group.We also utilized IN-1, an LRRK2 kinase inhibitor, to confirm the interplay between LRRK2 G2019S mutation and O-aS and found that IN-1 could significantly reduce the expression levels of NLRP3, LRRK2 G2019S contributed to the inflammatory response through NF-kB pathway NF-kB is an important transcription factor in the inflammatory pathway and mediates the pathological processes of neurodegeneration. 17To verify whether NF-kB was involved in the inflammation associated with LRRK2 G2019S mutation and O-aS, we detected the expression of phosphorylation-inhibitor of Kappa B Kinase (P-IKK), inhibitor kappa B alpha (IKB), and phosphorylation-p65 by western blotting.P-IKK and P-p65 were increased in the G2019S+O-aS group compared with O-aS group (Figures 4A-4C), and the level of IKB in the G2019S+O-aS group was significantly higher than that in O-aS group (Figures 4A and 4D).IN-1 reversed the changes in these protein levels affected by LRRK2 G2019S mutation (Figures 4A-4D).Immunofluorescence staining of p65 also revealed that LRRK2 G2019S mutation promoted the entry of p65 into the nuclei in both O-aS-treated and untreated groups, which could be blocked by IN-1 (Figure 4E).These results indicate that LRRK2 G2019S mutation might promote astrocytic neuroinflammation induced by O-aS via the NF-kB signaling pathway.

DISCUSSION
The penetrance of LRRK2 G2019S is incomplete, and thus it may be affected by the external environment or other genetic factors.The effects of LRRK2 G2019S are age dependent, and there was no loss of dopaminergic neurons in LRRK2 G2019S-Tg mice after a longterm observation. 18The present study showed that LRRK2 G2019S promoted dopaminergic cell loss and induced gait disturbance and motor learning deficit in O-aS-treated LRRK2 G2019S-Tg mice.However, no locomotion decline in G2019S+O-aS group was found in our study (data not shown).It is currently thought that classic motor symptoms of PD only become noticeable when the cell loss of dopaminergic neurons in SNpc reaches about 50% due to compensatory mechanisms.TH-positive cells were found to be decreased about 30% in G2019S+O-aS group compared with the control group, revealing that the animals presented early-stage parkinsonism at that time point.Meanwhile, no significant difference in TH-positive cells was found in the control+O-aS group compared with the control group.In A30P mice characterized by a rapid onset of a-syn pathology and the presence of O-aS across the whole brain, Anish et al. found that there was no loss of dopaminergic neurons, but dysregulation of the monoaminergic system was recorded in older mice (Behere et al., 2021).We speculate that the reduction of TH neurons induced by O-aS requires the combination of other disease-risk factors, such as aging, environmental effects, and PD-related variants.
0][21] Meanwhile, knockout of LRRK2 gene prevents the neurotoxic effects and neuroinflammatory reactions caused by paraquat. 22Although these studies have revealed the improving effect of LRRK2 G2019S mutation in neurotoxin-induced neuroinflammation, it is still difficult to speculate how LRRK2 is involved in the inflammatory processes.A clinical study has reported that there was a large concentration of O-aS in CSF of LRRK2 G2019S carriers. 9O-aS plays a crucial role in the pathological progression of PD, and in vitro experiments have shown that O-aS can be an inducing factor for neuroinflammation. 23The current study demonstrated that LRRK2 G2019S mutation exacerbates O-aS-induced inflammatory responses, and the effects could be rescued by LRRK2 kinase inhibitor.Although the numbers of microglia and astrocytes were significantly increased, only the morphology of astrocytes was found to be changed in the morphological analysis.Despite astrocytes taking up a great proportion of the brain cells, the contribution of astrogliosis to PD is still poorly understood.Although astrocytes perform a variety of physiological functions in the brain, they are pivotal mediators of a-syn toxicity since they internalize pathological a-syn released from damaged neurons. 24Astrocytes are considered to be instrumental amplifiers of neuroinflammation.Proinflammatory mediators such as IL-1b, TNF-a, and complement component 1q, as well as mitochondrial fragments released by activated microglia, can promote the activation of astrocytes to A1 proinflammatory state and lead to the further release of high levels of proinflammatory cytokines. 25abriel et al. proved that O-aS was rapidly engulfed by astrocytes and they were intracellularly stored, rather than degraded, resulting in impaired mitochondrial function. 26Meanwhile, oligomer-selective antibodies could prevent a-syn from accumulation and restore mitochondrial function in cultured astrocytes. 26These findings indicate that LRRK2 G2019S mainly increases the susceptibility to astrocytic neuroinflammation, leading to aggregation of NLRP3 and release of proinflammatory factors IL-1b and TNF-a.
The mice we used were LRRK2 G2019S systemic knockin, and in vivo research is difficult to explore the role of astrocyte-specific inflammatory response in neuroinflammation, excluding the role of microglia.Microglia are the primary immune cells in the brain, and microglial activation is correlated with PD severity. 272][33] Microglia from adult wild-type mice and LRRK2-overexpressing mice also did not show any LRRK2 immunoreactivity. 34In LPS-induced inflammatory models, LRRK2 expression could only be detected in activated microglia after intracranial injection of LPS, 34 while systemic administration of LPS did not induce microglial LRRK2 expression, 33 revealing that the expressing of LRRK2 in microglia may require a strong inflammatory stress.In the present study, we found that LRRK2 G2019S promoted the activation of astrocytes after striatal injection of O-aS which is considered to be the most toxic strain leading to PD pathogenesis.Hence, we speculate that therapeutic efforts targeting astrocytic neuroinflammation might be an effective strategy to alleviate LRRK2-related PD.
Pathogenic protein aggregates, such as a-syn fibrils, another hallmark of PD pathology and the primary component of Lewy bodies, were reported to activate NLRP3 inflammasomes in microglia through interaction with Toll-like receptors and activation of NF-kB. 17To investigate the underlying mechanisms of the effects of LRRK2 G2019S on inflammation induced by O-aS, the NF-kB signaling pathway was studied.We demonstrated that O-aS significantly upregulated the expression of P-P65, P-IKK, and IKB (markers of NF-kB pathway activation).This is consistent with the previously reported results.More importantly, we found that LRRK2 G2019S mutation exacerbated the activation of NF-kB pathway and NLRP3 inflammasome.LRRK2 kinase inhibitor IN-1 significantly reduced the expression of NLRP3, pro-IL-1b induced by LRRK2 G2019S, and the levels of IL-1b and TNF-a.But what's interesting is that inhibition of LRRK2 via IN-1 did not completely inhibit O-aS-mediated cytokine expression, which may support existence of LRRK2-independent inflammatory signaling pathway in PD-related neurodegeneration.
In conclusion, our study verified that LRRK2 G2019S, cooperating with O-aS, altered the morphology of astrocytes, maintained inflammatory response, and eventually aggravated the loss of dopaminergic neurons.Treatment with LRRK2 kinase inhibitor, IN-1, reduced NLRP3

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
This study does not involve any patients or healthy control participants.

Animals
LRRK2 transgenic mice with G2019S mutation (C57BL/6J-Tg LRRK2*G2019S 2AMjff/J) were purchased from Jackson Laboratory (Bar Harbor, ME, USA) and hemizygous mice were bred to noncarrier (wild-type) siblings.And male mice aged 6 months were used throughout this study.Mice were raised in a relatively stable environment: light/dark ratio 12/12 h, temperature (21 G 2 C), and relative humidity (55 G 5%). Five animals were placed in one cage and had unrestricted access to food and water.All animal experiment protocols were approved by the Institutional Animal Care and Use Committee of Soochow University (Suzhou, China).

Cultured cells
Primary astrocytes were acquired from P1-P3 pups with identified genotypes.The cortex was isolated and dissociated in 0.1 M PBS, then digested with 0.25% trypsin (GIBCO) for 15 min, and centrifuged (1,000 rpm) after filtration with a 70-mm strainer.Primary astrocytes were resuspended and maintained in Dulbecco's modified Eagle's medium/F12 (Gibco, Grand Island, NY, USA) containing penicillin (100 U/mL) and streptomycin (100 U/mL), supplemented with 10% fetal bovine serum (Gibco, Grand Island, NY, USA).Microglia were depleted from the mixed culture through shaking at 220 rpm/min for 3 h.Astrocytes sub-cultured for 36 generations were used in this study.LRRK2 kinase inhibitor IN-1 was purchased from MedChemExpress (HY-10875, Monmouth Junction, NJ, USA) and dissolved in DMSO.PC12 cells were purchased from Institute of Cell Biology (Chinese Academy of Sciences, Shanghai, China) and cultured in Dulbecco's modified Eagle's medium (Gibco, Grand Island, NY, USA).

METHOD DETAILS Preparation of a-syn and generation of O-aS
The plasmid pET21a containing human a-syn cDNA (addgene plasmid 51486, Michael J. Fox Foundation, USA) was transformed into BL21 (DE3) Escherichia coli (TransGen Biotech, China).Protein expression was induced at OD 600 0.8 with 0.25 mM isopropyl-1-thio-D-galactopyranoside (IPTG) for 4 h at 37 C.The cell pellet was harvested and lysed using a tip sonicator in 100 mM Tris-HCl (pH 8.0) with 1 mM EDTA and 1 mM phenylmethylsulfonyl fluoride, and then boiled for 15 min.After centrifugation at 20,000 3 g for 30 min, 20 mg/mL streptomycin was added to the supernatant to remove nucleic acids, followed by further centrifugation.Unwanted proteins were removed from the supernatant by adjusting pH to 3.5 using 1 M HCl.After centrifugation, the supernatant was dialyzed overnight against 20 mM Tris-HCl (pH 8.0).The sample was filtered with a 0.22-mm membrane (SLGP033RB, Millipore, USA) and purified by HiTrapâ Q FF column (GE Healthcare, USA) and the purity of each peak fraction was analyzed by Coomassie blue stained SDS-PAGE.The protein concentration was quantified by measuring the absorbance at 280 nm using a NanoDrop2000 spectrophotometer (Thermo, USA).Endotoxin was cleaned using Pierce high-capacity endotoxin removal spin columns (88276, Thermo, USA) and measured to be less than 0.05 EU/mL using the ToxinSensorä chromogenic LAL endotoxin assay kit (L00350, GenScript, USA).The purified protein was pooled, dialyzed with deionized distilled water, concentrated with a 3-kDa cutoff membrane (UFC900308, Millipore, USA), aliquoted and lyophilized.To generate b-sheet-rich O-aS, the protein was resuspended in PBS (pH 7.4) to a final concentration of 12 mg/mL, and statically incubated at 37 C for 22-24 h, followed by ultracentrifugation and filtration using a 100-kDa cutoff membrane (UFC910008, Millipore, USA), as previously described. 10tive-polyacrylamide gel electrophoresis (PAGE) Coomassie blue-stained native-PAGE were performed using 4-20% Mini-PROTEAN Precast Protein Gels (4561093, BIO-RAD, USA) according to manufacturer instruction.

Transmission electron microscope (TEM) imaging
Protein samples were applied to carbon-coated copper grids and left to stand for 2 min followed by staining with 2% phosphotungstic acid (G1870, Solarbio, China) for 2 min.Micrographs were taken with a TEM (TECNAI G2 F20, FEI, USA) at 200 kV and 34,000 X magnification.

O-aS induced neuroinflammatory mouse model
Male mice aged 6 months were used throughout this study, consisting of LRRK2 G2019S-Tg and noncarrier (wild-type) from the cross above.The mice were anesthetized by 2% isoflurane inhalation and placed in a stoelting stereotaxic apparatus (68001, RWD Life Science, China).O-aS (1 mg/mL, 2 mL) was stereotactically injected into the right striatum according to the Elsevier brain atlas (compact 3rd edition) 0.5 mm anterior to the bregma, 2.0 mm to the midline, and 3.0 mm subdural.The same volume of PBS was injected into the left striatum as the control.
The rate was kept at 0.2 mL/min and left for 10 min, then slowly retracted.
automatically recorded by the software.The experiment was performed for 3 days and the average of the three trials in each day was calculated for further analysis.

Cytokine measurement
After treatment with O-aS, the cell supernatant was collected and centrifuged at 3,000 rpm for 5 min for detection.The concentration of interleukin (IL)-1b and TNF-a was determined by ELISA (Thermo Fisher Scientific, USA).IL-10 was quantified using ELISA from R&D Systems (Minneapolis, MN, USA).

Sholl analysis
We performed Sholl analysis of the morphology of microglia and astrocytes. 39,40In brief, images of mouse brain slices which immunostained with IBA1 or GFAP antibody were captured with OLYMPUS VS200.Using the Sholl Analysis plugin for automatic drawing analysis in Fiji software, immunoreactive areas in each region were thresholded, divided by area, and expressed in corresponding multiples.The ending radius and branching index were calculated for further analysis.

QUANTIFICATION AND STATISTICAL ANALYSIS
All analyses were performed using GraphPad Prism 8 (GraphPad Software, La Jolla, CA, USA).Numerical data were compared with one-way ANOVA, two-way ANOVA or repeated measures ANOVA.All values are displayed as mean G SEM. P < 0.05 was considered significant (*P < 0.05, **P <0.01, ***P < 0.001; ns, not significant).

Figure 1 .
Figure 1.LRRK2 G2019S promoted early parkinsonism-like behaviors and dopaminergic neuronal loss in O-aS-treated mouse model (A) Identification of a-syn protein (monomer and oligomer) with Coomassie blue-stained native-PAGE.(B) Representative electron microscopy image of a-syn oligomers (scale bar = 50 nm).(C and D) Stance time of control (CTR), G2019S, O-aS, and G2019S+O-aS mice in gait performance.Two-way ANOVA followed by Turkey's post-hoc test, n = 6.*p < 0.05; ns, not significant.(E) Latency to fall off the accelerating Rota-rod during the days training of the between G2019S group and G2019S+O-aS group.Two-way ANOVA followed by Turkey's post-hoc test, n = 6.# p < 0.05.(F) Representative images of IHC staining of TH-positive neurons in the substantia nigra pars compacta (SNpc) of mice in the control, G2019S, O-aS, and G2019S+O-aS groups.(scale bar = 1 mm).(G) Statistical comparison of TH-positive neurons of control, G2019S, O-aS, and G2019S+O-aS groups in the substantia nigra pars compacta.Two-way ANOVA followed by Sidak's post-hoc test, n = 3. *p < 0.05; ns, not significant.Data are represented as mean G SEM.

Figure 3 .
Figure 3. LRRK2 G2019S mutation exacerbated inflammatory response in O-aS-treated astrocytes, and IN-1 reduced the effects of LRRK2 G2019S and O-aS (A) Flow-process diagram of experimental treatment of primary astrocytes.(B and C) Western blotting assessment of NLRP3 protein expression in astrocytes in the control, G2019S, O-aS, and G2019S+O-aS groups.Two-way ANOVA followed by Sidak's post-hoc test, n = 3. *p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant.(D-F) The levels of IL-1b, IL-10, and TNF-a in supernatants of different treatments were determined by ELISA in the control, G2019S, O-aS, and G2019S+O-aS groups.Two-way ANOVA followed by Sidak's post-hoc test, n = 3. ***p < 0.001.Data are represented as mean G SEM.

Figure 4 .
Figure 4. LRRK2 G2019S regulated inflammatory activity through NF-kB pathway (A) Western blotting assessment of P-IKK, P-P65, and IKB protein expression in astrocytes in the control, G2019S, O-aS, and G2019S+O-aS groups.(B-D) Quantitative expression of P-IKK, P-P65, and IKB in astrocytes in the control, G2019S, O-aS, and G2019S+O-aS groups.Two-way ANOVA followed by Sidak's post-hoc test, n = 3. *p < 0.05; ns, not significant.(E) Immunostaining for primary astrocyte on the substantia nigra of mice with anti-P65 antibody in the control, G2019S, O-aS, and G2019S+O-aS groups.Nuclei are stained with DAPI.(Scale bar = 50 mm， n = 3 per group).Data are represented as mean G SEM.

Figure 5 .
Figure 5. Schematic diagram for the mechanism of LRRK2 regulating inflammation in astrocytes induced by O-aS LRRK2 G2019S enhanced O-aS induced inflammatory levels in astrocytes through NF-kB pathway.Treatment with LRRK2 kinase inhibitor, IN-1, reduced NLRP3 inflammasome activation and the level of IL-1b and TNF-a and increased the level of IL-10.