Neuroinflammation Upregulated Neuronal Toll-Like Receptors 2 and 4 to Drive Synucleinopathy in Neurodegeneration

Background: Parkinson’s disease (PD) is characterized by intraneuronal α-synuclein aggregation called Lewy bodies and progressive dopaminergic neurodegeneration. Toll-like receptor (TLR) signaling is a major pathway mediating inflammation. The molecular link on how neuroinflammation upregulates neuronal TLRs and induces accumulation of α-synuclein aggregates to drive synucleinopathy remains to be determined. Objective: Despite conditioned medium from microglia and TLR agonists were utilized to study their effects on neuronal cells, a Transwell coculture system, comprising lipopolysaccharide-activated microglia on top and retinoic acid-differentiated SH-SY5Y cells at the bottom more mimicking in vivo neuroinflammation, was employed to elucidate the mechanism of activated microglia on neuronal cells. Methods: Genetic variants of TLRs in PD patients were genotyped and the multiplex cytokines, sRAGE, and HMGB1were assessed. A coculture system was employed to measure α-synuclein aggregates and neurite shortening by confocal microscope. The expression of TLR2/4 and autophagy flux was detected by western blot and immunofluorescence. Results: PD patients showed higher plasma levels of proinflammatory cytokines and genetic TLR4 variant, c.896 A > G (p. D299G). Elevated proinflammatory cytokines in coculture medium was also seen. Phosphorylation and aggregation of α-synuclein, shortening of neurite, upregulation of TLR2/4 expression, activation of downstream p38 and JNK, and dampening of autophagic flux were seen in SH-SY5Y cells cocultured with activated microglia. Those were prevented by inhibiting TLR2/4 and p38/JNK signaling. Conclusion: Activated microglia-derived neuroinflammation induced neuronal TLR2/4-p38/JNK activation to perturb autophagy, causing accumulation of α-synuclein aggregates and neurite shortening. Targeting neuronal TLR2/4 pathway might be a mechanistic-based therapy for neurodegenerative disease, such as PD.

Since TLRs are expressed in microglia and neurons, whether microglia-derived neuroinflammation upregulated neuronal TLRs to drive α-synuclein aggregation, leading to neurite shortening, was examined by activated microglia cocultured with differentiated dopaminergic neuronal cells. We also analyzed plasma inflammatory cytokines and genetic variants of TLRs in a cohort of PD patients and healthy controls.

Part I: Human Study
Participants A total of 1,029 participants, including 516 patients with PD and 513 healthy controls, were recruited from the National Taiwan University Hospital. All PD patients fulfilled the United Kingdom PD Society Brain Bank diagnostic criteria of PD and received regular evaluations of motor and cognitive functions (Hughes et al., 1992). The age and sex were comparable between enrolled PD patients and controls. Participants who have underlying immune-related disorders, including autoimmune diseases, or who have used immune suppressants, or nonsteroid antiinflammatory drugs within 1 year before the recruitment were excluded. Motor symptom severity was evaluated using the Unified Parkinson's Disease Rating Scale (UPDRS) motor subscale and Hoehn-and-Yahr staging. All participants provided informed consent before participating the study, and the institutional ethics review board of the National Taiwan University Hospital approved this study (the protocol number is 201703010RINA). All consent was acquired in accordance with the ethical standard in the Declaration of Helsinki. untranslated region variants including rs1927911, rs1927914, and rs10116253 in TLR4, and rs5743810 (p. S249P) in TLR6 (TaqMan ® assay ID: C___1180648_20) were genotyped through real-time polymerase chain reaction (PCR) using TaqMan ® Genotyping Assays on a StepOnePlus Real-Time PCR machine (Applied Biosystems).

Multiplex Cytokines, Soluble Forms of Receptors for Advanced Glycation End Products and High-Mobility Group Box 1 Assessments
Of the 1,029 participants, plasma samples from 241 participants, including 118 patients with PD and 123 controls, were measured for individual cytokines. Plasma from 10 ml of peripheral venous blood was isolated as previously described (Lin et al., 2019). Cytokine levels, including GM-CSF, IFN-γ, TNF-α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-12p70, IL-13, and IL-18, were measured in a fixed volume of plasma (25 μl) using a Th1/Th2 Cytokine 11-Plex Human ProcartaPlex ™ Panel (Thermo Fisher Scientific Cat. No. EPX110-10810-901) kit following the instructions of the manufacturer. The plasma levels of total soluble forms of receptors for advanced glycation end products (sRAGE) and high-mobility group box 1 (HMGB1) were determined using a commercially available ELISA kit (RAGE, R&D Systems, Minneapolis, MN, USA; HMGB1, IBL, Hamburg, Germany) according to the protocol of the manufacturer. Measurements were performed in duplicate, and the results were averaged.

Statistical Analyses
Hardy-Weinberg equilibrium between the expected and observed genotype distributions of PD patients and healthy controls was assessed by Chi-square tests. To assess the relative risk conferred by a particular allele and genotype, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. A value of p < 0.05 was considered statistically significant. All data are presented as the mean ± SD. Variables that followed a Gaussian distribution were compared with two-tailed t-tests or analysis of variance. All statistical analyses were performed with Stata 8.0 (StataCorp LP, College Station, TX, USA) software.

Transwell Coculture System
Transwell is a well-characterized 3D model extensively used in cocultures (Renaud and Martinoli, 2016). Microglia and SH-SY5Y cells were cocultured in two chambers: on top were activated microglia in Transwell insert carrying a 0.4-μm semipermeable membrane, and at the bottom were differentiated SH-SY5Y cells (Transwell plate, Corning Incorporated) (Supplementary Figure S1). Prior to coculture, 10 μM retinoic acid were employed to differentiate SH-SY5Y cells plated at a density of 50,000 cells/well on glass coverslips in sixwell plates, then washed with PBS and refreshed with 10% FBScontaining DMEM equilibrating for 30 min (Corning CLS3452, Corning, NY, USA). Human microglial cells plated in Transwell insert (1.5 × 10 5 cells per well) activated by 0.5 or 1 μg/ml of LPS (Sigma Chemical Co. Aldrich, MO, USA) for 24 h were washed by PBS, then placed on top of the differentiated GFP-tagged αsynuclein SH-SY5Y cells (Supplementary Figure S1). After 24 h of coculture, supernatants were collected for measuring cytokines, and SH-SY5Y cells were harvested for further analysis.

Real-Time Live Cell Confocal Imaging
Live cell confocal imaging of SH-SY5Y cells was performed using an inverted confocal laser scanning microscope (LSM780; Carl Zeiss microimaging, Inc.) equipped with an incubation system and a Plan-Apochromat 20x/0.8M27 dry immersion objective. Live imaging was performed under an incubation system at 37°C and 5% CO 2 , and two to five images (×20) per sample were taken using the phasecontrast and green fluorescence mode. For assessment of neurite length in SH-SY5Y cells, phase-contrast live cell imaging was used. The image scale was converted from pixel units into micrometers (µm) using ImageJ software (version 1.41, Rasband WS, ImageJ, National Institutes of Health, Bethesda, MD, USA). The length of neurites was traced and measured from the distal end of the neuron growth cone to the tip of the neurite as previously described (Lin et al., 2016). A total of 80-100 neurites were analyzed per group. The somatic GFP-tagged α-synuclein punctate signal represented αsynuclein aggregations from individual cells. A total of 100-150 cells were analyzed per group.

Immunofluorescence Labeling Analysis
SH-SY5Y cells stably expressing GFP-tagged α-synuclein were placed at a density of 50,000 cells/well in six-well plates, grown on glass coverslips, and then fixed in 4% PFA (Sigma Chemical Co. Aldrich, St. Louis, MO, USA). Cells were then incubated in phosphate-buffered saline (PBS) containing 0.2% Triton X-100, 1% BSA, and perspective antibodies at 4°C overnight, followed by secondary antibody incubation at room temperature for 30-60 min. After rinsing, the glass coverslips were mounted using ProLong Gold Antifade with DAPI (Invitrogen, CA, USA).

Image Analysis
Fluorescence intensity was measured using corrected total cell fluorescence (CTCF) and analyzed by ImageJ software (version 1.41; National Institute of Health, Bethesda, Maryland, United States) and calculated using the following formula: CTCF = Integrated Density-(Area of selected cell × Mean fluorescence of background signal). Co-localization was analyzed using IMARIS x64 v9.5.4 software (BITPLANE, Oxford Instruments) (Costes et al., 2004) provided by the First Core of College of Medicine, National Taiwan University.

RNA Isolation and Quantitative Real-Time Polymerase Chain Reaction
The mRNA levels of IL-1β, IL-12, and IL-18 in microglial cells before and after stimulation with LPS were determined by quantitative RT-PCR. Total RNA from human microglial cells was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA) and reverse-transcribed into cDNA using SuperScriptTM III reverse transcriptase (Invitrogen, Carlsbad, CA, USA). Quantitative RT-PCR was performed on a cDNA amount equivalent to 100 ng of total RNA using SYBR Green PCR master mix (Applied Biosystems, Foster City, CA, USA) and was performed on ABI Prism 7,900 (Applied Biosystems, Foster City, CA, USA). The housekeeping gene is GAPDH.

Western Blot Analysis
Cocultured SH-SY5Y cells were collected by trypsinization from a Transwell system, and cell lysates as well as western blot analysis were conducted as previously described (Lin et al., 2015). Antibodies and qPCR primer sequences are listed in Supplementary Materials.

Statistical Analyses
Statistical analysis for more than two groups, one-way ANOVA multiple comparison was used. Between the two groups, t-test was used. All data are presented as the mean ± SEM. Statistical significance was defined as a p-value of less than 0.05. All statistical analyses were performed with Stata 8.0 (StataCorp LP, College Station, TX, USA) software.

Plasma Levels of Cytokines are Increased in Parkinson's Disease Patients
Patients with PD and healthy controls were recruited from National Taiwan University Hospital, and their genetic variants for TLRs and plasma levels of cytokines were examined (Supplementary Table S1 and Table 1). In the multiple comparison shown in Table 1, the cutoff p-value for the statistical significance was 0.05/13 = 0.0038. In this regard, plasma level IL-2, IL-4, IL-18, IFN-γ, TNF-α, and GM-CSF were increased in the PD group compared with those of controls ( Table 1), while that of IL-1β (p = 0.01) and IL-12p70 (p = 0.01) between PD and controls did not reach the statistical difference. In the study cohort, most of the genetic variants of TLR1, TLR2, TLR4, and TLR6 were wild type, except rs4986790 (c.896A > G p. D 299G) in TLR4 and rs5743810 (c.745 T > C S249P) in TLR6 (Supplementary Table S2). There was a modest difference in the genotype distribution of PD patients compared with the age-and sex-matched control subgroup (p = 0.04) in the nonsynonymous variant of TLR4 (c.896A > G p. D 299G; rs4986790), and the frequency of the G allele was higher in the PD group (odds ratio (OR) = 3.03, 95% CI j = 1.03-9.45, p = 0.04) (Supplementary Table S2). These results suggest that the systemic inflammatory response was increased in PD patients and that TLR-4 genetic variants might increase the risk of PD.

DISCUSSION
In this study, TLR4 genetic variants and an elevation of the systemic inflammatory response in PD patients were revealed. A Transwell coculture system comprising activated microglia and differentiated SH-SY5Y cells, which more mimics the in vivo condition is employed to explore the underlying mechanism on how neuroinflammation induced upregulation of neuronal TLRs to drive synucleinopathy. Activated microglia-mediated neuroinflammation upregulated neuronal TLR2 and TLR4 to activate the p38/JNK pathway resulting in the inhibition of autophagic flux. This cascade leads to α-synuclein aggregation and neurite shortening as illustrated in Figure 6. Postmortem and animal models of PD have revealed an involvement of neuroinflammation in the neurodegeneration of PD (Hirsch and Hunot, 2009;Lin et al., 2019). Current increase in plasma IL-1β, IL-2, TNF-α, and IFN-γ in PD patients is consistent with the previous findings in spinal cerebrospinal fluid and postmortem brain (Nagatsu et al., 2000;Qin et al., 2016). Previous studies have shown that the plasma levels of HMGB1 and sRAGE were increased in patients with neurodegenerative disorders, including PD (Santoro et al., 2016;Yang et al., 2018). However, we failed to observe the difference between the PD and the control groups. Future studies enrolling more participants with a longitudinal study design are needed to clarify their role in the pathogenesis of PD.
Neuronal TLR2 was expressed and activated in differentiated dopaminergic neuron-like SH-SY5Y cells, primary human neural progenitor cells, and mature neurons (Dzamko et al., 2017). In contrast to microglia and astrocytes, the role of neuronal TLR in neuroinflammation and downstream signaling cascades still needs to be investigated (Tang et al., 2007;Rietdijk et al., 2016). The increased TLR2 in postmortem brain neuron of PD patients was confirmed as well as the correlation and localization of TLR2 with the accumulation of pathological αsynuclein (Dzamko et al., 2017). They studied the pathway mediating disease pathogenesis by treating TLR2/TLR1 dimer agonist PAM3CSK4 to RA-differentiated SH-SY5Y cells in the FIGURE 6 | Schematic representation of a potential mechanism that neuroinflammation upregulated neuronal TLR2/4 expression to drive α-synuclein aggregation in neurodegeneration disease, such as Parkinson's disease (PD).
Frontiers in Pharmacology | www.frontiersin.org March 2022 | Volume 13 | Article 845930 absence of microglia. Although the increase in α-synuclein protein at 3-and 7-day treatments in SH-SY5Y cells was found, the western blots were not consistent (Dzamko et al., 2017). Authors admitted that intrinsic production of inflammatory cytokines by neuronal SH-SY5Y cells was insufficient to promote cell death and suggested that activation of TLR2 on neurons requires amplification of the inflammatory response following the recruitment and activation of microglia (Dzamko et al., 2017). Actually, activated microglia are prevalent in the pathologically effected areas in brains of patients with PD and Alzheimer's disease (McGeer et al., 1988;Hughes et al., 1992). Our Transwell coculture system comprising microglial cells and differentiated SH-SY5Y cells demonstrated that microglial cells are critical for the expression of neuronal TLR2/TLR4. Only activated microglia mediating neuroinflammation can upregulate neuronal TLR2 and TLR4, while non-activated microglia (in the absence of LPS-activation) fail to exhibit such effects after 24-h coculture. In addition, our human study revealed that PD patients show increased inflammatory responses as well as TLR4 genetic variant, consistent with the report that a genetic variant of TLR4 is associated with the risk of PD (Zhao et al., 2015). TLR activation is reported to inhibit autophagic activity (Oh and Lee, 2014;Kim et al., 2015), and fibrillar α-synuclein was also shown to reduce autophagic flux activity (Tanik et al., 2013). TLR2 ablated autophagy activity via accumulation of autophagosomes, and p62 was reported to result in the accumulation of neuronal α-synuclein (Kim et al., 2015). Our findings further demonstrated that neuroinflammation upregulated TLR2 and TLR4 in SH-SY5Y cells to activate the p38/JNK pathway, leading to the increase in phosphorylation and accumulation of α-synuclein proteins and shortening neurites. Under neuroinflammation, we also found an increase in the autophagy marker, beclin-1, and LC3b-I/II, indicating the formation of autophagosomes. Furthermore, an elevation of p62 parallel with a decrease in the cytosolic co-localization of LC3b-I/II and LAMP-1 were seen ( Figures 4A, B), suggesting the disruption of autolysosome formation by dampening the fuse of lysosome with autophagosomes. In the presence of TLR2/4 inhibitors, only p62 but not LC3b (autophagosomes formation) was reduced, indicating that dampened autophagic flux might occur and explain the aggregation of α-synuclein, and neurite shortening in neuronal cells. However, in the absence of microglia, RA-differentiated SH-SY5Y cells upon PAM3CSK4 treatment only showed elevation of p62 but not beclin 1 and LC3 (Dzamko et al., 2017).
We demonstrated that p38 and JNK activations were downstream of TLR2/4 to reduce autophagy. JNK and p38 were reported to be involved in different neurodegenerative diseases through mediating apoptosis (Oh-Hashi et al., 1999). In contrast, pharmacological inhibition of p38 increases autophagy (Boland et al., 2008;Schnoder et al., 2016), and downregulation of p38α induces lysosomal activation to consequently reduce amyloid-β accumulation in an animal model of Alzheimer's disease (AD) (Schnoder et al., 2016). We found that the specific inhibitor, BIRB-0796, inhibited neuronal p38 activation and decreased p62 level as well as phosphorylated α-synuclein. However, among a collection of small-molecule inhibitors targeting TLR signal transduction pathway, BIRB-0796 had no effect on p62 level despite reducing α-synuclein protein (Dzamko et al., 2017). A recent report also indicated that p38 mediated microglial activation through inhibiting autophagy (He et al., 2018). Consistent with a previous finding that neuronal TLR2 is involved in the inhibition of autophagic activity to accumulate αsynuclein aggregates in neurons (Kim et al., 2015), our findings extend current knowledge to reveal that both neuronal TLR2 and TLR4 act through the p38/JNK pathway to disrupt the formation of autolysosome, and also partly explain the recent report that TLR4 mediates inflammation in the gut and neuronal α-synuclein accumulation in the enteric nervous system (Perez-Pardo et al., 2019).

CONCLUSION
Our study demonstrates the role of neuroinflammation-induced upregulation of neuronal TLR2 and TLR4 in α-synuclein pathology, and neurodegeneration through regulating p38/JNK pathway to perturb the autophagy flux (summarized schematically in Figure 6). The TLR2/4 pathway could be a mechanism-based therapeutic strategy for neurodegenerative disease such as PD.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the institutional ethics review board of the National Taiwan University Hospital (Taipei, Taiwan; (the protocol number is 201703010RINA). The patients/participants provided their written informed consent to participate in this study.