Suppression of the JAK/STAT Pathway Inhibits Neuroinflammation in the Line 61-PFF Mouse Model of Parkinson’s Disease

Parkinson’s disease (PD) is characterized by neuroinflammation, progressive loss of dopaminergic neurons, and accumulation of a-synuclein (a-Syn) into insoluble aggregates called Lewy pathology. The Line 61 a-Syn mouse is an established preclinical model of PD; Thy-1 is used to promote human a-Syn expression, and features of sporadic PD develop at 9–18 months of age. To accelerate the PD phenotypes, we injected sonicated human a-Syn preformed fibrils (PFFs) into the striatum, which produced phospho-Syn (p-a-Syn) inclusions in the substantia nigra pars compacta and significantly increased MHC Class II-positive immune cells. Additionally, there was enhanced infiltration and activation of innate and adaptive immune cells in the midbrain. We then used this new model, Line 61-PFF, to investigate the effect of inhibiting the JAK/STAT signaling pathway, which is critical for regulation of innate and adaptive immune responses. After administration of the JAK1/2 inhibitor AZD1480, immunofluorescence staining showed a significant decrease in p-a-Syn inclusions and MHC Class II expression. Flow cytometry showed reduced infiltration of CD4+ T-cells, CD8+ T-cells, CD19+ B-cells, dendritic cells, macrophages, and endogenous microglia into the midbrain. Importantly, single-cell RNA-Sequencing analysis of CD45+ cells from the midbrain identified 9 microglia clusters, 5 monocyte/macrophage (MM) clusters, and 5 T-cell (T) clusters, in which potentially pathogenic MM4 and T3 clusters were associated with neuroinflammatory responses in Line 61-PFF mice. AZD1480 treatment reduced cell numbers and cluster-specific expression of the antigen-presentation genes H2-Eb1, H2-Aa, H2-Ab1, and Cd74 in the MM4 cluster and proinflammatory genes such as Tnf, Il1b, C1qa, and C1qc in the T3 cluster. Together, these results indicate that inhibiting the JAK/STAT pathway suppresses the activation and infiltration of innate and adaptive cells, reducing neuroinflammation in the Line 61-PFF mouse model.


INTRODUCTION
Parkinson's disease (PD) is the most common motor neurodegenerative disorder, affecting 2-3% of the population over 65 years of age (1,2).Immune system dysfunction, genetic mutations, and environmental factors contribute to its pathogenesis (3), while polymorphisms in the HLA-DR (MHC Class II) locus increase risk of its development (4).Studies show that anti-tumor necrosis factor (TNF) and nonsteroidal anti-in ammatory treatments are associated with a reduced incidence of PD symptoms (5)(6)(7).
The core pathological features of PD are neuroin ammation (8), a loss of dopamine-producing neurons, and the accumulation of misfolded, aggregated α-synuclein (α-Syn) in the substantia nigra pars compacta (SNpc), called Lewy pathology (1).In postmortem tissue, Lewy pathology is associated with enhanced MHC Class II expression on microglia and in ltration of macrophages and T-cells in the midbrain (9)(10)(11).Neuroimaging studies con rm chronic myeloid activation in the brains of PD patients (12), and increased numbers of α-Syn-reactive T-cells circulating in the blood of PD patients demonstrate immune cell activation (13)(14)(15).
Research has focused on the neuroin ammatory features of PD, particularly endogenous microglia (MG) and monocytes/macrophages (MM), which in ltrate from the periphery (3,16).A recent study associated an increase in classical and Toll-like receptor-positive monocytes in early-moderate PD with rapid progression, greater clinical impairment, and early cognitive decline (17).Animal models and clinical studies of PD correlate cytokine dysregulation, speci cally, elevated levels of IL-6 and IFN-γ, with the degeneration of dopamine neurons (16).Genomic studies associate PD with mutations/variants of LRRK2, MHC Class II (18), and PINK1 (19,20), genes implicated in immune cell function.
Studies show that the adaptive immune system is essential to PD pathogenesis (3,(21)(22)(23)(24).Two major classes of lymphocytes, T-cells and B-cells, induce adaptive immunity, conferring antigen speci city and immunological memory (16).Patients with PD have elevated numbers of CD4 + and CD8 + T-cells in the ventral midbrain compared to healthy controls (9).T-cells from PD patients initiate an immune response to post-translationally modi ed α-Syn, implicating an autoimmune reaction to intrinsic antigens (15,21).
Moreover, a recent study associated α-Syn-speci c T-cell reactivity with preclinical and early PD, and proin ammatory CD4 + T-cells are most abundant shortly after diagnosis of motor symptoms (13).
The JAK/STAT signaling pathway, in which Janus Kinases (JAKs) and Signal Transducers and Activators of Transcription (STATs) proteins interact with cytokine receptors, plays a critical role in the activation and regulation of immune responses (25,26).Its dysregulation (i.e., over-activation) is associated with many pathological conditions, including multiple sclerosis (MS), rheumatoid arthritis, in ammatory bowel disease, and many cancers (27,28).Emerging evidence indicates that it polarizes myeloid cells and T-cells to pathogenic phenotypes (29).We previously demonstrated that inhibiting the JAK/STAT pathway ameliorates disease severity in several preclinical models of MS (30) and reduces neuroin ammation in a PD model in which adeno-associated virus (AAV) human α-Syn mediates overexpression of α-Syn (31).However, virally induced α-Syn overexpression can cause up to a 30-fold increase in α-Syn levels and thus does not model idiopathic PD.A pre-clinical model that better mimics PD neuroin ammatory processes is needed to de ne how inhibiting the JAK/STAT pathway mediates protection against neuroin ammation.The Thy1-α-Syn transgenic mouse, or Line 61, uses the Thy1 promoter to express full-length, wild-type, human synuclein.It reproduces many features of sporadic PD, including progressive changes in dopamine release in striatal content, α-Syn pathology, motor and nonmotor de cits, neuroin ammation, and biochemical and molecular changes similar to those observed in PD (32).However, the mice do not develop these phenotypes until they are 9-18 months old (32,33).In another PD model, small, brillar seeds of mouse α-Syn induce wild-type mice to form inclusions that biochemically and morphologically resemble those found in PD brains.However, while microglia are activated, this preformed bril (PFF) model shows minimal expression of the MHC Class II-positive cells that characterize PD (34,35).
Hypothesizing that introducing human PFFs into mice expressing human α-Syn would better replicate human PD, we injected PFFs generated from puri ed, recombinant human α-Syn into Line 61 mice (Line 61-PFF).Injection of hu-α-Syn PFFs produced α-Syn inclusions in the SNpc and striatum, induced MHC Class II expression and enhanced innate and adaptive immune cell in ltration and activation in the midbrain, in contrast to injection of monomeric hu-α-Syn.
The Line 61-PFF model enables us to evaluate the contribution of the JAK/STAT pathway to neuroin ammatory responses, and the impact of inhibiting this pathway.Administering the JAK1/2 inhibitor AZD1480 suppressed neuroin ammation and reduced p-α-Syn inclusions.Single-cell RNA Sequencing (scRNA-Seq) revealed that PFF injection induced a speci c monocyte/macrophage (MM) cluster expressing antigen-presentation genes and a T-cell cluster expressing numerous proin ammatory genes which correlate with neuroin ammation.Inhibiting the JAK/STAT pathway abrogated the presence of these two PFF-induced immune cell clusters.These ndings elucidate neuroin ammatory mechanisms and may inform the development of more speci c therapeutic approaches and/or diagnostic biomarkers for patients with PD.

MATERIALS AND METHODS
Mice.The Thy1-α-Syn (Line 61) mouse overexpresses full-length, human, wild-type α-Syn under the Thy-1 promoter.The model is bred by crossing female Line 61 mice with male hybrid B6D2F1 (BDF1) mice and maintained on a congenic background (32).Since the transgene is inserted in the X chromosome, and random inactivation of the X chromosome carrying the mutation in female mice may occur (36), experiments used only littermate males with a Thy1-α-Syn-positive transgene.The University of Alabama at Birmingham (UAB) Institutional Animal Care and Use Committee approved all animal research protocols.
Human α-Syn Monomer and PFFs.Human monomeric α-Syn (monomer) was puri ed using sizeexclusion chromatography followed by anion exchange.The Pierce high-capacity endotoxin kit was used to remove endotoxins to < 0.1 EU/µg.α-Syn PFFs were generated as described (37).Monomer concentration was measured using A280 and the extinction coe cient of 5960 M − 1 cm − 1 .Monomer was diluted to 5 mg/ml in 150 mM KCl and 50 mM Tris-HCl buffer and shaken for seven days to generate PFFs.After seven days, PFF protein concentration was determined and previously noted buffer was used to bring concentration to 5 mg/ml.On the day of stereotaxic injections, PFFs were sonicated using a cup horn sonicator (QSonica) with a 15°C water bath.Dynamic light scattering (Wyatt Technology) con rmed < 50 nm fragmentation of PFFs.Immediately before injection, 5 mg/ml of monomer was spun at 20,000 x g, and only the supernatant was injected to prevent aggregate formation.
Intracranial Stereotaxic Injections and Treatment with the JAK1/JAK2 Inhibitor AZD1480.Male Line 61 mice between the age of 12-15 weeks were anesthetized using iso urane and received stereotaxic, unilateral, intrastriatal injections using the coordinates relative to bregma: 1 mm A/P, 2 mm M/L, and − 3.2 mm D/V relative to the skull (37).Those used in the ow cytometry, scRNA-Seq, and phospho-STAT experiments received bilateral intrastriatal injections for immune cell collection.Two microliters of sonicated PFFs (5 mg/ml) or monomeric α-Syn (5 mg/ml) were injected at 0.5 µl/min with a Hamilton syringe.Mice received 10 units of 1.5 mg/ml carprofen subcutaneously at the time of surgery and the following day.
Two weeks after PFF injections, Line 61-PFF mice were administered either AZD1480 by oral gavage at 25 mg/kg or 0.1% DMSO (Sigma-Aldrich) as a vehicle control every day.The mice were sacri ced two or four weeks later (four-or six-weeks post injection [wpi]), depending on the analysis performed.For examination of TH positive neurons, the mice were sacri ced at a 12-week time point.At time of sacri ce, they were transcardially perfused with cold PBS (pH 7.4), followed by 4% paraformaldehyde for immuno uorescence or cold PBS alone for other experiments.
Immuno uorescence and Immunohistochemistry. Prior to sectioning, brains were post xed in 4% PFA overnight at 4°C, followed by cryoprotection in 30% sucrose/PBS until tissues sank to the bottom.Whole brains were frozen in N-methylbutane at -50°C.Coronal sections 40-µm thick representing the entire brain were serially collected and stored in 50% glycerol in 0.01% sodium azide in Tris-buffered saline (TBS) at -20°C.
Free-oating immunohistochemistry was performed as detailed previously (38).Brain sections were labeled with anti-tyrosine hydroxylase (Millipore, ab152), a marker of dopaminergic neurons in mouse midbrain.On day two the sections were labeled with a biotinylated secondary followed by HRPconjugated ampli cation complex.DAB chromogenic substrate deposition was timed and visualized by eye (Vector, SK-4100) before mounting on positively charged slides, which were dehydrated and coverslipped using Permount (Fisher, SP15-500).
Unbiased Stereology.To quantify TH neurons in the SNpc, unbiased stereology was performed as described (38).A reviewer blinded to condition coded and analyzed 5-6 TH-DAB-stained slides encompassing the rostro caudal SNpc using an Olympus BX51 microscope and the optical fractionator probe in the StereoInvestigator software (MBF Bioscience).Both ipsilaterally injected and contralaterally uninjected sides of the SNpc were quanti ed.TH + neurons within the SNpc contours on a 100 µm x 100 µm grid were counted at an optical dissector height of 22 µm.Weighted section thickness was used to account for variations in tissue thickness.Bright eld images of TH + neurons in the SNpc were acquired using the Olympus BX51 microscope.
Phospho-Synuclein Aggregate Quantitation.After immuno uorescence staining with anti-α-phospho-Serine129 (p-α-Syn) as described (37,39), 20X tiled, ipsilateral SNpc confocal images were obtained using the Nikon Ti2-C2 microscope.An ImageJ cell counter was used to count each p-α-Syn-positive neurite and inclusion in the SNpc, which was delineated by drawing a contour using TH + neurons as a reference and then splitting the channels so only p-α-Syn channel uorescence remained.Two to six SNpc sections were analyzed per animal, depending on their availability and quality after staining.Mononuclear Cell Isolation.Six weeks after bilateral injection of monomer or PFFs, mononuclear cells in the ventral midbrain were isolated as described (31,40).Midbrain tissue was passed through a 100-µm lter to obtain a single-cell suspension, and mononuclear cells were isolated using a 30/70% Percoll gradient.
Flow Cytometry.To detect surface proteins, mononuclear cells were incubated with Fc Block (Bio X Cell, 2.4G2) for 15 min and washed, followed by incubation with viability dye.
Immunoblotting.Thirty µg of midbrain mononuclear cell homogenate was separated by electrophoresis and probed with antibodies as described (30).All immunoblots represent three individual experiments.
Single-cell RNA Sequencing (scRNA-Seq).Mononuclear cells were isolated for sequencing as described above with 3-5 ventral midbrains pooled per sample and sorted for CD45 + live cells on a BD FACsAria.
Sorted cells were loaded onto the 10X Chromium platform (10X Genomics), and libraries constructed using the Single Cell 3′ Reagent Kit V3.1 according to the manufacturer's instructions.At least three biological replicates for each group were processed separately (PBS ).Samples were sequenced at an average depth of 20,000 reads per cell using Illumina NextSeq 500.Raw base call les were demultiplexed into FASTQ les.Sequencing les were processed and mapped to mm10, and count matrices were extracted using the Cell Ranger Single Cell Software (v 7.1.0)(42).
scRNA-Seq Analysis.The count matrices in the h5 le format were imported into the Partek Flow (Partek, Inc.) pipeline (43).Single-cell quality control was performed by applying an inclusion lter on counts per cell (500-15000) and detected genes per cell (250-5000).Cells with more than 10% mitochondrial gene expression were excluded to eliminate apoptotic or dying cells (44).The noisereduction lter was also applied to exclude features where the value ≤ 0 is in at least 99.9% of cells.The ltered dataset was normalized and scaled with SCTransform work ow.
Principal component analysis (PCA) was performed on the SC-scaled data to graph clustering based on the Louvain algorithm, with the PCA number set to 20.The data were visualized by 3D uniform manifold approximation and projection (UMAP) dimensional reduction using the rst 20 principal components.Cell annotations for each cluster were determined using the top differentially expressed genes (DEGs) in computed biomarkers and canonical markers following the classi cation work ow in Partek Flow (45).ScRNA-Seq data will be available online.The single-cell data have been deposited in the GEO under the accession number GSE264525.Raw les supporting our ndings are available from the corresponding authors upon reasonable request.

RESULTS
Novel Pre-clinical PD Model (Line 61-PFF).We previously showed that the JAK1/2 inhibitor AZD1480 prevents neuroin ammation and protects against dopaminergic neuron loss in response to AAV-humanα-Syn overexpression in rats (31).To better mimic idiopathic PD, we tested mouse PFFs and found that injection into mouse striatum activated microglia but did not induce the accumulation of MHC Class IIexpressing immune cells (data not shown) that occur in human PD brains (49).We therefore tested whether injecting human α-Syn PFFs into the striatum of Line 61 mice, which use the Thy1 promoter to express human α-Syn (32), would induce more robust in ammatory phenotypes.Using an antibody to α-Syn phosphorylated at serine 129 (p-α-Syn), we identi ed α-Syn inclusions in the SNpc four weeks after injecting Line 61 mice with human PFFs (Figs. 1A, B).We also observed abundant MHC Class II-positive immune cells in the SNpc (Figs. 1A, 1C).Injecting monomeric α-Syn into Line 61 mice did not produce pα-Syn inclusions or MHC Class II-positive cells (Fig. 1A).
We tested the JAK1/2 inhibitor AZD1480 in the Line 61-PFF model as described (31).Two weeks after PFF injections, AZD1480 or VH was administered for two weeks, then mice were sacri ced for immuno uorescence studies.Figures 1A-C show that AZD1480 treatment signi cantly reduced α-Syn inclusions and the number of MHC Class II-positive cells in the SNpc.
Sections from the SNpc were stained for TH to determine the number of dopaminergic neurons in mice injected with monomer or PFFs and treated with VH or AZD1480.Line 61-PFF mice did not exhibit a loss of TH-positive neurons compared to monomer-injected mice at a 12-week time point, and AZD1480 treatment produced no signi cant differences (Supplemental Figs.1A, B).
Neuroin ammation in Line 61-PFF Mice.Injecting human PFFs into the striatum of Line 61 mice increased MHC Class II positive immune cells in the SNpc, and AZD1480 treatment reduced the abundance of these cells (Fig. 1).Immune cell phenotypes in the midbrain were further characterized by ow cytometry (Figs.2A, B).In Line 61-PFF mice, total immune cells in the midbrain were signi cantly higher than in monomer α-Syn-injected mice (Fig. 2C).We identi ed microglia as CD45 Mid CD11b + , macrophages as CD45 Hi CD11b + , dendritic cells (DCs) as CD45 Hi CD11b + CD11c + , and lymphocytes as CD45 + CD11b − .In Line 61-PFF mice, macrophages, DCs, and lymphocytes showed signi cant increases compared to monomer-injected mice, but the total number of microglia did not increase (Fig. 2C).
We assessed the activation status of innate immune cell subsets by determination of MHC Class II expression.Absolute numbers of MHC Class II positive microglia, macrophages, and DCs were signi cantly higher in the midbrain of Line 61-PFF mice than in the monomer group (Fig. 2D).
We next determined how AZD1480 treatment in uenced speci c subsets of immune cells (Figs. 2A, B).AZD1480 or VH was administered two weeks after PFF injections, and treatment continued for four weeks, when mice were sacri ced for analysis.Compared to VH, AZD1480 signi cantly suppressed absolute numbers of immune cells (microglia, macrophages, DCs, and lymphocytes), the numbers of MHC Class II positive microglia, macrophages and DCs, and the number of CD4 + T-cells, CD8 + T-cells, and CD19 + B-cells in the midbrain of Line 61-PFF mice (Figs.2C-E).These data show that inhibiting the JAK/STAT pathway suppresses the in ltration and activation of innate and adaptive immune cells.
PFF Induction of STAT3 Activation.Activation of the JAK/STAT pathway results in phosphorylation of STAT proteins (28).We assessed activation in Line 61-PFF mice by measuring tyrosine phosphorylation of STATs, particularly STAT1 and STAT3 (30,50).Immunoblotting was performed to measure protein expression levels of total STAT1, phosphorylated STAT-1 (p-STAT1), total STAT3, and p-STAT3 in mononuclear cells isolated from the midbrain.We observed a signi cant increase in p-STAT3 in Line 61-PFF mice treated with VH compared to monomer, which was inhibited by AZD1480 treatment (Figs. 3A,  B).There were no statistical differences in total STAT3, total STAT1, and p-STAT1 expression (Figs. 3A-C).Immunohistochemistry con rmed the signi cant increase in p-STAT3 + cells in Line 61-PFF mice treated with VH compared to monomer treatment, and suppression after AZD1480 treatment (Fig. 3D).These data indicate that PFF injection activates the JAK/STAT pathway as demonstrated by phosphorylated STAT3, and AZD1480 inhibits STAT3 activation in mononuclear cells.
scRNA-Seq Characterization of Immune Cell Clusters in Line 61-PFF Mice.Several scRNA-Seq studies point to the heterogeneity of cells involved in PD pathogenesis (51)(52)(53)(54).To identify cell-speci c contributions in the Line 61-PFF model, we performed scRNA-Seq on sorted CD45 + leukocytes obtained from Line 61 mice injected with PBS, monomer, PFF plus VH, or PFF plus AZD1480 (Supplemental Fig. 2).Cell clusters were annotated with the top differentially expressed biomarkers and canonical markers for microglia (MG), monocytes/macrophages (MM), T-cells (T), B-cells (B) and neutrophils (Neu) (Supplemental Fig. 3A).We observed no difference in the percentage of each cluster or total cell numbers between the PBS and monomer groups, indicating that monomer injection did not affect the immune cell subsets found in PBS-treated Line 61 mice (Supplemental Figs.3B, C).Most cells (> 80%) were identi ed as MG and separated into 9 clusters (MG1-9; Supplemental Fig. 3A).Upon PFF injection, either the percentage or absolute numbers of MG2, MG4, and MG5 clusters were higher than those in PBS-and monomer-injected mice (Supplemental Figs.3B, C).Percentages and cell numbers of MM clusters and the T-cell cluster were higher in PFF injected mice than in PBS-and monomer-injected mice (Supplemental Figs.3B, C).Neu and B-cell clusters were not examined due to low (< 1%) numbers.
We examined whether the MM clusters changed with PFF injection.MM1, MM2, MM3, and MM4 cell numbers increased in Line 61-PFF mice compared to monomer, with the MM4 cluster showing the greatest increase (Figs. 4A, B), and DEG analysis revealed upregulated expression of MM4 genes (Fig. 4C).
Collectively, these results reveal that a novel in ammatory T-cell cluster, T3, is induced in response to PFF injection, and is associated with neuroin ammatory responses.AZD1480 treatment signi cantly decreased T3 cell numbers and suppressed proin ammatory gene expression.AZD1480 also restored T1 and T5 transcriptional pro les, suggesting that the JAK/STAT pathway affects T-cells in the Line 61-PFF model of PD.

DISCUSSION
Developing a pre-clinical model of PD that closely resembles human PD is crucial in understanding the mechanisms that in uence onset and progression of the disease and identifying potential therapeutic targets.We have generated a new pre-clinical PD model by injecting human PFF into Line 61 mice that express wild-type human α-Syn.This model, Line 61-PFF, is characterized by increased expression of MHC Class II and in ltration and activation of innate and adaptive immune cells in the midbrain, thereby exhibiting a strong neuroin ammatory response.We utilized single-cell transcriptomics, a powerful research tool that enables high-resolution analysis of gene expression (45,(73)(74)(75) to evaluate the transcriptional pro les of immune cells in ltrating into the midbrain of Line 61-PFF.Most strikingly, we identi ed two unique clusters, MM4 and T3, that were induced by PFF injection.A JAK1/2 inhibitor, AZD1480, was used to investigate the effect of blocking the JAK/STAT pathway on neuroin ammation.Our results revealed an immunosuppressive effect of JAK1/2 inhibition by reducing immune cell in ltration in the brain, and inhibiting proin ammatory transcriptional pro les.These ndings indicate the therapeutic potential of JAK/STAT blockade in the treatment of neurodegenerative diseases with pronounced neuroin ammation.
To de ne the heterogeneity of immune cells involved in neuroin ammation in Line 61-PFF mice, scRNA-Seq was utilized.We identi ed 5 monocyte/macrophage (MM) clusters and a signi cant effect of PFF injection on the MM4 cluster, enhancing expression of the antigen-presentation genes H2Eb1, H2-Aa, H2-Ab1, and Cd74 as well as genes related to neurological diseases and pro-in ammatory M1 polarization, including Nav2, Nav3, Malat1, Csmd3, Ldlrad4, and Dennd4a (58-63).Neuron navigator 2 (Nav2) is highly expressed in patients with rheumatoid arthritis (RA); it is considered a pathogenic gene for RA (76), and STAT3 activation has been shown to upregulate its expression in RA synoviocytes (77).Several single-nucleotide polymorphisms (SNPs) in the Nav2 gene are associated with the risk and age at onset of AD (78).Note that STAT3 is activated in Line 61-PFF mice (Fig. 3), and AZD1480 treatment suppresses Nav2 expression in the MM4 cluster (Fig. 4C).In addition, GSEA revealed that the MM4 cluster was enriched in TNF-α signaling (Supplemental Fig. 4C), and TNF-α has been shown to signi cantly increase Nav2 expression (76).In Line 61-PFF mice, we found enriched Tnf gene expression in the PFF-induced T3 cluster (Supplemental Fig. 6A), suggesting that the increase in MM4 Nav2 expression may be induced by T3-derived TNF-α.The expression of CUB and Sushi multiple domains 3 (Csmd3) and neuron navigator 3 (Nav3), genes related to ASD (58-61), is also elevated in the MM4 cluster and suppressed by AZD1480 treatment (Fig. 4C).
Recent studies demonstrated that monocyte-derived disease in ammatory macrophages (DIMs) accumulate during aging, in ammation, and AD (79)(80)(81).The DIM-conserved transcriptional signature includes genes such as Il1a, Il1b, Tnf, Cd49f, Cd54 and Cd83 (79).The MM4 cluster induced by PFF injection expresses many of these genes.Enrichment analysis identi ed that the MM4 cluster was enriched in in ammatory response, ROS signaling, and TNF-α signaling pathways (Supplemental Fig. 4C), indicating a pro-in ammatory phenotype.These observations indicate the MM4 cluster identi ed in Line 61-PFF mice exhibits in ammatory gene signatures of DIMs (79).Importantly, AZD1480 treatment abrogated the presence of the MM4 cluster.These ndings identify the MM4 cluster as a potential therapeutic target for PD.
The MM3 cluster identi ed in Line 61-PFF mice resemble the disease-activated border-associated macrophages (DaBAMs) identi ed in another α-Syn model of PD (57).Both are characterized by high expression of genes such as Cd163, F13a1, Pf4, Mrc1, Ms4a7, and Apoe.ApoE mediates neuroin ammation and neurodegeneration in AD (82, 83).The APOE genotype directly in uences the development of α-synuclein pathology in PD dementia (84, 85).Mouse models of α-synucleinopathy showed that APOE4 exacerbated α-synuclein pathology in the absence of amyloid (86, 87).Apoe expression was elevated in the MM3 cluster (Supplemental Fig. 4A) and suppressed by AZD1480 treatment (Fig. 4C).Collectively, our results point to MM4 and MM3 as potentially pathogenic clusters and demonstrate that AZD1480 treatment inhibits proin ammatory gene expression to reduce neuroin ammation.(orthologous to human KLRB1), which resembles the human T-cell cluster C11 in peripheral blood of patients with PD (89).Another scRNA-Seq study using human PD peripheral blood showed that PDassociated cytotoxic CD4 + T-cells exhibited a signi cant increase in proportion and enhancement of IFNγ responses (53).Consistent with this nding, GSEA demonstrated that the IFN-γ response pathway was enriched in T2 and T4 clusters from Line 61-PFF mice (Supplemental Fig. 6E).
Our scRNA-Seq analysis demonstrated a novel PFF-induced T-cell cluster, T3, which speci cally expressed proin ammatory genes, including Cst3, Csmd3, C1qc, Tnf, and Il1b.AZD1480 treatment abolished the T3 cluster (Fig. 6B) and, surprisingly, restored the naïve T-cell cluster (T1) as well as expression of the Th2-related genes Gata3 and Il-4 in the T5 cluster (Fig. 6E).Il4 and Gata3 are critical for the differentiation of CD4 + Th2 cells, which have immunosuppressive functions (68, 69).Interestingly, idiopathic PD patients show low Gata3 mRNA levels (90), consistent with another study reporting signi cantly few Th2 cells in PD patients (91).IL-4 attenuates the in ammatory responsiveness of macrophages, which limits in ammasome activation, IL-1β production, and pyroptosis (92).We surmise that JAK inhibition may promote immunosuppressive functions in adaptive T-cells.However, the exact role of Th2 cells, Gata3, and Il-4 in PD remains largely unknown.The JAK-STAT pathway has a critical role in the fate of T helper cell (Th) differentiation (26, 93).STAT3 plays a crucial role in T-cell pathogenicity, promoting in ammatory responses (94)(95)(96)(97).In Line 61-PFF mice, STAT3 is activated in mononuclear cells from the midbrain, which may lead to the promotion of proin ammatory T-cells.AZD1480 treatment reduced STAT3 activation, which was associated with reduced neuroin ammation and abrogation of the T3 cluster.
In Line 61 mice, there is considerable pathology at 9-18 months of age, including accumulation of α-Syn, degeneration of TH positive neurons, microglial and astrocytic activation, elevation of in ammatory markers and CD4 + T-cell in ltration into the brain (36, 98).In Line 61-PFF mice, we observed a much earlier demonstration of α-Syn inclusions and neuroin ammatory responses at 4-6 weeks.However, no TH positive neuronal loss was observed in this model at 12 weeks following PFF injection, thus we were not able to assess the impact of inhibiting the JAK/STAT pathway on the neurodegenerative process.
Future studies will require examination of longer time points or the use of other models with signi cant dopamine neuron loss in the SNpc to determine if neurodegeneration can be in uenced by JAK/STAT pathway inhibition.

Conclusion
In summary, studies in the novel Line 61-PFF model, which displays a strong neuroin ammatory response, identi ed two speci c immune cell clusters, MM4 and T3, that express high levels of genes related to antigen presentation and neuroin ammation.We also showed the clinical potential of inhibiting the JAK/STAT pathway critical to immune cell in ltration and expression of signature genes in these pathogenic macrophage and T-cell clusters.Future studies targeting these genes/clusters may inform diagnostic markers and/or therapeutic approaches to prevent or ameliorate currently incurable neuroin ammatory and neurodegenerative diseases, such as PD.The abundance of MHC Class II-positive cells or p-a-Syn inclusions was quanti ed using Fiji.B. An independent t-test showed signi cant differences in the number of MHC Class II-positive cells between VH-and AZD1480-treated groups.C.An independent t-test showed signi cant differences in the number of p-a-Syn-positive aggregates between groups.Scale bar = 100 µm.*p < 0.05. Figure 5 Studies demonstrate the essential role of T-cells in the neuropathogenesis of PD(9,15, 16,21) and associate α-Syn-speci c, proin ammatory CD4 + T-cell reactivity with preclinical and early PD(13).Several scRNA-Seq studies have focused on T-cells isolated from the blood and cerebrospinal uid of PD patients, demonstrating the differentiation and expansion of peripheral CD8 + T-cells and CD4 + T-cells, as well as the interaction of peripheral CD4 + T-cells with endothelial cells in PD patients(53, 88, 89).A recent snRNA-Seq study revealed an increased frequency of T-cells within the SNpc from postmortem PD patients(52).However, little is known about the transcriptional features of T-cells in PD, especially Tcells in ltrating the CNS.We identi ed increased percentages and numbers of T-cell clusters in Line 61-PFF mice (Supplemental Fig.3and Figs.6A, B).Wang et al revealed activated and expanded T-cell populations in the blood of PD patients compared to healthy controls (89).The T-cell cluster 11 in their study was enriched in KLRB1 (kill cell lectin like receptor B1) expression with clonal expansion in the blood.Our scRNA-Seq analysis showed that the T2 cluster had elevated expression of Klrb1b and Klrb1c

Figures
Figures

Figure 1 Increased
Figure 1

Figure 3 AZD1480
Figure 3 (47)g samples were determined by the Hurdle model on log2 normalized counts.Dot and violin plots were generated with sc.pl.dotplot and sc.pl.violin functions in Scanpy (1.9.1) (46) using annotated h5ad les exported from Partek work ow.The MM1 and MM2 clusters were subsets of the original annotated data node in Partek.They were reclustered as MM1 to MM5 following standard PCA and graph-based processes on SCTransformscaled data and visualized with UMAP.The T-cell cluster was a subset of the original annotated data node in Partek and reclustered as T1 to T5(47).
Pathway Enrichment Analysis.GSEA (gene set enrichment analysis): DEG between individual MG, MM, or T clusters was determined via the Partek gene-speci c analysis (GSA) test.The exported DEG list was ranked by -log(P) and converted to an RNK le uploaded to GSEA software (version 4.3.2,BROAD Institute).GSEAPreRanked chooses a hallmark gene-set database (48).Pathway analysis results were plotted in terms of a normalized enrichment score (NES) and a false discovery rate (FDR) using the ggplot2 (version 3.4.0)package in RStudio.Statistical Analysis.Flow cytometry experiments used 3-5 independent samples per group, with two ventral midbrains per sample (6-10 mice per experiment).Data were analyzed using an unpaired t-test (two-tailed) or two-way ANOVA with Tukey's multiple comparison test with nested design.Graphs display the individual values and mean ± SEM, with *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.Data Set Availability.
AZD1480 Treatment Reduces MG Cluster Cell Numbers but Does Not In uence Transcriptional Pro les in Line 61-PFF Mice.Microglia play a critical role in PD pathogenesis (64-66), so we assessed their transcriptional pro les in Line 61 mice.MG1, MG3, MG4, and MG6 were the major MG clusters found in the PBS and monomer groups, with no differences in the percentage or cell numbers of the 9 MG clusters (Supplemental Figs.3B, C).All MG clusters expressed the canonical gene markers P2ry12 and Cx3cr1.MG1 strongly expressed Klf2 and Egr3, MG3 expressed Maf and Slc2a5, and MG5 expressed Spp1 and Apoe.Although MG2 primarily upregulated Gm42418 and Cmss1, and MG4, Gng10 and Alox5ap, neither does so exclusively.MG6 upregulated two genes, Snx29 and Anks1, and downregulated Maf and Slc2a5.The MG7 cluster upregulated Stat3 and Cd83 and downregulated Klf2 and Egr3.MG8 expressed Top2a and Mki67, and MG9 uniquely expressed I t2, I t3, Oasl, and Irf7 (Supplemental Figs.5A