Skip to main content
NCBI home page
NPJ Parkinson's Disease logoLink to NPJ Parkinson's Disease
. 2026 Mar 6;12:99. doi: 10.1038/s41531-026-01310-1

LRRK2 kinase mediates increased GCase activity in microglia in response to IFNγ-induced proinflammatory stimulation

Emma J MacDougall 1, Carol X-Q Chen 2, Eric Deneault 2,3, Zhipeng You 2, David Kalaydjian 2, Narges Abdian 2, Thomas M Durcan 2, Konstantin Senkevich 1,4, Ziv Gan-Or 1,4, Edward A Fon 1,✉
PMCID: PMC13087243  PMID: 41786767

Abstract

Variants in the LRRK2 and GBA1 genes are among the most common risk factors associated with Parkinson’s disease (PD). Both patients carrying PD-associated variants in GBA1, encoding lysosomal enzyme glucocerebrosidase (GCase), and a subset of non-carrier patients have been shown to have reduced GCase enzymatic activity, suggesting that reduced GCase activity may be a feature of both genetic and a subset of sporadic PD. However, the effect of PD-associated variants in LRRK2, encoding a serine/threonine kinase, on GCase activity remains controversial, with conflicting results in various tissues and cell types. Moreover, rare patients carrying both GBA1 and LRRK2 risk alleles seem to have a more benign disease course than carriers of GBA1 variants alone, suggesting a complex interplay between these two genes in PD. Here, we evaluate the effect of LRRK2 kinase activity on GCase activity in human induced pluripotent stem cell (iPSC)-derived microglia (iMGs), a PD-relevant brain cell type expressing high levels of LRRK2. Using CRISPR editing, isogenic control iPSC lines were generated to match PD patient-derived iPSC lines harbouring the LRRK2 p.G2019S, p.M1646T, or p.N551K-p.R1398H protective haplotype variants. Whereas iMGs harbouring the p.M1646T variant, and the protective haplotype, respectively increased and decreased phosphorylation of canonical LRRK2 substrate, Rab10, GCase protein levels and activity were not altered in any of the LRRK2 variant lines. Additionally, whereas pharmacological inhibition of LRRK2 kinase activity had no impact on GCase activity in iMGs under basal conditions, it attenuated the increase in GCase activity elicited in response to interferon γ (IFNγ) treatment. Moreover, GCase activity induced by IFNγ was reduced in PD risk LRRK2 p.M1646T iMGs and increased in p.N551K-p.R1398H protective haplotype iMGs compared to their isogenic corrected controls, congruent with their respective effects on LRRK2 kinase activity and PD risk. Thus, our data suggest a role for LRRK2 kinase activity in regulation of GCase activity in response to neuroinflammation.

Subject terms: Cell biology, Diseases, Genetics, Molecular biology, Neurology, Neuroscience

Introduction

Parkinson’s disease (PD) is marked by death of dopamine neurons in the substantia nigra pars compacta region of the midbrain, resulting in the characteristic motor symptoms of the disease1. Current PD therapeutics are symptom-modifying, lose efficacy over time, and do not alter disease progression1. Among the most common genetic causes and risk factors for PD are mutations in the LRRK2 and GBA1 genes2.

LRRK2 encodes a dual-function GTPase—Ser/Thr kinase, leucine-rich repeat kinase 2 (LRRK2). LRRK2 PD pathogenic variants, including the most common p.G2019S, lead to increased kinase activity in vivo3. Additionally, LRRK2 harbours variants that alter PD risk. The p.M1646T variant increases risk, reaching significance in the most recent PD genome wide association study (GWAS), while the p.N551K-p.R1398H protective haplotype is associated with decreased risk of PD4–7. The p.M1646T variant is associated with increased kinase activity, and the protective haplotype with reduced kinase activity8–10. Substrates of LRRK2 kinase activity include a subset of Rab GTPases11–13. LRRK2 has been implicated in many different cellular processes; notably, along with downstream Rabs, in the autophagy lysosomal pathway (ALP)14–20. There is emerging evidence for a role of LRRK2 in immune function. LRRK2 is highly expressed in immune cells and has been implicated in their response to inflammatory stimuli21–25. Further, proper function of the ALP and response to lysosomal damage induced by pathogens are key to mediating the immune response14,26,27. Additionally, a common non-coding genetic variant in the 5’ region of LRRK2, found to be associated with increased PD risk by GWAS, has been shown to increase LRRK2 expression in microglia but not in other brain cell types28.

Biallelic variants in GBA1, the gene encoding lysosomal enzyme glucocerebrosidase (GCase), may lead to Gaucher disease (GD). In GD, a loss of GCase function results in varied and, at times, neurologic disease manifestations. Heterozygous GBA1 variants have been associated to PD29–32. Unlike GD, the mechanism of pathogenicity of GBA1 variants in PD remains less clear; however, reduced GCase activity has been reported in PD patients with and without GBA1 variants33–36.

GCase activity has been shown to be increased in whole blood of carriers of the LRRK2 p.G2019S or p.M1646T variants, while the p.N551K-p.R1398H protective haplotype is nominally associated with decreased GCase activity7,33. Conversely, decreased GCase activity has been observed in iPSC-derived dopamine and cortical neurons carrying LRRK2 hyperactive variants37,38. Additional studies utilizing LRRK2 knockout mice, GBA1 mutant mice, or double LRRK2 and GBA1 mutant Drosophila also support an interaction between the two enzymes39–41. However, assessment of GCase activity in PD peripheral blood mononuclear cells (PBMCs) showed no effect of LRRK2 kinase activity on GCase activity42. Work from Kedariti and colleagues indicates that the outcome may be cell type dependent, observing a positive correlation between LRRK2 kinase activity and GCase activity in the majority of models surveyed43. Moreover, the discordance in these reports could also stem from the use of varied methods to assess GCase activity.

Clinically, PD patients carrying LRRK2 variants have, on average, a milder disease course than those carrying GBA1 variants2. Rare PD patients carrying both a LRRK2 variant and GBA1 variant have a less severe phenotype than those carrying a GBA1 variant alone44–46. This points to a potential protective effect of LRRK2 variants in GBA1-PD, and bolsters the idea that LRRK2 activity and GCase activity are positively correlated. Both LRRK2 kinase activity and GCase activity are being pursued as therapeutic targets in PD2,47,48; thus, it is crucial to understand the interplay between these two enzymes, and which patients will benefit from therapies targeting one or the other.

Here we assess the impact of the LRRK2 p.G2019S, p.M1646T and p.N551K-p.R1398H variants on GCase activity in PD-patient derived human induced pluripotent stem cell (iPSC)-derived microglial cells (iMGs)—a PD relevant cell type expressing high levels of both LRRK2 and GCase. We confirm that the p.M1646T PD risk variant leads to increased LRRK2 kinase activity, and the p.N551K-p.R1398H protective haplotype to decreased kinase activity. However, we see no effect of any of these LRRK2 variants, or of pharmacological LRRK2 kinase inhibition, on GCase protein levels or lysosomal GCase activity under basal conditions. In contrast, treatment of iMGs with interferon γ (IFNγ) increased lysosomal GCase activity in a manner that could be attenuated by LRRK2 kinase inhibition. Further, when treated with IFNγ, LRRK2 p.M1646T PD risk variant iMGs showed increased GCase activity, whereas the p.N551K-p.R1398H protective haplotype showed decreased GCase activity, when compared to their respective isogenic control iMGs. This suggests that LRRK2 kinase activity and GCase activity are positively correlated in iMGs under proinflammatory conditions.

Results

Generation of LRRK2 isogenic control iPSCs and differentiation to iMGs

iPSCs were generated by reprogramming PD patient derived PBMCs heterozygous for the LRRK2 p.G2019S, p.M1646T, and p.N551K-p.R1398H (protective haplotype) variants (Fig. 1a). A previously generated healthy control (LWT) line was also used in this study49. LRRK2 knockout (LKO) and isogenic control lines with correction of LRRK2 variants were generated using CRISPR/Cas9 editing. Two isogenic control lines were generated for the LRRK2 protective haplotype—one with correction of the p.N551K variant (p.N551KCORR-p.R1398H), and one with correction of the p.R1398H variant (p.N551K-p.R1398HCORR). Correcting each variant individually allows for dissection of the effects of each variant of the haplotype. iPSC lines were subjected to quality control measures, including karyotyping, and assessment of expression of pluripotency markers (Figs. S1 and S2).

Fig. 1. Differentiation and characterization of iMGs.

Fig. 1

Open in a new tab

a Schematic of LRRK2 domain structure and variants of interest. Variants predicted to increase kinase activity in red, and to decrease kinase activity in blue. Parental and matching isogenic iPSC lines used in this study. b Differentiation of iPSCs to iMGs. iPSCs are differentiated through mesoderm to produce hematopoietic progenitor cells (iHPCs) using Stem Cell Technologies STEMdiff™ Hematopoietic Kit (Day 0–12). iHPCs are further differentiated to iMGs via M-CSF, IL-34, and TGF-β1 supplementation (Day 12–37). Primitive iMGs are matured by the addition of CX3CL1 and CD200 (Day 37–40). Scale bars 200 μm. c iMGs express microglial markers Iba1 and PU.1, as detected by IF staining. Scale bar 50 μm. n = 1. d iMGs from all lines are > 90% double positive for macrophage markers CD45 and CD11b by flow cytometry analysis. n = 1.

All iPSC lines were differentiated into iMGs following the protocol published by McQuade and colleagues (Fig. 1b)50. Microglial identity was confirmed by assessing expression of canonical microglial (Iba1 and PU.1) and macrophage (CD11b and CD45) markers by immunostaining and flow cytometry, respectively (Fig. 1c, d). iMGs generated from all lines were at least 90% double positive for CD11b and CD45 expression, indicating little to no contamination from non-myeloid lineage cells.

LRRK2 variants alter kinase activity in iMGs

To assess the effects of the LRRK2 p.G2019S and p.M1646T risk variants, and the p.N551K-p.R1398H protective haplotype on LRRK2 kinase activity in iMGs, we performed western blotting using phospho-specific antibodies against LRRK2 and well-characterized LRRK2 substrates Rab10 and Rab12. Phosphorylation of LRRK2 was assessed both at its autophosphorylation site S1292 and its biomarker site S935. A time-course and dose-response for LRRK2 inhibitor treatment was optimized (Fig. S3), and LRRK2 knockout iMGs were employed as controls for antibody specificity and to establish levels of LRRK2-independent Rab phosphorylation. No significant differences in LRRK2 phosphorylation at either site were observed in LRRK2 variant iMGs compared to their isogenic controls (Fig. S4a–c). A trend towards decreased LRRK2 S935 phosphorylation was observed with an acute 6 h treatment using LRRK2 kinase inhibitor MLi-2; however was not significant except in LRRK2 p.M1646T iMGs (Fig. S4c). MLi-2 treatment did, however, lead to a significant decrease in pT73 Rab10 levels in all lines, excluding the knockout iMGs (Fig. 2a, b). Despite the increased kinase activity associated with the LRRK2 p.G2019S variant as a recombinant protein in vitro11,51,52, in over-expression models53, and some endogenous contexts54, there was no difference in Rab10 phosphorylation in the LRRK2 p.G2019S iMGs compared to their isogenic controls. This is consistent with reports in mouse lung and kidney tissue and human neutrophils55,56. In concordance with its risk variant status, and a previously published in vitro kinase assay8, an increase in Rab10 phosphorylation was observed in the LRRK2 p.M1646T variant iMGs compared to isogenic control iMGs. As the p.R1398H variant is thought to drive the association of the protective haplotype to PD, and is known to impact both LRRK2 GTPase and kinase function57,58, we chose to normalize both the protective haplotype iMGs and the p.N551KCORR iMGs to the p.R1398HCORR iMGs. Individual correction of both the p.N551K and p.R1398H variants of the LRRK2 protective haplotype increased phosphorylation of Rab10 (Fig. 2a, b). Phosphorylation of Rab12 at S106 was also assessed. There was a trend towards decreased Rab12 phosphorylation with MLi-2 treatment, which reached significance in some lines, but no significant differences in LRRK2 variant iMGs compared to their isogenic controls (Fig. S4a, d). Thus, the LRRK2 p.M1646T risk variant and the p.N551K-p.R1398H protective haplotype, respectively, increase and decrease Rab10 phosphorylation in iMGs.

Fig. 2. Rab10 phosphorylation is altered in LRRK2 variant iMGs, while GCase and α-synuclein levels are not.

Fig. 2

Open in a new tab

a LRRK2 expression and Rab10 phosphorylation in LRRK2 variant iMGs with or without 6 h 100 nM MLi-2 treatment as measured by WB. Membrane images have been cropped to remove additional LRRK2 inhibition paradigm and to separate isogenic pairs. n = 3. b Quantification of WB of pT73 Rab10 normalized to total Rab10, levels normalized to isogenic control. c GCase and α-synuclein protein levels in LRRK2 variant iMGs with or without 6 h 100 nM MLi-2 treatment as measured by WB. n = 3. d, e Quantification of WB normalized to isogenic control. d Quantification of GCase. e Quantification of α-synuclein. One Way ANOVA with Bonferroni post-hoc test *p < 0.05, **p < 0.01, ****p < 0.0001.

LRRK2 variants do not alter basal GCase or α-synuclein protein levels in iMGs

LRRK2 kinase activity has been reported to lead to altered levels of GCase protein38,43, thus expression of GCase was assessed by western blotting. No significant differences in GCase protein levels were observed between LRRK2 variant iMGs and their isogenic controls. Additionally, acute LRRK2 inhibition had no effect on GCase levels (Fig. 2c, d).

As LRRK2 variants and loss of GCase activity have both been shown to influence accumulation of α-synuclein in neurons59–61, western blotting was used to assess changes in basal expression or accumulation of α-synuclein. There were no marked changes in α-synuclein protein levels in the LRRK2 variant iMGs or with acute LRRK2 kinase inhibition (Fig. 2c, e). Overall, these data indicate that the changes in LRRK2 kinase activity or other potential effects associated with the p.G2019S, p.M1646T, and p.N551K-p.R1398H LRRK2 variants do not alter GCase or α-synuclein protein levels in iMGs under basal conditions.

LRRK2 kinase activity has no effect on basal GCase activity in iMGs

To assess the effects of LRRK2 variants and their altered kinase activity on GCase activity in live cells, cleavage of the lysosome-specific fluorogenic GCase substrate 5-(Pentafluorobenzoylamino)Fluorescein Di-β-D-Glucopyranoside (PFB-FDGlu) was monitored by high-content microscopy (Figs. 3a and S5). The mean PFB-FDGlu fluorescence signal per cell was not significantly different in LRRK2 variant or knockout iMGs compared to their isogenic controls at any point during the 140-min incubation period (Fig. 3b). Additionally, there were no differences in the rate of GCase activity, as represented by the slope of the linear portion (20 to 90 min) of PFB-FDGlu fluorescence curves, in variant iMGs compared to their isogenic controls; or with use of two different LRRK2 inhibitors for 6 h periods – MLi-2 or PF-475 (Figs. 3c and S6a). LRRK2 inhibition by PF-475 treatment in iMGs was confirmed by a reduction in Rab10 phosphorylation observed via western blot (Fig. S6b). Mean total lysotracker area per cell, mean lysotracker spot area (representative of lysosome size) per cell, and mean lysotracker spot number per cell area were quantified. No differences across these lysosomal measurements were observed between LRRK2 variant iMGs and their isogenic controls (Fig. 3d), or upon LRRK2 inhibition (Fig. S7a–c). Additionally, no difference in total lysosomal area was observed in LKO iMGs compared to LWT iMGs. However, significant differences in lysotracker spot area and number were observed, with a higher number of smaller lysosomes present in LKO iMGs (Fig. 3e, f). Varied effects of LRRK2 knockout on lysosome size and number have been reported in a number of models (summarized by Kuwahara and Iwatsubo)20. In line with our findings, multiple reports observe an increase in number of lysosomes17; although often accompanied by an increase in lysosomal size62,63. Taken together, these data indicate that there is no effect of LRRK2 kinase activity on GCase activity in iMGs under basal conditions. LRRK2 knockout does impact lysosomal size and number in iMGs, however, an effect of LRRK2 variants or inhibition was not observed. This may be due to the subtle effect of variants as opposed to a knockout, the relatively acute period of inhibition used, or the changes observed in the knockout may be driven by the loss of a domain of LRRK2 other than the kinase.

Fig. 3. Lysosomal GCase activity is unaltered in LRRK2 variant iMGs under basal conditions.

Fig. 3

Open in a new tab

a Representative images of LWT iMGs stained with lysotracker deep-red and PFB-FDGlu at 0, 50, 100 and 140 min after dye-loading. Scale bar 50 μm. b Percent change in mean PFB-FDGlu fluorescence per cell per well. n = 4. c Slope of the linear region of PFB-FDGlu fluorescence curves (20–90 min) with or without 6 h 100 nM MLi-2 or 500 nM PF-475 treatment. d Mean total lysotracker fluorescence area per cell per well (px2) at baseline (0 min). e Mean lysotracker spot area per cell per well (px2) at baseline (0 min). f Mean lysotracker spot number normalized to cell area per cell per well (px−2) at baseline (0 min). c–f Repeated Measures One Way ANOVA Tukey post-hoc test *p < 0.05, **p < 0.01, ****p < 0.0001.

LRRK2 kinase inhibition reduces GCase activity in iMGs in response to IFNγ-induced proinflammatory stimulation

Neuroinflammation is a hallmark of PD; levels of pro-inflammatory cytokines, including IFNγ, are found to be increased in PD post-mortem brain64. Additionally, LRRK2 expression levels are known to be increased in response to inflammatory stimuli22. As such, the impact of IFNγ treatment on GCase activity in iMGs was assessed. In healthy control (LWT) iMGs, IFNγ treatment led to an increased concentration of a number of pro-inflammatory cytokines in the culture media, as measured by xMAP protein quantification assay (Fig. S8). Several of these cytokines have been reported as increased in PD post-mortem brain (IL-2)65, or in blood of PD patients (IL-2, IL-10, Pentraxin-3)66–68, to correlate with disease severity (RANTES)69, or are encoded by genes carrying single nucleotide polymorphisms linked to increased PD risk in specific populations (IL-16, IL-18, IL-13)70–72. Other cytokines also reported to be elevated in PD brain had a mild, but insignificant increase in release (TNFα, IL-6, IL-1β)65,66,73. IFNγ treatment also led to an increase in GCase activity, reflected by an increased slope of the linear portion of PFB-FDGlu fluorescence (Figs. 4a, c and S9). This is consistent with published data reporting increased GCase activity in PBMCs following IFNγ treatment42. Interestingly, a relatively acute 1 h pretreatment of iMGs with MLi-2 followed by 24 h co-treatment with IFNγ and MLi-2 partially reversed the increase in GCase activity. The complete loss of PFB-FDGlu fluorescence upon treatment with lysosomal GCase inhibitor (CBE) demonstrates the specificity of the assay (Fig. 4b, c). The impact of IFNγ on lysosomal mass, and how it may vary across cell types has yet to be fully elucidated. Both activation and inactivation of the TFEB lysosomal biogenesis pathway by IFNγ signalling have been reported74,75. Here, no changes were observed in mean total lysotracker area or mean lysosomal size, as represented by lysotracker spot area (Fig. 4d, e). A subtle, yet significant decrease in mean lysosome number was observed upon treatment with IFNγ, but it was not sensitive to MLi-2 treatment. Western blotting was performed to further assess lysosomal content, α-syn levels, and LRRK2 activity in iMGs upon IFNγ stimulation. In line with previous studies21,22, LRRK2 expression, phosphorylation, and downstream Rab10 phosphorylation were increased in response to IFNγ treatment. Increased LRRK2 phosphorylation at S935 and Rab10 phosphorylation could both be blocked by MLi-2 treatment (Fig. 5a–d). There were no marked changes in GCase or LAMP1 protein levels upon IFNγ or MLi-2 treatment (Fig. 5a, e, f), indicating that the LRRK2-dependent increase in GCase activity observed under these conditions is unlikely to be driven by increased GCase protein or lysosomal mass. As loss of GCase activity has been reported to impact α-syn accumulation in neurons59–61, we investigated whether the LRRK2-dependent increase in GCase activity observed with IFNγ treatment altered basal α-syn levels. However, no changes in α-syn levels were observed (Fig. 5a, g). Increases in lysosomal cathepsin activity have also been reported to occur in response to IFNγ treatment42, thus global lysosomal proteolytic capacity was assessed by monitoring cleavage of fluorogenic protease substrate DQ Red BSA. No significant changes in global lysosomal proteolytic capacity were observed in response to IFNγ or MLi-2 treatment in LWT iMGs (Fig. 5h, i). As a lysosomal enzyme, GCase functions optimally at a low pH (~4.5–5.9)76–78. Both LRRK2 kinase hyperactivity and inflammatory stimuli have been reported to influence lysosomal pH17,79. To assess if the IFNγ stimulated, LRRK2-dependent increase in GCase activity that we observe is driven by a change in lysosomal pH we employed the pH-sensitive lysosomal-targeted pHLys Green probe. Increased fluorescence intensity of this probe is indicative of increased lysosomal acidity (lower pH). A subtle but significant increase in lysosomal acidity was observed after 24 h of IFNγ treatment. However, LRRK2 inhibition via MLi-2 had no effect on this change in pH (Fig. 5j, k). Together, these data suggest that LRRK2 kinase signaling is involved in mediating an increase in GCase activity in response to inflammatory stimulation, independent of gross changes in overall lysosomal mass, degradative capacity, and pH, or GCase protein levels.

Fig. 4. Lysosomal GCase activity is regulated by LRRK2 kinase activity under IFNγ-induced inflammatory stimulation.

Fig. 4

Open in a new tab

a Percent change in mean PFB-FDGlu fluorescence per well per cell in LWT iMGs with and without 24 h 20 ng/mL IFNγ and 25 h 100 nM MLi-2 treatment. n = 4. b Representative images of LWT iMGs, with and without IFNγ and MLi-2 treatment, stained with lysotracker deep-red after 140 min PFB-FDGlu incubation. Scale bar 50 μm. c Slope of linear region of PFB-FDGlu fluorescence curve (20–90 min). d Mean total lysotracker fluorescence area per cell per well (px2) at baseline (0 min). e Mean lysotracker spot area per cell per well (px2) at baseline (0 min). f Mean lysotracker spot number normalized to cell area per cell per well (px−2) at baseline (0 min). c–f Repeated Measures One Way ANOVA Tukey post-hoc test *p < 0.05, **p < 0.01, ****p < 0.0001.

Fig. 5. IFNγ-induced inflammatory stimulation does not induce LRRK2-dependent changes in GCase or α-syn protein expression, lysosomal proteolytic capacity, or pH.

Fig. 5

Open in a new tab

a Western blot analysis to assess LRRK2 levels and activity, and lysosomal content in LWT iMGs with and without 24 h 20 ng/mL IFNγ and 25 h 100 nM MLi-2 treatment. n = 3. b–f Quantification of western blot. b Quantification of pS935 normalized to total LRRK2. c Quantification of LRRK2. d Quantification of pT73 normalized to total Rab10. e Quantification of GCase. f Quantification of LAMP1. g Quantification of α-syn. h Representative images of DQ Red BSA signal in LWT iMGs following IFNγ and MLi-2 treatment 24 h after dye-loading. Acquired using a 40X water immersion objective. i Mean DQ Red BSA fluorescence intensity normalized to mean total lysotracker area (px2) per cell per well, normalized to DMSO treated control, 24 h after dye-loading. n = 3. j Representative images of pHLys Green signal in LWT iMGs following IFNγ and MLi-2 treatment. Acquired using a 40X water immersion objective. k Mean pHLys Green fluorescence intensity normalized to mean total LysoPrime Deep Red area (px2) per cell per well, normalized to DMSO treated control. n = 3. b–k One Way ANOVA with Bonferroni post-hoc test *p < 0.05, **p < 0.01, ****p < 0.0001.

PD-associated LRRK2 variants modulate GCase activity in iMGs in response to IFNγ-induced proinflammatory stimulation

We hypothesized that the LRRK2 kinase dependent increase in GCase activity observed in response to IFNγ treatment may reveal differences in GCase activity driven by PD-associated LRRK2 variants. Thus, we assessed GCase activity in LRRK2 variant iMGs and their isogenic controls after 24 h of IFNγ treatment, with or without pre- and co-treatment with MLi-2 (Figs. 6 and S10). GCase activity was increased in all iMGs treated with IFNγ compared to non-treated controls, and consistent with our results in LWT cells, there was a trend for this increase to be attenuated by MLi-2 treatment—although only reaching significance in p.G2019S and p.M1646T variant iMGs (Fig. 6c). In line with previous findings7, IFNγ-treated p.M1646T iMGs exhibited increased GCase activity—illustrated by an increased slope of the linear region (20–90 min) of PFB-FDGlu fluorescence, compared to isogenic control iMGs. Additionally, correction of the p.N551K variant, but not the p.R1398H variant, of the LRRK2 protective haplotype led to an increase in GCase activity after IFNγ treatment. No effect of the LRRK2 variants or IFNγ treatment was observed on total lysotracker area or lysosome size (Fig. 6d, e). Similar to our results in LWT cells, there is a trend towards decreased lysosome number upon IFNγ treatment, however, this only reaches significance in p.M1646T variant iMGs. This indicates that in p.M1646T and protective haplotype iMGs LRRK2 activity and GCase activity are positively correlated, with the change in GCase activity being driven by the p.N551K variant in the latter case. The LRRK2 p.G2019S variant has been reported to both increase and decrease GCase activity. Here, we were unable to detect changes in GCase activity in these cells compared to their isogenic controls, even under IFNγ stimulation.

Fig. 6. LRRK2 variants impact GCase activity under IFNγ-induced inflammatory stimulation.

Fig. 6

Open in a new tab

a Percent change in mean PFB-FDGlu fluorescence per cell per well in LRRK2 variant and isogenic control iMGs upon 24 h 20 ng/mL IFNγ and 25 h DMSO treatment. n = 4. b Representative images of LRRK2 variant and isogenic control iMGs treated with IFNγ and DMSO 140 min after dye-loading. c Slope of linear region of PFB-FDGlu fluorescence curve (20–90 min) in LRRK2 variant and isogenic control iMGs upon 24 h 20 ng/mL IFNγ and 25 h 100 nM MLi-2 treatment. d Mean total lysotracker fluorescence area per cell per well (px2) at baseline (0 min). e Mean lysotracker spot area per cell per well (px2) at baseline (0 min). f Mean lysotracker spot number normalized to cell area per cell per well (px−2) at baseline (0 min). c–f Repeated Measures One Way ANOVA Tukey post-hoc test *p < 0.05, **p < 0.01, ****p < 0.0001.

Discussion

There is growing evidence for an interaction between LRRK2 and GCase activity. However, there is little agreement on the outcome of this interaction and how it may manifest in the context of PD. The aim of this work was to assess the effect of the LRRK2 p.G2019S, p.M1646T, and p.N551K-p.R1398H variants, all previously associated with altered GCase activity7,37,80, on LRRK2 kinase activity and on GCase activity in iPSC-derived microglial cells. We demonstrate, for the first time in human PD-patient derived microglia, that in response to proinflammatory stimulation, LRRK2 kinase and GCase activities are positively correlated.

Phosphorylation of Rab10 was not increased in p.G2019S iMGs compared to isogenic control iMGs. This may be due to a subtle increase in kinase activity resulting from the p.G2019S mutation—altered Rab10 phosphorylation may be more readily detected using higher sensitivity methods such as mass spectrometry or mesoscale detection assays. These subtle changes may also be amplified by increased LRRK2 expression in response to inflammatory stimuli81. Alternatively, there is evidence for differential phosphorylation of Rabs by various LRRK2 hyperactive variants across cell and tissue types55,56. Consistent with a previously published in vitro kinase assay and its risk variant status, Rab10 phosphorylation was increased in p.M1646T iMGs. Like the p.R1441C/G/H and p.Y1699C PD pathogenic variants, the p.M1646T variant falls in the ROC-COR GTPase domain (Fig. 1a), which is well-known to influence the kinase activity of LRRK262,82. We confirm, in iMGs, previous reports of decreased kinase activity associated with the LRRK2 protective haplotype10,57. Interestingly, we find that correction of either the p.N551K or p.R1398H variant leads to increased phosphorylation of Rab10—indicating that both variants can contribute to decreased kinase activity of the haplotype. The p.R1398H variant, localized in the GTPase domain, has been shown to be the primary driver of the association of the protective haplotype to PD, and to have increased GTPase activity and reduced kinase activity57,58. An effect of the p.N551K variant alone on LRRK2 kinase activity has not been previously described. The p.N551K variant is localized to the N-terminal armadillo repeat (ARM) domain of LRRK2, which has recently been implicated in the binding of LRRK2 to Rabs. Binding of Rab29 or Rab12 to the ARM domain drives membrane recruitment and activation of LRRK283,84, providing a potential mechanism for the altered kinase activity of the p.N551K variant in vivo.

Despite changes in kinase activity, the LRRK2 variants assessed had no effect on GCase protein levels or activity in iMGs under basal conditions. Likewise, LRRK2 KO and kinase inhibition had no effect on GCase expression or activity. LRRK2 kinase inhibition has been reported to have no effect on GCase activity, to increase GCase activity, and to decrease GCase activity across different studies and models37,40,42,43. Notably, we employed relatively acute 6 or 24 h MLi-2 or PF-475 treatments (sufficient for dephosphorylation of Rab10 in iMGs. Figs. 2a, b, S3a, and 5a, d), whereas longer treatment periods (3–12 days) were utilized in studies where LRRK2 kinase inhibition led to increased GCase activity37,40. Transcriptional changes induced by LRRK2 kinase inhibition over multiple days may underlie the altered GCase activity observed, and the same changes may be less likely to occur during the inhibition periods used here. Use of different methods of measuring GCase activity and cell models could also contribute to the inconsistencies in the effect of LRRK2 inhibition.

The lysosome is known to play a key role in mediating the immune response, but there is no consensus as to how the lysosome responds under IFNγ treatment85. We find that GCase activity is increased with IFNγ stimulation, and that this increase can be blocked by LRRK2 kinase inhibition. Similarly, Wallings and colleagues report increases in GCase activity and pan-cathepsin activity in PBMCs stimulated with IFNγ; although these increases were not sensitive to LRRK2 kinase inhibition42. Taken together, these data could point to a global increase in lysosomal degradative capacity in response to IFNγ. However, using a pan-lysosomal protease substrate, we did not observe a change in global lysosomal proteolytic activity in iMGs treated with IFNγ. IFNγ treatment did induce an increase in lysosomal acidity, but this was not attenuated by LRRK2 kinase inhibition. Moreover, we did not observe changes in gross lysosomal mass in iMGs, as measured by total lysotracker area, mean lysosome size, and LAMP1 protein levels, in response to IFNγ. A subtle decrease in the number of lysosomes was observed with IFNγ treatment, but was not rescued by LRRK2 kinase inhibition. Taken together, these findings indicate that it is unlikely that the LRRK2-driven increase in GCase activity is mediated by increased lysosomal mass, degradative capacity, or altered pH. GCase protein levels remained unchanged in response to IFNγ treatment, therefore, the increased GCase activity observed is not driven by increased protein expression, whereas LRRK2 levels and Rab10 phosphorylation are increased upon IFNγ treatment. Whether increased GCase activity is driven solely by the increased expression of LRRK2, or by some other LRRK2-dependent consequence of IFNγ stimulation, should be investigated. Additionally, although no change in GCase protein expression was detected, determination of lysosomal levels of GCase protein would be valuable given prior implication of LRRK2 in proper trafficking of GCase transporter LIMP2 to the lysosome86. Overall, we find no impact of LRRK2 kinase activity on GCase activity under basal conditions. However, LRRK2 kinase activity and expression level were positively correlated with GCase activity in LWT iMGs in response to proinflammatory stimulation with IFNγ. We further assessed GCase activity in the LRRK2 variants of interest after stimulation with IFNγ. Here we observed some, although not absolute, positive correlation between increased LRRK2 kinase activity and increased GCase activity in iMGs. The p.G2019S variant did not affect kinase activity in our hands, and also did not affect GCase activity. Whereas the p.M1646T variant and correction of the p.N551K variant both led to increased kinase activity and increased GCase activity upon IFNγ stimulation. The exception to this trend was the correction of the p.R1398H variant, which led to increased kinase activity as measured by phosphorylation of Rab10, but did not increase GCase activity under inflammatory conditions, suggesting again that the p.N551K drives the protective effect seen in the p.N551K-p.R1398H protective haplotype.

We would like to acknowledge several limitations of our study. Microglia are a highly dynamic cell type that arises from a distinct biological origin, yolk-sac derived early erythromyeloid progenitor cells87–89. These factors make generation of iPSC-derived microglia uniquely challenging. Here we follow a widely accepted protocol for generation of iPSC-derived microglia from hematopoietic precursor cells50, however, it is still difficult to determine where our iMGs fall on the continuum from monocyte to macrophage to microglia. The high expression of LRRK2 in our iMGs may indicate that they more closely resemble peripheral monocytes, which express more LRRK2 compared to brain resident microglia90, or that the iMGs are in a state of activation in which LRRK2 expression is upregulated. Further, here we use only one iPSC line per each LRRK2 variant studied. Although this is counterbalanced by the use of isogenic corrected control lines for each variant, our findings would be strengthened by confirmation in additional variant iPSCs derived from distinct donors. Additionally, while we observe that LRRK2 kinase inhibition blocks an IFNγ-driven increase in GCase activity, we did not assess GCase activity in IFNγ treated LKO iMGs. Thus, we cannot exclude ‘non-kinase’ effects potentiated by the other domains of LRRK2 on GCase activity. Lastly, our study is also limited by the fact that we do not observe an impact on GCase activity in iMGs from the LRRK2 p.G2019S variant, even though it is the most common LRRK2 pathogenic variant associated with PD in the Caucasian/European population3.

There is evidence that the effect of LRRK2 kinase activity on GCase activity is cell type dependent; our data implies that it is also dependent on inflammatory state, adding another layer of complexity to the interplay between these two enzymes. Inflammation, and more specifically, neuroinflammation are key hallmarks of PD. Inflammation and neuroinflammation resulting from pathogen exposure have been proposed to influence PD pathogenesis (reviewed by Tansey and colleagues)91, and further have been proposed to be triggered by substrate accumulation in GD and other lysosomal storage disorders92,93. Here, we present evidence that LRRK2 kinase activity mediates an increase in lysosomal GCase activity in response to IFNγ-induced inflammatory stimulation in microglial cells. This LRRK2-mediated increase in GCase activity under inflammatory conditions could allow microglia to better degrade excess glucosylceraminde and glucosylsphingosine, thus underlying a protective effect of LRRK2 kinase activity in GBA1-PD. Further investigation of the involvement of LRRK2 kinase activity in mediating GCase activity under inflammatory conditions is warranted, and a better understanding of the interaction of these two enzymes is crucial for guiding use and development of LRRK2 and GCase targeted PD therapeutics.

Methods

Generation and CRISPR editing of iPSC lines

The use of human iPSCs and iPSC-derived cells in this research was approved by the McGill University Research Ethics Board (IRB Study Number A03-M19-22A). All study procedures adhered to the Declaration of Helsinki. PD patient derived PBMCs heterozygous for the LRRK2 p.G2019S, p.M1646T, and p.N551K-p.R1398H (protective haplotype) variants were reprogrammed as indicated in earlier studies49, and provided to The Neuro’s C-BIG Open Biobank for storage and dissemination. All patient-derived samples underwent full LRRK2 sequencing using targeted Molecular Inversion Probe (MIP) based sequencing, as previously described7, to exclude the presence of any additional LRRK2 variants prior to reprogramming. The healthy control (LWT) line underwent whole genome sequencing to confirm no disease associated variants were present. These iPSC lines were subjected to quality control measures, including karyotyping and assessment of expression of pluripotency markers (Fig. S1). LRRK2 KO and isogenic control lines with correction of LRRK2 variants were generated using CRISPR/Cas9 editing and a ddPCR-based screening system. Two isogenic control lines were generated for the LRRK2 protective haplotype—one with correction of the p.N551K variant, and one with correction of the p.R1398H variant. CRISPR editing was confirmed by Sanger sequencing, and LKO and isogenic control lines were subjected to the same quality control measures as their parental lines (Fig. S2). All iPSC lines used in this study are registered with the hPSCReg repository (https://hpscreg.eu/browse/provider/1355). Cell line identification numbers are as follows: LWT—IPSC0063 (https://hpscreg.eu/cell-line/CBIGi001-A), LKO—IPSC0117 (https://hpscreg.eu/cell-line/CBIGi001-A-45), p.G2019S—IPSC0006 (https://hpscreg.eu/cell-line/CBIGi006-A), p.G2019SCORR—IPSC0007 (https://hpscreg.eu/cell-line/CBIGi006-A-1), p.M1646T—IPSC0017 (https://hpscreg.eu/cell-line/CBIGi013-A), p.M1646TCORR—IPSC0018 (https://hpscreg.eu/cell-line/CBIGi013-A-1), p.N551K-p.R1398H—IPSC0058 (https://hpscreg.eu/cell-line/CBIGi044-A), p.N551KCORR-p.R1398H—IPSC0059 (https://hpscreg.eu/cell-line/CBIGi044-A-1), p.N551K-p.R1398HCORR—IPSC0060 (https://hpscreg.eu/cell-line/CBIGi044-A-2).

Differentiation of iMGs

iPSCs were differentiated to iMGs following the protocol published by McQuade and colleagues50. Briefly, iPSCs were seeded at varying densities onto Matrigel (Corning 8774552) coated 6-well plates in mTESR1 (STEMCELL Technologies 85850). iPSCs were differentiated to hematopoietic precursor cells (iHPCs) using the STEMCELL Technologies STEMdiff™ Hematopoietic Kit (05310). iHPCs were collected and re-seeded in microglial differentiation media (MDM) + three cytokine cocktail (M-CSF, IL-34, and TGF-β1), onto matrigel-coated 6-well plates on days 10 and 12 of iHPC differentiation. Cells were supplemented with MDM + three cytokine cocktail every other day to mediate microglial differentiation. A full media change was performed on day 12, and a media change to MDM + five cytokine cocktail (M-CSF, IL-34, TGF-β1, CD200, and CX3CL1) on day 25. Cells were supplemented with MDM + five cytokine cocktail on day 27 and considered mature on day 28. Human recombinant M-CSF (300-25), IL-34 (200-34), TGF-β1 (100-21), and CX3CL1 (300-31) purchased from Peprotech. Human recombinant CD200 (BP004) was purchased from Bonopus Bioscience. For immunofluorescence, PFB-FDGlu, pHLys, and DQ-BSA assays, mature iMGs were replated. iMGs were dissociated by scraping in PBS, and replated onto 12 mm diameter glass coverslips at a density of 100,000 cells per coverslip, or 96-well Flat Clear Bottom Black Polystyrene TC-treated Microplates (Corning 3904) without Matrigel coating at a density of 20,000 cells per well. iMGs were supplemented with MDM + five cytokine cocktail every other day and assayed five days after replating.

Flow cytometry detection of CD45 and CD11b

Mature iMGs were collected; non-adherent cells were collected with media, while adherent cells were scraped in PBS. iPSCs were harvested in a single-cell suspension using Accutase (STEMCELL Technologies 07922) dissociation. iPSCs and iMGs were stained with Live/Dead fixable aqua stain (ThermoFisher L34965) in PBS for 30 min, washed with FACS buffer (1X PBS, 1% FBS, 0.1% NaN3) and blocked with TrueStain blocking reagent (Biolegend 422302) for 10 min. Cells were stained with anti- CD45—AlexaFluor 700 (1/40 Biolegend 304024) and CD11b—PE (1/20 BD Biosciences 555388) fluorophore-conjugated antibodies for 15 min, washed and resuspended in FACS buffer for analysis. iPSCs and iMGs were analyzed on a Thermo Attune NxT cytometer (Thermo) equipped with 405, 488, and 561 nm lasers and 610/20, 620/15, 530/30, and 525/50 filters (NeuroEDDU Flow Cytometry Facility, McGill University). For each sample, 50,000 events were collected, and single live cells were subsequently gated for CD45 and CD11b. Data were analyzed using FlowJo (BD Biosciences).

Immunofluorescence staining of Iba1 and PU.1

iMGs replated onto coverslips were fixed with 4% paraformaldehyde for 20 min. iMGs were washed three times with PBS, permeabilized with 0.2% triton-X 100 in PBS for 10 min, and blocked with 5% normal donkey serum (NDS) in PBS for 1 h at room temperature. iMGs were incubated with primary antibodies against Iba1 (1/1000, Synaptic Systems 234009) and PU.1 (1/500, Cell Signalling 2266S) diluted in 5% NDS in PBS overnight at 4 °C. iMGs were then washed with PBS three times, and incubated with a 1/500 dilution of donkey anti rabbit AlexaFluor 647 conjugated secondary antibody (Invitrogen A-32795), a 1/500 dilution of donkey anti-chicken AlexaFluor 488 conjugated secondary antibody (Invitrogen A-78948) and 1 mg/mL Hoechst in 5% NDS in PBS for 1 h at room temperature. iMGs were washed three times in PBS, mounted onto slides using Aqua-Poly/Mount (Polysciences Inc. 18606-20) and imaged on a Leica SP8 confocal microscope.

iMG treatments

For LRRK2 inhibition, iMGs were treated with 100 nM MLi-2 (Tocris 5756) or 500 nM PF-475 (MedChemExpress HY-12477) for 6 h. For IFNγ stimulation experiments, iMGs were pretreated with DMSO or 100 nM MLi-2 for 1 h. iMGs were then treated with 20 ng/mL IFNγ (Peprotech 300-02) or vehicle and DMSO or 100 nM MLi-2 for 24 h. GCase inhibition was achieved with 25 nM Conduritol B Epoxide (Sigma C5424) treatment for 25 h.

Western blotting

Mature iMGs were detached by scraping in PBS (without MLi-2), and resuspended in lysis buffer (50 mM Tris-HCl, pH 7.4, 1% v/v Triton-X 100, 10% Glycerol, 1X Halt phosphatase inhibitor cocktail (Thermo 78,428), 0.1 mg/mL microcystin-LR (Enzo Life Sciences ALX-350-012), 1X Complete protease inhibitor cocktail (Sigma 11873580001)). Samples were then incubated at 4 °C with rotation for 30 min, followed by three rounds of sonication in a water bath (30 s in water, 30 s on ice). Finally, samples were centrifuged at 20,800 × g for 20 min at 4 °C. Supernatants were collected and protein concentration measured using Detergent Compatible Protein Assay (Bio-Rad 5000112). Samples were prepared with a total of 25 μg of protein, and resolved on 4–15% or 15% SDS-PAGE gels, then transferred to PVDF using a Trans-Blot Turbo System (Bio-Rad). Membranes used for detection of α-synuclein were fixed in 4% paraformaldehyde, 0.1% glutaraldehyde for 30 min at room temperature, and washed three times in TBS with 0.1% Tween-20 (TBS-T). Membranes were blocked in 5% bovine serum albumin (BSA) in TBS-T. Membranes were incubated with primary antibodies diluted in 5% BSA in TBS-T overnight at 4 °C with shaking. Primary antibodies used and dilutions are as follows: LRRK2 pS1292 1/200 (Abcam ab203181), LRRK2 pS935 1/500 (Abcam ab133450), total LRRK2 1/500 (Abcam ab133474), Rab10 pT73 1/500 (Abcam ab230261), total Rab10 1/500 (Cell Signaling 8127S), Rab12 pS106 1/1000 (Abcam ab256765), total Rab12 1/500 (Proteintech 18843-1-AP), GCase 1/1000 (Abcam ab55080), α-synuclein 1/1000 (BD Biosciences 610787), LAMP1 1/1000 (Cell Signaling 9091S), α-actin 1/50,000 (Millipore MAB1501). Membranes were washed three times with TBS-T, incubated with HRP-conjugated secondary antibodies diluted at 1/5000 in 5% BSA in TBS-T for 1 h at room temperature, and washed as before. Western blots were visualised with Clarity Western ECL Substrate or Clarity Max Western ECL Substrate (Bio-Rad 170–5061, 170-5062) on a ChemiDoc MP Imaging System (Bio-Rad). Analysis performed using FIJI with α-actin as a loading control.

PFB-FDGlu GCase assay

Replated iMGs were treated with LRRK2 inhibitor and/or IFNγ as described above. iMGs were stained with lysotracker deep-red (1/20,000, Invitrogen L12492) for 30 min. Media was then exchanged for fluorobrite imaging media (Thermo A1896701) with 37.5 μM PFB-FDGlu (Invitrogen P11947). LRRK2 inhibition was continued throughout. iMGs were imaged with a 40X water immersion objective on an Opera Phenix high-content confocal microscope (Revvity) every 10 min for a total of 140 min. Columbus software was used to identify cells (using lysotracker signal) for quantification of mean PFB-FDGlu fluorescence per cell per well, and mean total lysotracker area per cell per well across three technical replicate wells (12 fields acquired per well) per biological replicate.

DQ Red BSA lysosomal proteolysis assay

Replated iMGs were treated with LRRK2 inhibitor and/or IFNγ as described above. Concomitant with IFNγ treatment, cells were loaded with 1 μg/mL DQ Red BSA (Invitrogen D12051) for 24 h. iMGs were stained with lysotracker deep-red (1/20,000, Invitrogen L12492) for 30 min. Media was exchanged for fluorobrite imaging media, and cells were imaged on an Opera Phenix high-content confocal microscope using a 40X water immersion objective. Columbus software was used to identify cells (using lysotracker signal), and mean DQ-BSA fluorescence per cell per well was quantified and normalized to mean total lysotracker area per cell per well across three technical replicate wells (30 fields per well) per biological replicate.

pHLys green lysosomal pH assay

Lysosomal pH was assayed using the Lysosomal Acidic pH Detection Kit-Green/Deep Red (Dojindo L286-10). Replated iMGs were treated with LRRK2 inhibitor, MLi-2 and/or IFNγ as described above. iMGs were then loaded with LysoPrime Deep Red (1/1000) for 30 min. iMGs were then washed once with Hank’s balanced salt solution (HBSS, Gibco 14175095), and loaded with pHLys Green diluted 1/1000 in HBSS for 30 min. pHLys Green was removed, and cells were imaged in HBSS with a 40X water immersion objective on an Opera Phenix high-content confocal microscope. Columbus software was used to identify cells (using LysoPrime Deep Red signal), and mean pHLys Green fluorescence per cell per well was quantified and normalized to mean total LysoPrime Deep Red area per cell per well across three technical replicate wells (27 fields per well) per each biological replicate.

xMAP protein quantification assay

Cell culture media from three parallel differentiations (n = 1, three technical replicates) of mature LWT iMGs treated with vehicle or 20 ng/mL IFNγ for 24 h were collected, spiked with bovine serum albumin (Wisent 800-095-EG) to a final concentration of 0.1% and stored at -80°C until assayed. Samples were assayed using Bio-Plex Pro Human Cytokine Screening Panel (48-Plex) (BioRad 12007283), Bio-Plex Pro Human Chemokine Panel (40-Plex) (BioRad 171AK99MR2), and Bio-Plex Pro Human Inflammation Panel 1 (37-plex) (BioRad 171AL001M) and a Luminex 200 Instrument (R&D Systems). The mean of three standard curves was used to determine the concentration in pg/mL of each analyte. In cases where the fluorescence intensity (FI) was out of the range of detection, the concentration of the upper or lower limit of detection was subbed in as appropriate. The mean log2 fold change of analyte concentration in media from IFNγ treated iMGs versus vehicle treated iMGs was calculated. P-values were calculated using Welch’s t-tests on the FI signal of each analyte from the three technical replicates of IFNγ versus vehicle treated iMG media. Analytes with a log2 fold change greater than or equal to one, and a p-value less than or equal to 0.05 were considered significantly differently regulated in IFNγ versus vehicle treated iMG media.

Statistical Analysis

Statistical analysis was conducted in GraphPad Prism9 software. Biological replicates are defined as experiments conducted using distinct differentiations from iPSCs to iMGs. All plots depict the mean value across biological replicates ± standard deviation. Statistical comparisons were computed using one-way ANOVA with Bonferroni post test for western blotting experiments, DQ-BSA, and pHLys assays. Repeated Measures One way ANOVA with Tukey post test was utilized for PFB-FDGlu assays. Significance levels are depicted in figure legends.

Supplementary information

Supplementary Information (11MB, pdf)

Acknowledgements

Thanks to Wolfgang Reintsch and Julien Sirois for training and support with high content imaging and flow cytometry experiments, respectively. Thanks to Marie-France Dorion for training and advice on differentiation and characterization of iPSC-derived microglia. Thanks to Jace Jones-Tabah, Roxanne Larivière, and Andrea Krahn for establishing and optimizing live-cell GCase, lysosomal proteolytic, and pH assays. E.J.M. has been supported by a Fonds de Recherche du Québec-Santé Doctoral Fellowship and Jeanne Timmins Costello Fellowship awarded by the Integrated program in Neuroscience at McGill University. This work was funded by grants from The Michael J. Fox Foundation for Parkinson’s Research (MJFF-019045) and from the Canadian Institutes of Health Research (FDN-154301). G-Can (GBA1 Canada) Initiative, an open-science collaborative initiative aimed at addressing GBA1 associated neurodegeneration, has contributed to this research. G-Can is supported by The Hilary & Galen Weston Foundation, J. Sebastian van Berkom, Ghislaine Saucier and Silverstein Foundation. E.A.F. is supported by a Canada Research Chair (Tier 1) in Parkinson’s disease.

Author contributions

E.J.M. generated p.M1646T and protective haplotype isogenic correction cell lines, differentiated iMGs, designed and performed experiments, analysed data, and prepared the manuscript. C.X-.Q.C. and N.A. performed quality control experiments on iPSC lines used in this study. E.D. designed CRISPR editing strategies and ddPCR screening strategies to generate isogenic correction lines, and generated the p.G2019S isogenic correction line. Z.Y. designed CRISPR KO strategy and generated the LKO line. D.K. performed xMAP experiments and analysed data. T.M.D., K.S., and Z.G-.O. designed experiments. E.A.F. designed experiments, supervised the project, and prepared the manuscript.

Data availability

All data generated or analysed during this study are included in this published article.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

The online version contains supplementary material available at 10.1038/s41531-026-01310-1.

References

  • 1.Kalia, L. V. & Lang, A. E. Parkinson’s disease. Lancet386, 896–912 (2015). [DOI] [PubMed] [Google Scholar]
  • 2.Senkevich, K., Rudakou, U. & Gan-Or, Z. New therapeutic approaches to Parkinson’s disease targeting GBA, LRRK2 and Parkin. Neuropharmacology202, 108822–108822 (2022). [DOI] [PubMed] [Google Scholar]
  • 3.Alessi, D. R. & Sammler, E. LRRK2 kinase in Parkinson’s disease. Science360, 36–37 (2018). [DOI] [PubMed] [Google Scholar]
  • 4.Program, T. G. P. s. G. & Leonard, H. L. Novel Parkinson’s disease genetic risk factors within and across European populations. medRxiv10.1101/2025.03.14.24319455 (2025).
  • 5.Ross, O. A. et al. Association of LRRK2 exonic variants with susceptibility to Parkinson’s disease: a case-control study. Lancet Neurol.10, 898–908 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Heckman, M. G. et al. Population-specific frequencies for LRRK2 susceptibility variants in the genetic epidemiology of Parkinson’s disease (GEO-PD) consortium. Mov. Disord.28, 1740–1744 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sosero, Y. L. et al. LRRK2 p.M1646T is associated with glucocerebrosidase activity and with Parkinson’s disease. Neurobiol. Aging103, 142.e141–142.e145 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Refai, F. S., Ng, S. H. & Tan, E. K. Evaluating LRRK2 genetic variants with unclear pathogenicity. BioMed. Res. Int.2015, 678701–678701 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Tan, E. K. et al. Multiple LRRK2 variants modulate risk of Parkinson's disease: a Chinese multicenter study. Hum. Mutat.31, 561–568 (2010). [DOI] [PubMed] [Google Scholar]
  • 10.Wang, X. et al. Understanding LRRK2 kinase activity in preclinical models and human subjects through quantitative analysis of LRRK2 and pT73 Rab10. Sci. Rep.11, 1–17 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Steger, M. et al. Phosphoproteomics reveals that Parkinson’s disease kinase LRRK2 regulates a subset of Rab GTPases. eLife5, e12813–e12813 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Steger, M. et al. Systematic proteomic analysis of LRRK2-mediated Rab GTPase phosphorylation establishes a connection to ciliogenesis. eLife6, e31012–e31012 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Waschbüsch, D. et al. Structural basis for Rab8a recruitment of RILPL2 via LRRK2 phosphorylation of Switch 2. Structure28, 406–417.e406 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Herbst, S. et al. LRRK2 activation controls the repair of damaged endomembranes in macrophages. EMBO J.39, 10.15252/embj.2020104494 (2020). [DOI] [PMC free article] [PubMed]
  • 15.Eguchi, T. et al. LRRK2 and its substrate Rab GTPases are sequentially targeted onto stressed lysosomes and maintain their homeostasis. Proc. Natl. Acad. Sci. USA115, E9115–E9124 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Eguchi, T. et al. The V-ATPase–ATG16L1 axis recruits LRRK2 to facilitate the lysosomal stress response. J. Cell Biol.223, 10.1083/JCB.202302067 (2024). [DOI] [PMC free article] [PubMed]
  • 17.Henry, A. G. et al. Pathogenic LRRK2 mutations, through increased kinase activity, produce enlarged lysosomes with reduced degradative capacity and increase ATP13A2 expression. Hum. Mol. Genet.24, 6013–6028 (2015). [DOI] [PubMed] [Google Scholar]
  • 18.Yadavalli, N. & Ferguson, S. M. LRRK2 suppresses lysosome degradative activity in macrophages and microglia through MiT-TFE transcription factor inhibition. Proc. Natl. Acad. Sci. USA120, e2303789120–e2303789120 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Madureira, M., Connor-Robson, N. & Wade-Martins, R. “LRRK2: autophagy and lysosomal activity”. Front. Neurosci.14, 10.3389/FNINS.2020.00498 (2020). [DOI] [PMC free article] [PubMed]
  • 20.Kuwahara, T. & Iwatsubo, T. The emerging functions of LRRK2 and Rab GTPases in the endolysosomal system. Front. Neurosci.14, 528270–528270 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Cook, D. A. et al. LRRK2 levels in immune cells are increased in Parkinson’s disease. npj Parkinson Dis.3, 1–12 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Gardet, A. et al. LRRK2 is involved in the IFN-γ Response and Host Response to Pathogens. J. Immunol.185, 5577–5585 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Moehle, M. S. et al. LRRK2 inhibition attenuates microglial inflammatory responses. J. Neurosci. Off. J. Soc. Neurosci.32, 1602–1611 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Ahmadi Rastegar, D. et al. Effect of LRRK2 protein and activity on stimulated cytokines in human monocytes and macrophages. npj Parkinson Dis.8, 1–11 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Panagiotakopoulou, V. et al. Interferon-γ signaling synergizes with LRRK2 in neurons and microglia derived from human induced pluripotent stem cells. Nat. Commun.11, 5163–5163 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Deretic, V. Autophagy in inflammation, infection, and immunometabolism. Immunity54, 437–437 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Cao, M., Luo, X., Wu, K. & He, X. Targeting lysosomes in human disease: from basic research to clinical applications. Signal Transduct. Target. Ther.6, 1–28 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Langston, R. G. et al. Association of a common genetic variant with Parkinson’s disease is mediated by microglia. Sci. Transl. Med.14, eabp8869-eabp8869 (2022). [DOI] [PMC free article] [PubMed]
  • 29.Sidransky, E. et al. Multicenter analysis of glucocerebrosidase mutations in Parkinson’s disease. N. Engl. J. Med.361, 1651–1661 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Mitsui, J. et al. Mutations for Gaucher disease confer high susceptibility to Parkinson's disease. Arch. Neurol.66, 571–576 (2009). [DOI] [PubMed] [Google Scholar]
  • 31.Nalls, M. A. et al. Identification of novel risk loci, causal insights, and heritable risk for Parkinson’s disease: a meta-genome wide association study. Lancet Neurol.18, 1091–1091 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Gan-Or, Z. et al. Differential effects of severe vs mild GBA mutations on Parkinson's disease. Neurology84, 880–887 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Alcalay, R. N. et al. Glucocerebrosidase activity in Parkinson’s disease with and without GBA mutations. 10.1093/brain/awv179 (2015). [DOI] [PMC free article] [PubMed]
  • 34.Gegg, M. E. et al. Glucocerebrosidase deficiency in substantia nigra of parkinson disease brains. Ann. Neurol.72, 455–463 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Atashrazm, F. et al. Reduced glucocerebrosidase activity in monocytes from patients with Parkinson’s disease. Sci. Rep.8, 1–12 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hughes, L. P. et al. Glucocerebrosidase activity is reduced in cryopreserved parkinson’s disease patient monocytes and inversely correlates with motor severity. J. Parkinson Dis.11, 1157–1165 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Ysselstein, D. et al. LRRK2 kinase activity regulates lysosomal glucocerebrosidase in neurons derived from Parkinson’s disease patients. Nat. Commun.10, 5570–5570 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Chedid, J. et al. Dopamine and cortical neurons with different Parkinsonian mutations show variation in lysosomal and mitochondrial dysfunction. npj Parkinson Dis.11, 177–177 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Ferrazza, R. et al. LRRK2 deficiency impacts ceramide metabolism in brain. Biochem. Biophys. Res. Commun.478, 1141–1146 (2016). [DOI] [PubMed] [Google Scholar]
  • 40.Sanyal, A. et al. Lysosome and inflammatory defects in GBA1-mutant astrocytes are normalized by LRRK2 iInhibition. Mov. Disord.35, 760–773 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Serebryany-Piavsky, V. et al. The modifying effect of mutant LRRK2 on mutant GBA1-associated Parkinson disease. Hum. Mol. Genet.34, 1184–1203 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Wallings, R. L. et al. WHOPPA enables parallel assessment of leucine-rich repeat kinase 2 and glucocerebrosidase enzymatic activity in Parkinson’s disease monocytes. Front. Cell. Neurosci.16, 10.3389/FNCEL.2022.892899/FULL (2022). [DOI] [PMC free article] [PubMed]
  • 43.Kedariti, M. et al. LRRK2 kinase activity regulates GCase level and enzymatic activity differently depending on cell type in Parkinson’s disease. npj Parkinson Dis.8, 1–14 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Omer, N. et al. A possible modifying effect of the G2019S mutation in the LRRK2 gene on GBA Parkinson’s disease. Mov. Disord.35, 1249–1253 (2020). [DOI] [PubMed] [Google Scholar]
  • 45.Ortega, R. A. et al. Association of dual LRRK2 G2019S and GBA variations with Parkinson disease progression. JAMA Netw. Open4, e215845–e215845 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Yahalom, G. et al. Carriers of both GBA and LRRK2 mutations, compared to carriers of either, in Parkinson’s disease: risk estimates and genotype-phenotype correlations. Parkinsonism Relat. Disord.62, 179–184 (2019). [DOI] [PubMed] [Google Scholar]
  • 47.Kluss, J. H., Lewis, P. A. & Greggio, E. Leucine-rich repeat kinase 2 (LRRK2): an update on the potential therapeutic target for Parkinson’s disease. Expert Opin. Ther. Targets26, 537–546 (2022). [DOI] [PubMed] [Google Scholar]
  • 48.Menozzi, E., Toffoli, M. & Schapira, A. H. V. Targeting the GBA1 pathway to slow Parkinson disease: insights into clinical aspects, pathogenic mechanisms and new therapeutic avenues. Pharmacol. Ther.246, 108419–108419 (2023). [DOI] [PubMed] [Google Scholar]
  • 49.Chen, C. X. Q. et al. A multistep workflow to evaluate newly generated iPSCs and their ability to generate different cell types. Methods Protoc.4, 50–50 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.McQuade, A. et al. Development and validation of a simplified method to generate human microglia from pluripotent stem cells. Mol. Neurodegeneration13, 67–67 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.West, A. B. et al. Parkinson’s disease-associated mutations in leucine-rich repeat kinase 2 augment kinase activity. Proc. Natl. Acad. Sci. USA102, 16842–16847 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Greggio, E. et al. The Parkinson’s disease kinase LRRK2 autophosphorylates its GTPase domain at multiple sites. Biochem. Biophys. Res. Commun.389, 449–454 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Kalogeropulou, A. F. et al. Impact of 100 LRRK2 variants linked to Parkinson’s disease on kinase activity and microtubule binding. Biochem. J.479, 1759–1783 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Kluss, J. H. et al. Detection of endogenous S1292 LRRK2 autophosphorylation in mouse tissue as a readout for kinase activity. npj Parkinson Dis.4, 13–13 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Iannotta, L. et al. Divergent effects of G2019S and R1441C LRRK2 mutations on LRRK2 and Rab10 Phosphorylations in Mouse Tissues. Cells9, 2344–2344 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Fan, Y. et al. R1441G but not G2019S mutation enhances LRRK2 mediated Rab10 phosphorylation in human peripheral blood neutrophils. Acta Neuropathol.142, 475–494 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Hui, K. Y. et al. Functional variants in the LRRK2 gene confer shared effects on risk for Crohn’s disease and Parkinson’s disease. Sci. Transl. Med.10, 7795–7795 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Nixon-Abell, J. et al. Protective LRRK2 R1398H variant enhances GTPase and Wnt signaling activity. Front. Mol. Neurosci.9, 179717–179717 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.O’Hara, D. M., Pawar, G., Kalia, S. K. & Kalia, L. V. LRRK2 and α-synuclein: distinct or synergistic players in Parkinson’s disease? Front. Neurosci.14, 541220–541220 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Mazzulli, J. R. et al. Gaucher disease glucocerebrosidase and α-synuclein form a bidirectional pathogenic loop in synucleinopathies. Cell146, 37–52 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Muñoz, S. S., Petersen, D., Marlet, F. R., Kücükköse, E. & Galvagnion, C. The interplay between Glucocerebrosidase, α-synuclein and lipids in human models of Parkinson’s disease. Biophys. Chem.273, 106534–106534 (2021). [DOI] [PubMed] [Google Scholar]
  • 62.Herzig, M. C. et al. LRRK2 protein levels are determined by kinase function and are crucial for kidney and lung homeostasis in mice. Hum. Mol. Genet.2010.1093/hmg/ddr348 (2011). [DOI] [PMC free article] [PubMed]
  • 63.Baptista, M. A. S. et al. Loss of leucine-rich repeat kinase 2 (LRRK2) in rats leads to progressive abnormal phenotypes in peripheral organs. PLoS ONE8, 10.1371/journal.pone.0080705 (2013). [DOI] [PMC free article] [PubMed]
  • 64.Mogi, M., Kondo, T., Mizuno, Y. & Nagatsu, T. p53 protein, interferon-γ, and NF-κB levels are elevated in the Parkinsonian brain. Neurosci. Lett.414, 94–97 (2007). [DOI] [PubMed] [Google Scholar]
  • 65.Mogi, M. et al. Interleukin (IL)-1β, IL-2, IL-4, IL-6 and transforming growth factor-α levels are elevated in ventricular cerebrospinal fluid in juvenile Parkinsonism and Parkinson’s disease. Neurosci. Lett.211, 13–16 (1996). [DOI] [PubMed] [Google Scholar]
  • 66.Brodacki, B. et al. Serum interleukin (IL-2, IL-10, IL-6, IL-4), TNFα, and INFγ concentrations are elevated in patients with atypical and idiopathic Parkinsonism. Neurosci. Lett.441, 158–162 (2008). [DOI] [PubMed] [Google Scholar]
  • 67.Rentzos, M. et al. Circulating interleukin-10 and interleukin-12 in Parkinson’s disease - PubMed. Acta Neurol. Scandinavica119, 10.1111/j.1600-0404.2008.01103.x (2009). [DOI] [PubMed]
  • 68.Lee, H. W., Choi, J. & Suk, K. Increases of pentraxin 3 plasma levels in patients with Parkinson’s disease. Mov. Disord.26, 2364–2370 (2011). [DOI] [PubMed] [Google Scholar]
  • 69.Tang, P. et al. Correlation between Serum RANTES levels and the severity of Parkinson’s disease. Oxid. Med. Cell. Longev.2014, 208408–208408 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Zhang, X. et al. Association of IL-16 gene polymorphisms with sporadic Parkinson’s disease in a Han Chinese population. Neurosci. Lett.724, 10.1016/j.neulet.2020.134877 (2020). [DOI] [PubMed]
  • 71.Xu, X. et al. Interleukin-18 promoter polymorphisms and risk of Parkinson’s disease in a Han Chinese population. Brain Res.1381, 10.1016/j.brainres.2011.01.025 (2011). [DOI] [PubMed]
  • 72.Aguirre, C. A. et al. Two single nucleotide polymorphisms in IL13 and IL13RA1 from individuals with idiopathic Parkinson’s disease increase cellular susceptibility to oxidative stress. Brain Behav. Immun.88, 10.1016/j.bbi.2020.04.007 (2020). [DOI] [PMC free article] [PubMed]
  • 73.Mogi, M. et al. Tumor necrosis factor-α (TNF-α) increases both in the brain and in the cerebrospinal fluid from parkinsonian patients. Neurosci. Lett.165, 208–210 (1994). [DOI] [PubMed] [Google Scholar]
  • 74.Wu, M. et al. Immunomodulators targeting MARCO expression improve resistance to postinfluenza bacterial pneumonia. Am. J. Physiol. Lung Cell. Mol. Physiol.313, L138–L153 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75.Singh, N. et al. Antimycobacterial effect of IFNG (interferon gamma)-induced autophagy depends on HMOX1 (heme oxygenase 1)-mediated increase in intracellular calcium levels and modulation of PPP3/calcineurin-TFEB (transcription factor EB) axis. Autophagy14, 972–991 (2018). [DOI] [PMC free article] [PubMed]
  • 76.Lieberman, R. L. et al. Structure of acid β-glucosidase with pharmacological chaperone provides insight into Gaucher disease. Nat. Chem. Biol.3, 101–107 (2007). [DOI] [PubMed] [Google Scholar]
  • 77.Offman, M. N., Krol, M., Silman, I., Sussman, J. L. & Futerman, A. H. Molecular basis of reduced glucosylceramidase activity in the most common gaucher disease mutant, N370S. J. Biol. Chem.285, 42105–42114 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78.Karatas, M. et al. Enzyme kinetics and inhibition parameters of human leukocyte glucosylceramidase. Heliyon6, e05191–e05191 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 79.Quick, J. D. et al. Lysosomal acidification dysfunction in microglia: an emerging pathogenic mechanism of neuroinflammation and neurodegeneration. J. Neuroinflammation20, 1–13 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 80.Alcalay, R. N. et al. Glucocerebrosidase activity in Parkinson’s disease with and without GBA mutations. Brain138, 2648–2658 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 81.Ohtonen, S. et al. Human iPSC-derived microglia carrying the LRRK2-G2019S mutation show a Parkinson’s disease related transcriptional profile and function. Sci. Rep.13, 1–16 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 82.Deniston, C. K. et al. Structure of LRRK2 in Parkinson’s disease and model for microtubule interaction. Nature588, 344–349 (2020). [DOI] [PMC free article] [PubMed]
  • 83.Vides, E. G. et al. A feed-forward pathway drives LRRK2 kinase membrane recruitment and activation. eLife11, 10.7554/ELIFE.79771 (2022). [DOI] [PMC free article] [PubMed]
  • 84.Dhekne, H. S. et al. Genome-wide screen reveals Rab12 GTPase as a critical activator of Parkinson’s disease-linked LRRK2 kinase. eLife12, 10.7554/ELIFE.87098 (2023). [DOI] [PMC free article] [PubMed]
  • 85.Irazoqui, J. E. Key roles of MiT transcription factors in innate immunity and inflammation. Trends Immunol.41, 157–171 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 86.Kuwahara, T. et al. LRRK2 and RAB7L1 coordinately regulate axonal morphology and lysosome integrity in diverse cellular contexts. Sci. Rep.6, 1–12 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 87.Gosselin, D. et al. An environment-dependent transcriptional network specifies human microglia identity. Science356, 1248–1259 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 88.Ginhoux, F. et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science330, 841–845 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 89.Kierdorf, K. et al. Microglia emerge from erythromyeloid precursors via Pu.1- and Irf8-dependent pathways. Nat. Neurosci.16, 273–280 (2013). [DOI] [PubMed] [Google Scholar]
  • 90.Wallings, R. L., Herrick, M. K. & Tansey, M. G. Frontiers | LRRK2 at the interface between peripheral and central immune function in Parkinson’s. Front. Neurosci.14, 10.3389/fnins.2020.00443 (2020). [DOI] [PMC free article] [PubMed]
  • 91.Tansey, M. G. et al. Inflammation and immune dysfunction in Parkinson disease. Nat. Rev. Immunol.22, 657–673 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 92.Boddupalli, C. S. et al. Neuroinflammation in neuronopathic Gaucher disease: role of microglia and NK cells, biomarkers, and response to substrate reduction therapy. eLife11, 10.7554/ELIFE.79830 (2022). [DOI] [PMC free article] [PubMed]
  • 93.Wang, A. et al. Innate immune sensing of lysosomal dysfunction drives multiple lysosomal storage disorders. Nat. Cell Biol.26, 219–234 (2024). [DOI] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information (11MB, pdf)

Data Availability Statement

All data generated or analysed during this study are included in this published article.

Articles from NPJ Parkinson's Disease are provided here courtesy of Nature Publishing Group

RESOURCES