Membrane Dysfunction as a Central Mechanism in LRRK2-Associated Parkinson’s Disease: Comparative Analysis of G2019S and I1371V Variants

Highlights

What are the main findings?

• GTPase domain LRRK2 mutation (I1371V) drives severe membrane lipid dysregulation:
  The I1371V mutation induces enhanced LRRK2 autophosphorylation and Rab8A and Rab10 hyperphosphorylation, leading to impaired sterol trafficking, selective membrane cholesterol depletion, increased membrane fluidity, disrupted lipid microdomains, altered membrane topology, and consequent defects in dopamine transporter localization and dopamine uptake—effects that are substantially milder in the kinase domain G2019S mutation.
• LRRK2 mutation-specific pharmacological responses reveal mechanistic heterogeneity:
  Membrane and dopaminergic defects caused by the I1371V mutation are preferentially rescued by a non-selective LRRK2 modulator (GW5074) rather than a kinase-selective inhibitor (MLi-2), demonstrating that GTPase domain-driven pathology depends on broader LRRK2 regulatory mechanisms and highlighting the need for variant-specific therapeutic strategies.

What are the implications of the main findings?

• Membrane lipid dysregulation emerges as a core cell biological mechanism in LRRK2-associated PD, particularly for GTPase domain mutations, linking aberrant Rab phosphorylation to sterol trafficking defects, membrane disorganization, and impaired dopaminergic function.
• Therapeutic strategies for LRRK2-PD must be mutation-specific, as GTPase domain variants respond preferentially to broader LRRK2 modulation rather than kinase-selective inhibition, underscoring the need for precision targeting based on domain-specific pathogenic mechanisms.

Abstract

Mutations in leucine-rich repeat kinase 2 (LRRK2) are among the most common genetic causes of Parkinson’s disease (PD), yet substantial heterogeneity exists among pathogenic variants. How mutations in distinct functional domains of LRRK2 differentially perturb cellular homeostasis remains incompletely understood. Here, we compared two pathogenic LRRK2 mutations—G2019S in the kinase domain and I1371V in the GTPase domain—across multiple cellular models, including SH-SY5Y and U87 cells, and healthy human iPSC-derived floor plate cells. We demonstrate that the I1371V mutation induces markedly more severe cellular dysfunction than G2019S. I1371V-expressing cells exhibited elevated LRRK2 autophosphorylation at S1292 and robust hyperphosphorylation of Rab8A and Rab10, indicating enhanced downstream signaling. These alterations impaired sterol trafficking, leading to selective depletion of membrane cholesterol without changes in total cellular cholesterol. Consequently, I1371V cells displayed increased membrane fluidity, disrupted microdomain organization, altered membrane topology, reduced caveolin-1 expression, and impaired dopamine transporter surface expression and dopamine uptake. Lipidomic profiling further revealed a broad disruption of lipid homeostasis, including reductions in cholesteryl esters, sterols, sphingolipids, and glycerophospholipids, whereas G2019S cells showed comparatively modest changes. Pharmacological intervention revealed mutation-specific responses, with the non-selective LRRK2 modulator GW5074 outperforming the kinase-selective inhibitor MLi-2 in restoring Rab8A phosphorylation, membrane integrity, and dopaminergic function. Collectively, these findings identify membrane lipid dysregulation as a central cell biological mechanism in LRRK2-associated PD and underscore the importance of variant-specific therapeutic strategies.

1. Introduction

An increasing number of clinical reports on Parkinson’s disease (PD) associated with LRRK2 highlight the heterogeneity of clinical parameters across mutant variants, including disease severity [1,2,3], progression [4,5,6], age of onset [7], brain pathology [8,9,10], drug-response [11,12], autophosphorylation, and substrate phosphorylation [13,14]. Among the LRRK2 variants, the G2019S mutation is the most prevalent and most extensively studied, exhibiting a typical PD syndrome with relatively mild non-motor symptoms and slower disease progression, along with low rates of REM sleep behavior disorder (RBD) [15,16]. In contrast, patients with LRRK2 GTPase mutations, particularly the I1371V variant, experience severe motor symptoms along with cognitive impairment. Importantly, individuals with the I1371V mutation also demonstrate a significantly lower response to levodopa and deep brain stimulation (DBS) therapy compared with those with the G2019S mutation [11,12]. Furthermore, these mutant variants also differ in terms of age of onset, with patients with the I1371V mutation tending to have an earlier presentation of symptoms [7,12].

The I1371V variant of LRRK2, initially identified by Paisán-Ruíz and colleagues in 2005 within an East Indian family, has since been reported in East Asian PD patients, as well as in two individuals from an Italian family and two other French families [1,17]. In contrast, the G2019S variant is more prevalent among Caucasian and Ashkenazi Jewish populations, with prevalence in the Indian population below 0.1% [18]. Additionally, other mutations in the GTPase domain of LRRK2, such as R1441C/G/H (with R1441G being prevalent in the Basque population), further highlight the complex genetic landscape of LRRK2-associated PD [19,20].

PD-causing LRRK2 mutations result in gain-of-function effects by activating kinase activity. Mutations in the kinase domain (such as G2019S) directly elevate kinase activity, while ROC domain mutations sustain LRRK2 in a GTP-bound kinase-active state by impeding GTP hydrolysis, thus prolonging kinase activity [21,22]. Studies have also highlighted differences in phosphorylation patterns including autophosphorylation and substrate phosphorylation between kinase and GTPase domain mutations, demonstrated in knock-in mice models and human samples [13]. Differences in α-synuclein phosphorylation patterns have been observed in rodent models expressing different LRRK2 variants (G2019S and R1441C), affirming distinct phosphorylation profiles among LRRK2 variants. There are no published studies on substrate and autophosphorylation in the I1371V variant as of yet. Cooper et al. (2013) described the differences in impact of LRRK2 variants (G2019S and R1441C) on mitochondrial function and drug response in vitro, underscoring the diverse effects of LRRK2 mutations on cellular processes [23].

Heterogeneity in LRRK2 function extends beyond mutant variants to variability across different cell types [24], emphasizing the importance of selecting appropriate cell types for cause-and-effect studies. For instance, Liou et al. (2008) observed no differential effects between wild-type (WT) LRRK2 and mutants in HEK293T cells, while overexpression of mutant LRRK2 led to a 20% reduction in cell viability compared with WT-LRRK2 in SH-SY5Y cells [25]. In addition, there are stark differences in membrane composition between humans and rodents, including variations in phospholipid percentage and levels of lipid metabolites [26]. Moreover, the stability of the LRRK2 protein and its degree of binding to substrates and effector proteins differ between rodents and humans, further complicating the study of PD pathology in these models [27].

Our previous research has demonstrated the PD pathology in dopaminergic neurons [28] and astrocyte dysfunction [22], both differentiated from LRRK2-I1371V mutation-carrying PD patient iPSCs, which were in coherence with other clinical findings [29]. The toxic gain-of-function of this variant in astrocytes was also replicated in U87 cells transfected with I1371V LRRK2 [22].

In addition to Rab10, LRRK2 phosphorylates Rab8A, which is crucial for LDL cholesterol delivery to the plasma membrane [30,31], potentially contributing to the reduced plasma cholesterol observed in human LRRK2 R1441G mutation carriers [32]. Rab8A, along with its interaction partners, also regulates endosomal egress. During lysosomal stress, LRRK2 is recruited to damaged lysosomes, where it is potentially involved in membrane repair and traffic reactions [33,34,35]. Recent studies using purified LRRK2 have also demonstrated its membrane-remodeling properties [36], though the impact of LRRK2 mutants on membrane cholesterol and fluidity remains poorly understood. Reyes et al. (2013) reported variant-specific differences in membrane protein (D2 and DAT) expression [37], suggesting potential modification of vesicle docking and membrane composition. Considering the crucial role of membrane cholesterol in the dynamics and cell surface expression of receptors and transporters involved in dopamine sensing and uptake, it is essential that the membrane dynamics of these mutants, particularly their Rab8A phosphorylation, is investigated further. Here, we aimed to assess and compare the impact of G2019S and I1371V mutant variants on membrane cholesterol content, fluidity, and lipid raft expression, along with substrate and autophosphorylation, as well as the corresponding cell surface expression of receptors and transporters. Gain-of-function studies were conducted in SH-SY5Y, U87, and floor plate cells (FPCs) derived from healthy control iPSCs to investigate these parameters.

2. Materials and Methods

2.1. Ethics Approval

The study received approval from the Institutional Committee of Stem Cell Research (IC-SCR) at the National Institute of Mental Health and Neurosciences (NIMHANS), under approval number SEC/01/005/B.P. Informed consent was obtained in accordance with IC-SCR guidelines. The study was also cleared by the Institutional Ethics Committee under IEC approval number NIMHANS/IEC/2022.

2.2. Generation of Floor Plate Cells (FPCs) from Induced Pluripotent Stem Cells (iPSCs)

Floor plate cells (FPCs) were derived from iPSCs following previously established protocols [29,38,39]. Briefly, healthy control iPSCs (NIMHAi006-A) were cultured on mouse embryonic fibroblast feeder layers treated with mitomycin C (Sigma, St. Louis, MO, USA, M4287-5X2MG). The iPSCs were maintained in a medium consisting of DMEM/F-12, 20% knockout serum replacement (Gibco, Waltham, MA, USA, 10828028), 1% penicillin–streptomycin (Gibco, Waltham, MA, USA, 15070063), 1% Glutamax (Gibco, Waltham, MA, USA, 35050061), 1% non-essential amino acids (Gibco, USA, 11140050), 0.1 mM 2-mercaptoethanol (Gibco, Waltham, MA, USA, 31350010), and 20 ng/mL FGF-2 (Immunotools, GmbH, Friesoythe, Germany, 11343627). Once the iPSCs formed compact colonies with clear edges, they were manually picked and transferred using an EVOS M5000 microscope (Thermo Fisher Scientific, Waltham, MA, USA, AMF5000). The cells were then seeded onto Geltrex™-coated 6-well plates at a density of 2.5–3.0 × 10⁵ cells per well and cultured in StemFlex medium (Gibco, Waltham, MA, USA, A3349401). After the cells adhered, they were switched to neural induction medium, which included neurobasal medium (Gibco, USA, 10888022) with neural induction supplement (Gibco, Waltham, MA, A1647701), 1% penicillin–streptomycin, and 1% Glutamax, with medium changes every other day. On the seventh day of neural induction, the resulting neural progenitors (NPs) were passaged enzymatically using StemPro™ Accutase™ (Gibco, Waltham, MA, USA, A1110501) and replated at a density of 0.5–1.5 × 10⁵ cells/cm². The medium was then changed to neural expansion medium (NEM), which consisted of a 1:1 mix of neurobasal medium and advanced DMEM/F-12 (Gibco, Waltham, MA, USA, 12634010), supplemented with 1X neural induction supplement, 1% penicillin–streptomycin, and 1% Glutamax. The neural progenitors were cultured in NEM with 25 ng/mL FGF8 (Immunotools, GmbH, 11344836) for an additional 7 days to promote midbrain patterning and generate FPCs.

2.3. Cell Culture and Transfection and LRRK2 Inhibitor Treatments

The SH-SY5Y and U87 cell lines were cultured in Dulbecco’s modified Eagle’s medium/F12 (DMEM/F12) enriched with 10% fetal bovine serum (FBS) (Himedia, Mumbai, Maharashtra, India), 1X Glutamax, and 100 μg/mL penicillin–streptomycin and incubated at 37 °C in a 5% CO₂.

The SH-SY5Y and U87 transfection was performed as earlier [22,39], wherein cells were cultured and maintained in DMEM/F12 supplemented with 10% fetal bovine serum (FBS), 1% penicillin–streptomycin, and Glutamax (all from Invitrogen, Waltham, MA, USA). Cells were transfected using the jetPRIME DNA transfection reagent (Polyplus-transfection_SA, Illkirch-Graffenstaden, France) with the pDEST51-LRRK2-WT (Addgene, Watertown, MA, USA, plasmid, Catalog No.: #25080), pDEST51-LRRK2-G2019S (Addgene, Watertown, MA, USA, plasmid, Catalog No.: #29401), and pDEST51-LRRK2-I1371V plasmids (Addgene, Watertown, MA, USA, plasmid #29399) and an empty vector (EV) for a duration of 18 h. After transfection, cells were selected against 2 mg/mL blasticidin (Invitrogen, Waltham, MA, USA) for 48 h. The transfection efficiency was assessed by examining the expression of V5-tagged LRRK2 protein using fluorescence-activated cell sorting (FACS) along with the expression for LRRK2 (Supplementary Figure S1A–F). Transfected cells successfully showed higher LRRK2 expression along with >80% immunopositive population for the V5 tag in comparison with the empty vector (EV) (Supplementary Figure S1B,E; p > 0.05 for SH-SY5Y cells and p > 0.001 for U87 cells). WT transfected cells were used as controls for the G2019S and I1371V mutant groups to ensure comparable LRRK2 expression across conditions, thereby minimizing effects arising from WT overexpression and enabling accurate assessment of mutant-specific phenotypes.

During the seventh day of the FPCs stage, the FPCs were transfected with jetPRIME_DNA transfection reagent (Polyplus-transfection_SA, Illkirch-Graffenstaden, France) using the pDEST51-LRRK2-WT (Addgene plasmid Catalog No.: #25080, USA), pDEST51-LRRK2-G2019S (Addgene, Watertown, MA, USA, plasmid Catalog No.: #29401, USA), and pDEST51-LRRK2-I1371V plasmids (Addgene, Watertown, MA, USA, plasmid #29399, USA) and an EV for a duration of 4 h. Following transfection, the cells were cultured in NEM, supplemented with a basal medium comprising a 1:1 ratio of neurobasal medium and advanced DMEM/F12 (Gibco, Waltham, MA, USA, 12634010), along with 1· neural induction supplement, 25 ng/mL FGF8, and 1% Glutamax. Cultures were maintained at 37 °C in a 5% CO₂ atmosphere. To ensure stable expression of the V5-tagged EV and LRRK2-WT, LRRK2-G2019S, and LRRK2-I1371V, the cells underwent selection against 1.5 mg/mL blasticidin (Invitrogen, Waltham, MA, USA) for 2 h after transfection. Subsequently, the adherent cells were maintained in NEM. The transfection efficiency of LRRK2-I1371V tagged with V5 in the cells was assessed using FACS and immunofluorescence techniques (Supplementary Figure S2).

Treatments: LRRK2-I1371Vtransfected SH-SY5Y and U87 cells were treated with 10 µM MLi-2 (Cayman chemical, USA, Catalog No.: 19305) and GW5074 (Cayman chemical, Ann Arbor, MI, USA, Catalog No.: 10010368); the final concentration was made in complete culture media and the treatment was given for 24 h at 37 °C in 5% CO₂ [40].

2.4. Semiquantitative RT-PCR

Reverse transcriptase PCR was performed on the transfected cell groups to analyze LRRK2 gene expression. Total RNA was extracted using TRIzol-LS Reagent (Invitrogen, Waltham, MA, USA, Catalog No.: 15596018), and 1 μg of RNA was reverse-transcribed into complementary DNA (cDNA) with a TaKaRa PrimeScriptTM RT Reagent Kit (Takara, Shimogyo-ku, Kyoto, Japan, Catalog No.: RR037A). TaKaRa EmeraldAmp PCR Master Mix (Catalog No.: RR310A) was used to amplify the LRRK2 gene using the following primers: for LRRK2: forward: 5′-ATGAGATATGCACTCTTCTG-3′, reverse: 5′-GCATGGATCCCAATGC-3′; and for the housekeeping gene 18S: forward: 5′-CGGCTACCACATCCAAGGAA-3′, reverse: 5′-GCTGGAATTACCGCGGCT-3′.

2.5. Immunophenotyping

Transfected cells were harvested and suspended in phosphate saline buffer (PBS), then centrifuged at 300 g for 5 min. The cell pellets were fixed overnight at 4 °C in 2% paraformaldehyde (Sigma P6148). For intracellular staining, cells were permeabilized with 0.1% Triton X-100 (Himedia, Mumbai, Maharashtra, India, RM845) for 20 min; this step was omitted when staining for cell surface (DAT) and lipid raft (caveolin-1) markers. The cells were subsequently blocked with 3% BSA (Himedia, Mumbai, Maharashtra, India, MB083) for 45 min. After blocking, cells were incubated overnight at 4 °C with primary antibody diluted 1:100. The primary antibodies against LRRK2, LRRK2 (phosphorylated at S935), LRRK2 (phosphorylated at S1292), V5 tag, Rab8A, Rab10, Rab8A (phosphorylated at T72), Rab10 (phosphorylated at T73), Caveolin-1, and DAT were used, followed by secondary antibodies (Alexa Fluor^®405, Alexa Fluor^®488 or Alexa Fluor^® 647) against respective primary antibodies and staining (dilution: 1:200) at room temperature for 90 min. RRID details of the antibodies are present in Supplementary Table S1. The stained cells were resuspended in 500 µL of sheath fluid (BD Biosciences, 342003) and analyzed on a BD FACSVerse flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA), recording 10,000 events per sample, and cells stained only with secondary antibody was used for gating. Data were analyzed using FACS Suite software (BD Biosciences, Version 1.0) to generate scatter plots.

2.6. Immunocytochemistry

SH-SY5Y cells were grown on 12 mm coverslips and, once confluent, were fixed with 4% paraformaldehyde (PFA) (Sigma P6148) and permeabilized using 0.1% Triton X-100 (Himedia RM845) for 15 min; this step was omitted when staining for cell surface (DAT) and lipid raft (Caveolin-1) markers. Afterward, cells were blocked for 45 min with 3% BSA (HIMEDIA, Mumbai, Maharashtra, India, MB083) in PBS. Primary antibody staining was performed at a 1:100 dilution and incubated overnight at 4 °C. The primary antibodies against LRRK2, LRRK2 (phosphorylated at S935), LRRK2 (phosphorylated at S1292), V5 tag, Rab8A, Rab10, Rab8A (phosphorylated at T72), Rab10 (phosphorylated at T73), caveolin-1, and DAT were used, followed by secondary antibodies (Alexa Fluor^®405, Alexa Fluor^®488 or Alexa Fluor^® 647) against respective primary antibodies and staining (dilution: 1:200) at room temperature for 90 min. Nuclei were counterstained with DAPI (4′,6-diamidino-2-phenylindole dihydrochloride; Sigma, St. Louis, MO, USA) for 2 min at a 1:400 dilution. RRID details of the antibodies are present in Supplementary Table S1. Coverslips were mounted onto glass slides using DABCO, and imaging was carried out with an Axio Observer.Z1/7 and ZEISS LSM 980 confocal microscope equipped with a Plan Apochromat 40X/1.3 Oil DIC (UV) VIS-IR M27 objective. Images were acquired using ZEISS ZEN 3.7 software (RRID: SCR_013672) with fluorescence contrast, the detector gain set to 650V, and pinhole sizes of 1 AU/35 μm for Track 1 (AF488) and AU/29 μm for Track 2 (DAPI). The excitation/emission wavelengths used were 488/517 nm for AF488 and 405/495 nm for DAPI. Images were captured at a resolution of 2791 × 2791 pixels with 8-bit depth and an effective numerical aperture (NA) of 1.3 and analyzed using ImageJ v1.52f (RRID:SCR_003070). Immunocytochemical controls were performed using only secondary antibodies, with nuclei counterstained with DAPI, to verify the absence of non-specific secondary antibody binding (Supplementary Figure S6).

2.7. Immunoblotting

Protein lysates were prepared from cell cultures, and 40 µg of total protein was loaded per lane, consistent with previously published protocols [22]. Proteins were separated using SDS–PAGE and transferred onto PVDF membranes using a semidry transfer system along with a PageRuler Prestained NIR Protein Ladder (Thermo Fisher Scientific #26635). Membranes were blocked with 5% BSA and incubated with primary antibodies (1:1000 dilution) against p-Rab8A, Rab8A, p-Rab10, Rab10, p-LRRK2 (S1292), LRRK2, and β-actin. Secondary detection was performed using AzureSpectra 800 anti-mouse for β-actin and 700 anti-rabbit antibodies (Azure Biosystems, Dublin, CA, USA) for p-Rab8A, Rab8A, p-Rab10, Rab10, phospho-LRRK2 (S1292), and LRRK2 at 1:10,000 dilution. Membranes were washed with TBST (0.1% Tween-20) between each step. Fluorescent signals were visualized using the Azure Biosystems Sapphire FL Imaging System, USA. Densitometric quantification was performed for phospho-Rab8A/total Rab8A, phospho-Rab10/total Rab10, and phospho-LRRK2 (S1292)/total LRRK2, each normalized to β-actin and expressed as relative density. Data analysis was carried out using ImageJ v1.52f (RRID:SCR_003070). RRID details of the antibodies are present in Supplementary Table S1. Full immunoblot images are provided in Supplementary Figures S7 and S8.

2.8. Measurement of Membrane Fluidity

A total of 6 × 10⁵ cells was harvested in Tris–KCl buffer (pH 7.4), and 10 µM TMA-DPH (Catalog No.: T204) was added to the cells, followed by incubation for 10 min at room temperature. The cells were then transferred to a Thermo-Nunc Flat Black 96-well plate, and fluorescence polarization was measured using a Spark^® Multimode Reader (Tecan, Männedorf, Switzerland) at an excitation/emission wavelength of 340/430 nm. The anisotropy readings were analyzed using the following formula [41]:

Anisotropy (r) = (Ivv − Gf × Ivh)/(Ivv + 2 × Gf × Ivh)

Where Gf = Ivv/Ivh;

Ivv = emitted parallel light intensity;

Ivh = emitted perpendicular light intensity.

Quantification of total cellular cholesterol

A total of 5 × 10⁶ cells was harvested in a complete lysis buffer (RIPA) and incubated on ice for 90 min. To the cells, 300 µL of RIPA lysis buffer (containing 150 mM NaCl, 50 mM Tris pH 8, 2 mM DTT, 1 mM EDTA, and 1% Triton X, along with SDS, DNase, a protease inhibitor cocktail, and PMSF) was added. The organic fraction containing cellular cholesterol was extracted using Folch’s lipid extraction method [42]. Briefly, 800 µL of chloroform and 400 µL of methanol were added to the isolated cell lysate, and the mixture was then centrifuged at 13,500 g for 15 min at 4 °C. The lower organic phase, which contained cholesterol, was collected and analyzed using a cholesterol ELISA with the total cholesterol (TC) colorimetric assay kit (Elabscience, Houston, TX, USA, E-BC-K109-S) being used to quantify total cellular cholesterol. Optical density (OD) of the colored compound benzoquinone imine phenazone was measured using a spectrophotometer (Tecan Spark multimode microplate reader) at 510 nm. The calculated cholesterol concentration in the sample was normalized to the cell number.

2.9. Quantification of Membrane Cholesterol

Quantification of membrane cholesterol was done as per our pervious study [39]. Briefly, 10 × 10⁶ cells were harvested in 1 mL of homogenization buffer (containing 250 mM sucrose, 1 mM EDTA, and 10 mM Tris–HCl, pH 7.2) and homogenized using a Cole-Parmer homogenizer. The cells were then sonicated at 30% amplitude with 3–10 s pulses, allowing for 5 s intervals between pulses, and subsequently centrifuged at 700 g for 10 min at 4 °C. The supernatant, which contained the soluble cytosolic fraction, was collected and subjected to ultracentrifugation at 100,000 g for 1 h at 4 °C. The resulting pellets, which contained the cellular membrane fraction, were dissolved in complete lysis RIPA buffer, and the organic fraction containing cellular cholesterol was extracted using Folch’s lipid extraction method [42]. Cholesterol quantification was performed from the isolated organic phase using a total cholesterol (TC) colorimetric assay kit (Elabscience, Houston, TX, USA, E-BC-K109-S) as described in the above section. The calculated cholesterol concentration in the sample was normalized to the cell number.

2.10. Quantification of Cellular Dopamine Uptake

WT, G2019S, and I1371V transfected SH-SY5Y cells were plated in a 24-well plate and, upon confluency, treated with 50 µM dopamine (Catalog #H8502G, Sigma, St. Louis, MO, USA) for 2 h at 37 °C. Extracellular dopamine was then removed, and the cell lysate was collected and used to measure cellular dopamine content using a dopamine ELISA kit (Catalog #K12-1302, KINESISDx, Wilmington, DE, USA) according to the manufacturer’s instructions. Untreated cell lysate was used as a blank control for the ELISA assay to differentiate and evaluate extracellular dopamine uptake from the intracellular dopamine already present in the cells of WT, G2019S, and I1371V. The obtained colorimetric values were then plotted on a cubic spline curve of standard values, and the unknown concentrations were calculated.

2.11. Atomic Force Microscopy

WT, G2019S, and I1371V transfected SH-SY5Y cells were plated onto 22 mm coverslips and, upon confluency, fixed with 4% PFA. After fixation, cells were air-dried overnight. The atomic force microscopy (AFM) images were taken in a non-contact scanning mode with a scanning size of 5.00 × 5.00 µm (X × Y) in NX20 (PARK Systems, Suwon, Republic of Korea), using a SCOUT_70_RAI probe. To find the nature of observed microdomains, WT transfected SH-SY5Y cells were treated with 5 mM methyl-beta-cyclodextrin (MβCD, Sigma, USA) for 15 min at 37 °C for cholesterol depletion, and cells were then subjected to atomic force microscopy.

2.12. Membrane Lipid Extraction

A total of 10 × 10⁶ cells was harvested in 1 mL of homogenization buffer (containing 250 mM sucrose, 1 mM EDTA, and 10 mM Tris–HCl, pH 7.2) and homogenized using a Cole-Parmer homogenizer. The cells were then sonicated at 30% amplitude with 3–10 s pulses, allowing for 5 s intervals between pulses, and subsequently centrifuged at 700 g for 10 min at 4 °C. The supernatant, which contained the soluble cytosolic fraction, was collected and subjected to ultracentrifugation at 100,000 g for 1 h at 4 °C. The resulting pellets, which contained the cellular membrane fraction, were dissolved in complete lysis RIPA buffer, and the organic fraction containing cellular cholesterol was extracted using Folch’s lipid extraction method [42]. An amount of 10 µL of SPLASH™ LIPIDOMIX™ Mass Spec Standard (Catalog No.: 330707, USA) was added to each sample before lipid extraction.

2.13. LC-MS/MS Protocol

The isolated organic phase was sent for lipidomics (LIPIDOMIX™, Alabaster, AL, USA) study at Centre for Cellular and Molecular Platforms (C-CAMP), GKVK Post, Bengaluru. Chromatography was performed using a Dionex Ultimate3000 UHPLC LC setup; injection volume: 5 µL. The column used was C 18 Waters Aquity BEH 2.1 × 100, 1.7 μm, where mobile phase A and B was 60: 40 ACN: water in 10 mM ammonium formate (0.1% FA) and 80: 20 IPA: ACN in 10 mM ammonium formate (0.1% FA), respectively. Flow rate was 0.15 mL/min and run time 25 min. Column oven and auto-sampler temperatures were set at 45 °C and 10 °C, respectively. The gradient was run as follows: 0–1 min: 0%B, 1–5 min: 0–40%B, 5–7.5 min: 40–64%B, 7.5–12 min: 64%B, 12–12.5 min: 64–82.5%B, 12.5–19 min: 82.5–85%B, 19–20 min: 85–95%B, 20–20.1 min: 95–0%B, 20.1–24.9 min: 0%B. Lipid extracts were analyzed using a Thermo Fisher Q Exactive mass spectrometer operated in both positive and negative ion acquisition modes with a survey scan over the mass range 200–1200 at resolution of 70 K and cycle time of about 1 s. Data was analyzed in ‘mzmine’ software version 4.8. Lipid peak areas (n = 3) were normalized to their respective internal standards, and intergroup comparisons were performed to assess statistical significance. Aligned data (including lipid identity, retention time, peak area, p-values, etc.) were exported to Excel. Lipids with statistically significant differences (p < 0.05) were selected for heatmap visualization.

2.14. Statistical Analysis

Data are reported as the mean ± standard deviation (SD). A one-way analysis of variance (ANOVA) was used for statistical comparisons, followed by Bonferroni post hoc analysis, using R software (R Foundation; R Project for Statistical Computing, RRID:SCR_001905). A p-value less than 0.05 was considered significant. Graphs were prepared using GraphPad Prism 6 (GraphPad Software; GraphPad Prism, RRID:SCR_002798) or Sigma Plot 12.5. Statistical significance was denoted by &: EV vs. WT/GS/IV; ^: WT vs. GS/IV/IV MLi-2/ IV GW5074; #: GS vs. IV; *: IV vs. IV MLi-2/IV GW5074; $: IV MLi-2 vs. IV GW5074; and @: WT vs. WT-MβCD. Single symbol: p  <  0.05; double symbol: p < 0.01; triple symbol: p  <  0.001. For all experiments, a sample size of N = 5 biological replicates (from five independent sets of transfected cells) was used, except for the lipidomic study and Western blot, which had N = 3 biological replicates. Data from the lipidomics study are represented as heatmaps showing relative intensity of lipid peak areas, showing significant p-values upon comparison between groups: WT: WT-LRRK2 transfected SH-SY5Y; GS: G2019S-LRRK2 transfected SH-SY5Y; and IV: I1371V-LRRK2 transfected SH-SY5Y, with internal standard-based normalization.

3. Results

3.1. Differential Effect of LRRK2 Genetic Variants, G2019S and I1371V, on Membrane Cholesterol and Membrane Fluidity

Rab8A, one of the substrates for LRRK2, is recognized for its role in sterol trafficking to the membrane [30,31]. We measured total cholesterol and membrane cholesterol levels in WT, LRRK2-G2019S, and LRRK2-I1371V transfected SH-SY5Y and U87 cells. The total cholesterol content in SH-SY5Y and U87 cells did not differ significantly among WT, LRRK2-G2019S, and LRRK2-I1371V variants (Figure 1B,D; p > 0.05). However, membrane cholesterol content was notably lower in LRRK2-I1371V compared with LRRK2-WT (p < 0.001) and LRRK2-G2019S (p < 0.001) transfected cells (Figure 1A,C). Additionally, LRRK2-G2019S showed lower membrane cholesterol compared with WT in both SH-SY5Y and U87 cells (Figure 1A,C, p < 0.05 in SH-SY5Y and p < 0.001 in U87).

Membrane cholesterol plays a crucial role in influencing membrane fluidity [43,44], which we assessed using membrane anisotropy measured using TMA-DPH labelling and polarization spectroscopy in live cells. Membrane anisotropy was significantly lower in SH-SY5Y, U87, and FPCs transfected with LRRK2-G2019S and LRRK2-I1371V variants compared with WT-LRRK2 transfected cells (Figure 1E–G; p < 0.001). Notably, cells expressing the LRRK2-I1371V variant exhibited the lowest membrane anisotropy, indicating the highest membrane fluidity among the genetic variants tested. This observation suggests that while cholesterol synthesis remained consistent across variants, differences in membrane cholesterol content significantly influenced membrane fluidity.

3.2. Differential Effect of LRRK2 Genetic Variants, G2019S and I1371V, on Cell Surface Expression of Caveolin-1

Cholesterol, being a key component of lipid rafts [45], was evaluated through the lipid raft marker Caveolin-1 in LRRK2-WT, LRRK2-G2019S, and LRRK2-I1371V transfected SH-SY5Y cells, U87 cells, and FPCs. Immunocytochemical analysis revealed substantially lower fluorescence intensity of caveolin-1 cell surface expression in LRRK2-I1371V transfected cells compared with WT and G2019S variants (Figure 2A–C). FACS analysis showed a significantly lower percentage of Caveolin-1 immunopositive cells in SH-SY5Y cells, U87 cells, and FPCs transfected with LRRK2-I1371V compared with those transfected with LRRK2-WT and LRRK2-G2019S (Figure 2D–F; p < 0.001).

3.3. Differential Effect of LRRK2 Genetic Variants, G2019S and I1371V, on Plasma Membrane Topology

The plasma membrane topology of LRRK2-WT, LRRK2-G2019S, and LRRK2-I1371V transfected SH-SY5Y cells was studied using non-contact mode atomic force microscopy (AFM) [46,47,48]. The AFM images revealed distinct microdomains in cells transfected with LRRK2-WT and -G2019S, whereas well-defined microdomains were absent in cells transfected with LRRK2-I1371V (Figure 3A, Figure 3D and Figure 3G, respectively). The color bar on the left edge of the 2-D AFM image provides a height scale for the membrane, with color intensity corresponding to surface elevation or depth. Figures on the right side of the 2-D images show high-resolution 3-D images detailing the ultrastructure of the plasma membrane topology (Figure 3B,E,H). Cells transfected with LRRK2-I1371V exhibited a less ruffled plasma membrane compared with WT and G2019S (Figure 3C,F,I). The height of the microdomains, denoted by Δz (nm), was measured using Gwyddion software (Version 2.66) and was significantly lower in I1371V cellular membranes compared with WT and G2019S (Figure 3J; p < 0.01 for WT vs. IV and p < 0.05 for GS vs. IV).

To investigate the effect of cholesterol on the microdomains, WT transfected SH-SY5Y cells were treated with methyl-beta-cyclodextrin (MβCD) for cholesterol depletion. AFM images of these treated cells showed a visible decrease in the number (Figure 3K), membrane ruffles (Figure 3L,M), and height (Figure 3N; p < 0.01) of microdomains. The membrane topology of these cholesterol-depleted cells closely resembled that of I1371V transfected cells, suggesting that cholesterol depletion affects microdomain formation. This indicates that the LRRK2-I1371V genetic variant impacts the formation of microdomains or lipid rafts in the membrane.

3.4. Altered Membrane Lipid Composition in LRRK2 Mutant Transfected SH-SY5Y Cells

To investigate the differential impact of LRRK2 mutant variants on membrane composition, a lipidomics analysis was performed on isolated membrane fractions from transfected SH-SY5Y cells. Both LRRK2-G2019S and I1371V transfected cells showed a significant reduction in lipid peak areas corresponding to cholesteryl esters (CEs), sterols (STs), and fatty acids (FAs), with the most pronounced decrease observed in I1371V cells (Figure 4A). Among sphingolipids, including ceramides (Cer), sphingomyelin (SM), steryl esters (SEs), and hexosyl ceramides (HexCer, Hex2Cer, and Hex3Cer), a marked decline was evident particularly in LRRK2-I1371V transfected cells (Figure 4B). Interestingly, HexCer species, critical for lipid raft structure, were significantly elevated in LRRK2-G2019S transfected cells compared with WT and I1371V. Additionally, LRRK2-I1371V transfected cells showed a notable decrease in glycerophospholipids—phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylinositol (PI), phosphatidylserine (PS), and lysophospholipids [lysophosphatidylcholine (LPC), lysophosphatidylethanolamine (LPE) lysophosphatidylinositol (LIP)] (Figure 4C)—as well as glycerolipids such as triglycerides (TGs) and diacylglycerol (DAG) (Figure 4D). Overall, the I1371V mutation resulted in the most severe disruption of membrane lipid composition, suggesting a distinct and potentially more deleterious effect on membrane integrity and function compared with the G2019S and WT.

3.5. Differential Effect of LRRK2 Genetic Variants, G2019S and I1371V, on Cell Surface Expression of DAT and Dopamine Uptake

Changes in the membrane composition and fluidity can influence the cell surface expression of membrane transport proteins [49,50]. As the dopamine transporter (DAT) is crucially important for dopamine uptake, we assessed its expression. Immunocytochemistry analysis revealed significantly lower fluorescence intensity of DAT in LRRK2-I1371V transfected SH-SY5Y cells compared with LRRK2-WT and LRRK2-G2019S transfected cells (Figure 5A–D; p < 0.001). FACS analysis confirmed a significantly lower population of DAT-immunopositive cells in LRRK2-I1371V transfected SH-SY5Y cells compared with WT and G2019S (Figure 5F; p < 0.001). Dopamine uptake was also significantly reduced in LRRK2-I1371V transfected SH-SY5Y cells compared with LRRK2-WT (p < 0.001) and LRRK2-G2019S (p < 0.05) transfected cells (Figure 5E). Additionally, LRRK2-G2019S transfected cells exhibited lower DAT cell surface expression (Figure 5D–G; p < 0.01) and dopamine uptake (Figure 5E; p < 0.01) compared with LRRK2-WT transfected cells. These findings suggest that the LRRK2-I1371V genetic variant has a more pronounced effect on dopamine uptake from the extracellular milieu due to reduced DAT expression.

3.6. Differential Effects of G2019S and I1371V Variants on LRRK2 Substrate Phosphorylation

Mutations in the kinase domain increase kinase activity, while mutations in the GTPase domain prolong kinase activity, leading to increased autophosphorylation and substrate phosphorylation [21,22]. Immunocytochemistry revealed significantly higher expression of phosphorylated Rab8A (pRab8A) and Rab10 (pRab10) in SH-SY5Y cells, U87 cells, and HC FPCs transfected with LRRK2-I1371V compared with the WT and the G2019S variant (Figure 6A,B; p < 0.001). FACS analysis of the pRab8A/Rab8A and pRab10/Rab10 ratios consistently showed the highest levels in cells transfected with LRRK2-I1371V across SH-SY5Y, U87, and FPCs (Figure 6C–H; p < 0.001). Representative histograms illustrating the percentage of immunopositive cells for pRab8A, Rab8A, pRab10, and Rab10 in WT, LRRK2-G2019S, and LRRK2-I1371V transfected cells are provided in Supplementary Figures S3, S4 and S5A–D. SH-SY5Y, U87, and FPCs transfected with LRRK2-G2019S exhibited a significantly higher pRab8A/Rab8A ratio (Figure 6C,E,G; p < 0.01 for SH-SY5Y, p < 0.001 for U87, and FPCs) and pRab10/Rab10 ratio compared with their respective WT transfected cells (Figure 6D,F,H; p < 0.001). Further, immunoblotting analysis confirmed the significantly higher expression of pRab8A/Rab8A and pRab10/Rab10 in LRRK2-I1371V transfected SH-SY5Y and U87 cells in comparison with WT and LRRK2-G2019S (Figure 7A–F; p < 0.001). Here too, the LRRK2-G2019S transfected cells showed significantly higher pRab8A/Rab8A and pRab10/Rab10 than their respective WT transfected cells (Figure 7A–F; p < 0.001). Therefore, substrate phosphorylation of Rab proteins (Rab8A and Rab10) consistently showed the highest levels in cells expressing LRRK2-I1371V, indicating a distinct pattern of substrate phosphorylation influenced by LRRK2 mutations. Expression of autophosphorylated LRRK2 at the S1292 residue was also assessed in LRRK2-WT, LRRK2-G2019S, and LRRK2-I1371V transfected SH-SY5Y and U87 cells using immunoblotting. Densitometric analysis revealed that phospho-LRRK2 (S1292) levels were significantly higher in G2019S- and I1371V-LRRK2 transfected cells compared with WT (Supplementary Figure S1G–J, p < 0.001). Notably, phospho-LRRK2 (S1292) levels were significantly elevated in LRRK2-I1371V transfected SH-SY5Y and U87 cells relative to LRRK2-G2019S, indicating a gain-of-function effect of the I1371V mutation on LRRK2 activity.

3.7. Differential Response of LRRK2-I1371V Transfected Cells to LRRK2 Inhibitors on Rab8A and Rab10 Phosphorylation

Response to LRRK2 inhibitors for LRRK2-I1371V transfected SH-SY5Y and U87 cells resulted in reduced phosphorylation of Rab8A (pRab8A) and Rab10 (pRab10), as measured by immunocytochemistry (ICC) (Figure 8A–H) and flow cytometry (Figure 8I–L). CTCF analysis revealed a significant decrease in pRab8A fluorescence following MLi-2 treatment in both SH-SY5Y and U87 cells compared with untreated LRRK2-I1371V transfected cells (Figure 8B,D; p-value < 0.001). Upon GW5074 addition, the reduction in pRab8A fluorescence intensity was even more pronounced in SH-SY5Y (p < 0.001) and U87 cells (p < 0.05) compared with their respective MLi-2-treated counterparts.

In SH-SY5Y cells, CTCF analysis of pRab10 showed a similar pattern. GW5074-treated cells exhibited lower pRab10 fluorescence intensity than both untreated LRRK2-I1371V transfected cells (Figure 8F; p < 0.001) and MLi-2-inhibited cells (p < 0.01). Similarly, U87 cells treated with inhibitors showed a significant decrease in pRab10 fluorescence intensity compared with untreated LRRK2-I1371V transfected cells (Figure 8H; p < 0.001). Flow cytometry further confirmed these results. FACS analysis of LRRK2-I1371V transfected SH-SY5Y cells demonstrated a significant decrease in both pRab8A/Rab8A and pRab10/Rab10 immunopositive population ratios following treatment with MLi-2 (p < 0.001) and GW5074 (Figure 8I,J; p < 0.001) compared with untreated controls. Notably, GW5074-treated cells showed a significantly greater decrease than the MLi-2-treated group (p < 0.001). LRRK2-I1371V transfected U87 cells showed similar results, with significant decreases in both pRab8A/Rab8A and pRab10/Rab10 immunopositive population ratios following MLi-2 and GW5074 treatment compared with untreated controls (Figure 8K,L; p < 0.001). However, while the pRab8A/Rab8A ratio decrease was greater with GW5074 than MLi-2 (Figure 8K; p < 0.001), the pRab10/Rab10 ratio was comparable between the two inhibitor treatments (Figure 8L). These findings demonstrate that GW5074 exhibits superior efficacy compared with MLi-2 in reducing phosphorylation of both Rab8A and Rab10 in LRRK2-I1371V transfected cells, making it a more potent inhibitor for this particular mutation. The drug-treated LRRK2-I1371V transfected cells were compared against LRRK2-I1371V transfected drug-untreated cells and LRRK2-WT transfected cells. Confocal images showing immunocytochemistry for pRab8A and pRab10 in LRRK2-WT and untreated LRRK2-I1371V transfected SH-SY5Y and U87 cells are shown in Figure 6A and B, respectively. Representative histograms illustrating the percentage of immunopositive cells for pRab8A, Rab8A, pRab10, and Rab10 in MLi-2- and GW5074-treated LRRK2-I1371V transfected SH-SY5Y and U87 cells are provided in Supplementary Figures S3 and S4E–H.

3.8. Differential Response of LRRK2-I1371V Transfected Cells to LRRK2 Inhibitors on Membrane Fluidity, Expression of Caveolin-1, and Membrane Topology

Treatment with both the LRRK2 inhibitors MLi-2 and GW5074 led to a substantial increase in membrane anisotropy for LRRK2-I1371V transfected SH-SY5Y and U87 cells (Figure 9A,B). GW5074 treatment induced a significantly greater increase in membrane anisotropy than MLi-2 treatment (p < 0.001) across both cell types. Thus, we conclude that GW5074 was much more efficient in restoring the lost membrane rigidity as compared with MLi-2.

GW5074 treatment significantly increased membrane cholesterol content in both SH-SY5Y (Figure 9C; p < 0.001) and U87 (Figure 9D; p < 0.001) cells compared with untreated LRRK2-I1371V transfected cells. Furthermore, GW5074-treated cells exhibited markedly higher membrane cholesterol levels than MLi-2-treated cells in both cell types (SH-SY5Y: p < 0.01; U87: p < 0.001). MLi-2 treatment also elevated membrane cholesterol content in U87 cells relative to untreated LRRK2-I1371V transfected controls (p < 0.001).

Both GW5074 and MLi-2 treatments significantly increased surface expression of the lipid raft marker Caveolin-1 in LRRK2-I1371V transfected SH-SY5Y and U87 cells compared with untreated LRRK2-I1371V transfected cells (Figure 9E–H; p < 0.001). Specifically, MLi-2 treatment increased Caveolin-1 expression in SH-SY5Y cells (p < 0.05) and U87 cells (p < 0.001), while GW5074 treatment produced even greater increases in both cell types (p < 0.001). Again, GW5074-treated cells exhibited significantly higher Caveolin-1 expression than MLi-2-treated cells (p < 0.05). The drug treated LRRK2-I1371V transfected cells were compared against LRRK2-I1371V transfected drug-untreated cells and LRRK2-WT transfected cells. The confocal images showing immunocytochemistry for caveolin-1 in LRRK2-WT and untreated LRRK2-I1371V transfected SH-SY5Y and U87 cells are shown in Figure 2A and B, respectively.

Unlike in MLi-2 treated LRRK2-I1371V SH-SY5Y cells, AFM images showed a higher number of microdomains in GW5074-treated LRRK2-I1371V transfected cells (Figure 10). These microdomains resembled those observed in wild-type and G2019S-LRRK2 transfected cells. GW5074 treatment successfully restored the microdomain presence that was lost in the I1371V mutation (Figure 3G–I), whereas MLi-2 treatment failed to achieve this restoration. Furthermore, GW5074 treatment restored the characteristic ruffled membrane morphology in LRRK2-I1371V transfected SH-SY5Y cells, as demonstrated in the 3-D representations (Figure 10A–F) and corresponding graphs (Figure 10C,F). This treatment also restored microdomain height to normal levels (Figure 10G, p < 0.001), an effect not observed with MLi-2 treatment.

4. Discussion

This study reveals that two pathogenic LRRK2 mutations—G2019S (GS) in the kinase domain and I1371V (IV) in the GTPase domain—produce strikingly different cellular phenotypes, with I1371V causing a more severe disruption of membrane dynamics, lipid homeostasis, and dopaminergic function. Using multiple cell models including SH-SY5Y neuroblastoma cells, U87 glioblastoma cells, and healthy iPSC-derived floor plate cells (FPCs), we demonstrate that these mutations lead to divergent pathogenic mechanisms, highlighting the complexity of LRRK2-associated disease.

A broad range of studies have demonstrated that even a single conservative substitution between isoleucine and valine can significantly alter protein behavior—including reduced NO synthesis in iNOS [51,52,53], altered inhibitor sensitivity in prostaglandin G/H synthase [54], relaxed ligand dependence in erbB [55,56], disrupted orexin receptor signaling [57,58] and enzyme properties of human glutathione S-transferase GSTP1-1 [59], modified ligand selectivity in PYL receptors [60,61], impaired coagulation in factor IX–related hemophilia B [62], pathogenic shifts in prion protein processing [63,64,65] in Creutzfeldt–Jakob disease, attenuated viral virulence in Newcastle disease virus [66], and conformational changes in E. coli ACP [67,68]. Collectively, these findings highlight that an Ile↔Val substitution, despite being chemically conservative, can profoundly influence enzyme kinetics, receptor function, protein stability, pathogenicity, and disease phenotypes across biological systems—a phenomenon replicated for the LRRK2 protein as well.

Overall, LRRK2 GTPase domain mutations (R1441C/G/H, I1371V, and Y1699C) present clinically with a typical PD motor phenotype and a good levodopa response [69]. Notably, I1371V, N1437H, N1437D, and Y1699C have been found to feature a younger age at onset of ~50 years, compared with R1441C/G/H [12]. With respect to DBS treatment, the LRRK2 N1437H carrier, another ROC domain mutation like I1371V, has reported a poor outcome in DBS treatment, likely resulting from the aggressive nature of the disease in that specific case [70]. A study including four LRRK2 R1441G PD patients in the Basque country of Spain stated that they too had a limited DBS response on motor function, daily life activities, and quality of life [71].

The picture with respect to Lewy body pathology appears to be variable. Similar to I1371V, several members of an affected family harboring the R1441C mutation were found to exhibit either Lewy pathology, neurofibrillary tangle pathology, or ubiquitin pathology [8]. Ubiquitin inclusions in the absence of Lewy pathology were also observed in two patients with the Y1699C mutation [8]. Intriguingly, no inclusions were detected in a patient with the R1441G mutation [72]. The V1447M variant was also associated with Lewy body pathology in the brainstem [73].

From a biochemical standpoint, LRRK2 is a bifunctional enzyme with a GTPase (Roc/COR) domain that regulates its C-terminal kinase domain. GTP binding promotes kinase activity, whereas GTP hydrolysis helps return the protein to a basal (less active) state. Mutations that decrease GTPase activity typically increase the proportion of GTP-bound, kinase-active LRRK2, thereby prolonging kinase activity [74,75]. In line with earlier studies on pathogenic mutations in the Roc/COR domain (e.g., N1437H, R1441C/G/H, Y1699C), we have also observed increased S1292 autophosphorylation and phosphorylation of Rab8A and Rab10 substrates in cells [39,76,77,78,79,80].

In our study, the GTPase domain variant I1371V showed robust S1292 phosphorylation in SH-SY5Y and U87 cells, with a higher pS1292/LRRK2 ratio compared with G2019S, suggesting elevated kinase activation. This S1292 autophosphorylation pattern following GTPase domain gain-of-function aligns with earlier studies [39,81,82,83]. Substrate phosphorylation patterns revealed pronounced mutation-specific effects as well. Phosphorylation of Rab8A and Rab10 was consistently highest in I1371V-expressing cells across all tested cell types, indicating that GTPase domain mutations exert more pronounced downstream effects on LRRK2 substrate targeting. The substrate phosphorylation finding carries particular significance given the established roles of Rab8A and Rab10 in vesicular trafficking and membrane dynamics. LRRK2 mediates lipid storage via Rab8A phosphorylation at the T72 residue [84], which is also required for LDL cholesterol delivery to the plasma membrane [30,31]. Recent work has further demonstrated that Rab10 plays a crucial role in membrane reservoir formation [85]. Phosphorylation of Rab8A and Rab10 at T72 and T73 residues, respectively, prevents binding of their subsequent GEFs and GAPs [86], resulting in functional arrest. These findings suggest that I1371V may exert stronger effects on vesicular trafficking and membrane remodeling via hyperactivation of Rab GTPases, potentially contributing to altered endolysosomal and recycling pathways.

Iannotta et al. (2020) [13] showed that kinase domain (G2019S) and Roc/COR domain (R1441C) LRRK2 mutations elicit distinct tissue- and age-dependent signaling profiles in mouse models. While Ser935 phosphorylation was reduced in both mutants, Ser1292 autophosphorylation was selectively increased in G2019S but not R1441C mice, whereas Rab10 phosphorylation was elevated in R1441C despite unchanged Ser1292 levels. Consistent with this context dependence, our biochemical data in human cells expressing the GTPase domain I1371V variant reveal heightened Rab phosphorylation along with autophosphorylation at Ser1292. The autophosphorylation pattern may differ from that of R1441C due to differences in stability, activity, and mutation effects between human and mouse [27]. Together with recent reports of cell type-specific LRRK2 signaling in mouse microglia [87], these findings underscore the importance of human cellular models and caution against direct cross-model comparison of Ser1292 and Rab phosphorylation as universal readouts of LRRK2 activity.

In line with Rab8A’s well-documented role in sterol trafficking, we observed that I1371V-expressing cells exhibited significantly lower membrane cholesterol compared with G2019S and wild-type controls, while total cholesterol levels remained unchanged. This pattern suggests that hyperphosphorylation of Rab8A by mutant LRRK2 impairs its function in cholesterol transport to the plasma membrane, resulting in cholesterol depletion and subsequent membrane destabilization.

The functional consequences of cholesterol depletion manifested as increased membrane fluidity, evidenced by reduced anisotropy measurements. This finding aligns with established biophysical principles where cholesterol acts as a membrane rigidity-promoting agent [43,44]. As membrane cholesterol is essential for maintaining lipid raft integrity [45], its depletion leads to disruption of critical cellular signaling domains [88,89]. The greater membrane fluidity observed with I1371V points to a more substantial compromise in membrane structure and function, which may have clinical relevance given that higher blood cholesterol levels in elderly individuals are associated with reduced risk of cognitive decline [90]. This connection is particularly noteworthy as patients carrying GTPase domain mutations in LRRK2 often exhibit severe cognitive impairment, with some cases reporting cognitive symptoms even before motor dysfunction [12].

A comprehensive lipidomics analysis corroborated the membrane cholesterol observations and revealed the broader scope of lipid alterations. LRRK2-I1371V transfected cells showed significant reductions in cholesteryl esters (CEs) and sterols (STs), with less pronounced declines in G2019S-expressing cells. Furthermore, essential lipid raft structural components—including sphingolipids, glycerophospholipids, and glycerolipids—were markedly depleted in LRRK2-I1371V cells. These findings align with those of Galper et al. (2022), who reported altered sphingolipid and glycerolipid levels in cerebrospinal fluid from PD patients carrying LRRK2 GTPase domain mutations, supporting disrupted lipid metabolism as a key pathogenic mechanism in PD [32].

In contrast, G2019S cells showed selective elevation of hexosyl ceramides, indicating a potential compensatory response that may partially maintain lipid raft structural integrity. These results demonstrate that I1371V exerts a significantly greater disruptive effect on lipid biosynthesis and distribution, which may be central to its pathogenic mechanism.

The biochemical changes in lipid composition translated into observable structural alterations in membrane architecture. Caveolin-1 expression, a marker for lipid raft domains [91], was significantly lower in I1371V-expressing cells across all examined cell types. AFM imaging confirmed a characteristically flattened membrane topology in I1371V-expressing cells with absent microdomains—a phenotype very similar to that observed following experimental cholesterol depletion. In contrast, G2019S-expressing cells showed minimal differences in membrane topology compared with wild-type controls.

The concurrent reduction of Caveolin-1 expression and disrupted microdomain architecture suggest that the I1371V mutation fundamentally impairs plasma membrane organization through cholesterol-dependent mechanisms. These structural disruptions carry significant functional implications, as lipid rafts serve as essential platforms for receptor clustering [92,93,94], signal transduction cascades [91], and protein trafficking networks [95]. The lipid alterations further implicate I1371V in compromising membrane architecture and raft-associated signaling.

The structural membrane changes resulted in clear functional consequences for dopaminergic signaling. Both DAT surface expression and dopamine uptake were significantly impaired in I1371V cells, with G2019S showing similar but less severe effects compared with wild-type controls. The reduced cell surface expression of DAT could result from hyperphosphorylation of Rab10, which is required for membrane docking of proteins [96].

These findings have important clinical implications as studies suggest a strong connection between cognitive decline in PD and reduced DAT and dopamine uptake [97]. Additionally, lower DAT levels have been associated with increased risk of impulse control disorder (ICD) in these patients [98]. The D2 receptor functions as a postsynaptic receptor involved in dopamine signal transmission, while DAT controls the spatial and temporal dynamics of dopamine neurotransmission by driving reuptake of extracellular transmitters into presynaptic neurons [99]. Our findings suggest a mechanistic link between loss of membrane integrity through disrupted raft domains and impaired transporter trafficking.

To ensure that the observed phenotypes are not due to non-physiological levels of LRRK2, all experiments used WT-LRRK2-overexpressing cells as the main control for the G2019S and I1371V mutants instead of untransfected or empty vector controls. This approach helps us account for the overall effects of LRRK2 overexpression—such as pathway saturation, changes in membrane association, or lipid remodeling—and specifically identify effects caused by the mutations. LRRK2 expression levels were quantitatively matched across WT, G2019S, and I1371V variants in all cell types used (Supplementary Figures S1 and S2).

Given that membrane structure factors such as lipid composition and fluidity can influence drug response, and as I1371V showed more substantial effects on membrane topology, composition, and fluidity, we next evaluated the effects of LRRK2 inhibitors MLi-2 and GW5074 on I1371V-expressing cells.

MLi-2 is a potent and selective inhibitor of LRRK2 kinase activity [100], while GW5074 is a relatively non-selective inhibitor that effectively attenuates both LRRK2 autophosphorylation and substrate phosphorylation [40]. Importantly, GW5074 has been shown to mitigate LRRK2-induced neuronal loss across multiple model systems, including C. elegans, Drosophila, murine models, and cellular assays [101]. We specifically selected GW5074 for treating I1371V-expressing cells based on earlier evidence demonstrating its superior efficacy over other LRRK2 GTPase domain mutations such as R1441C [23]. While the exact mechanism underlying this preferential effectiveness remains unclear, the recent lack of success of LRRK2 kinase inhibitors in clinical trials highlights the significant heterogeneity among the different variants of LRRK2 mutations. We therefore tested the response of both these inhibitors on LRRK2-I1371V-expressing cells.

LRRK2-I1371V transfected cells showed significant differential responses to these inhibitors, with GW5074 being markedly more effective than MLi-2 in reducing especially Rab8A hyperphosphorylation in both SH-SY5Y and U87 cells transfected with LRRK2-I1371V. GW5074 treatment also resulted in significant restoration of plasma membrane rigidity in LRRK2-I1371V transfected cells, showing effects superior to MLi-2 treatment. The observation that Rab8A knockdown in parental SH-SY5Y cells recapitulated the increased membrane fluidity phenotype provides additional evidence that Rab8A is essential for maintaining proper membrane lipid organization, presumably through its role in cholesterol trafficking. Collectively, these findings suggest that pathogenic LRRK2-I1371V mutation disrupts membrane homeostasis by hyperphosphorylating Rab8A, functionally inactivating its capacity to regulate cholesterol distribution and membrane fluidity. Additionally, membrane cholesterol content improved more effectively with GW5074 treatment, corroborated by increased Caveolin-1 expression and improved membrane topology, as analyzed by immunophenotyping and AFM imaging. AFM 2-D images revealed increased microdomain numbers following GW5074 treatment, with cellular membranes exhibiting significantly increased ruffling in 3-D representations—effects not observed with MLi-2 treatment. In LRRK2-I1371V transfected SH-SY5Y, Rab10 hyperphosphorylation was also decreased with GW5074.

These results suggest that the pathogenic mechanism of I1371V depends more on GTPase dysfunction than kinase hyperactivation, highlighting the need to broaden LRRK2 modulation strategies. The ability of GW5074 to restore microdomains and Caveolin-1 expression suggests that membrane defects associated with LRRK2 mutations may be potentially reversible, supporting the therapeutic potential of comprehensive LRRK2 inhibition.

Our findings support a mechanistic model (graphical abstract) where the LRRK2-I1371V GTPase domain mutation leads to enhanced autophosphorylation at S1292 and hyperphosphorylation of Rab substrates, particularly Rab8A. This hyperphosphorylation impairs sterol trafficking, leading to membrane cholesterol depletion, increased membrane fluidity, and lipid raft disruption. The resulting membrane alterations compromise the function of membrane-associated proteins like DAT, potentially contributing to cellular dysfunction and neurodegeneration.

The more severe disease phenotype associated with I1371V compared with G2019S suggests that mutations affecting GTPase activity may have broader cellular consequences than those primarily affecting kinase activity. This distinction has important implications for understanding disease progression and developing targeted therapies for familial PD patients carrying different LRRK2 mutations.

While our study provides comprehensive characterization of membrane alterations associated with two LRRK2 mutations, several limitations must be acknowledged. LRRK2 is also expressed in immune cells, and future studies should validate whether the differential effects between G2019S and I1371V variants are recapitulated in those cell types. Overexpression-based models represent a reductionist approach and may not fully recapitulate endogenous expression levels, regulatory mechanisms, or long-term cellular adaptations observed in patient-derived or in vivo systems. Given differences in LRRK2 stability, activity, and membrane composition between rodents and humans, it would be valuable to study the differential effects of these inhibitors using patient iPSC-derived midbrain organoids and assembloids, targeting both inflammatory signatures and dopaminergic neuronal survival and function. Additionally, the molecular mechanisms linking Rab protein hyperphosphorylation to specific membrane lipid changes require further investigation. Future studies employing radiolabeled dopamine, DAT inhibitors (e.g., GBR12935), and short-term kinetic assays will be required to directly quantify DAT-specific uptake. Understanding these pathways could reveal additional therapeutic targets and inform the development of combination therapies addressing both kinase activity and membrane integrity. Moreover, GW5074 is known to have broader kinase activity, and the potential off-target effects of GW5074 need to be investigated to determine whether other pathways beyond LRRK2 inhibition contribute to the rescue. Emerging evidence suggests that dysregulation of miRNA forms a crucial post-transcriptional regulatory layer affecting lipid metabolism, vesicular trafficking, and Rab GTPase signaling in PD, particularly pathways involving LRRK2 [102]. While miRNA profiling was not within this study’s scope, combining miRNA analysis with lipidomics and Rab-focused signaling in human iPSC-derived midbrain models represents a valuable direction for future research investigations.

5. Conclusions

Our findings elucidate distinct molecular and functional consequences of the LRRK2-G2019S and -I1371V mutations. While both variants enhance LRRK2 kinase activity, as evidenced by increased autophosphorylation, the I1371V mutation results in markedly elevated phosphorylation of downstream substrates Rab8A and Rab10 across all examined cell types. Crucially, the study establishes membrane dysfunction as a central pathological mechanism in LRRK2-associated PD, with the I1371V mutation demonstrating more severe effects than G2019S. The I1371V variant exerts broader and more severe effects on membrane lipid composition, topology, membrane fluidity, and dopamine transporter function. Significantly, we also demonstrate that these membrane abnormalities are reversible through pharmacological inhibition of LRRK2. The non-selective GTPase/kinase modulator GW5074 was particularly effective in rescuing the I1371V-induced deficits, outperforming the direct kinase inhibitor MLi-2, suggesting that GTPase domain variants may respond better to broader LRRK2 modulators than to kinase-selective inhibitors. These insights deepen our understanding of the diverse mechanisms driving LRRK2-associated neurodegeneration and pathobiology and underscore the importance of developing mutation-specific therapeutic approaches for PD treatment.