Leucine-rich repeat kinase 2 (LRRK2) mutations are recognized as a major contributing factor to Parkinson’s disease (PD) [1-3]. In fact, both familial and sporadic forms of PD share common pathological mechanisms associated with LRRK2 mutations, whether through genetic variations or non-coding alterations at its locus [4-7]. Pathogenic mutations in the LRRK2 gene have been demonstrated to enhance its kinase activity, and this abnormal increase in LRRK2 kinase activity is believed to play a role in the pathogenesis of PD [8-12]. Consequently, pharmacological inhibition of LRRK2 kinase activity has been explored as a potential therapeutic strategy for PD patients.

PD primarily manifests as the loss of dopaminergic neurons in the brain, leading to motor symptoms such as tremors, rigidity, and bradykinesia. However, PD is not limited to motor symptoms; it also affects the autonomic nervous system, which controls involuntary functions like heart rate, blood pressure, and digestion. The interplay between impaired sympathetic vasoconstriction and compensatory autonomic responses may disrupt blood flow regulation and increase the risk of stroke in PD patients. For instance, neurogenic orthostatic hypotension can result in reduced cerebral blood flow, rendering the brain more susceptible to ischemic events [13, 14].

Notably, recent studies have reported an elevated risk of stroke among PD patients [15, 16]. Specifically, individuals with PD attributed to mutations in the LRRK2 gene appear to have a higher stroke risk than those with PD caused by other factors. Moreover, when compared to PD patients with mutations in other genes such as parkin and glucocerebrosidase, who exhibit stroke risk levels similar to controls, it suggests that LRRK2-associated mechanisms may significantly contribute to the heightened stroke risk in PD patients with LRRK2 mutations. Hence, we posit that pharmacological inhibition of LRRK2 kinase activity may offer potential benefits in reducing the risk of stroke. To test this hypothesis, we conducted an investigation into the functional role of LRRK2 and the neuroprotective effects of an LRRK2 inhibitor in a mouse stroke model.

Rose Bengal (RB) photothrombosis mice model

Male C57BL/6 mice (7~8 weeks old) were obtained from Damul Science (Daejeon, Korea). To induce photothrombosis, a photosensitive dye, Rose Bengal (RB), dissolved in PBS, was injected through the tail vein at a dosage of 0.5 mg/100 μl. To aggregate the dye, an area of 0.6 mm in diameter in the right cerebral cortex was locally illuminated with green light (535±25 nm) using an LED lamp (ZEISS, CL 6000 LED) for 3 minutes. For pharmacological inhibition of LRRK2, starting from 2 days before RB photothrombosis modeling, 30 mg/kg of MLi-2 (Selleck Chemicals, TX; n=10) or the vehicle control (n=10) were administered via oral gavage once daily. Mice were euthanized 72 hours after modeling.

Immunohistochemistry

C57BL/6 male mice were anesthetized with isoprene + 1% O2 and perfused intracardially with 4% paraformaldehyde in PBS buffer. To prepare cryostat sections, post-fixed brains were cryoprotected in 30% sucrose (wt/vol) in PBS for 48 hours at 4°C, embedded in Tissue-Tek OCT compound (Sakura Finetek USA, CA), and frozen on dry ice. Cryostat sections were blocked for 1 hour at room temperature using a solution containing 5% normal goat serum (Vector Laboratories Inc., CA) in 0.01 M PBS with 0.5% Triton X-100. Subsequently, sections were incubated with primary antibodies diluted in blocking solution at 4°C overnight. The following primary antibodies were used: anti-LRRK2 (Abcam, MA), anti-p-LRRK2 (Ser1292) (Abcam), anti-NeuN (Millipore, MO), anti-GFAP (Dako, Glostrup Municipality, Denmark), anti-PINK1 (Novus Biologicals, CO), anti-Parkin (Abcam), anti-LC3B (Cell signaling, MA), anti-cytochrome C (Abcam), and JC-1 (Invitrogen, MA) dye. Sections were then incubated with Cy3- and Cy2-conjugated secondary antibodies (Jackson ImmunoResearch, PA) for 2 hours at room temperature. Following PBS washes, tissues were incubated with DAPI for 3 minutes. After three washes, tissues were mounted on slides and visualized under a fluorescence microscope. The Click-iT^TM TUNEL Alexa Fluor^TM 488 Imaging Assay (Thermo Fisher Scientific, MA) was used to analyze neuronal apoptosis following the manufacturer’s instructions.

Infarct volume quantification

Brains were removed 24 hours after RB photothrombosis, sectioned into 2 mm slices, and immersed in 1% TTC (Sigma-Aldrich, MO) solution at 37°C for 15 minutes, followed by fixation with 4% paraformaldehyde. ImageJ software was utilized to measure the areas of ischemic damage for each brain slice. Lack of TTC staining defined the core, while red staining indicated viable tissue. The total infarction volume was calculated as the sum of core and penumbra volumes. Damage volume was expressed as a percentage of the ipsilateral hemisphere volume (damage/total volume × 100). This calculation was performed for 4 hemispheric brain slices per animal.

Neurological deficits measurement

The neurological status of the experimental animals was assessed using an adapted five-point scale test based on Yang et al [17]. The criteria were as follows: Grade 0=normal neurological function, Grade 1=failure to extend the right forepaw, Grade 2=circling, Grade 3=leaning to the contralateral side at rest, Grade 4=no spontaneous motor activity, Grade 5=dead. Neurological deficit was evaluated for 24 hours after RB photothrombosis. Motor performance was assessed using the Wire Hang test, conducted 1 day and 3 days after RB photothrombosis. A wire cage top (18 inches×9 inches) with taped edges was employed for this experiment. Mice were placed on the metallic wire, the top was inverted, and latency to fall was recorded. This was repeated three times with a 5-minute interval between attempts. The cutoff time was set at 60 seconds, and the mean latency to fall was determined [18].

Image analysis

The puncta measurement data were analyzed with scripts written in ImageJ. The analyses were performed using a script written in ImageJ for batch processing to handle the hundreds of ROIs imaged per slide.

Statistical analysis

All data were presented as mean±SEM. Statistical differences among more than two groups were analyzed using one-way ANOVA followed by Tukey post hoc analysis. Differences between two groups were analyzed using a t-test, and a p-value<0.05 was considered statistically significant. Statistical calculations were performed using the GraphPad Prism program.

Induction of LRRK2 and p-LRRK2 in the Rose Bengal Photothrombotic Stroke Model

The functional involvement of LRRK2 in the risk of stroke was elucidated by its expression and activation in the Rose Bengal (RB) photothrombosis model, which simulates human stroke induced by blood clots [19, 20]. Histological analysis of cortex lesions 3 days after RB photothrombosis revealed an increased number of cells with LRRK2-positive puncta in infarct lesions, suggesting that the induction of LRRK2 protein expression might be associated with the risk of stroke (Fig. 1). Notably, LRRK2 puncta were observed in the penumbra (the boundary of the infarct), but not in the infarct core lesions. This is in line with previous studies indicating that endogenous LRRK2 is present at low levels in most cell types [21], LRRK2 expression was not observed in the contralateral areas. Similarly, p-LRRK2 (Ser1292), which represents LRRK2 kinase activity, was highly detected in the in the penumbra (the boundary of the infarct) but absent in the contralateral areas.

Pharmacological inhibition of LRRK2 kinase activity mitigates neurological deficits in the RB photothrombosis model

The hypothesis that LRRK2 activation under stroke conditions could exacerbate brain damage suggested the therapeutic potential of LRRK2 inhibition in treating stroke. To investigate this, we utilized MLi-2, a highly selective LRRK2 inhibitor capable of penetrating the blood-brain barrier [22]. Mice received 30 mg/kg of MLi-2 daily via oral gavage, starting 2 days before RB modeling and continuing for 2 days after modeling (Fig. 2A). The repeated administration of 30 mg/kg of MLi-2 in mice markedly inhibited p-LRRK2 (Ser1292) in the penumbra and exhibited the excellent tolerability and no adverse effects (Fig. 2B). In contrast, the control groups treated with the vehicle did not exhibit any LRRK2 activation. As seen in Fig. 1 and 2B, RB modeling strongly activated LRRK2 in the penumbra (the boundary of the infarct), while MLi-2 administration attenuated this RB-induced LRRK2 activation. Compared to the RB-only groups, the MLi-2-treated RB group displayed weaker p-LRRK2 signals that did not form puncta, suggesting that MLi-2 treatment effectively reduces LRRK2 kinase activity in the context of RB-induced stroke.

Importantly, the MLi-2 treatment group exhibited a one-third lower infarct volume than the vehicle treatment group following RB photothrombosis (infarct volume in the vehicle-treated RB groups, 32.00%±2.21%; in the MLi-2-treated RB groups, 21.60%±1.21%; p=0.003). Additionally, analysis of neurological deficit scores (vehicle-treated RB groups, 2.30±0.42; MLi-2-treated RB groups, 1.40±0.16; p=0.031) and results from the wire hang behavior test, conducted 3 days after RB photothrombosis (vehicle-treated RB groups, 8.23±1.01; MLi-2-treated RB groups, 14.77±1.65; p=0.0013), indicated that MLi-2 treatment significantly attenuated RB photothrombosis-induced neurological deficits (Fig. 2D, E).

p-LRRK2 co-localization with mitochondrial apoptosis signaling proteins

The punctate LRRK2 expression in the RB photothrombosis model suggested a potential connection to the autophagy process, which involves the delivery of intracellular components to the lysosome for degradation [23, 24]. One well-studied form of autophagy is macro-autophagy, and the degradation of mitochondria via recruitment to autophagosomes by the autophagy mediator, LC3, is known as mitophagy. Damaged mitochondria can trigger apoptotic stress and cell death through the release of cytochrome c. Importantly, LRRK2 mutations have been shown to impair PINK/Parkin-dependent mitophagy in several studies [25-28]. To investigate whether LRRK2 activation in the RB photothrombosis model could be related to PINK/Parkin-dependent mitophagy, we analyzed the co-localization of p-LRRK2 (Ser1292) with mitophagy-related proteins (PINK1, Parkin, LC3B) [29]. Additionally, we examined its co-localization with cytochrome c, a marker of mitochondrial apoptosis, as several studies have reported that LRRK2 mutations could induce cell toxicity via mitochondria-dependent apoptosis [23, 30]. As shown in Fig. 3A, p-LRRK2 (Ser1292) strongly co-localized with PINK1, Parkin, LC3B, and cytochrome c in cortical infarct penumbra at 72 hours post-stroke. Furthermore, late-stage apoptosis was assessed by detecting DNA & mitochondrial fragmentation using the TUNEL assay in core infarcts at 72 hours post-stroke (Fig. 3B), revealing an overlap between p-LRRK2-positive puncta and TUNEL-positive puncta.

Pharmacological inhibition of LRRK2 attenuated mitochondrial apoptosis

A distinctive feature of the early stages of apoptosis is the disruption of active mitochondria, including changes in mitochondrial membrane potentials. The JC-1 dye, capable of intercalating into the mitochondrial membrane, is commonly used to evaluate mitochondrial membrane potential changes in apoptosis [31, 32]. To elucidate the relationship between LRRK2 activation and mitochondrial apoptosis, brain tissues obtained from the Rose Bengal photothrombosis model were stained with JC-1. Consistent with Fig. 3, p-LRRK2 puncta were found to co-localize with JC-1-positive mitochondrial puncta, suggesting that elevated p-LRRK2 could be associated with mitochondrial apoptosis (Fig. 4A). Furthermore, the number of JC-1 and p-LRRK2-positive puncta was significantly lower in the MLi-2-treated RB group compared to the vehicle-treated RB group (Fig. 4A, B). This observation was in line with the results showing JC-1 and p-LRRK2-positive fluorescence intensity (Fig. 4C).

This study represents the first report of the possible involvement of LRRK2 expression and its activation in an in vivo stroke mouse model. RB photothrombosis induced the expressions of both LRRK2 and p-LRRK2 (Ser1292), which has been demonstrated as a physiological and direct marker of LRRK2 kinase activity [5, 6]. Moreover, LRRK2 and p-LRRK2-positive puncta were observed in the infarct boundary but not in infarct core lesions, suggesting that the induction/activation of LRRK2 in the progression of stroke may determine the cell's fate to die, as the functional abnormality of LRRK2 has been known to be pathogenic [10-12].

Interestingly, pharmacological inhibition of LRRK2 using MLi-2 demonstrated an alleviation of stroke-induced neurological deficits in the in vivo stroke mouse model. The MLi-2-treated RB group showed significant improvements in neurological deficit scores and behavioral test results. Considering the higher stroke risk in patients with PD with LRRK2 mutations [15] and the reports indicating that LRRK2 mutations induce an abnormal increase in its kinase activity contributing to PD pathology, our results suggest that LRRK2 inhibition in PD treatment could reduce the risk of stroke in patients with PD as well as slow the progression of PD.

So, how does pharmacological inhibition of LRRK2 work in stroke? The pattern of LRRK2/p-LRRK2-positive puncta suggests a link to physiological vesicle-mediated signaling pathways, such as autophagy. Increased autophagosomes and autolysosomes are often seen in degenerating or dying neurons, representing autophagic stress [33, 34]. These phenotypes, including autophagic vacuoles, have been observed in various neurodegenerative diseases, including PD [35-38]. Additionally, the functional abnormality of LRRK2 caused by mutations has been shown to result in mitochondrial dysfunction and impaired autophagy of mitochondria, known as mitophagy [25-28]. Mitophagy has strong ties to PD, as it was discovered that PINK1 and Parkin, two other genes mutated in familial PD, initiate mitophagy through mitochondrial depolarization [39-42]. Interestingly, several studies have reported that pathogenic LRRK2 mutants impair mitophagy by reducing mitochondrial trafficking [25, 26] and disrupting the recruitment of mitochondrial proteins like Drp1 and Parkin [27] and inhibited small GTPase RAB10 interaction with the autophagy receptor OPTN [26]. Consequently, impaired mitophagy has been proposed as a common factor in PD. The activation of LRRK2 in PINK/Parkin-dependent mitophagy in RB photothrombosis suggests that the LRRK2-driven PINK/Parkin-dependent mitophagy mechanisms observed in PD may also be involved in the progression of stroke.

Bae et al. [43], group found that various in vitro and in vivo models of traumatic brain injury (TBI) markedly enhanced LRRK2 expression in neurons and also increased the level of hypoxia-inducible factor (HIF)-1α. They also revealed direct binding of HIF-1α in LRRK2 proximal promoter and found that HIF-1α-dependent transcriptional induction of LRRK2 exacerbated neuronal cell death following injury. These studies are very similar to our findings, and the mechanism of HIF-1a dependent transcriptional induction of LRRK2 may also operate in stroke, and future research is needed. In addition, the co-localization of p-LRRK2-positive puncta with cytochrome c/TUNEL/JC-1-positive puncta in RB photothrombosis suggests that elevated p-LRRK2 may induce mitochondrial apoptosis in infarct lesions. Importantly, the number of JC-1 and p-LRRK2-positive puncta in RB photothrombosis was significantly reduced by MLi-2 treatment, indicating that pharmacological inhibition of LRRK2 can attenuate mitochondrial apoptosis and ultimately exhibit neuroprotective potential in the progression of stroke.

In summary, LRRK2 activation may contribute to neuro-pathogenesis through impaired mitophagy in the progression of stroke, and pharmacological inhibition of LRRK2 may hold promise in reducing the risk of stroke and delaying the progression of PD in patients with LRRK2 mutations.

This work was supported by the project grants (SI2131-10, SI2231-10 & KK2331-10 funded by KRICT), the National Research Foundation of Korea (NRF-2022R1A2B5B02001886), and by research fund of Chungnam National University.

The authors declare that they have no conflict of interest.

All protocols performed in accordance with the Chungnam National University hospital Guide for the Care and Use of Laboratory Animals and were approved by the University of Chungnam Institutional Animal Care and Use Committee.