The role of Rab GTPases in the pathobiology of Parkinson’ disease

Abstract

Rab GTPases are key regulators of vesicle-mediated transport and are proposed to play a crucial role in the pathobiology of Parkinson’s disease. As membrane trafficking seems to be a relevant pathway altered in Parkinson’ disease, understanding the role of Rab GTPases in the disease progression could open a window for therapeutic interventions. In this review, we focus on the recent advances on the role of Rab GTPases in the biology of two main proteins involved in Parkinson’s disease: LRRK2 and a-synuclein, given that mutations in their genes (LRRK2 and SNCA) cause familial and sporadic Parkinson’s disease.

INTRODUCTION

Vesicle-mediated transport is regulated by different families of proteins that play distinct roles to ensure an appropriate trafficking of cargos, including proteins, lipids and organelles, within the cell. Important mediators of these processes include the Ras-like, Rab GTPases. Currently, over sixty distinct Rabs have been identified in the human genome. Like the other members of the Ras superfamily, Rabs interconvert between their inactive, GDP bound, form to an active, GTP bound, state. Rab proteins are stably prenylated which is important for their association with membranes. Prenylation occurs on one or two cysteine residues at the C-terminus by geranylgeranyl transferases after binding with REPs (Rab escort proteins). Once attached to a membrane, Rab proteins are activated by the exchange of GDP for GTP, mediated by guanine nucleotide exchange factors (GEFs). GTP is subsequently hydrolyzed to GDP, the rate of which is enhanced by GTPase activating proteins (GAPs), thus enabling the recognition of Rab proteins by a GDP dissociation inhibitor (GDI) that retrieves Rabs from membranes and sequesters Rabs in the cytosol.

Membrane trafficking plays an essential role in maintaining cellular homeostasis. As such, trafficking is thought to be particularly critical in human diseases such as neurodegenerative disorders. Currently, vesicle-mediated transport has emerged as a key pathway altered in different types of neurodegenerative diseases[1]. In particular, Parkinson’s disease (PD) has been suggested to arise from defects in membrane trafficking using human genetic data [2],[3], suggesting that there is a causal relationship between altered membrane trafficking and PD pathogenesis. Specifically, mutations in PINK1, VPS13C, VPS35, DNAJC6, DNAJC13, PRKN, FBXO7, ATP13A2 or PLA2G6 have been discovered in familial cases of PD (for a review[4]) and linked to various aspects of trafficking. In addition, genomewide association studies (GWAS) in sporadic PD have identified risk variants present in loci that include several membrane trafficking genes such as TMEM175, RAB7L1, GAK, CHMP2B, CTSB, GALC, VPS13C, SH3GL2, SYT4, ATP6V0A1or VAMP4[5–7**]. Additionally previous work proposed a functional connection between Rab GTPases and PD[8][9].

Based on this shared biology between sporadic and familial PD, it is reasonable to think that there are potential therapeutic targets that hopefully could develop into pharmacological treatments coming from manipulation of these pathways. In this review, we will summarize the most recent data on the role of Rab GTPases in LRRK2 and ??-synuclein biology, two heavily studied PD proteins that are currently under investigation for potential therapeutic avenues to limit disease progression. We will discuss how insights into vesicular trafficking might lead to a deeper understanding of the relationship between these PD genes.

Leucine rich repeat kinase 2 (LRRK2)

Point mutations in LRRK2 cause familial late-onset PD and GWAS approaches have nominated variants in the LRRK2 locus as a risk factors for sporadic PD. LRRK2is therefore an example of a pleomorphic locus, in this case on human chromosome 12. The LRRK2 gene encodes a large protein (280 KDa) with seven proposed domains: an Armadillo domain at the N-terminus is followed by an Ankyrin domain then by Leucine-rich repeats (LRR). In the center of the sequence, a Ras of complex (ROC) and C-terminus of ROC (COR) domains collective encode and regulate GTPase activity and may be important in dimer formation of LRRK2. The ROC-COR bidomain is then followed by a Kinase and a WD40 domains at the C-terminus of the protein [10].

PD-causing variants in LRRK2 occur in the ROC-COR bi-domain (N1437H, R1441C/G/H, R1628P, Y1699C/G) and the kinase domain (I2012T, G2019S, I2020T) and are gain-of-function mutations leading to a toxic hyperactive protein[11],[12]. Among its proposed substrates, LRRK2 phosphorylates fourteen Rab GTPases (Rab3A/B/C/D, Rab5A/B/C, Rab8A/B, Rab10, Rab12, Rab29, Rab35 and Rab43) at a conserved residue (Serine or Threonine) in the switch II domain[13,14**] (Figure 1, Table 1). To date, it appears that all pathological mutations in LRRK2 result in enhanced Rab phosphorylation. In addition, LRRK2-mediated Rab phosphorylation is abrogated by LRRK2 kinase pharmacological inhibition. It is noteworthy that LRRK2-driven Rab GTPase phosphorylation occurs in different cell types and has been shown in mouse tissue at endogenous levels[15], which further strengthens the idea that Rab phosphorylation is a critical pathway to PD pathogenesis.

Figure 1. The Parkinson’s disease kinase LRRK2 and its Rab substrates.

LRRK2 with its seven domains (the Kinase domain colored in red). Upon activation, LRRK2 phosphorylates its auto-phosphorylation site (S1292) and 14 different Rab GTPases (Rab3A/B/C/D, Rab5A/B/C, Rab8A/B, Rab10, Rab12, Rab29, Rab35 and Rab43). In blue, pathogenic mutations found in familial cases of late-onset autosomal dominant PD. ARM, Armadillo. ANK, Ankyrin. LRR, Leucine-rich repeat. ROC, Ras of complex. COR, C-terminal of ROC. KIN, Kinase. P, phosphorylation.

TABLE 1.

Summary with the known Rab GTPases involved with LRRK2 and ??-synuclein

Rab GTPase	Function	References
Rab7A	Promotes ??-synuclein clearance	[40], [41]
Rab8A	Disrupts the centrosome upon LRRK2 phosphorylation Recruited by LRRK2 in overload LYS	[14], [14]**, [22], [23], [24]
Rab8B	Blocks ??-synuclein propagation Impairs ??-synuclein aggregation Promotes ??-synuclein clearance	[31]*
Rab10	Blocks ciliogenesis upon LRRK2 phosphorylation Recruited by LRRK2 in overload LYS	[13], [20]*, [14]**
Rab11A	Impairs ??-synuclein aggregation Promotes ??-synuclein clearance	[31]*
Rab13	Blocks ??-synuclein propagation Impairs ??-synuclein aggregation Promotes ??-synuclein clearance	[31]*
Rab29	Candidate for the PARK16 locus Recruits LRRK2 to the TGN Recruits LRRK2 to overload LYS	[7]**, [14]**, [16], [17], [18]*, [19]*
Rab35	Increases ??-synuclein propagation upon LRRK2 phosphorylation	[14]**, [49]**, [50]
Rab39B	Impairs ??-synuclein aggregation	[31]*, [33]

Independently, human genetic data has demonstrated that RAB29 (also known as RAB7L1) is a candidate for the PARK16 locus in the chromosome 1[5],[6]. Rab29 has been previously linked to LRRK2 by our group and others[16],[17]; indeed Rab29 physically interacts with LRRK2 and recruits it to the trans-Golgi network (TGN) membrane. In order to accomplish this relocalization, Rab29 needs to be in an active state (GTP-bound)[16]and prenylated, as mutation of cysteines in the C-terminal domain impairs LRRK2 recruitment to the TGN[18*]. Once forming a complex with Rab29 (possibly through the Ankyrin domain[19*]) at the TGN, LRRK2 kinase activity is enhanced, measured by auto-phosphorylation of pS1292. Consequently, active LRRK2 phosphorylates its Rab substrates (Rab10, Rab8A or Rab29) suggesting that Rab29-dependent TGN relocalization of LRRK2 is a key mediator for LRRK2 activity and the subsequent Rab phosphorylation[18*],[19*]. An important question that requires clarification is whether LRRK2 activation is dependent of Rab29 specific binding and recruitment to the TGN, or simply due to the presence of LRRK2 at a membranous compartment.

Although it is not clear at this stage how such molecular events lead to neurodegeneration, different groups have linked LRRK2 to specific Rab8A/Rab10 cellular functions, such as ciliogenesis, centrosome dynamics or lysosomal exocytosis (Figure 2, Table 1) that could be intermediary biological events between mutation and neuronal damage. For example, Rab8A and Rab10 phosphorylation by LRRK2 at T72/T73 mediates the interaction between the Rab proteins and RILPL1 and RILPL2. RILPL1 acts a primary cilia inhibitor. Therefore cells carrying a hyper-active LRRK2 pathogenic mutated protein (R1441G) present defective ciliogenesis[13][14,20*] through Rab10 phosphorylation and its enhanced binding to RILPL1[20]. In vivo data suggest that striatal cholinergic and cortical somatosensory neurons in the Lrrk2-R1441C knock-in mouse brain present defective ciliogenesis, which allowed the authors to propose that a lack of SHH-GDNF circuit could fail to provide neurotrophic support from the striatum to the midbrain dopaminergic neurons[20*]. However, it is important to note Lrrk2-R1441C knock-in mice fail to reproduce any PD-related pathological phenotype[21], suggesting that ciliation defects are not sufficient by themselves to result in dopaminergic cell loss, a key pathological event in PD.

Figure 2. LRRK2-dependent Rab GTPases cellular functions.

Schematic representation on the proposed functions of Rab GTPases that are substrates for LRRK2 (Rab8A, Rab10 and Rab29). Thin arrows indicate LRRK2-mediated phosphorylation. Phosphorylated Rab10 blocks ciliogenesis and phosphorylated Rab8A disrupts centrosome positioning. Overload late endosomes/lysosomes recruit LRRK2 that in turns phosphorylates Rab8A and Rab10. The proposed mechanism suggests a consequent recruitment of effectors EHBP1 and EHBP1L1 to promote lysosomal exocytosis. LE, late endosome. LYS, lysosome. P, phosphorylation. Question marks indicate unresolved aspects of the model.

Similarly, pathogenic LRRK2 causes centrosomal positioning defects by Rab8A phosphorylation in cultured cells[22]. Rab8A-mediated centrosomal defects are also induced by Rab29-dependent relocalization of LRRK2 to the TGN, which activates LRRK2 and consequently phosphorylates Rab8A[23]. Even though the link between LRRK2 and centrosomal positioning/ciliogenesis has therefore been described mechanistically, the connection between these processes and PD pathogenesis is less clear.

By pharmacologically overloading the lysosome through chloroquine addition, LRRK2 has been proposed to recruit Rab10 and Rab8A and its effectors (EHBP1 and EHBP1L1) to lysosomes to induce lysosomal exocytosis, in order to secrete non-degraded cargo to the extracellular space [24]. Although these results are initially compelling, caution must be taken with interpretation in the context of PD for several reasons. First, chloroquine-induced LRRK2 recruitment to the lysosomes has mainly been studied in cancer cells. Second, lysosomal exocytosis was evaluated exclusively by cathepsin D expression and mainly its intermediate form, which is present in late endosomes and not in lysosomes[25]. Finally, it is not yet clear that the effect of Rab8A, Rab10 and their effectors on lysosomal secretion in response to chloroquine is through LRRK2 phosphorylation per se or mediated through additional pathways, including toxicity. Further work is warranted to understand the relationships between lysosomal damage, LRRK2 activation and Rabs at different intracellular membranes.

These observations link LRRK2 to Rab biology and, from there, to vesicular trafficking events in cells. As discussed above, a major unresolved question is how these cellular alterations result in neurodegeneration. However, what we will next describe is the data linking Rab biology to a second PD-relevant protein, ??-synuclein, that is perhaps more readily linked to PD pathology in that it is a major deposited protein in the diseased human brain, as well as a second pleiomorphic risk locus for PD.

??-SYNUCLEIN

As well as neuronal cell death, a hallmark of PD is the presence of intracellular protein inclusions in surviving neurons called Lewy bodies that contain the protein ??-synuclein. ??-synuclein is encoded by the SNCA gene, and mutations in SNCA as well as duplications and triplications cause familial autosomal dominant PD (for a review[26]). In addition, the SNCA locus in the chromosome 4 has been nominated as a risk factor for sporadic PD through GWAS[27],[5,6]. ??-synuclein is a small protein (14 KDa) with a central amyloid domain that is crucial for its aggregation. Natively folded ??-synuclein forms protective tetramers that block aggregation[28]. In its monomeric form, ??-synuclein aggregates to form potentially toxic oligomers and fibrils that can further aggregate into Lewy bodies. Additionally, ??-synuclein inclusions can also propagate from cell to cell, potentially spreading the disease throughout the brain [29]. The balance between ??-synuclein aggregation, propagation and degradation may therefore determine a given neuron’s fate and become a determinant factor in PD pathogenesis. Consequently, promoting ??-synuclein clearance or blocking its propagation or aggregation have been proposed as therapeutic targets for PD[30].

In terms of understanding these phenomena at a molecular level, one recent shRNA screen of 1387 genes involved in vesicle-mediated transport identified nine modulators of ??-synuclein aggregation, propagation and degradation. Among them, four Rab GTPases were nominated as ??-synuclein regulators (Rab39B, Rab13, Rab11A and Rab8B). In cells, knockdown of any of these four Rabs lead to a higher propensity to aggregate ??-synuclein whereas cell-to-cell propagation was increased only upon RAB8B and RAB13KD. Endocytic recycling of ??-synuclein oligomers seems to be mediated by Rab11A and Rab13, whereas Rab11A, Rab13 and Rab8B may also be involved in ??-synuclein inclusion clearance[31*] (Figure 3, Table 1). Interestingly, loss-of-function mutations in RAB39B cause early onset PD with ??-synuclein pathology in humans, suggesting that this Rab may play a more specific role in PD pathogenesis [32],[33]. It would be very interesting to know if altering the expression of these Rabs can in turn affect ??-synuclein dynamics the brain, leading to new therapeutic approaches.

Figure 3. Rab GTPases as key regulators of ??-synuclein aggregation, propagation and clearance.

The cartoon depicts the role of certain Rab GTPases as important modifiers of ??-synuclein biology. MVB, multivesicular body. LE, late endosome. LYS, lysosome. P, phosphorylation. Question marks indicate unresolved aspects of the model.

Rab7A (also known as Rab7) has several different functions within the cell including membrane transport from early to late endosomes[34], lysosomal positioning[35], recycling components from the endosomes to the TGN[36], mitophagy[37], endosomal maturation[38] and autophagosome to lysosome fusion during autophagy[39]. In a C. elegans model of ??-synuclein-dependent dopaminergic neuron loss, over expression of the HOPS complex subunit Vps41 rescued neurotoxicity through Rab7 and AP-3[40]. It is likely that Rab7 and HOPS work together by recruiting an effector (possibly PLEKHM1) to fuse autophagosomal/endosomal material to the lysosome for degradation[41]. In contrast, in HEK293 cells, overexpression of Rab7 increases ??-synuclein clearance, therefore reducing apoptosis[42]. The protective role of Rab7 in clearing ??-synuclein aggregates seems to be mediated by its effector FYCO1[43]. Indeed, in a fly model, Rab7 and FYCO1 rescue motor deficits induced by mutant (A53T) ??-synuclein, thus proposing Rab7 as a possible therapeutic target in PD (Figure 3). FYCO1 promotes Rab7-dependent transport of autophagosomes to the cell periphery[44]while autophagosome fuse to lysosomes in the perinuclear area. Hence, it is unusual that autophagosome anterograde transport might cause ??-synuclein degradation and these results require a deeper mechanistic evaluation before we could consider Rab7 as a new therapeutic target.

Although there is clearly additional data needed, these results link ??-synuclein to Rab biology and, indirectly, to LRRK2-mediated processes. However, there is some additional data that suggests a more direct link between LRRK2 and ??-synuclein via Rab35.

RAB35 AS A LINK BETWEEN LRRK2 AND ??-SYNUCLEIN

Several independent sources of data suggest that Rab35 may be important in PD pathogenesis. For example, Rab35 protein levels are increased in the serum of PD patients compared to matched controls and patients with other parkinsonism disorders and Rab35 serum levels significantly correlate with age-at-onset of the disease [45]. Rab35 is also one of the potential LRRK2 Rab substrates. Mutations in the LRRK2 phosphorylation site for multiple Rabs cause neurotoxicity in primary neurons that is especially severe in the Rab35 phospho-mutants (both phosphomimetic and phospho-null)[46]. Additionally, over expression of Rab35 phospho-mutants in the murine substantia nigra by viral delivery causes neurodegeneration. One proposed mechanism for PD progression in the brain is the propagation of ??-synuclein aggregates [47], and LRRK2 pathogenic mutations have been proposed to increase ??-synuclein fibrils propagation in primary neurons[48]. Recently it has been suggested that LRRK2-mediated ??-synuclein propagation occurs through Rab35 phosphorylation at its T72 site[49**]. A proximity labelling screening for potential ??-synuclein interactors in rat cortical neurons showed that native ??-synuclein is likely to interact with Rab35, among other Rab GTPases (Rab3A, Rab3B, Rab3C, Rab4B, Rab6A, Rab8A and Rab15)[50]. Even though the molecular mechanism of how phosphorylated Rab35 is able to propagate ??-synuclein aggregates remains unclear, it is known that endocytosed ??-synuclein is degraded in the lysosome after transfer through the endosomal pathway and that ESCRT-III depletion blocks ??-synuclein degradation and increases its release from the cell[51]. ESCRT-III promotes the biogenesis of multivesicular bodies (MVB), and Rab35 stimulates the release of MVB material to the extracellular space [52],[53]. It is therefore possible that phosphorylated Rab35 impairs ??-synuclein clearance by promoting its release through exosomes thus enhancing ??-synuclein propagation (Figure 3, Table 1).

CONCLUDING REMARKS

Rab GTPases are crucial actors in cellular trafficking and recent data discussed here suggests a specific role for these proteins in PD. For instance, it is now known that LRRK2 phosphorylates fourteen different Rabs thus regulating their cellular function in a way that could be relevant for the disease progression. Understanding the molecular pathway underlying pathogenic LRRK2-driven toxicity could lead to the discovery of new therapeutic targets. It is safe to predict that in the close future more studies on LRRK2 cellular function through Rab GTPases will be published, which in turn could increase our knowledge on the role of LRRK2 in vesicle-mediated transport. However, it has not been established yet that all of the vesicular trafficking roles of LRRK2 are mediated by its Rab substrates given that LRRK2 has multiple protein-protein interaction domains. It is therefore important to remain open to the possibility that additional interactions of LRRK2 might also be important in PD.

The specific phosphorylation of Rabs by LRRK2 has also a potential use as biomarker of PD activity and of future therapeutic responses, especially considering that pathogenic LRRK2 is hyperactive and drugs are being developed to counteract kinase activity. It is therefore essential to evaluate these signaling events in endogenous tissues to establish their potential pathological relevance. In a similar way, ??-synuclein aggregation and propagation are proposed causes for PD progression. Different Rab proteins have been nominated as potential regulators for these events in cultured systems and invertebrate animals, thus suggesting its use for therapy. It would be of crucial interest to assess whether the modulation of these Rabs could actually modify ??-synuclein aggregation, clearance and propagation in rodent models.

HIGHLIGHTS.

• Recent advances in human genetics point to membrane trafficking as a key pathway in Parkinson’s disease (PD)

• LRRK2 phosphorylates fourteen different Rab proteins in a conserved residue in the Switch II domain

• By phosphorylating its Rab substrates, LRRK2 regulates ciliogenesis, centrosome dynamics and lysosomal exocytosis

• Different Rab GTPases act as α-synuclein regulators by controlling its aggregation, clearance and propagation

• By phosphorylating Rab35, LRRK2 promotes the propagation of α-synuclein aggregates (a proposed mechanism for PD progression).