This is my first post and I am happy to share with the world a bit of literature review on a protein involved in Parkinson’s disease that I wrote:
The Role of LRRK-2 in Parkinson’s Disease (PD)
Heather Van Tassel
December 15, 2017
University of British Columbia
Abstract
This review summarizes the current understanding of leucine-rich repeat kinase domain (LRRK2)’s role in Parkinson’s Disease (PD) progression regarding its biochemical activity, molecular interactions, and alterations in mutant phenotypes. Evidence from many types of animal models and cell lines is used to build general conclusions and speculations on the cellular pathways LRRK2 and its different mutations are involved in, and how these depict LRRK2’s role in PD. Emphasis on recent findings in c. elegans will be useful in determining future research avenues to help find viable cures. The occurring themes of vesicle trafficking, cytoskeletal dynamics, protein translation and autophagy disruption and their key pathogenic protein interactions and pathways are discussed. Contributions of alpha-synuclein to LRRK-2 PD symptoms are intriguing and deserve further investigation.
Keywords: Parkinson’s Disease, LRRK2, vesicle trafficking, autophagy, cytoskeletal dynamics, protein translation, kinase, genetic risk
The role of LRRK-2 in Parkinson’s Disease (PD) Parkinson’s Disease (PD) remains the second most common neurodegenerative disease. Given that age is the greatest risk factor for developing the disease (Collier et al., 2011), and that Canada has an aging population, Parkinson’s disease prevalence is expected to double over 25 years to reach 8.7-9.3 million by 2030 (Dorsey, 2007). PD is characterized clinically by resting tremor, rigidity and hypokinesia, and presents pathologically with loss of dopaminergic neurons in the substantia nigra, and usually lewy body formation containing aggregated alpha-synuclein and or tau proteins. There are currently no therapies that can slow or prevent the disease. While 90% of cases are sporadic, genetic factors have been found to play a role in 5-10% of cases. There are three familial genes known to be associated with PD: alpha-synuclein (SNCA), leucine-rich repeat kinase-2 (LRRK2), and vacuolar protein sorting-associated protein 35 (VPS35) (Zimprich et al., 2004)(Rideout, 2017). It is interesting that SNCA PD is fully penetrant, while LRRK2 mutations have reduced and age-dependent penetrance (Healy et al., 2008; Clark et al., 2006; Goldwurm et al., 2007; Ozelius et al., 2006; Ruiz-Martinez et al., 2010). Perhaps not surprisingly, the disease pathology most commonly found in LRRK2 carriers is lewy-bodies, composed of alpha-synuclein aggregates (Hardy, 2009). However, not all animal models and humans with classic PD symptoms present with lewy body pathology, making LRRK2 a promising candidate in the search for the root cause of most forms of PD. Mutations in the LRRK2 (Park8) gene are the most common cause of the genetic form of the disease and drive pathogenesis in both familial and non-familial PD (Kumari and Tan, 2009). Thus, this review focuses mainly on LRRK2 and how it may be involved with other genetic risk factors in driving PD pathogenesis. Both familial and non-familial forms of LRRK-2 PD have been found to have an almost indistinguishable clinicopathology regarding incomplete penetrance, age of onset, the likely (although not required) presence of Lewy bodies (LBs), and motor and nonmotor symptoms (Healy et al., 2008; Kasten et al., 2010; Silveira-Moriyama et al. 2008; Haugarvoll et al., 2008). Current findings suggest that LRRK-2 plays an important role in the dysregulation of protein translation, vesicle trafficking, neurite outgrowth, autophagy, and cytoskeletal dynamics. LRRK2 Structure and Function LRRK2 is a large, multi-domain protein composed of 2527 amino acids (289 kDa). It contains a kinase domain sequence, a Ras of complex catalytic protein domain (ROC) and the regulatory C-terminal of ROC (COR) domain that are predicted to bind and hydrolyze GTP similarly to the ROCO protein family (Gotthardt et al., 2008). These three domains are considered the catalytic core of LRRK2. Additionally, LRRK2 has Ankyrin repeats, Leucine-rich repeats (LRR) and a WD40 domain that is important for protein folding, and thus kinase activity, and predominantly serves as binding sites for protein-protein interactions and structural scaffolds for different signaling processes. Thus, unsurprisingly, the entire protein, including the C-terminal domain is required to produce acute toxic effects in neuronal culture (Jorgensen et al., 2009). This implies that only a functional LRRK2 protein will induce toxicity and that the C-terminal domain plays a key role in maintaining its structure. The normal function of LRRK2 is unknown, however is implicated in vesicle trafficking, protein synthesis, cytoskeletal dynamics, and immune function. LRRK2 Mutations LRRK2 mutations are predominantly found in the kinase (G2019S, I2020T) and the ROC-COR GTP-ase domains (R1441C/G/H, Y1699C), implying that these enzymatic activities are crucial for pathogenesis (Rudenko and Cookson, 2014). There are seven missense mutations known to be truly pathogenic: G2019S, R1441C/G/H, I2020T, Y1699C, and G2385R (Aasly et al., 2010). Defining the functional characteristics of theses mutants is ongoing; current findings are summarized in Table 1. The seven pathogenic mutations are shown in relation to the entire LRRK2 protein in Figure 1. The G2019S kinase domain mutation is the most common LRRK2 mutation leading to PD, and activates the kinase two- to threefold compared to wild-type (WT) (West et al., 2005; Khan, 2005; Jaleel et al., 2007; Healy et al., 2008). Increased LRRK2 kinase activity has been shown to retard neuritic outgrowth and extension in several primary neuronal cultures (Smith et al., 2006; Macleod et al., 2006; Smith et al., 2005; Greggio et al., Dachsel et al., 2010) Additionally, many in vivo invertebrate models have shown similar evidence, building support for the toxic effect of increased LRRK2 kinase activity (Imai et al., 2008; Yao et al., 2010; Liu et al. 2008). Transgenic G2019S LRRK2 mice display progressive degeneration of substantia nigra pars compacta dopaminergic neurons and motor dysfunction, as well as disrupted vesicle trafficking, suggesting that this mutation may be functionally relevant to the disease (Chen et al., 2009)(Pan et al., 2017). LRKK2 mutations associated with sporadic Parkinson’s Disease The R1441 residue is the second most common spot of pathogenic LRRK2 mutations with three substitutions: R1441C, R1441G, and R1441H (Healy, 2008). While the location of these mutations is in the ROC-COR GTPase domain, they have generally been observed to increase kinase activity, although results vary. This is likely due to intramolecular regulation between the GTPase and kinase domains. The R1441H mutation exhibits slowed GTP hydrolysis and increased affinity for GTP (Saha et al., 2014). Y1699C is also in the GTPase domain and exhibits similar kinase activity levels as other 1441 mutants. I2020T has been associated with both increased and decreased kinase activity which may depend on the substrate used in the assay (Ray S et al., 2014). All of these mutations have been shown to have enhanced vulnerability to mitochondrial dysfunction, inhibition of autophagy, and neurodegeneration, but their biochemical functional effects are not well characterized (Rideout, 2017). G2385R has recently been discovered to be a partial loss of function mutant that regulates tethering of synaptic vesicles (Carion et al., 2017). It binds many proteins with its WD40 C-terminal domain involved in synaptic vesicle exocytosis in pre-synaptic membranes. It is genetically linked to PD in Asian and Korean families and while it is associated with similar clinical symptoms to idiopathic PD, it significantly lowers the expected age of onset (Tan et al., 2009). This finding complicates the requirements for LRRK2 toxicity, and does not align with the G2019S toxic gain of function theory. Perhaps, a more intricate balance of LRRK2 activity is important for cell survival, as increased or decreased presence of the protein seems to influence toxicity. This is very perplexing and deserves further study. Effects of over-expression and knock-out of LRRK2 Over-expression of WT, mutants and knock-out models of LRRK2 have also been used to gain insight into its function. WT LRRK2 seems to have a toxicity protective effect; LRRK2 was found to have a protective effect on dopamine (DA) neurons in C. elegans from the toxicity of 6-hydroxydopamine and human α-synuclein (Yuaun et al., 2011). However, when LRRK2 WT is overexpressed, cellular dysfunctions such as fragmentation of the Golgi complex and dopaminergic degeneration occur (Lin et al., 2009; Xiong et al., 2012). Additionally, G2019S-LRRK2 overexpression results in mitochondrial uncoupling (Papkovskaia et al., 2012 ). The observations that over-expression of LRRK2, but not knock-out, leads to increased dopaminergic neurodegeneration (Xiong et al., 2012) suggest that PD-causing mutations of LRRK2 are primarily toxic gain of function mutations. Thus, the majority of findings support the idea that higher levels of the protein are more toxic, likely due to its kinase activity. Figure 1 LRRK2 protein domains
Figure 1: LRRK2 protein domains Table 1 LRRK2 pathogenic mutations and related symptoms and mechanisms of pathogenicity Mutation Location Symptoms Functional level References G2019S Kinase domain • Inhibition of neuritic outgrowth and extension. • Lewy body most common morphology • Enhanced vulnerability to mitochondrial dysfunction • Inhibition of autophagy • Neurodegeneration • Increased incidence of tumors. • Lysosomal enlargement and sorting deficits • Anxiety-like behaviour. • Increase in activated microglia. • Elevated glutamate release • increased phosphor-tau • Increased mTOR expression in kidneys • Increases kinase activity 2-3 fold • Toxic gain of function via various pathways Smith et al., 2006; Macleod et al., 2006; Smith et al., 2005; Greggio et al., Dachsel et al., 2010 Saha 2009 and 2014 Liu et al. 2011 Yao, 2010 Lin et al., 2009 Beccano-Kelly et al., 2015
R1141C ROC-COR GTPase domain • Enhanced vulnerability to mitochondrial dysfunction • Inhibition of autophagy • Impairs dopaminergic neurotransmission • Neurodegeneration • Decreased catecholamine release in cultured cells. Variable- no to moderate increase in kinase activity Saha 2009 and 2014. Liu et al., 2011. Yao et al., 2010 Tong, 2009 R1141G ROC-COR GTPase domain • Enhanced vulnerability to mitochondrial dysfunction • Inhibition of autophagy • Neurodegeneration Variable- no to moderate (20%) increase in kinase activity. Also may exhibit tauopathy Saha 2009 and 2014. R1141H ROC-COR GTPase domain • enhanced vulnerability to mitochondrial dysfunction • inhibition of autophagy • Neurodegeneration • slowed GTP hydrolysis and increased affinity for GTP • Variable- no to moderate increase in kinase activity Saha 2009 and 2014. Liao et al., 2014 Y1699C ROC-COR GTPase domain • enhanced vulnerability to mitochondrial dysfunction • inhibition of autophagy • Sometimes exhibit neurofibrillary tau tangles and increased alpha-synuclein levels Variable- no to moderate increase in kinase activity Kett and Dauer 2012 I2020T Kinase domain Sometimes exhibit neurofibrillary tau tangles Variable- decrease and increase in kinase activity shown Kett and Dauer 2012 Ray et al., 2014 G2385R C-terminal WD40 domain Causes increased SV fusion via decreased binding of SV loss of function variant. Reduced kinase activity Rideout, 2017 LRRK2 knockout n/a • no discernible effects on the dopaminergic system. • Immune response: increased sensitivity to inflammatory bowel disease • Increases translocation of NFAT to the nucleus • Increased mTOR expression in kidneys Loss of all activity Rideout, 2017
Liu et al., 2011
LRRK2 WT n/a • Protects against mitochondrial stress
• Age-dependent inhibition of autophagy, when alpha-synuclein is expressed.
Healthy Saha,et al., 2015, 2009
LRRK2 over-expression n/a • dopaminergic neurodegeneration
• fragmentation of Golgi complex
• retention of NFAT in the cytosol, which inhibits immune responses Increased kinase activity Liu et al., 2011
Papkovskaia et al., 2012
Lin et al., 2009; Xiong et al., 2012
WT: wildtype; SV: synaptic vesicles NFAT: nuclear factor of activated T-cells; GTP: Guanosine triphosphate mTOR: target of rapamycin
LRRK2 Phosphorylation and molecular regulation of kinase and GTPase domains In order to begin to understand LRRK2’s role in disease progression, it is important to determine how the enzyme is activated and how this relates to potential downstream proteins actions and pathways. Several key functions of LRRK2 are regulated by its phosphorylation status, including its dimerization and interactions with other proteins, such as 14-3-3, localization, and its turnover (Rideout, 2017). Phosphorylation of LRRK2 plays a role in its own dimerization (Deng et al., 2008). LRRK2 dimers are thought to possess greater kinase activity than monomers and may be enriched at membranes (Sen et al. 2009; Berger et al. 2010). Since LRRK2 mutations are thought to be toxic gain-of function, investigating how this protein dimerizes is important. It is also very interesting that over-expressed LRRK2 and LRRK1 hetero-dimerize in coimmunoprecipitation studies (Klein et al. 2009; Dachsel et al. 2010). Thus, LRRK2 kinase activity may be affected by LRRK1 and its relevance to phosphorylation levels and significance to modifying disease warrant further study. Phosphorylation status of LRRK2 mediates its interactions with many other proteins including 14-3-3 proteins, which regulate cellular localization. Loss of 14-3-3 protein binding results in cytoplasmic accumulations and filamentous skein-like structures (Nichols et al., 2010; , Kett et al 2012; Li et al., 2011). It seems that constitutive phosphorylation of LRRK2 regulated by upstream kinases modulates the recruitment of LRRK2 to the trans-Golgi network (Chia et al., 2014). Probable upstream kinases include casein kinase 1alpha, and PKA (Rideout, 2017). In C. elegans, the localization of LRRK-1 (homolog of LRRK2) is dependent on UNC-23-HSP-1 chaperone complex, which are homologous to mammalian BAG2 and HSC70/HSP70, respectively (Fukuzono et al., 2016). BAG2 is phosphorylated by MAPKAPK2, which is activated by oxidative stress. The importance of cellular localization should not be overlooked, as a study linked kinase activity and cellular localization to regulation of macroautophagy (Schapansky et al., 2014). Macroautopagy is important for orderly degradation and recycling of cellular components, and is impaired in LRRK2 PD models. The final known role of phosphorylation is in LRRK2’s turnover: dephosphorylation increases LRRK2 ubiquination and subsequent degradation with calyculin A-sensitive ubiquitination (Zhao et al., 2015). This regulation seems to be controlled by phosphatases including PP1 (Lobbestael et al., 2013), and is reversible via phosphorylation (Rideout, 2017). Different LRRK2 mutations show contradicting patterns in phosphorylation and kinase activity; The R1441C/G and Y1699C mutations exhibit increased kinase activity but decreased phosphorylation, leading to more ubiquitination, while G2019S displays increased kinase activity and no change in phosphorylation or ubiquitination compared to WT (Lobbestael et al., 2013). One site of autophophorylation has been validated (S1292) and varies in phosphorylation levels across variants (Sheng et al., 2012) and does not seem to correlate with kinase activity (Reynolds et al., 2014). However, most mutations increase overall phosphorylation levels (West et al., 2007). Since S1292 is conserved across species and several LRRK2 mutants increase autophosphorylation of this residue, it may be central to its activity and potentially a disease biomarker (Sheng et al., 2012). The molecular structure of LRRK2 has not been solved, and thus the debate over how the two enzymatic (kinase and GTPase) domains interact with one another to facilitate similar or opposing functions is still ongoing. The GTP-ase domain is thought to help activate the kinase domain, as several GTP-binding deficient mutants have reduced kinase activity in vitro (Smith et al., 2006). The GTPase domain can be activated in the monomeric or dimeric state, and R1441H mutant LRRK2 prolongs this active state (Liao et al., 2014). Supporting evidence for this hypothesis includes a study in which a c. elegans GTP-binding defective mutant rescues G2019S-LRRK2 neurodegeneration and locomotor dysfunction shown by basasl slowing response and speed (Yao et al., 2010). This evidence is also in line with the theory of kinase-induced toxicity of LRRK2, as higher levels of kinase and GTPase activity seem to increase aberrant phenotypes, while decreasing these levels attenuates them. Much further research is needed to determine the specific roles of autophosphorylation in intra and inter-molecular activation of LRRK2, and how this affects its enzyme kinetics and substrate binding.
LRRK-2 interactions with other proteins To look for potential protein substrates and regulators of LRRK-2, it has been useful to look at other less common genetic mutations found in PD such as PTEN-induced kinase (PINK1), Parkin, DJ-1, GBA and FBXO-7 (Hardy et al., 2009; Calahorro and Ruiz-rubio, 2011; Cookson and Bandmann, 2010). In fact, 28 independent risk variants for PD were identified across 24 loci (Nalls et al., 2014). Three of these genes are implicated in autosomal recessive (or early onset) parkinsonism: ubiquitin E3 ligase (parkin), PINK-1 kinase, and oncogene DJ-1 (Calahorro and Ruiz-rubio 2011) (Cookson and Bandmann, 2010). These loss of function mutations are all linked to defective mitochondrial function. Some of these genetic risk factors interact with LRRK2 and each other. In the past, there was not much convincing evidence that these genetically identified PD risk factors shared any direct functional connection on a molecular level (Yue and Lachenmeyer, 2011). However, more recently, evidence of some of these proteins acting in similar pathways has been found. There are two leading disease-altered pathways- protein translation and vesicle trafficking, which will be discussed below. A brief summary of all known disrupted pathways in LRRK2 PD is shown in Table 2, and their correlating protein components are summarized in Table 3.