Research Project 1

(1)

Internship, master Brain and Cognitive Sciences (UvA)

Date 10-7-2019

Student name Sebastiaan van Bruchem

Student number 10213929

Email srvanbruchem@gmail.com

First assessor Dr. Wilma D.J. van de Berg Second assessor Prof. Dr. Paul Lucassen Daily supervisor Hanneke Geut, M.D.

Internship institution Vrije Universiteit medisch centrum

Department Anatomy and Neurosciences

Section Clinical Neuroanatomy and Biobanking

Address De Boelelaan 1108

1081 HZ Amsterdam

Autophagy-lysosome pathway dysfunction in

GBA-related Parkinson’s disease and dementia

with lewy bodies

(2)

(3)

2

Abstract

Dysfunction of the autophagy-lysosome pathway (ALP) has been linked to Parkinson’s disease (PD) and dementia with Lewy bodies (DLB). In addition the GBA gene, coding for the lysosomal enzyme glucocerebrosidase (GCase), has been shown to be the most important risk factor in both PD and DLB. Multiple links have been found between GBA mutations, ALP dysfunction and lysosomal pH. This research aims to further elucidate this link.

Postmortem substantia nigra tissue from 15 PD (6 GBA mutation carriers, GBA+), 15 DLB (4 GBA+), and 15 non-demented controls (NDC) (2 GBA+) was included in this study. A proteomics study was performed to identify differentially expressed proteins related to the ALP between GBA mutation carriers and GBA wildtype. Immunohistochemical experiments were performed with LAMP-2 as a lysosomal marker and LC3B as an autophagosome marker. Staining area of LAMP-2 and LC3B was used as an indicator of ALP function and lysosomal pH. A cathepsin-D maturation assay was performed to give additional information about lysosomal pH in the subjects.

A total of 36 ALP related differentially expressed proteins between GBA+ and GBA- groups were found. Staining area of LAMP-2 and LC3B has not been quantified; quantitative and semi-quantitative analyses have been considered. Cathepsin-D maturation was not significantly different between NDC GBA-, PD/DLB GBA-, NDC GBA+, and PD/DLB GBA+ as determined by ANOVA analysis (p =0.917). There was also no difference in cathepsin-D maturation between GBA- and GBA+ (p = 0.918) and between NDC and PD/DLB (p = 0.492).

In conclusion, multiple ALP related proteins were found to be differentially expressed in GBA mutation carriers, possibly hinting at specific ALP dysfunction in this group. Immunohistochemical and cathepsin-D maturation experiments were inconclusive. Future experiments have been proposed: quantifying LC3B staining area as an indicator of ALP function and lysosomal pH, determining cathepsin-D maturation in the frontal cortex, and measuring lysosomal pH in cell models containing a GBA mutation.

(6)

5

Introduction

Parkinson’s disease and dementia with Lewy bodies

Parkinson’s disease (PD) is a neurodegenerative disease where patients show symptoms related to motor dysfunction and in later stages cognitive defects. Dementia with Lewy bodies (DLB) patients mostly show cognitive defects like dementia, and parkinsonism in later stages of the disease. The pathological hallmark for both of these diseases is the presence of Lewy bodies (Davie 2008), neuronal inclusion bodies found in neurons. It has been shown a protein called α-synuclein is a core component of these Lewy bodies (Spillantini et al. 1997). Mutations in the gene encoding for α-synuclein have been linked to familial type PD (Polymeropoulos et al. 1997, Singleton et al. 2003). What causes PD and DLB is not known, but a strong link has been found between dysfunction of the autophagy-lysosome pathway (ALP) and these diseases (Levine & Kroemer 2008). The link between ALP and PD and DLB is further substantiated by the fact that mutations of the GBA gene, which codes for the lysosomal enzyme glucocerebrosidase (GCase), have been shown to be the most important risk factor in PD (Westbroek et al. 2011, Gegg et al. 2012) and DLB (Tsuang et al. 2012). This research focuses on the link between ALP dysfunction, GCase and PD and DLB.

Autophagy system

Autophagy is the process whereby the cell degrades cellular components like organelles or proteins. Autophagy is subdivided in macro autophagy, micro autophagy and Chaperone Mediated Autophagy (CMA) (Glick et al. 2010), see figure 2.

Figure 1, a schematic representation of the different types of autophagy (Fujiwara et al. 2016)

Macro autophagy starts with the formation of a phagophore; a bilipid vesicle formed by membranes originating from the ER, Golgi, and/or lysosome. The phagophore engulfs cytoplasmic cellular material that is to be degraded and the resultant vesicle is called an autophagosome. The autophagosome fuses with the lysosome where enzymes, activated by low lysosomal pH, degrade the delivered material. With micro autophagy, the lysosome itself engulfs cytoplasmic material to be

(7)

6 degraded. Lastly, CMA makes use of chaperone proteins, like hsc70, to transport proteins across the membrane and bind LAMP-2A (Cuervo & Dice 1996) which causes the uptake of the protein into the lysosome. After uptake the protein is degraded inside the lysosome.

Autophagy function is often quantified by looking at autophagic flux, which is defined as autophagic degradation activity (Loos et al. 2014). Autophagic flux can be measured by assessing the amount of autophagosomes and lysosomes present in the cell (Klionsky et al. 2012). An impairment of fusion between autophagosomes and lysosomes is characterized by an increase of autophagosomes and an enlargement of lysosomes, which is indicative of ALP dysfunction.

Lysosomal trafficking and function

Lysosomal proteins are synthesized in the endoplasmic reticulum and transported to the cis-Golgi where they are post-translationally modified (see figure 2-1 on the next page). Most lysosomal enzymes will be tagged with a mannose-6-phosphate (M6P) residue in the Golgi network by GlcNAc-1-phosphotransferase (Gnptg). The M6P tagged proteins are recognized by the M6P receptor (M6PR) located on the membrane of clathrin coated vesicles bound to the Golgi. Next, the M6P tagged proteins are transported to the late endosome where they dissociate from the M6PR under low pH conditions (Nishi & Forgac, 2002) (figure 2-2). Some lysosomal proteins undergo M6P-independent transport. For example, GCase is transported to the late endosome by LIMP2 (Jović et al. 2012), cathepsin-D by CIMPR (Vagnozzi et al. 2018), while prosaposin is transported by both M6P dependent transport and by sortilin.

The pH of the late endosome is lowered to enable fusion with a lysosome. Lysosomal pH is maintained by ATPase proton pumps (figure 2-x). The endosome/lysosome hybrid will subsequently reform into a lysosome. Under low pH conditions, some lysosomal enzymes will mature and become enzymatically active. For example, cathepsin-D will mature from a 52kDa protein via an intermediate 44kDa protein into its mature double chain form consisting of a ~32kDa and a ~14kDa chain (Erikson et al. 1981). Cathepsin-D has been shown to be the main degrading enzyme of α-synuclein inside the lysosome (Sevlever et al. 2008). Besides degrading α-synuclein, cathepsin-D is also responsible for the conversion of prosaposin into saposins (figure 2-3). These saposins are activators of enzymes that are part of the sphingolipid pathway. For example, saposin A activates galactocerebrosidase (GALC) (Hill et al. 2018) while saposin C activates GCase (Alattia et al. 2007). After delivering their cargo at the lysosome, certain receptors are recycled back to the Golgi by retromer transport.

The vesicles are coated with a retromer complex (VPS25, VPS29 and VPS35), which is recognized by multiple proteins on an actin filament on the Golgi. The retromer complex, specifically VPS35, has been shown to be linked to PD (Zimprich et al. 2011). VPS35 knockdown in a drosophilia model impaired retromer trafficking, impaired CIMPR recycling, subsequent improper delivery of cathepsin-D to the lysosome and eventually impaired lysosomal function (Miura et al. 2014). The vesicle binds the GARP complex on the Golgi and fuses with the Golgi membrane with the aid of SNARE proteins.

(8)

7

Figure 2, schematic representation of ER-Golgi-Endo/lysosome transport (1 and 2), prosaposin cleavage (3), sphingolipid metabolism (4), retromer transport (5) and lysosomal acidification (x). Multiple links have been found between sphingolipid metabolism and lysosomal function, for example ceramide activates cathepsin-D and sphingolipids are necessary for V-ATPase V1 subunit formation.

(9)

8 Lysosomal pH

Lysosomal pH is maintained by the (H+_{)-ATPase proton pumps. The vacuolar (V-type) ATPase proton}

pump resides specifically on cellular vacuoles, the P-type ATPase resides on all cellular membranes, including cellular vacuoles (Bublitz et al. 2011). V-ATPase consists of two domains, the V0 and V1

domain (Cipriano et al. 2008). The V0 domain is a membrane bound domain consisting of 6 protein

subunits and is responsible for proton translocation. The V1 domain consists of eight protein subunits

and is responsible for ATP hydrolysis. Both domains reversibly couple with each other to form a functional V-ATPase proton pump, and this reversible coupling is a mechanism by which V-ATPase activity is regulated.

The maintenance of low lysosomal pH is of great importance for lysosomal function. First and foremost, lysosomal enzyme activity is highly pH- dependent. For example, cathepsin-D matures in the lysosome and it has been shown acidic lysosomal pH is necessary for this maturation process (Kokkonen et al. 2004). ATP13A2 deficiency, which causes a raise in lysosomal pH, causes a decrease in cathepsin-D activity (Matsui et al. 2013). Second, acidic pH is necessary for the hydrophobicity of saposins (Vaccaro et al. 1995) and it has been shown that the binding of saposin A to galactocerebrosidase (GALC, a lysosomal enzyme) is pH dependent (Hill et al. 2018). Also, high lysosomal pH interferes with M6PR recycling to the Golgi and lysosomal protein trafficking (Van Weert et al. 1995). Lastly, fusion of autophagosomes with lysosomes is strongly pH dependent and raising lysosomal pH blocks fusion which results in accumulation of autophagosomes (Kawai et al. 2007, Lu et al. 2013).

Sphingolipid metabolism

Multiple links have been found between sphingolipid metabolism and lysosomal acidification. For example, certain sphingolipids are required for the formation of a functional V-ATPase V1 subunit (Chung et al. 2003). Other enzymes involved in sphingolipid metabolism like Orm1p and Orm2p are required for V-ATPase function in yeast (Finnigan et al. 2011).

An important task of the lysosome is sphingolipid metabolism. Sphingolipids are a class of lipids that reside in the cell membrane and have broad functions within the cell including cell signaling and mediation of cell signaling, as well as supporting the structure of the cellular membrane (Merrill 2008). Sphingolipids reside in the cellular membrane which prevents enzymes from binding to these sphingolipids. Therefore, enzymes need activators, saposins, in order to metabolize these sphingolipids. Two models have been proposed for this activation: a liftase model whereby saposins place themselves inside the lysosomal membrane enabling the enzyme to bind to a sphingolipid, and a solubility model whereby saposins capture a sphingolipid from the lysosomal membrane and presents it to the enzyme (see figure 3-4).

Enzymes that take part in sphingolipid metabolism include GCase and GALC (Hill et al. 2018). GCase catalyzes the breakdown of glucosylceramide and glucosylsphingosine, two sphingolipids, to ceramide and sphingosine respectively. Homozygous mutations of GCase can result in Gaucher’s disease (Hruska et al. 2008). In Gaucher’s disease patients accumulation of GCase substrate results in severe physical and neurological defects. GALC catalyzes the breakdown of galactosylceramide into ceramide and galactosylsphingosine (psychosine) into sphingosine. Mutations of the GALC gene can cause the lysosomal storage disorder Krabbe’s disease which causes a buildup of unmetabolized lipids and psychosine (Debs et al. 2013).

(10)

9

Autophagy-lysosome pathway disorder in PD and DLB

A strong link has been found between ALP dysfunction and PD (Pan et al. 2008; Xilouri & Stefanis 2011). Besides synuclein being degraded by the proteasome system (Bennet et al. 1999), α-synuclein is also degraded by chaperone mediated autophagy (CMA) (Cuervo et al. 2004). Two mutant forms of α-synuclein, A30P and A53T, which cause familiar forms of PD (Nussbaum & Plymeropoulos, 1997), have been shown to have a high affinity for the CMA related lysosomal receptor LAMP-2A and have the ability to block this receptor. This causes a subsequent block in CMA and buildup of α-synuclein. The link between ALP dysfunction and PD is substantiated by the fact that a large part of PD patients have mutations in one or more genes related to lysosomal storage diseases (Robak et al. 2017).

GALC is also a genetic risk factor for sporadic PD (Chang et al. 2017; Marshall et al. 2018). Mutations of GALC cause misfolding and subsequent ERAD of the protein (Spratley & Deane 2016). Lower GALC activity causes accumulation of its substrates. Psychosine, a substrate of GALC, has been shown to accelerate the fibrillization of α-synuclein in brains of Twitcher mice, a mouse model of Krabbe’s disease, (Smith et al. 2014) which might be caused by direct binding of psychosine to α-synuclein altering its conformation (Santos & Bongarzone, unpublished results, referred to by Smith et al. 2014). Immunohistochemical analysis also showed that α-synuclein colocalized with thioflavin-S, a component of Krabbe’s disease inclusion bodies, in the brains of Twitcher mice. This means that Krabbe’s disease might be categorized as a synucleinopathic disease.

Disturbed lysosomal pH in PD and DLB

Two ATPase related genes have been identified as PD risk genes: ATP13A2 (PARK9) (Ramirez et al. 2006) and ATP6AP2 (Korvatska et al. 2013). ATP13A2 codes for a P-ATPase subunit and besides being a genetic risk factor has also been found to be reduced in Lewy body dementia GBA mutation carriers (Kurzawa-Akanbi et al. 2012). ATP6AP2 has been shown to be necessary for V-ATPase assembly (Kinouchi et al. 2010), providing a link between lysosomal pH disturbances and PD. The link between lysosomal pH and PD is further substantiated by the fact that multiple environmental PD risk factors (e.g. methamphetamine, rotenone, MPTP) raise lysosomal pH (Colacurcio & Nixon 2016). Cells which have a mutant form ATP13A2 show an elevated lysosomal pH and a disturbed cathepsin-D maturation process (Dehay et al. 2012). In this study, ATP13A2 mutant cells also show accumulation of autophagic vacuoles (autophagosomes) and enlarged lysosomes. Since cathepsin-D degrades α-synuclein, the disturbed cathepsin-D maturation seen with ATP13A2 deficiency might provide a causal link between ATP13A2 and PD. Indeed it has been shown ATP13A2 deficiency can cause α-synuclein accumulation in mouse brain (Schultheis et al. 2013).

(11)

10

Glucocerebrosidase

As mentioned before, the GBA gene, coding for GCase, is the most important genetic risk factor in PD and DLB. GCase consists of three domains (see figure 3) (Dvir et al. 2004). Domain I (residues 1-27 and 383-414) is targeted for glycosylation which is necessary for GCase function in vivo (Berg-Fussman et al. 1993). Domain II (residues 30-75 and 431-497) resembles an immunoglobulin fold and

may therefore be involved in protein-protein and protein-protein-ligand interactions (Williams & Barclay 1988). Domain 3 (residues 76-381 and 416-430) comprises the catalytic domain of GCase. Mutations in domains II and III have been found in PD patients: the most common GCase mutations found in PD patients are N370S and L444P (Aharon-Peretz et al. 2004; Clark et al. 2005; De Marco et al. 2008; Toft et al. 2006), and E326K (Pankratz et al. 2012). Mutations that are less prevalent in GBA-related PD include K198T, R329C, T369M (Lwin et al. 2004), R262H, K303K (Nichols et al. 2009) and F231I and R353W (Sun et al. 2010). GCase mutations can result in multiple kinds of dysfunction of the enzyme. For example, the N370S mutation makes GCase especially susceptible to changes in pH causing the enzyme to lose efficiency (Van Weely et al. 1993). Other mutations like the L444P mutation can cause misfolding of GCase and subsequent endoplasmic-reticulum associated protein degradation (ERAD) (Tan et al. 2014; Ron & Horowitz 2005; Sidransky et al. 2012). Impaired ER-Golgi-Lysosomal trafficking of GCase has also been noted as a result of GCase mutations (Schmitz et al. 2005).

Two models have been proposed to link GBA mutations to PD and DLB: a loss of function model and a gain of function model.

Loss of function

In the loss of function model, mutations of GCase resulting in a decrease of catalytic activity increase the susceptibility to PD and DLB. Decreased activity of GCase causes substrate (glucosylceramide and glucosylsphingosine) accumulation and a decrease in the product of GCase catalytic activity: ceramide. Eblan and colleagues (2005) found that α-synuclein possesses a high affinity for glucosylceramide containing glycosphingolipids. Because of this, it has been hypothesized that an increase in GCase substrate results in sequestration and subsequent accumulation of α-synuclein. It has been shown that GCase deficiency promotes α-synuclein aggregate propagation in SH-SY5Y cells (Bae et al. 2014). Also, disturbed lipid homeostasis by GCase deficiency might also cause α-synuclein tetramer instability (Kim et al. 2018). This instability can cause oligomerization of α-synuclein. Low GCase activity has indeed been linked to the formation of toxic α-synuclein oligomers and GCase depletion has been shown to decrease lysosomal proteolysis (Mazzulli et al. 2011). In contrast, GCase therapy in mice and rats overexpressing α-synuclein prevents α-synucleinopathy (Rocha et al. 2015).

Figure 3, schematic representation of glucocerebrosidase (Dvir et al. 2004)

(12)

11 Elevated α-synuclein inhibits intracellular trafficking (Mazzulli et al. 2016) and lysosomal function of GCase, thus creating a positive feedback loop between GCase and α-synuclein in the loss of function model. As a result of impaired GCase activity, levels of its product ceramide will also be lowered. Interestingly, ceramide has been shown to enhance cathepsin-D maturation (Heinrich et al. 2002) and as mentioned before cathepsin-D is the main degradative enzyme of α-synuclein.

Gain of function

The gain of function model postulates improperly folded GCase undergoes ERAD, thereby increasing the burden of misfolded proteins that need to be degraded. Additionally, mutant GCase might affect lysosomal synucleinase activity. In support of this model, Cullen and colleagues (2011) found that multiple GCase mutants (N370S, L444P, D409H, D409V, E235A and E340A) that showed no decrease in enzymatic activity significantly raised α-synuclein levels compared to controls in neural MES23.5 and PC12 cells.

Relation to lysosomal pH

A relation between mutated GCase and lysosomal pH has been established in earlier research. For example Bourdenx and colleagues (2016) found a raised lysosomal pH in fibroblasts from PD patients with a N370S or G377S point mutation. This resulted in an impairment of cathepsin-D maturation in these cells. The addition of acidic nanoparticles that lowered lysosomal pH restored cathepsin-D maturation, revealing a link between GCase, lysosomal pH and cathepsin-D maturation. Inhibition of GCase by conduritol B-expoxide causing a subsequent increase of glucosylceramide, as well as direct addition of glucosylceramide, has been noted to raise lysosomal pH in human lymphoblasts (Sillence 2013). In some mutants this could possibly cause a positive feedback loop. For example, the catalytic activity of the N370S mutant is highly sensitive to changes in lysosomal pH (Van Weely et al. 1993).

Research aim

The GBA gene has been shown to be the most important risk factor in both PD and DLB and it is therefore important to elucidate the biochemical link between mutated GCase and the development of PD and DLB. Previous research provides a strong link between ALP dysfunction and PD/DLB. This research aims to further elucidate the link of GBA mutations with ALP dysfunction in the context of PD and DLB. First of all, a proteomics study has been performed to identify ALP proteins that are downregulated in GBA mutation carriers. This in itself can provide insight into ALP function in GBA mutation carriers, for example if important lysosomal enzymes are downregulated in this group. ALP function is further quantified by looking at autophagic flux, specifically fusion between autophagosomes and lysosomes. Immunohistochemical experiments will be performed using two markers: LC3B, which is a specific autophagosome marker, and LAMP-2, which is a specific lysosome marker. Impaired ALP will result in an accumulation of autophagosomes and an enlargement of lysosomes. This can be detected using immunohistochemical techniques as an increased staining area of LC3B and LAMP-2 in the cell. Lysosomal pH is also of great importance in lysosomal function and certain PD/DLB risk genes like ATP13A2 have been shown to raise lysosomal pH. Since certain GBA mutants (N370S and G377S) have been shown to raise lysosomal pH in human PD fibroblasts, this study will also explore a possible link between GBA related PD and DLB and elevated lysosomal pH. As mentioned before, fusion between autophagosomes and lysosomes is dependent on lysosomal pH and this method can therefore also be used as an indirect measure of lysosomal pH. Also, since cathepsin-D maturation is strongly pH dependent, a cathepsin-D maturation assay will be

(13)

12 used to give additional information about lysosomal pH. In summary, two research questions have been formulated for this research:

• Do GBA-mutation carriers show lysosomal dysfunction determined by downregulation of ALP proteins and impaired fusion between autophagosomes and lysosomes?

• Do GBA-mutation carriers show impaired lysosomal acidification determined by impaired cathepsin-D maturation and impaired fusion between autophagosomes and lysosomes?

(14)

13

Methods

Study group

Table 1 shows subject demographics of the cases included in this study. Postmortem brain tissue was obtained from 15 PD, 15 DLB and 15 non-demented controls. GBA status was determined for this study group in an earlier study (Moors et al. 2019); a GBA variant was detected in 4 DLB patients, 6 PD patients and 2 NDC. One case, TM25, contained a GBA splice variant Fresh frozen and paraffin embedded substantia nigra (SN) tissue from each patient was obtained from the Netherlands Brain Bank. The fresh frozen SN tissue was lysed and used for LC-MS/MS analysis, Western Blot and ELISA, the paraffin embedded SN was cut and used for immunohistochemical analysis. For the Western Blot and ELISA pilots, fresh frozen putamen tissue was used.

Table 1, demographics of subjects

Patient Diagnosis GBA variant Age Sex Patient Diagnosis GBA variant Age Sex TM01 DLB 78 F TM23 PD 73 M TM02 NDC 83 F TM24 NDC 84 F TM03 NDC 83 F TM25 PD c.762-18T>A 80 F TM04 NDC 76 F TM26 NDC 76 M TM05 DLB 70 M TM27 NDC E326K 64 F TM06 NDC 77 F TM28 DLB E326K 80 F TM07 DLB 76 F TM29 DLB 81 M TM08 NDC 79 M TM30 DLB E326K 69 M TM09 PD 81 F TM31 PD 81 M TM10 PD E326K, D140H, 83 F TM32 DLB 79 M T369M (triple) TM33 PD 80 M TM11 DLB T369M 81 F TM34 PD 81 F TM12 NDC 79 F TM35 DLB 83 M TM13 DLB 75 M TM36 PD P319L 84 M TM14 DLB 74 M TM37 PD 76 M TM15 DLB E326K 72 M TM38 DLB 78 M TM16 NDC 78 F TM39 NDC 79 M TM17 NDC 70 F TM41 NDC 83 M TM18 NDC 70 F TM42 PD 78 M TM19 DLB 72 F TM43 PD L444P 65 M TM20 PD E326K 76 M TM44 PD E326K 69 M TM21 DLB 83 M TM45 NDC E326K 73 F TM22 PD 72 M TM46 PD 74 M

(15)

14

Chemicals used

The following chemicals were used for the experiments:

Table 2, chemicals used in experiments

Type Description/target Supplier Article number

Primary antibody LC3 MBL PM036

Primary antibody LC3B Novus Biologicals NB600-1384

Primary antibody LAMP-2 Santa Cruz SC-8100

Primary antibody Cathepsin-D Abcam Ab75852

Primary antibody α-synuclein, directly labeled with AlexaFluor 488

Roche 11A5

Secondary antibody Alexa Fluor 546 Invitrogen A10040

Secondary antibody Alexa Fluor 647 Invitrogen A21447

Secondary antibody STAR 580 Aberrior 2-0002-005-1

Secondary antibody STAR 635P Aberrior 2-0012-007-2

Secondary antibody IRDye 800cw Li-cor 826-32211

ELISA kit GALC MyBiosource MBS760168

ELISA kit ATP6V0C MyBiosource MBS9329169

In addition, the following protease inhibitors were used: aprotinin, leupeptin, PMSF, pepstatin.

Lysing tissue

Brain tissue was pulverized and aliquoted in 1.5 ml eppendorf cups. GCase buffer (50 mM citric acid, 50 mM KPi, 110 mM KCl, 10 mM NaCl, 1 mM MgCl2 pH 6) containing 0.1% Triton X-100 and protease inhibitors were added to the tissue, 1 ml per 250 mg tissue. One ball per cup was added and tissue was lysed using a Qiagen TisseuLyser LT for 4 min. The mixture was incubated for 30 min on ice and subsequently centrifuged for 10 min, 4°C, 21255 rcf. The supernatant was separated from the pellet and protein concentration was measured by using a BCA-assay. The supernatant was captured as an insoluble fraction.

Proteomics study

Triton X-100 soluble and insoluble brain tissue samples were precipitated with cold acetone and washed twice. Pellets were dissolved in Biognosys Denature buffer and reduced/alkylated by addition of reduction/alkylation solution for 1 h at 37°C. Proteins were trypsin digested overnight at 37°C, a 1:50 ratio of trypsin-protein was used. Proteins were fractionated into 6 fractions by high pH reversed-phase chromatography using a Dionex Ultimate 3000 RS pump on an Acquity UPLC CSH C18 1.7 µm, 2.1x150 mm column (from 1-40% acetonitrile in 20 min). A tissue specific spectral library was generated shotgun LC-MS/MS, mass spectrometric data was analyzed using the SpectroMine 1.0 search engine (Biognosys). A human UniProt database (retrieved on 2018-07-01) was used for the search engine. Postmortem samples were analyzed by ID+ HRMTM_{LC-MS/MS using a Thermo}

Scientific Easy nLC 1200 nano-liquid chromatography system on a Dr. Maisch Reprosil Pur C18 1.9 µm particle size, 120 Å pore size, 75 µm inner diameter, 50 cm length column. The chromatography system was connected to a Thermo Scientifc QExactive HF mass spectrometer (nano-electrospray source).

(16)

15

Solubilizing insoluble fraction

Three protocols to solubilize the insoluble fraction were compared. The following protocol is based on a protocol developed by Geumann et al. (2010): to the pellet captured after tissue lysis 1 µl/mg of solubilizing buffer was added (10 mM sodium phosphate pH 7.4, 130 mM NaCl, 5 mM EDTA, 1.2% SDS, protease inhibitors). The mixture was vortexed and subsequently incubated at RT for 15 minutes. After incubation, 5 volumes of ice cold quenching buffer (PBS, 1.2% Triton X-100) were added and the mixture was centrifuged for 40 min, 4°C, 21255 rcf. The resultant supernatant contained the previously insoluble proteins. The second protocol followed the same steps as the previous protocol, with the addition of a sonication step after addition of SDS-buffer.

A third protocol utilized a UTC buffer for solubilization. The pellet was washed with GCase buffer containing 0.1% Triton X-100 and protease inhibitors and centrifuged for 10 min, 4°C, 21255 rcf. The pellet was resuspended in UTC buffer (50 mM Tris pH 7.5, 7 M urea, 2 M thiourea and 4% CHAPS), 1 ul/mg pellet. The mixture was sonicated with a Branson Sonifier 250, 3 times 10 pulses. The mixture was centrifuged for 10 min, 4°C, 21255 rcf and the supernatant was captured.

For the ELISA pilots, some samples were further concentrated ~10x using an Amicon© Ultra 0.5 mL Centrifugal Filter (Regenerated Cellulose 3000 NMWL). Filters were centrifuged for 10 min, 4°C, 21255 rcf and the residual liquid containing the protein of interest was captured.

Immunohistochemistry

A total of 16 cases were used for immunohistochemical analysis. Slides were cut at 10µm and stained; antibodies used for this staining are shown in table 3. Paraffin was removed from the brain slides by washing in xylene substitute and ethanol and subsequently heated in citrate buffer (pH 6) at 100°C for 30 minutes. Brain slides were blocked for 1 hour (2% Normal Goat Serum, 0.1% Triton X-100 in TBS). Brain slides were incubated with LAMP-2 (1:X-100) primary antibody and either LC3B (1:100) in blocking solution overnight at 4°C. Slides were washed with TBS (three times) and incubated with secondary antibodies (all diluted 1:400) for 2 hours at RT. Slides were washed three times in TBS and incubated with α-synuclein (1:100) primary antibody and DAPI (0.6 mg/mL, diluted 1:1000) for 2 hours at RT.

Table 3, antibodies used for immunohistochemical experiment

Antibody Target Host Label

Primary α-synuclein (pS129) Goat A488

Primary LAMP-2 Mouse N/A

Secondary Mouse (anti LAMP-2) Goat STAR 580

Primary LC3B Rabbit N/A

Secondary Rabbit (anti LC3B) Goat STAR 635p

Staining was verified using a Leica TCS SP8 confocal microscope. For quantification of staining area, slides were scanned using a PerkinElmer Vectra® Polaris™ slide scanner.

Western Blot

Protein samples were diluted to a concentration of 3 mg/ml total protein. To 5 volumes of protein sample, 1 volume of 6x sample buffer (25 mM Tris pH 6.8, 10% SDS, 40 mM DTT, 30% glycerol, 0.012% bromophenol blue) was added for an end concentration of 2.5 mg/ml total protein and

(17)

16 incubated at 95°C for 5 min. Next, the samples were loaded on an 18% or 10% SDS-Page gel, 10 ul per sample. After SDS-Page, the proteins were transferred onto a nitrocellulose or PVDF membrane by Western Blotting (semi-dry, 90 min at 30 V, 250 mA). The membrane was washed twice with TBS and blocked in blocking buffer (50% Odyssey© Blocking Buffer, 50% TBS) for 30 min at RT. Next, the membrane was incubated in blocking buffer containing 0.05% Tween and Cathepsin-D primary antibody (1:2000 or 1:5000) at 4°C overnight or 1 hour at RT. The membrane was subsequently washed three times in TBST and incubated in blocking buffer containing the secondary antibody (IRDye 800CW 1:5000) for 1 hour at RT. After incubation the membrane was washed three times with TBST and once with TBS. The membrane was analyzed with a Li-cor Odyssey© AS imaging system, images were analyzed using Li-Cor Image Studio 4 software.

ELISA

The protocol provided by the supplier (MyBiosource) was followed in performing the ELISA’s. Samples were prepared using the supplied dilution buffer. In short, for the ATP6V0C ELISA samples and standards were added to the wells and incubated with HRP-Conjugate Reagent, either for 1h at 37°C (following protocol) or overnight at 4°C (deviation from protocol). Wells were washed, Chromogen solution A and B were added and incubated for 15min at 37°C. Stop solution was added and optical density at 450nm was measured using a SpectraMax Plus 384 spectrophotometer.

Bioinformatic analysis

Proteomics data

Proteomics mass spectrometric data was analyzed using Spectronaut™ X and comparing to the generated tissue-specific spectral library. The false discovery rate was set at 1%. HRM measurements were normalized using local regression normalization. Peptide and protein quantities were calculated using a topN approach; protein expression for each case was quantified as sum of top 1-3 peptide intensities. Proteins were annotated with GO data (Gene Ontology data; protein data regarding biological process, molecular function and the cellular component the protein is present in) for all proteins.

Differentially expressed proteins

In order to identify differentially expressed proteins in the GBA mutation carrier group, a factorial ANCOVA was performed with GBA-status as factor and age and sex as covariant. Proteins were deemed significantly differentially expressed when p<0.05. A list of differentially expressed autophagy-related proteins was subsequently constructed by filtering proteins based on their cellular function (GO terms related to the autophagy pathway), and by comparing to two databases containing autophagy-related proteins (Homma et al. 2010; Wang et al. 2018).

Cathepsin-D maturation

Cathepsin-D maturation was analyzed using Graphpad Prism 8. Three group comparisons were done, with - indicating GBA WT and + indicating GBA mutation carrier. An ANOVA analysis was performed for the NDC- vs NDC + vs PD/DLB- vs PD/DLB + comparison. A student t-test was performed for the NDC vs PD/DLB and GBA- vs GBA+ comparison. Cathepsin-D maturation was considered significantly different between groups when p<0.05.

(18)

17

Results

Proteomics

In total, 6001 unique proteins have been quantified in the insoluble fractions and 4136 unique proteins have been quantified in the soluble fractions. In all comparisons combined, 730 proteins were found to be differentially expressed, 347 proteins were differentially expressed in the GBA+ group compared to the GBA- group. The 347 differentially expressed proteins were compared to both autophagy databases (see methods) and protein GO terms were screened for terms related to the autophagy pathway. These proteins encompass all proteins related to the autophagy process (e.g. trafficking, CMA receptors, lysosomal enzymes). This resulted in a total of 36 differentially expressed autophagy-related proteins, the complete list is shown in appendix 1. Functional protein associations were mapped and visualized using string-db.org, the results are shown in figure 4.

Figure 4, String analysis of differentially expressed proteins between GBA + and GBA – groups (string-db.org). Two clusters of proteins differentially expressed include the RAS superfamily and the heterotrimeric G-proteins. Also proteins involved in sphingolipid metabolism and lysosomal acidification were found to be differentially expressed.

Two superfamilies of proteins were identified: the RAS superfamily of proteins and the heterotrimeric G-proteins, most proteins of both families were significantly downregulated. The RAS superfamily is involved in cellular signal transduction and mainly regulates cell growth, differentiation and survival. The heterotrimeric G-proteins are involved in the transmembrane signaling system. Proteins AP1S2, AP3B1, AP3D1, and AP3M1 are part of the clathrin-associated adaptor protein complexes.

(19)

18 Three differentially expressed proteins that are present in the lysosomal pathway presented in the introduction section were selected for validation by ELISA: GCase, GALC, and ATP6V0C. All proteins were downregulated in the insoluble fraction in the GBA+ group, the p-value and fold change of these proteins are shown in table 4, scatterplots are shown in figure 5. The candidate proteins have been discussed in the introduction section; GCase and GALC are involved in sphingolipid metabolism and are both genetic risk factors in PD and DLB, ATP6V0C comprises the proton pore of the V-ATPase proton pump and is therefore directly involved in lysosomal acidification. Since all proteins were present in the insoluble fraction and thus it was necessary to follow the aforementioned protocols for solubilizing the insoluble pellet for use in the ELISA’s.

Table 4, significance and fold change of candidate proteins

Protein Fraction p value Fold Change ATP6V0C Insoluble 0.020 -1.366

GALC Insoluble 0.025 -1.299

GCase Insoluble 0.000 -1.230

Figure 5, scatterplots for candidate proteins: GCase, GALC, and ATP6V0C were significantly downregulated in the GBA+ group compared to GBA- group.

Solubilizing the insoluble fraction

The insoluble fraction was solubilized using three different methods. No BCA-assays were performed to determine protein concentration because of the high chance of interference by the chemicals used in solubilizing the pellet. For the first protocol using SDS/Triton without sonication, cathepsin-D presence was confirmed in the soluble fraction and the resolubilized insoluble fraction using Western Blot (see appendix 5).

Pilot ELISA

Two different buffers were tested to determine whether they were suitable in solubilizing the insoluble fraction for ELISA. The results of the performed ELISA pilots are shown in appendix 4. Firstly, a clear matrix effect was found for the UTC buffer: when spiking the UTC buffer with the internal standard a much lower signal was observed than expected. This indicates a strong denaturing effect of the UTC buffer, indicating this buffer is unsuitable for ELISA. No such matrix

ND C - GB A W T PD /DLB - G BA WT ND C - GB A v aria nt PD /DLB - G BA var iant 0 100000 200000 300000 400000 500000 GCase P ro te in i n te n s it ie s ( n o rm a li z e d ) ND C - GB A W T PD /DLB - G BA WT ND C - GB A v aria nt PD /DLB - G BA var iant 0 200000 400000 600000 800000 1000000 GALC P ro te in i n te n s it ie s ( n o rm a li z e d ) ND C - GB A W T PD /DLB - G BA WT ND C - GB A v aria nt PD /DLB - G BA var iant 0 200000 400000 600000 800000 ATP6V0C P ro te in i n te n s it ie s ( n o rm a li z e d ) p = 0.000 p = 0.025 p = 0.020

(20)

19 effect was found for the SDS/Triton X-100 buffer. In all re-solubilized fractions no signal was found except for a faint higher than blank signal in one sample (undiluted sample and incubated overnight at 4°C). In an attempt to amplify this signal, it was decided to concentrate the samples with a centrifugal filter. The sample was concentrated approximately 10 times. This did not, however, result in increased signal. In conclusion, all protocols used for resolubilizing the insoluble pellet are not suitable for ELISA’s. No ELISA was attempted to validate protein expression in the subjects of this research.

Immunohistochemical analysis with α-synuclein, LAMP-2 and LC3B markers

From this study’s group, 16 cases were used for immunohistochemical staining. Slides were stained for phosphorylated α-synuclein (pS129) which is a pathological form of α-synuclein present specifically in synucleinopathy lesions (Fujiwara et al. 2002) and can therefore be used to identify pathological cells in PD/DLB brain slides. Slides were also stained for LAMP-2, which is a lysosomal marker, and LC3B, which is an autophagosomal marker. Slides were first scanned using an SP8 confocal microscope. Nueromelanin positive cells (dopaminergic cells) of some cases were scanned as an example, the results are shown in figure 6 and figure 7 below.

All PD and DLB cases showed Lewy Bodies in the SN. The signal for LAMP-2 appeared ambiguous. In the TM14 (PD GBA-) and PD triple mutant cases there appeared to be some specific signal higher than background, while in case TM16 (NDC GBA-) there was no signal higher than background. LC3B however did show stronger selective cytoplasmic signal as indicated by the arrows in the figure. On visual inspection the cytoplasmic LC3B signal appeared to be stronger in the disease cases compared to controls. In the TM18 (NDC GBA-) cell there does appear to be some cytoplasmic LC3B signal, however this appears to be lower than the disease cases. There was strong overlap in signal between LAMP-2 and LC3B indicating possible colocalization. Interestingly, there also appears to be overlap in the signals of LAMP-2, LC3B, and α-synuclein.

Whole slides were scanned using a PerkinElmer Vectra® Polaris™ slide scanner, however no quantification has been done yet when this report was written. Both a quantitative and a semi-quantitative method have been considered for quantification but not executed due to lack of time.

(21)

20

Figure 6, examples of staining for LAMP-2, LC3B and α-synuclein using a SP8 confocal microscope. Arrows indicate specific LC3B signal in cell cytoplasm. DAPI and α-synuclein showed strong staining, LAMP-2 had some specific signal but with a relatively high background, and LC3B had clear specific cytoplasmic signal.

(22)

21

Figure 7, examples of staining for LAMP-2, LC3B and α-synuclein using a SP8 confocal microscope. Arrows indicate specific LC3B signal in cell cytoplasm. DAPI and α-synuclein showed strong staining, LAMP-2 had some specific signal but with a relatively high background, and LC3B had clear specific cytoplasmic signal.

(23)

22

Cathepsin-D maturation pilots

The results for the Western Blot pilots for determining cathepsin-D maturation are shown in appendix 6. Clear bands can be seen for the mature form of cathepsin-D (32 kDa), 52 kDa immature cathepsin-D consistently , while the 44 kDa immature was not always visible. Therefore, the percentage of immature 52 kDa cathepsin-D was used as a measure of cathepsin-D maturation. Another pilot using a dilution range of sample was performed to determine the optimal dilution for determining cathepsin-D maturation. The results of this are shown in appendix 6 figure 13, the results of the quantification are shown in appendix 7. No saturation effect was observed for the mature form (32 kDa) of cathepsin-D. In order to have the highest signal for immature (52 kDa) cathepsin-D, and since no saturation effect for mature cathepsin-D was observed, it was decided to load the samples undiluted (2.5 mg/ml total protein).

Cathepsin-D maturation

A cathepsin-D maturation assay is able to give an indication of lysosomal dysfunction, since cathepsin-D maturation can only be performed successfully when both ER-Golgi-Lysosomal transport and lysosomal acidification are working properly. Samples from the complete study group, except for TM24 and TM25 because of lack of available sample, were analyzed by Western Blot with cathepsin-D antibody as described in the method section. The results of the Western Blot are shown in figure 8.

Figure 8, Western Blot results for determining cathepsin-D maturation. 18 samples had visible bands for immature (52 kDa) cathepsin-D. For all samples without signal for immature (52 kDa) cathepsin-D, the percentage of immature (52 kDa) cathepsin-D was set at 0%. All samples showed clear bands for mature (32 kDa) cathepsin-D.

On visual inspection all cases had visible bands for mature (32 kDa) cathepsin-D, 18 cases showed visible bands for immature (52 kDa) cathepsin-D. TM34 showed a significantly lower signal for mature cathepsin-D. It was determined an error was made in diluting this sample, therefore TM34 was excluded from the analysis. Quantification of all bands, including non-visible bands, was done in Li-Cor Image Studio 4; the results of this quantification are shown in appendix 9. Quantified areas for

Cath-D (32 kDa) Cath-D (52 kDa) Cath-D (32 kDa) Cath-D (52 kDa) Cath-D (52 kDa) Cath-D (32 kDa)

(24)

23 immature cathepsin-D that had negative values were set to 0%. Areas selected for quantification are available in appendix 8. For each case, the percentage of immature (52 kDa) cathepsin-D relative to total cathepsin-D (52 kDa and 32 kDa) was determined. Scatterplots were made in Graphpad Prism 8, the results are shown in figure 9. Three different comparisons were made. An ANOVA was performed for the NDC- vs NDC+ vs Disease- vs Disease+ which indicated there was no significant difference between groups (p = 0.917). A t-test was performed for the NDC vs Disease comparison (p = 0.492) and GBA- vs GBA+ comparison (p = 0.918), both were non-significant. However, excluding the 0% values resulted in a non-significant trend of disturbed cathepsin-D maturation in the disease group compared to control (p = 0.072). All groups had a mean immature (52 kDa) cathepsin-D of around 1%. However, the disease group did show 4 clear outliers with a relatively high percentage of immature (52 kDa) cathepsin-D. In conclusion, no significant difference in cathepsin-D maturation was found in all group comparisons.

Figure 9, scatterplots showing cathepsin-D maturation (as percentage of immature cathepsin-D compared to total cathepsin-D). An ANOVA for comparison between NDC GBA WT, PD/DLB GBA WT, NDC GBA variant, and PD/DLB GBA variant showed no significant difference between groups. T-tests for the NDC vs PD/DLB groups and the GBA- vs GBA+ groups also showed no significant difference. There were however multiple outliers in the disease group that showed a higher percentage of immature (52 kDa) cathepsin-D present compared to the other samples, indicating disturbed cathepsin-D maturation.

ND C - GB A W T PD /DLB - G BA WT ND C - GB A v aria nt PD /DLB - G BA var iant 0 2 4 6 8 10 Cath-D maturation Group % i m m a tu re ( 5 2 k D a ) c a h t-D NDC PD/DLB 0 2 4 6 8 10 Diagnosis comparison Group % i m m a tu re ( 5 2 k D a ) c a h t-D GBA- GBA+ 0 2 4 6 8 10 GBA comparison Group % i m m a tu re ( 5 2 k D a ) c a h t-D p = 0.917 p = 0.492 p = 0.918

(25)

24

Discussion

The goal of this research was to elucidate the connection between GBA mutations and ALP dysfunction in the context of PD and DLB. A proteomics study was performed that identified 36 ALP related proteins that were differentially expressed in GBA mutation carriers compared to GBA wildtype. Immunohistochemical experiments have been performed with LC3B as an autophagosome marker and LAMP-2 as a lysosome marker, staining area of LC3B and LAMP-2 has not been quantified. Cathepsin-D maturation was not found to be significantly impaired in either GBA mutation carriers or PD/DLB cases compared to controls.

The proteomics study identified 36 ALP related differentially expressed proteins between GBA- and GBA+ groups. Multiple superfamilies of proteins were identified. An interesting subfamily were the subunits of the clathrin-associated adaptor protein complexes. These proteins are directly involved in ER-Golgi-lysosome trafficking of proteins. All these proteins were downregulated, suggesting a possible impairment of trafficking which could give rise to ALP dysfunction.

One resolubilization protocol was based on a protocol by Geumann and colleagues (2010). In this protocol, a 1.2% SDS buffer is used to solubilize the insoluble proteins. Because SDS attaches to the protein, thereby denaturing it, 5 volumes of Triton X-100 are added to quench the SDS from the protein. The resolubilization of the insoluble pellet to make it suitable for ELISA analysis was ineffective. The protocol utilizing the SDS buffer as described by Geumann and colleagues was effective in solubilizing protein as determined by Western Blot. The UTC buffer did show a strong matrix effect and was therefore deemed to be unsuitable for ELISA analysis. All resolubilized insoluble fractions showed no signal for ATP6V0C in the ELISA. The soluble fraction did have a signal in the ELISA indicating the failed ELISA’s were not because of a problem with the cases. The solubilization protocols for the insoluble fraction were probably unsuitable for the ELISA because of its denaturing effect on the proteins. Multiple methods can be attempted to refold the protein, for example dialysis or refolding in a chromatography column. However, the samples were to be used for protein quantification and the addition of refolding steps would introduce too many variables, drastically reducing the reliability of quantification.

The results of the immunohistochemical experiments show LC3B gives a clear staining of autophagosomes, the staining of lysosomes by LAMP-2 was less clear. Because lysosomal dysfunction by for example disturbed lysosomal acidification results in both accumulation of autophagosomes and enlargement of lysosomes, it is not necessary to use both markers (LC3B and LAMP-2) as an indicator of lysosomal function. Therefore, LC3B staining area can be used to give ample information about lysosomal function in human postmortem tissue. Also, earlier research has shown elevated LAMP-2 staining levels in GBA-related Lewy Body Disease (LBD) and healthy controls.

Whole slides were scanned using a PerkinElmer Vectra® Polaris™ slide scanner. The resultant files were too big to be opened in ImageJ for quantitative analysis. An option that has been explored was to open a small section of the whole image in ImageJ and perform quantification on that specific section. However, the quantification of staining area may be less reliable when arbitrary regions are selected for analysis. Another option would be a semi-quantitative analysis, whereby select cells in all slides are scanned using the SP8 STED microscope. Staining of LAMP-2 and LC3B would be scored based on relative staining on visual analysis. Due to lack of time, both options were not explored

(26)

25 further. The immunohistochemical results presented in this paper give no evidence of autophagosome accumulation in GBA mutation carriers.

However there were some interesting results from the immunohistochemical experiments. Firstly, LC3B and LAMP-2 were found to be strongly colocalized. Colocalization of LC3B with LAMP-2 is to be expected and has been observed in earlier research (e.g. Kilpatrick et al. 2015). Autophagosomes fuse with lysosomes to deliver cargo so it is to be expected LC3B and LAMP-2 colocalize during this fusion process. Also, LC3B and LAMP-2 has been shown to colocalize with Lewy Bodies in this research, colocalization of LC3B with Lewy Bodies in DLB patients has been shown in earlier research as well (Higashi et al. 2011).

Cathepsin-D maturation has been used in earlier research as an indicator of lysosomal function (e.g. Naslavsky et al. 2009; Dehay et al. 2012). The percentage of immature (52 kDa) cathepsin-D compared to total cathepsin-D (32 kDa and 52 kDa) serves as a measure of maturation, the higher this percentage the more disturbed the maturation process. Because the maturation is determined by the relative difference of proteins within one sample, it was not necessary to use loading controls. Also, loading controls for Western Blot like β-actin have been shown to be unreliable for quantification of low-abundance protein (Dittmer & Dittmer 2006). However, the cathepsin-D maturation assay is also indicative of other cellular pathology besides problems with lysosomal acidification. For example, because ceramide enhances maturation of cathepsin-D, a lower amount of ceramide by GCase dysfunction might result in disturbed cathepsin-D maturation. Also, impaired trafficking of cathepsin-D will also result in impaired cathepsin-D maturation.

Cathepsin-D expression was not found to be significantly lowered in GBA mutation carriers in this study group as determined by the proteomics experiment. However, expression is not indicative of maturation. Disturbed cathepsin-D function has been found in this study’s subjects in earlier research. Moors and colleagues (2019) found a significantly lower (p = 0.05) cathepsin-D activity in PD and DLB patients compared to controls in the frontal cortex. However, no such significant difference (p = 0.43) was found in the substantia nigra of these patients. It would be interesting to measure cathepsin-D maturation in the frontal cortex of these subjects. In this research, a lot of samples showed no visible bands for immature cathepsin-D, possibly because amount of immature cathepsin-D in these samples was below the detection limit. This resulted in a high number of samples with 0% immature (52 kDa) cathepsin-D. Without these 0% values there was a trend of impaired cathepsin-D maturation in the PD/DLB group. Also, the PD/DLB group did have four clear outliers that were indicative of disturbed cathepsin-D maturation. Previous research, for example Dehay et al. 2012, utilizing cathepsin-D maturation as an indicator for lysosomal function did so in cellular models. This method in determining cathepsin-D maturation might not be suited for human postmortem tissue since this tissue contains a homogenized mixture of cells, not only the cells showing pathology and impaired cathepsin-D maturation.

Strengths and limitations

The study group used in this study is unique in its size with regards to human postmortem research. Also, GBA status of each subject was known giving a unique opportunity to specifically research the influence of GBA mutations in the context of PD and DLB. The proteomics study also provides a huge valuable dataset in studying protein expression in GBA mutation carriers.

(27)

26 The difficulty in studying the effect of GBA mutations on cellular function stems from the fact that GBA mutations can result in a variety of issues with GCase like misfolding, impaired transport, or impaired function. For example the E326K mutation, which is highly prevalent in this study’s subjects (7 of 11 GBA+ cases have this mutation), has been shown to have a relatively minor impact on enzyme activity. In a study by Alcalay and colleagues (2005) they found that the E326K mutation resulted in GCase activity of 9.81 µmol/l/h compared to 11.93 µmol/l/h of non-mutated GCase. In contrast, the N370S mutation resulted in an activity of 6.42 µmol/l/h. Also, the prevalence of this mutation is not exceptionally high: 1.3% in GD patients compared to 0.9% in non GD cases, and 1.9% in PD patients compared to 1.37% in healthy controls (Horowitz et al. 2011). Also, the P319L mutation has not earlier been described as a risk factor for PD, and only the P319A mutation is known to cause Gaucher’s disease (Hruska et al. 2008). This mutation did result in observably disturbed cathepsin-D maturation: immature (52 kDa) cathepsin-D comprised 3.6% of total cathepsin-D compared to all group means of around 1%.

In this research, cathepsin-D maturation and accumulation of autophagosomes/lysosomes have been used as an indicator of lysosomal pH. As mentioned before, cathepsin-D maturation can be impaired due to other causes. Also, disturbed lipid homeostasis might prohibit fusion between autophagosomes and lysosomes, resulting in accumulation of autophagosomes. Substrate accumulation might result in enlargement of lysosomes.

Future research

Firstly, a quantification of LC3B staining area of the stained slides in this study should be performed to determine if there is significant ALP dysfunction in GBA mutation carriers. Since cathepsin-D activity was found to be lowered in the frontal cortex of this study’s subjects but not in the substantia nigra, it would be interesting to see whether cathepsin-D maturation is impaired in the frontal cortex of these subjects. Since ceramide is related to cathepsin-D maturation, a measurement of ceramide levels in GBA mutation carriers could further explore the link between mutated GCase and ALP dysfunction. Also, it would be interesting to see whether GBA mutations result in higher lysosomal pH in cellular models; lysosomal pH can be accurately measured in live cells.

Conclusion

This research postulated two research questions:

• Do GBA-mutation carriers show lysosomal dysfunction determined by downregulation of ALP proteins and impaired fusion between autophagosomes and lysosomes?

• Do GBA-mutation carriers show impaired lysosomal acidification determined by impaired cathepsin-D maturation and impaired fusion between autophagosomes and lysosomes? Multiple subunits of the clathrin-associated adaptor protein complexes were found to be downregulated. Proteins involved in sphingolipid metabolism and lysosomal acidification were also found to be downregulated. Downregulation of these proteins give an indication of possible ALP dysfunction. The cathepsin-D maturation assay indicated no significant difference in cathepsin-D maturation in GBA mutation carriers. Future experiments quantifying LC3B staining area could give additional information about ALP function. In conclusion, this research gives an additional indication of a link between GBA mutations and ALP dysfunction and can aid in identifying the exact role of GBA mutations in PD and DLB.

(28)

27

Acknowledgements

First of all I would like to thank Hanneke Geut for supervising me during this internship, giving continuous intellectual input with regards to this study’s design, and providing feedback in writing this report. I would also like to thank Wilma van de Berg for giving me the opportunity to follow this internship and for providing additional guidance during this internship.

Special thanks to Tim Moors, Cesc Bertran-Cobo and Joost Heuvelink for providing interesting discussions with regards to this study.

Lastly I would like to thank John Bol, John Breve, Kees Jongenelen, Angela Ingrassia, Evelien Timmermans-Huisman, Yvon Galis-de Graaf and others who helped greatly in successfully executing all experiments done in this study.

(29)

28

References

Aharon-Peretz, J., Rosenbaum, H., & Gershoni-Baruch, R. (2004). Mutations in the glucocerebrosidase gene and Parkinson's disease in Ashkenazi Jews. New England Journal of Medicine, 351(19), 1972-1977.

Alattia, J. R., Shaw, J. E., Yip, C. M., & Privé, G. G. (2007). Molecular imaging of membrane interfaces reveals mode of β-glucosidase activation by saposin C. Proceedings of the National Academy of Sciences, 104(44), 17394-17399.

Alcalay, R. N., Levy, O. A., Waters, C. H., Fahn, S., Ford, B., Kuo, S. H., ... & Rouleau, G. A. (2015). Glucocerebrosidase activity in Parkinson’s disease with and without GBA mutations. Brain, 138(9), 2648-2658.

Bae, E. J., Yang, N. Y., Song, M., Lee, C. S., Lee, J. S., Jung, B. C., ... & Lee, S. J. (2014). Glucocerebrosidase depletion enhances cell-to-cell transmission of α-synuclein. Nature communications, 5, 4755.

Bennett, M. C., Bishop, J. F., Leng, Y., Chock, P. B., Chase, T. N., & Mouradian, M. M. (1999). Degradation of α-synuclein by proteasome. Journal of Biological Chemistry, 274(48), 33855-33858.

Berg-Fussman, A., Grace, M. E., Ioannou, Y., & Grabowski, G. A. (1993). Human acid beta-glucosidase. N-glycosylation site occupancy and the effect of glycosylation on enzymatic activity. Journal of biological chemistry, 268(20), 14861-14866.

Bourdenx, M., Daniel, J., Genin, E., Soria, F. N., Blankhard-Desce, M., Bezard, E., & Dehay, B. (2016). Nanoparticles restore lysosomal acidification defects: Implications for Parkinson and other lysosomal-related diseases. Autophagy, 12(3), 472-483.

Bublitz, M., Morth, J. P., & Nissen, P. (2011). P-type ATPases at a glance. J Cell Sci, 124(15), 2515-2519 Chang, D., Nalls, M. A., Hallgrímsdóttir, I. B., Hunkapiller, J., van der Brug, M., Cai, F., ... & Hinds, D. (2017). A meta-analysis of genome-wide association studies identifies 17 new Parkinson's disease risk loci. Nature genetics, 49(10), 1511.

Chung, J. H., Lester, R. L., & Dickson, R. C. (2003). Sphingolipid requirement for generation of a functional v1 component of the vacuolar ATPase. Journal of Biological Chemistry, 278(31), 28872-28881.

Cipriano, D. J., Wang, Y., Bond, S., Hinton, A., Jefferies, K. C., Qi, J., & Forgac, M. (2008). Structure and regulation of the vacuolar ATPases. Biochimica et Biophysica Acta (BBA)-Bioenergetics, 1777(7-8), 599-604.

Clark, L. N., Nicolai, A., Afridi, S., Harris, J., Mejia‐Santana, H., Strug, L., ... & Ford, B. (2005). Pilot association study of the β‐glucocerebrosidase N370S allele and Parkinson's disease in subjects of Jewish ethnicity. Movement disorders: official journal of the Movement Disorder Society, 20(1), 100-103.

(30)

29 Colacurcio, D. J., & Nixon, R. A. (2016). Disorders of lysosomal acidification—The emerging role of

v-ATPase in aging and neurodegenerative disease. Ageing research reviews, 32, 75-88.

Cuervo, A. M., & Dice, J. F. (1996). A receptor for the selective uptake and degradation of proteins by lysosomes. Science, 273(5274), 501-503.

Cuervo, A. M., Stefanis, L., Fredenburg, R., Lansbury, P. T., & Sulzer, D. (2004). Impaired degradation of mutant α-synuclein by chaperone-mediated autophagy. Science, 305(5688), 1292-1295. Cullen, V., Sardi, S. P., Ng, J., Xu, Y. H., Sun, Y., Tomlinson, J. J., ... & Rochet, J. C. (2011). Acid β‐

glucosidase mutants linked to Gaucher disease, Parkinson disease, and Lewy body dementia alter α‐synuclein processing. Annals of neurology, 69(6), 940-953.

Davie, C. A. (2008). A review of Parkinson's disease. British medical bulletin, 86(1), 109-127.

De Marco, E. V., Annesi, G., Tarantino, P., Rocca, F. E., Provenzano, G., Civitelli, D., ... & Nicoletti, G. (2008). Glucocerebrosidase gene mutations are associated with Parkinson's disease in southern Italy. Movement disorders: official journal of the Movement Disorder Society, 23(3), 460-463.

Debs, R., Froissart, R., Aubourg, P., Papeix, C., Douillard, C., Degos, B., ... & Vanier, M. T. (2013). Krabbe disease in adults: phenotypic and genotypic update from a series of 11 cases and a review. Journal of inherited metabolic disease, 36(5), 859-868.

Dehay, B., Ramirez, A., Martinez-Vicente, M., Perier, C., Canron, M. H., Doudnikoff, E., ... & Bezard, E. (2012). Loss of P-type ATPase ATP13A2/PARK9 function induces general lysosomal deficiency and leads to Parkinson disease neurodegeneration. Proceedings of the National Academy of Sciences, 109(24), 9611-9616.

Dittmer, A., & Dittmer, J. (2006). β‐Actin is not a reliable loading control in Western blot analysis. Electrophoresis, 27(14), 2844-2845.

Dvir, H., Harel, M., McCarthy, A. A., Toker, L., Silman, I., Futerman, A. H., & Sussman, J. L. (2003). X‐ ray structure of human acid‐β‐glucosidase, the defective enzyme in Gaucher disease. EMBO reports, 4(7), 704-709.

Eblan, M. J., Walker, J. M., & Sidransky, E. (2005). The glucocerebrosidase gene and Parkinson's disease in Ashkenazi Jews. The New England journal of medicine, 352(7), 728-31.

Erikson, A. H., Conner, G. E., & Blobel, G. (1981). Biosynthesis of a lysosomal enzyme. Partial structure of two transient and functionally distinct NH2-terminal sequences in cathepsin DJ Biol. Chem, 25, 9608-9615.

Finnigan, G. C., Ryan, M., & Stevens, T. H. (2011). A genome-wide enhancer screen implicates sphingolipid composition in vacuolar ATPase function in Saccharomyces cerevisiae. Genetics, 187(3), 771-783.

Fujiwara, H., Hasegawa, M., Dohmae, N., Kawashima, A., Masliah, E., Goldberg, M. S., ... & Iwatsubo, T. (2002). α-Synuclein is phosphorylated in synucleinopathy lesions. Nature cell biology, 4(2), 160.