Screen for Huntington’s disease; The development and validation of a robust method for compound library screening in Huntington’s disease

(1)

4

Screen for Huntington’s disease

The development and validation of a robust method for compound library

screening in Huntington’s disease

J.G. Meulmeester Bachelor thesis

Bèta-gamma, major Biomedical Science track Neurobiology Supervisor Prof. E.A.J. Reits

Examinator dr. ir. H.A. Van den Burg University of Amsterdam, UvA 1 July 2018

(2)

Abstract

Huntington’s disease is a neurodegenerative disorder caused by a polyglutamine stretch in the Huntingtin gene which results in protein aggregation in neurons. The aim of this research is to set up a screening system where compound and siRNA libraries can be used to identify targets that will improve mutant Huntingtin degradation. Before this screen can be implemented a robust and reliable method should be developed and validated. The libraries will be tested on a striatal cell line expressing mutant Huntingtin with a polyglutamine expansion of 97 and a GFPQ16 reporter,

separated with an internal ribosome entry site (IRES). GFPQ16 is diffusely expressed but is recruited to mHttQ97 aggregates and is therefore a reporter for the otherwise non-fluorescent polyQ

aggregates. The aggregation of mHtt and diffuse GFP signal are measured with fluorescent microscopy. In addition to the screen setup first tests were run with the DUB6RK63 which showed improvement of the degradation of mutant Huntingtin through the ubiquitin proteasome system. This microscopy data shows reliable and reproducible output which is validated by biochemistry assays. To conclude large library screens can be implemented with this screening setup.

(3)

General introduction

Huntington’s disease is an autosomal dominant inherited neurodegenerative disorder caused by a polyglutamine (CAG) repeat in the Huntingtin gene that results in a polyglutamine (polyQ) expansion in the Huntingtin protein (Hannan, 2018). It is characterized by motor dysfunction, cognitive decline such as dementia and psychological dysfunction (Ortega, Díaz-Hernández, & Lucas, 2007). The diagnosis is most frequently certified by motor dysfunction while cognitive and psychiatric symptoms along with progressive weight loss and muscle-wasting are detectable prior to motor symptoms (Mo, Hannan, & Renoir, 2015; Saudou & Humbert, 2016). The disease leads to death 15-25 years after onset due to complications such as cardiac or respiratory failure.

The extended polyglutamine tract in genes is found to be responsible for at least eight neurodegenerative disorders such as spinocerebellar ataxia (SCA) and x-linked spinobulbar muscular atrophy (SBMA) (Hannan, 2018; Ortega, Díaz-Hernández, & Lucas, 2007; Sherzinger, et al.,1997). Neurodegeneration in Huntington is most severe in medium-sized spiny neurons in the striatum, which plays a role in initiating and controlling movements (Gusella & MacDonald, 1995; Jana, et al., 2001; Orr, 2012). Furthermore the cortical pyramidal neurons in the cerebral cortex are affected and the degradation will eventually result in cerebral cortex, thalamus and hippocampus damage (Jansen, 2017; Ortega, Díaz-Hernández, & Lucas, 2007; Rozas, et al., 2010; Schulte & Littleton, 2011).

In Huntington’s disease the CAG expansion lies in exon 1 at the N-terminus of the Huntingtin gene, starting at amino acid 18 (Schulte & Littleton, 2011). The repeat length varies among healthy humans with a repeat length of less than 35 glutamines and an average of 17-20 repeats (Imarisio, et al., 2008). PolyQ expansions with over 40 repeats in the Huntingtin protein is associated with disease and a juvenile onset when the repeats exceed above 50 (Schulte & Littleton, 2011). The mutant Huntingtin expanded polyQ stretch initiates misfolding of the protein causing the formation of aggregates in neurons.

Clinically, Huntington’s disease causes a reduced axonal transport, mitochondrial traffic disruption, transcriptional deregulation, dysfunction of the ubiquitin-proteasome system, disruption of endocytosis and synaptic transmission (Kim et al., 2009; Rozas et al., 2010; Schulte & Littleton, 2011). However the exact cause of neurodegeneration is still unclear. First the gene was only associated with gain-of-function properties but more recently it is proposed that both gain-of-function and loss-of-function cause the disease (Cattaneo et al., 2001; Gatchel & Zoghbi, 2005; Imariso et al., 2008; Kratter & Finkbeiner, 2010). The gain-of-function model does not preclude the possibility of modification by loss-of-function (Imariso et al., 2008).

(5)

5 Whether diffuse mutant Huntingtin (mHtt), the formation of intermediate oligomers or eventually the formation of aggregates is toxic for neurons remains elusive (Kim et al., 2009; Martin, Ladha, Ehrnhoefer, & Hayden, 2015; Saudou & Humbert, 2016). Some studies suggest mHtt and oligomers are toxic for cells while the aggregates protect the cells by reducing the amount of diffuse mHtt and oligomers (Kim et al., 2009; Saudou & Humbert, 2016). Other studies suggest the aggregation is interfering with cellular processes (Nucifora et al., 2001; Saudou & Humbert, 2016). Furthermore neurons appear to be especially sensitive to protein misfolding and aggregation due to their postmitotic and non-dividing properties (Cortes & Spada, 2014).

Several different approaches to reduce the neurodegeneration in Huntington’s disease are proposed (Sah & Aronin, 2011). Some strategies focus on indirect neuroprotection which aims at repairing suffered damage. Another therapeutic approach is lowering the mutant Huntingtin protein by gene suppression such as RNAi and antisense oligonucleotides (ASOs) (Kordasiewicz et al., 2012). These strategies target the causative gene directly. Another Huntingtin lowering strategy is enhancing the degradation of the Huntingtin protein by modifying the ubiquitin-proteasome system, autophagy or different chaperones. Both strategies prevent cellular damage and represent a strategy upstream the pathogenetic process in contrast to the neuroprotective strategies (Sah & Aronin, 2011). The numerous physiological activities of Huntingtin make it desirable to not suppress the Htt production completely (Sah & Aronin, 2011). Ideally, only the degradation of the aggregation prone mutant Huntingtin should be enhanced to prevent formation of toxic species while the soluble wild type Huntingtin, and thereby its function, is not affected.

The degradation of the Htt protein is generally done by the ubiquitin-proteasome system (UPS) (Jansen, Reits & Hol, 2017). The UPS regulates the destruction of short-lived and misfolded or damaged proteins in the cytoplasm (Rozas et al., 2010). The proteasome recognizes the target protein by a polyubiquitin chain signal on the lysine residues.

This so-called ubiquitination is accomplished by three different ligases. The first ligase is the ubiquitin-activating enzyme (E1), which activates the ubiquitin in an ATP-requiring reaction (Popovic, Vucic, & Dikic 2014). This activated ubiquitin is transferred to the ubiquitin-conjugating enzyme (E2), which binds to a specific ubiquitin-ligase (E3), and catalyzes the reaction where ubiquitin binds to the protein (Popovic, Vucic, & Dikic 2014). Ubiquitination is, besides by different E-ligases, also influenced by deubiquitinating enzymes (DUBs) (Mevissen & Komander, 2017). These DUBs can remove or disassemble the ubiquitin chains, having an editing function. Hereby DUB’s can control the ubiquitination and thereby the proteasomal degradation system (Jansen, 2017; Ortega, Díaz-Hernández, & Lucas, 2007).

(6)

6 The Htt protein is targeted by ubiquitination and the UPS is capable of degrading the polyQ tract of Htt when it is ubiquitinated (Juenemman, et al., 2013). However the ubiquitination of mutant Huntingtin seems to be partly impaired which results in a reduced recognition by the proteasome and is as a consequence less efficient degraded (Bennet et al., 2007; Juenemman, et al., 2013). Improved degradation could prevent formation of oligomers and aggregates and decrease the toxicity for neurons and possibly the neurodegeneration (Martin, Ladha, Ehrnhoefer, & Hayden, 2015).

The aim of this research is to set up a screening system where compounds and siRNAs libraries can be used to identify targets that will improve mutant Huntingtin degradation. Large compound libraries which might interfere in various cellular processes will be used, and smaller compound and siRNA libraries which directly target the UPS. However, before using these libraries a robust method should be developed and validated.

The libraries will be tested on striatal cells expressing mHtt with a polyQ expansion of 97 glutamines (mHttQ97) and a fluorescent GFP reporter coupled to a short glutamine expansion (GFPQ16). These proteins are separated by an internal ribosome entry site (IRES), for simultaneous and thereby equal expression levels. Importantly, the mHttQ97 protein does not have any (detectible) tag as this interferes with the stability of the protein. GFPQ16 is diffusely expressed but is recruited to aggregates induced by the mHttQ97 and is therefore a reporter for the otherwise non-fluorescent polyQ aggregates. In this way, aggregates can be scored by using fluorescent microscopy. For the screen it is important that the mHtt forms visible aggregates and the wild type Htt does not. Furthermore a robust way to measure GFP levels should be available to compare the expression levels of the mHttQ97-IRES-GFPQ16 construct.

Besides accurately counting aggregates and measuring GFP levels it is important for the validity of the screen to have a positive control. Therefore published compounds and a additionally set of DUB inhibitors provided by Leiden University Medical Center will be used. In this study a setup for large library screening is validated with reproducible measure methods and a positive control.

(7)

7

Methods

Cell lines

Three striatal cell lines were made with the mHttQ97-IRES-GFPQ16 pInducer plasmids provided by J. Janzen (Dept. of Medical Biology, AMC). The plasmid is inducible by tetracycline-controlled transcriptional activation, via doxycycline, and resistant to antibiotics.

The virus particles were generated in Human Embryonic Kidney 293T cells with a 3rd generation lentiviral packaging system of Addgene using pMD2G, pRSV-Rev, and pMDLg/pRRE (Appendix 1). The concentration of the virus was measured with an enzyme-linked immunosorbent assay for HIV-1 p24 (Appendix 2). Immortalized striatal neuronal cells (STHdh) were infected with the three different virus constructs; mHttQ97-IRES-GFPQ16, mHttQ97(3xR)-IRES-GFPQ16, and HttQ25-IRES-GFPQ16 and put on selection with Blasticidin 10 mg/ml. After the selection period the different cell lines were sorted with fluorescence-activated cell sorting (FACS) for the highest GFP fluorescent cell population. The cell lines were maintained in Dulbecco’s Modified Eagle Medium (DMEM, GIBCO) supplemented with 10% fetal calf serum (FBS) and 1% Penicillin Streptomycin glutamine (Thermo Fisher) in a humidified incubator with 5% atmospheric CO2 at 30°C. The characteristics of the cell lines

were tested with fluorescent microscopy.

Microscopy

The mHttQ97-IRES-GFP-Q16 cells were plated and induced with doxycycline (1:1000, 1 µg/ml) in a 6- or 12-wells plate for 48h. The cells were fixed with 4% paraformaldehyde (PFA) for 15 minutes and washed with phosphate-buffered saline PBS (GIBCO, Thermo Fisher). The nuclei were stained with Hoechst (1:500) combined with 0.1% Triton. The Leica LAS X imaging system was used to make at least 10 pictures of both the GFP and Hoechst fluorescence. The analysis was first done with manual counting of the aggregates and an ImageJ script.

MATLAB analysis

The previous method of manual counting and ImageJ analysis was replaced by a MATLAB script written in co-operation with R. Hoebe (Dept. of Medical Biology, AMC).

SDS-Soluble/Insoluble Fractionation

The mHttQ97-IRES-GFP-Q16 cells were plated with doxycycline (1:1000, 1 µg/ml) in a 6-wells plate. After 48 hours the cells were harvested in 1x PBS. The cell pellet was lysed with 1x TEX buffer (70 mM Tris/HCL pH 6.8, 1.5% SDS, 20% glycerol). The samples were sonicated with an ultrasonic disintegrator until the samples were not slurry anymore. After 10 µl DTT (1 M) was added, the samples were boiled for 10 min at 99°C in a thermoblock and centrifuged for 60 min at 14,000 rpm. The supernatant (SDS-soluble fraction) was transferred to a new tube containing 4 µl concentrated bromphenol blue solution (spatula-tip of bromphenol blue solved in water) without touching the SDS-insoluble pellet. The

(8)

8 insoluble pellet was solubilized with 10 µl 100% formic acid and incubated for 40 min at 37 °C while shaking. The formic acid was evaporated overnight at 30°C using a Speedvac system (Eppendorf). The remaining protein pellet was solved in 50 µl 1x TEX buffer supplemented with 0.05% bromophenol blue and boiled for 10 minutes at 99°C. The soluble and insoluble samples were loaded on an 12.5% SDS-page gel. The proteins were transferred to a nitrocellulose membrane by electrophoresis and blocked with 5% milk for 30 minutes. The membrane was stained with antibodies anti-Htt (1:1000, Abcam, N17) and anti-actin (1:1000, Santa Cruz, Santa Cruz, CA, USA) overnight. The signal was detected and quantified with the Odyssey imaging system V3.0 (Licor).

Filter Trap

The three cell lines were plated in the same way as the SDS-soluble/insoluble fractionation to perform a Filter Trap assay. The cells were harvested and lysed on ice with Triton X-100 buffer for 30 min and spinned down for 15 min at 4°C (14000 rpm). The pellet was resuspended in 100 µl Benzonase buffer (1 mM MgCL2, 50 mM Tris/HCL pH 8.0) and incubated for 1h at 37°C. A Bradford assay was performed on the supernatant to determine the protein concentration. The supernatant was loaded on a 12% SDS gel. The Benzonase reaction on the pellet was stopped with 2x Termination Buffer (40 mM EDTA, 4% SDS, 100 mM DTT). Samples were diluted with 2% SDS wash buffer. The 0.2 µm pore size cellulose acetate membrane (Schleicher and Shuell) was pre-washed in 2% wash buffer (2% SDS, 150 mM NaCl, 10 mM Tris/Hcl pH 8.0). The filter was placed in an BioRad filtration system and the samples were loaded in duplo. Low vacuum ensured the filtration of the sample. The filter was washed twice with 0.1 % SDS buffer (0.1% SDS, 150mM NaCl, 10mM Tris pH8.0) and blocked with 5% milk. The blot was stained with anti-Htt (1:1000, Abcam, N17) and anti-actin (1:1000, Santa Cruz, Santa Cruz, CA, USA) overnight. The signal was detected and quantified with the Odyssey imaging system V3.0 (Licor).

(9)

9

FITC intensity

Results

Validation of the cell lines

As mentioned before three cell lines were made to validate the screening method: the mHttQ97-GFPQ16 cell line and two control cell lines. First a HttQ25-mHttQ97-GFPQ16 cell line was made to confirm efficacy of the reporter as it should not aggregate in the absence of an extended polyglutamine repeat. And secondly, we included a lysine-dead variant of the mHtt protein (mHttQ97(3xR)-GFPQ16). In this construct we replace all lysine’s for arginine’s (R) to confirm or test whether a compound or siRNA is indeed acting via the UPS (Martin, Ladha, Ehrnhoefer, & Hayden, 2015). These cell lines were FACS sorted to determine the expression levels and the most fluorescent cells were selected. The HttQ25 and mHttQ97(3xR) cell line seem to be less fluorescent than the mHttQ97 cell line (Figure 1 & Appendix 3). In line with this data the mHttQ97(3xR) also forms less aggregates compared to the mHttQ97 measured with fluorescence microscopy and MATLAB analysis (Unpaired t-test, p = 0.0201; Appendix 4). The HttQ25 as expected did not form any visible aggregates, showing that the GFPQ16 does not initiate aggregates as a background signal (Figure 2). Furthermore it shows that not all cells are green fluorescent. These cells do not express the Htt GFPQ16 construct, or such low amounts that it is not detectible. This data shows that the cell lines function as expected and can be used for the large library screens.

Figure 1. FACs sorting data shows difference in fluorescence for mHttQ97, mHttQ97(3xR) and HttQ25. The first peaks show the not induced cells with as expected no fluorescence. The mHttQ97 (red) shows the highest fluorescence compared to the mHttQ97(3xR) (orange) and HttQ25 (green) when induced.

FACs sort graph

Ev e n ts d et ec ted

(10)

10

Setting up an analysing method for aggregate scoring

In the library screens the aggregation of the mHttQ97 will be measured with fluorescent microscopy to identify targets that improve the degradation of mHtt. The aggregates will be counted automatically by fluorescence microscopy analysis. The first tests however were done with manual counting. Diffuse and aggregated mHtt can be seen due to the GFPQ16 reporter as previously described. To have a good read-out, cells with too little expression of the mHtt GFPQ16 construct and thus GFP negative were not taken into account. The percentage of aggregates is therefore only calculated with the GFP positive cells. The threshold for green positive cells is determined by eye and then aggregates in the pool of GFP positive cells were counted. The percentage of aggregates mHttQ97 lies around 20%. As expected

Figure 2. Fluorescent microscopy pictures of mHttQ97, mHttQ97(3xR) and HttQ25. The three cell lines in the Hoechst (blue) and GFP (green) channel. The mHttQ97 has more aggregates and appears brighter than the mHttQ97(3xR) which corresponds to the FACs sorting data (Appendix 4). Furthermore the HttQ25 forms no visible aggregates.

(11)

11 from the FACS sort data the mHttQ97(3xR) cell line forms less aggregates with a percentage around 10%. The analysis with manual counting was replaced with a MATLAB script described below.

The previous method of counting manually was replaced by a self-written MATLAB analysis to make it more efficient and to include more parameters. This analysis includes (1) The mean fluorescence of all cells, (2) the GFP positive cells and (3) the percentage aggregates of the GFP positive cells. The mean fluorescence and percentage GFP positive cells can indicate if the compounds influence the expression levels of Huntingtin through a nonspecific effect, as GFP should not be affected.

For the analysis a picture was taken in the blue channel to visualize the Hoechst stained nuclei and in the green channel to visualize the GFP signal. The nuclei were counted with MATLAB software provided by R. Hoebe (Dep. Medical Biology). Next, The MATLAB script draws a line around the nucleus and grows this circle with a fixed pixel rate to represent the whole cell. This pixel rate is determined by the average size of the cells. This nuclear mask is than used as an overlay for the green channel picture. Because cells are not exactly round the black background is subtracted from this nucleus mask to minimize the influence on the measurements.

In this nuclear mask the mean GFP fluorescence is measured for each cell (1). The threshold for a GFP positive cell (2) is set with the standard deviation of this mean fluorescence (1) in the control condition. This standard deviation is chosen by analysing the data table outputs and trial and error with the visualisation of the counted GFP positive cells (Figure 3). The maximum value Is than used to detect an aggregate (3). This threshold is again determined by analysing the data tables from a number of control pictures and trial and error. The maximum fluorescence of a cell with an aggregate seem to be at least four times higher than the mean fluorescence. Visualisation of the counted aggregates showed this ratio detects aggregates accurately (Figure 3). The percentage of aggregates are again calculated with the GFP positive cells.

Figure 3. MATLAB analysis visualisation. The red nuclei are counted as GFP negative cells and the green nuclei are counted as GFP positive cells. Furthermore the counted aggregates are shown as a coloured ring around the nuclei.

(12)

12 This MATLAB script and threshold are compared with the previous manual counting method for the same pictures (Figure 4). Comparison between the two methods showed corresponding data.

In conclusion the percentage aggregates corresponds and the GFP

positive cells are counted accurately which suggest the thresholds are accurately set and this method can be used for further experiments with test compounds.

DUB inhibitor 6RK63 decreases aggregation in mHttQ97

Next, we tested the compounds Oleuropein, Trehalose, Methylene Blue, Epigallocatechin-3-gallate (EGCG), Curcumine and Congo Red that were described in literature to reduce aggregation

(respectively,

Rigacci, et al., 2015; Fernandez-Estevez, et al., 2014; Sontage et al., 2012; Hudson, Ecroyd, Dehle, Musgrave & Carver, 2009; Kumar, Padi, Naidu & Kumar, 2007; Heiser et al., 2000). By using this described compounds we were aiming to validate the method and to include a positive control. However, these compounds did not show a clear effect nor by microscopy or by biochemical assays on the mHttQ97 cell line (Results Appendix 5).

Therefore, as an alternative we included a small set of DUB inhibitors. Deubiquitinating enzyme inhibitors prevent the removal of ubiquitin chains and thereby might improve degradation. Since these compounds were only tested in vitro and not in living cells, we first preformed A MTT tetrazolium reduction assay to determine toxicity of the inhibitors. All inhibitors had a non-toxic working concentration of 2 µM (Appendix 6). At least 10 pictures were taken for each condition in the Hoechst (blue) and GFP (green) channel.

The pictures were first analysed by counting the amount of aggregates and green positive cells manually. The results are shown in Figure 5. There was a significant difference between the control DMSO and the DUB inhibitor 6RK63 (Paired t-test, p < 0.05). The other DUB inhibitors did not differ significantly from the control.

Figure 4. The percentage aggregates in different experiments is with manual counting (C) similar to MATLAB analysis (M). The percentage of aggregates of green cells counted with the first manual method (C) and with the MATLAB analysis (M). The percentages are similar which suggest the MATLAB analysis is accurate.

(13)

13 To again validate the MATLAB analysis and reproduce this result, more experiments with the DUB inhibitor 6RK63 are done (Figure 6 & Table 1). In total thirteen experiments were run with MATLAB, three of which were already included in the previous results.

The percentage of aggregates with the DUB6RK63 in the mHttQ97 is again significantly lower than in the control DMSO (Paired t-test, p < 0.001). In contrast the mean GFP fluorescence of all cells and the percentage of green cells did not differ significantly between the DUB63 and control. This results shows that the DUB inhibitor 6RK63 decreases the amount of aggregates in the mHttQ97 cells while the mean GFP fluorescence is not altered. This suggest the compound does not influence expression of the Htt GFPQ16 construct but effects the degradation of Huntingtin.

Figure 5. DUB6RK63 decreases the percentage aggregates of GFP positive cells, counted manually. The DUB6RK63 ((M = 13.70, SD = -4.44) differed significantly from the control DMSO (M = 25.62, SD = -7.26;

t(30), p = 0.0189). The other DUB inhibitors did not differ significantly. Mean ± SEM., one-way ANOVA

Dunnett’s multiple comparison test; * p<0.05

Figure 6. DUB6RK63 decreases the percentage aggregates and does not influence the mean GFP

fluorescence and percentage green cells in mHttQ97. A significant difference in percentage aggregates is found between the control DMSO and the DUB6RK63. This difference is not seen for the mean

fluorescence. And no significant result was found for the percentage of green cells. Mean ± SEM., Paired t-test.; ***p<0.001

(14)

14

DUB inhibitor 6RK63 acts via the ubiquitin proteasome system (UPS)

To determine if the DUB inhibitor 6RK63 influences the ubiquitination proteasome system the mHttQ97(3xR) cell line was analysed. In contrast to the mHttQ97, the mHttQ97(3xR) shows no significant difference in aggregation between the control and DUB63, nor for the mean GFP fluorescence and percentage green cells (Paired t-test; Figure 7 & Table 2). Even though, the percentage aggregates in the DMSO samples of the mHttQ97(3xR) cell line is lower than the aggregation in the DMSO sample of the mHttQ97 cell line which is in line with the FACs sorting data (Figure 3). All in all this data shows that the reduced aggregation by the DUB6RK63 occurs through ubiquitination and does not influence the protein production.

Variable Condition Mean Standard Deviation T(12) p-value Percentage aggregates DMSO DUB63 18.55 12.91 4.69 3.67 5.48 < 0.001 Mean GFP Fluorescence DMSO DUB63 16.87 16.91 6.60 6.66 0.10 0.9197 Percentage green cells DMSO DUB63 63.51 63.80 9.99 11.09 0.18 0.8633

Figure 7. Percentage aggregates, mean GFP fluorescence and percentage green cells in mHttQ97(3xR) does not differ for DUB6RK63. There is no significant difference found in either of the parameters. The

percentage of aggregates and mean GFP fluorescence is however less in the mHttQ97(3xR) compared to the mHttQ97. Mean ± SEM., Paired t-test.

(15)

15

DUB6K63 reduces insoluble Htt levels and does not affect soluble Htt levels in mHttQ97

To test the robustness of our screening method, we used the DUB inhibitor 6RK63 in biochemistry assays to further confirm the validity of the setup. Cells were treated similar as in previous experiments and subjected to a SDS-soluble/insoluble assay. This assay enables the measurement of the soluble Htt fraction compared to the insoluble Htt fraction. Previous described data suggest the percentage of aggregates decreases. This could be due to the fact that the mHtt is degraded and does not form aggregates, or the compound increases the amount of soluble mHtt by disassembling aggregates or oligomers. In the latter case the amount of soluble Htt would increase and the insoluble Htt would decrease.

In Figure 8 a SDS-soluble and insoluble gel is shown with the three different cell lines treated with the DUB6RK63. The gel is stained for Huntingtin and actin acts as a loading control. The bands around 46 kDa show the Htt signal. When quantified for six experiments no clear difference is observed between soluble levels for the control and DUB6RK63 in mHttQ97 (Figure 9). The DUB6RK63 does however decrease the insoluble Htt levels (Paired t-test, p < 0.01). This shows that the mHtt forms less oligomers and aggregates without affecting the soluble Huntingtin.

To again determine if the effect on the insoluble levels of Htt occurs through ubiquitination the mHttQ97(3xR) cell line was also tested. Here both the soluble and insoluble Htt levels are not affected by the DUB6RK63 compared to the control DMSO. Furthermore the insoluble level in the mHttQ97(3xR) is less compared to the mHttQ97 as expected from the FACs data and microscopy analysis. The HttQ25 cell line does not form aggregates and therefore has no insoluble Htt.

Altogether this data is in line with the microscopy analysis which both suggest the DUB6RK63 improves degradation through ubiquitination of mHtt aggregates. The counting of aggregates with fluorescent microscopy is therefore a reproducible and viable method.

Variable Condition Mean Standard Deviation T(5) p-value Percentage aggregates DMSO DUB63 12.54 13.06 4.890 5.599 0.791 0.4647 Mean GFP Fluorescence DMSO DUB63 12.96 12.52 3.602 3.402 2.298 0.0699 Percentage green cells DMSO DUB63 71.60 67.89 5.838 10.33 1.407 0.2185 Table 2. Statistical data analysis DUB6RK63 effect on mHttQ97(3xR)

(16)

16 Figure 8. DUB6RK63’s effect on SDS-soluble and insoluble in the thee cell lines. An 12% SDS gel

with Htt bands around 46 kDa stained with anti-Htt N17 (abcam) is shown. In the mHttQ97(3xR) the soluble Htt bands are also found around 58 kDa. The HttQ25 cell line shows the Htt signal around 25 kDa. As expected no Htt signal for the HttQ25 is found on the insoluble blot and the mHttQ97(3xR) shows less signal.

Figure 9. DUB6RK63’s effect on SDS-soluble and insoluble in the thee cell lines. The results are quantified and normalized to the soluble actin. mHttQ97 show no significant difference between the control and DUB6RK63 (M = 98.04, SD = 14.96) in soluble levels (paired t-test: t(5) = 0.32, p = 0.7617). However the insoluble levels in mHttQ97 show a significant decrease with DUB6RK63 (M = 60.61, SD = 21.28; Paired t-test: t(5) = 4.53, p = 0.0062). While more measurements should be performed the mHttQ97(3xR) showed no significant difference between the DMSO (M = 100.0, SD = 0.00) and DUB6RK63 (soluble: M = 104, SD = 1.41, insoluble: M = 97.0, SD = 12.73) with the Wilcoxon-Mann-Whitney test (soluble: p = 0.334, insoluble: p = 0.1)

(17)

17

DUB6K63 reduces insoluble Htt levels in mHttQ97

The Filter Trap is another way of separating mHtt aggregates from soluble Htt levels and to measure the effect of compounds on mHtt aggregation. Again this will be used to confirm the setup of counting aggregates by fluorescent microscopy and reproduce the previous fluorescent microscopy and SDS-soluble/insoluble results. After 48h of treatment with the DUB inhibitor, cells were harvested and subjected to the Filter Trap protocol. The Filter Trap blot shows again that in general less aggregates are formed in the mHttQ97(3xR) compared to the mHttQ97 and no aggregates are formed with the HttQ25 cell line (Figure 10). The quantification shows that the DUB63 contains significantly less mHtt aggregation than the control DMSO in mHttQ97 (Paired t-test, p < 0.05; Figure 11). In the mHttQ97(3xR) this result is not seen.

From this we can conclude that the DUB6RK63 decreases the mHtt aggregation by influencing the UPS system. In addition the microscopy analysis is reproducible in other robust proven methods, confirming its validity.

Figure 10. DUB6Rk63 reduces the insoluble Htt levels in mHttQ97. The three cell lines are treated with DMSO control, the DUB6RK63 and a non-induced condition.

Figure 11. Filter Trap quantification of DUB6RK63 for mHttQ97 shows a decrease in insoluble Htt levels. Significant less Htt aggregation is seen in the DUB6RK63 (M = 13.04, SD = 3.52) compared to the control DMSO (M = 23.94, SD = 3.52; t(2) = 6.56, p = 0.0238). For the mHttQ97(3xR) this result is not seen (DMSO: M = 7.38, SD = 5.31; DUB6RK63: M = 10.04, SD = 4.61; t(2) = 1.65, p = 0.1766). Mean ± SEM., Paired t-test.; *p<0.05

(18)

18

Discussion

This research provides a robust and reliable screening method where compounds and siRNAs libraries can be used to identify targets that will improve the degradation of Huntingtin. This screening method shows reproducible effects on aggregation validated with biochemistry follow-up experiments. The microscopy analysis seems to be even more robust than the conventional biochemistry assays. Implementation of this setup into a large scale screen, where 250.000 compounds and 700 DUB inhibitors will be tested can be proceeded.

The mHttQ97-IRES-GFPQ16 and mHttQ97(3xR)-IRES-GFPQ16 form measurable protein aggregates and the HttQ25-reporter does not both measured with fluorescent microscopy and biochemistry assays. Concluded from the FACs, microscopy and biochemistry data the mHttQ97(3xR) cell line forms less aggregates than the mHttQ97. For library screens it is ideal that these cell lines have the same Htt and GFP expression levels. This is accomplished by FACs sorting the three cell lines again with the same standards for GFP expression levels.

In addition, a DUB inhibitor that reduces mHtt levels is identified. The DUB inhibitor 6RK63 shows a decrease in aggregation and a decrease of insoluble Htt. Furthermore there is no difference found in GFP fluorescence and soluble Htt levels. This confirms that the DUB6RK63 does not influence the expression level. GFP and Htt are expressed in 1:1 ratio, a drop in Htt protein levels but not in GFP levels therefore indicates that the decrease happens at protein level and is specific for the Huntingtin protein.

Furthermore, the decrease in aggregation of mHtt is not seen in the mHttQ97(3xR) cell line. This indicates that the effect of the DUB6RK63 is due to its interference with the ubiquitin proteasome system. All in all the DUB6RK63 improves the degradation of mHtt through the ubiquitin proteasome system.

To determine where in the process DUB63 interferes an agarose gel electrophoresis for resolving aggregates could be used. The agarose gel shows a separation of different species and therefore could be used to analyse the different phases of oligomerization (Appendix 9). It is important to know if DUB63 enables degradation of soluble mHtt levels before it aggregates or of if it increases the ubiquitination of oligomers and aggregates.

It is yet unclear why the different variants of DUB6RK61, 62, 64 and 73 do not influence aggregation, further research will be needed to determine whether the DUB6RK63 has a specific or a-specific effect. The first test that should be done is activity labelling with active-site-directed probes or a proteasome inhibitor. The activity probes are used to measure DUB activity via fluorescent labelling of the active sites of all DUB’s present in the cell, which could give more insight in the specificity of the used inhibitors (Leestemaker, Jong & Ovaa, 2017) . If the inhibitor is specific for the

(19)

19 UPS system a reversed effect will be seen after proteasome inhibition which confirms the interaction of the DUB6RK63 with the UPS system.

DUB6RK61-64 supposedly targets USP30, a mitochondrial-associated deubiquitylate enzyme (Bingol, et al., 2014). If the DUB inhibitor affects the mitochondrial function more research is needed into mitochondrial functioning in Huntington’s disease. The measured effect could be due to a combination of improved proteasomal degradation and altered mitochondrial function.

(20)

20

Literature

Bennett, E. J., Shaler, T. A., Woodman, B., Ryu, K. Y., Zaitseva, T. S., Becker, C. H., Bates, G.P., Schulman H., & Kopito, R. R. (2007). Global changes to the ubiquitin system in Huntington's disease. Nature,

448(7154), 704.

Bingol, B., Tea, J. S., Phu, L., Reichelt, M., Bakalarski, C. E., Song, Q., Foreman, O., Kirkpatrick, D.S., & Sheng, M. (2014). The mitochondrial deubiquitinase USP30 opposes parkin-mediated mitophagy. Nature,

510(7505), 370.

Cortes, C. J., & La Spada, A. R. (2014). The many faces of autophagy dysfunction in Huntington's disease: from

mechanism to therapy. Drug discovery today, 19(7), 963-971.

East, D. A., & Campanella, M. (2016). Mitophagy and the therapeutic clearance of damaged mitochondria for neuroprotection. The international journal of biochemistry & cell biology, 79, 382-387.

Fernandez-Estevez, M. A., Casarejos, M. J., Sendon, J. L., Caldentey, J. G., Ruiz, C., Gomez, A., Perucho, J., Yebenes, J. G., & Mena, M. A. (2014). Trehalose reverses cell malfunction in fibroblasts from normal and Huntington's disease patients caused by proteosome inhibition. PloS one, 9(2), e90202.

Gusella, J.F. & MacDonald, M.E. (1995). Huntington’s disease. Cell Biology. 6, 21-28.

Hannan, A. J. (2018). Tandem repeats mediating genetic plasticity in health and disease. Nature Reviews

Genetics. doi:10.1038

Heiser, V., Scherzinger, E., Boeddrich, A., Nordhoff, E., Lurz, R., Schugardt, N., Lehrach, H., & Wanker, E. E. (2000). Inhibition of huntingtin fibrillogenesis by specific antibodies and small molecules: implications for Huntington's disease therapy. Proceedings of the National Academy of Sciences, 97(12), 6739-6744. Hudson, S. A., Ecroyd, H., Dehle, F. C., Musgrave, I. F., & Carver, J. A. (2009). (−)-Epigallocatechin-3-gallate (EGCG)

maintains κ-casein in its pre-fibrillar state without redirecting its aggregation pathway. Journal of molecular biology, 392(3), 689-700.

Imarisio, S., Carmichael, J., Korolchuk, V., Chen, C. W., Saiki, S., Rose, C., Krishna, G., Davies, E.J., Ttofi, E., Underwood, B.R., & Rubinsztein, D. C. (2008). Huntington's disease: from pathology and genetics to potential therapies. Biochemical Journal, 412(2), 191-209.

Jana, N. R., Zemskov, E. A., Wang, G. H., & Nukina, N. (2001). Altered proteasomal function due to the expression of polyglutamine-expanded truncated N-terminal huntingtin induces apoptosis by caspase activation through mitochondrial cytochrome c release. Human molecular genetics, 10(10), 1049-1059.

Jansen, A. H., Reits, E. A., & Hol, E. M. (2014). The ubiquitin proteasome system in glia and its role in neurodegenerative diseases. Frontiers in molecular neuroscience, 7, 73.

Jansen, A. H. P. (2017). Degrade or degenerate: Glial protein degradation in Huntington’s disease. Amsterdam, Nederland.

Juenemann, K., Schipper-Krom, S., Wiemhoefer, A., Kloss, A., Sanz, A. S., & Reits, E. A. (2013). Expanded polyglutamine-containing N-terminal huntingtin fragments are entirely degraded by mammalian proteasomes. Journal of Biological Chemistry, 288(38), 27068-27084.

Kim, M. W., Chelliah, Y., Kim, S. W., Otwinowski, Z., & Bezprozvanny, I. (2009). Secondary structure of Huntingtin amino-terminal region. Structure, 17(9), 1205-1212.

Kordasiewicz, H. B., Stanek, L. M., Wancewicz, E. V., Mazur, C., McAlonis, M. M., Pytel, K. A., Artates, J.W., Weiss, A., Cheng, S.H., Shihabuddin L.S., Hung, G., Bennet, F.C. & Cleveland, D.W.(2012). Sustained therapeutic reversal of Huntington's disease by transient repression of huntingtin synthesis. Neuron,

74(6), 1031-1044.

(21)

21 Kumar, P., Padi, S. S. V., Naidu, P. S., & Kumar, A. (2007). Possible neuroprotective mechanisms of curcumin in attenuating 3-nitropropionic acid-induced neurotoxicity. Methods and findings in experimental and clinical pharmacology, 29(1), 19-26.

Leestemaker, Y., de Jong, A., & Ovaa, H. (2017). Profiling the activity of deubiquitinating enzymes using chemically synthesized ubiquitin-based probes. In Activity-Based Proteomics (pp. 113-130). Humana Press, New York, NY

Martin, D. D., Ladha, S., Ehrnhoefer, D. E., & Hayden, M. R. (2015). Autophagy in Huntington disease and huntingtin in autophagy. Trends in neurosciences, 38(1), 26-35.

Mevissen, T. E., & Komander, D. (2017). Mechanisms of deubiquitinase specificity and regulation. Annual

review of biochemistry, 86, 159-192.

Mo, C., Hannan, A. J., & Renoir, T. (2015). Environmental factors as modulators of neurodegeneration: Insights from gene–environment interactions in Huntington's disease. Neuroscience & Biobehavioral Reviews,

52, 178-192.

Nucifora, F. C., Sasaki, M., Peters, M. F., Huang, H., Cooper, J. K., Yamada, M., Takahashi, H., Tsuji, S., Troncoso, J., Dawson, V.L., Dawson, T.M., & Ross, C.A., (2001). Interference by huntingtin and atrophin-1 with cbp-mediated transcription leading to cellular toxicity. Science, 291(5512), 2423-2428.

Orr, H. T. (2012). Polyglutamine neurodegeneration: expanded glutamines enhance native functions. Current

opinion in genetics & development, 22(3), 251-255.

Ortega, Z., Díaz-Hernández, M. & Lucas, J.J., (2007). Is the ubiquitin-proteasome system impaired in Huntington’s disease? Cellular and Molecular Life Science. 64 (17), 2245-2257

Popovic, D., Vucic, D., & Dikic, I. (2014). Ubiquitination in disease pathogenesis and treatment. Nature

medicine, 20(11), 1242.

Rigacci, S., Miceli, C., Nediani, C., Berti, A., Cascella, R., Pantano, D., Nardiello, P., Luccarini, I., Casamenti, F., & Stefani, M. (2015). Oleuropein aglycone induces autophagy via the AMPK/mTOR signalling pathway: a mechanistic insight. Oncotarget, 6(34), 35344.

Rozas, J. L., Gómez-Sánchez, L., Tomás-Zapico, C., Lucas, J. J., & Fernández-Chacón, R. (2010). Presynaptic dysfunction in Huntington's disease.

Sah, D. W., & Aronin, N. (2011). Oligonucleotide therapeutic approaches for Huntington disease. The Journal of

clinical investigation, 121(2), 500-507.

Saudou, F., & Humbert, S. (2016). The biology of huntingtin. Neuron, 89(5), 910-926.

Scherzinger, E., Lurz, R., Turmaine, M., Mangiarini, L., Hollenbach, B., Hasenbank, R., Bates, G.P., Davies, S.W., Lehrah, H., & Wanker, E. E. (1997). Huntingtin-encoded polyglutamine expansions form amyloid-like protein aggregates in vitro and in vivo. Cell, 90(3), 549-558.

Schulte, J., & Littleton, J. T. (2011). The biological function of the Huntingtin protein and its relevance to Huntington’s Disease pathology. Current trends in neurology, 5, 65-78.

Sontag, E. M., Lotz, G. P., Agrawal, N., Tran, A., Aron, R., Yang, G., Necula, M., Lau, A., Finkbeiner, S., Glabe, C., Marsh, J. L., Muchowski, P. J. & Thompson, L. M., (2012). Methylene blue modulates huntingtin aggregation intermediates and is protective in Huntington's disease models. Journal of Neuroscience, 32(32), 11109-11119.

(22)

22

Appendix

Appendix 1. Protocol: Virus production

Split Human Embryonic Kidney 293T cells (37°C, 5% CO2) with Dulbecco’s Modified Eagle Medium

(DMEM) supplemented with 10% fetal calf serum (FBS) and no antibiotics. Cells should be 80% confluent on transfection day. On transfection day use 5,56 µg pMD2G (1 µg/µl), 9,26 µg PRE/MDLg (1 µg/µl), 4,63 µg pRSV/REV (0,8 µg/µl), 18,53 µg shVECTOR (1 µg/µl) and 125 l PEI in a 15 cm petri dish with 1 ml Optimem. Add the transfection mix to HEK cells. Incubate for 24h at 37 °C. Change the medium to DMEM with 10% fetal calf serum (FBS) and 1% Penicillin Streptomycin glutamine. Harvest the supernatant on the third day and store it at 4 °C and add new DMEM to the HEK cells. Harvest again at day four. Combine the supernatants from day three and four, spin it down 5 min at 32000 rpm. Filtrate the solution through 0.45 µM filter. Spin down the pooled virus in SW27 tubes for 65 minutes at 32000 rpm at 4 °C. Eliquite 200 µl per tube and freeze it at -80 °C.

(23)

23

Appendix 2. Protocol: Antigen capture enzyme-linked immunosorbent assay (Elisa)

Prepare 96-wells plate

1. Add 200 ul Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal calf serum (FBS) and 1% Penicillin Streptomycin glutamine in duplo to the first wells

2. Mix 882 µl DMEM +/+ and 18 µl HIV1 P24 stock = 900 µl 200 pg 3. Add 450 µl mix and 450 ul DMEM +/+ = 900 µl 100 pg 4. Add 450 µl mix and 450 ul DMEM +/+ = 900 µl 50 pg 5. Add 450 µl mix and 450 ul DMEM +/+ = 900 µl 25 pg 6. Add 450 µl mix and 450 ul DMEM +/+ = 900 µl 12.5 pg 7. Add 200 µl of the virus dilution to the wells: 500x, 100x, 50x, 10x Elisa

● Dispense 20 µl of lysis buffer into each well.

● Dispense 200 µl of each standard curve dilution, supernatant sample, and culture medium into appropriately labelled duplicate wells.

● Incubate at 37(±1)°C for 60 (±5) minutes.

● Aspirate the contents of the wells, and wash the microtiter plate (6x).

● Dispense 100 µl of Anti-p24 (Biotin conjugate) detector antibody into each well. ● Incubate at 37(±1)°C for 60 (±5) minutes.

● Aspirate the detector antibody from the wells, and wash the microtiter plate (6x). ● Dispense 100 µl of Streptavidin-HRP conjugate into each well.

● Incubate at room temperature (18–25°C) for 30 (±5) minutes.

● Aspirate the conjugate from the wells, and wash the microtiter plate (6x).

● Without delay, dispense 100 µl of Substrate Solution into each well. A multichannel pipet should be used for best results.

● Protect the plate from direct light/sunlight, and incubate at room temperature (18–25°C) for 20 (±2) minutes.

● Stop the reaction by adding 100 µl of Stop Solution to each well including the culture medium blanks. The blue solution should change to a uniform yellow color. Ensure that the undersides of the wells are dry and that there are no air bubbles in the well contents.

● Immediately after adding the Stop Solution, read the absorbance values at 450 nm using a microtiter plate reader blanked on the negative control well.

● Washing: Add 300 µl 1x wash buffer (60ml/8wells) with multi pipet to wells, directly remove with aspirator. Repeat 6 times.

Absorbance at 480 nm measured with a spectrophotometer and exported in excel. Generate a standard curve by calculating the average absorbance of the duplo standard curve dilution with a range of 0 to 200 pg/ml p24. Use the formula of the standard curve to calculate the amount of pg/ml of the diluted virus concentrations.

(24)

24

Appendix 3. FACs sorting data

Cell line % GFP negative % GFP positive Geometric mean %

GFP positive Geometric mean whole curve Q97 new 1.5 98.5 12156 11423 Q97 old 10.1 89.9 7607 5372 3xR Q97 21.1 78.9 5658 2991 Q25 32.3 67.7 9704 3053 Wildtype 100 0 127

(25)

25

Appendix 4. Microscopy data in cell lines mHttQ97 and mHttQ97(3xR)

The mHttQ97 and mHttQ97(3xR) are compared in Figure 4.1. The percentage of aggregates for the mHttQ97 (M = 18.55, SD = 4.692) and mHttQ97(3xR) (M = 12.53, SD = 4.91) differ significantly (t(17) = 2.56, p = 0.0201). A slight decrease in mean GFP fluorescence is found in mHttQ97(3xR) (M = 12.98, SD = 3.60) compared to the mHttQ97 (M = 16.89, SD = 6.60) but this is not significant (t(17) = 1.34, p = 0.1965). The percentage of green cells is calculated with a threshold set with the control pictures of each cell line. Therefore this parameter cannot be statistically compared and the trend in expressing the mHttQ97 GFPQ16 construct cannot be seen.

Figure 4.1. mHttQ97 (n=13) and mHttQ97(3xR) (n=6) comparison for the percentage of aggregates, mean GFP fluorescence and percentage green cells.

(26)

26

Appendix 5. Tested compounds from literature

The compounds found in literature are tested with their own control and several different concentrations suggested by the literature (Table 5.1). The results are shown in Figure 5.1 and Table 5.2.

Compound Literature

Congo Red Heiser et al., 2000

Curcumine Kumar, Padi, Naidu

& Kumar, 2007

Epigallocatechin-3-gallate (EGCG

Hudson, Ecroyd, Dehle, Musgrave & Carver, 2009

Methylene Blue Sontage et al., 2012

Oleuropein Rigacci, et al., 2015

Trehalose Fernandez-Estevez,

et al., 2014

Figure 5.1. Congo Red, Curcumine, EGCG, Methylene Blue, Oleuropein and Trehalose tested for percentage of aggregates. No clear result was found in either of the used concentrations.

Table 5.1. Literature overview of the tested compounds.

(27)

27 Table 5.2. Congo Red, Curcumine, EGCG, Methylene Blue, Oleuropein and Trehalose tested for percentage of aggregates. Compound Control / Concentration % Aggregates Counted cells

Congo Red Water

0.5 µM 5 µM 50 µM 33 24 33 41 129 1654 133 131 Curcumine DMSO 0.5 µM 1.0 µM 5 µM 32 47 52 46 147 136 152 142 EGCG Water 1 mM 5 mM 10 mM 33 30 44 47 129 161 128 121

Methylene Blue Water

10 nM 100 nM 1000 nM 33 36 40 38 129 195 185 132 Oleuropein DMSO 1 µM 5 µM 10 µM 32 36 34 39 147 176 213 217 Trehalose Water 5 mM 25 mM 50 mM 33 26 29 22 129 182 284 322

(28)

28

Appendix 6. Mtt assay DUB inhibitors

To determine the right compound concentration for the screens on STHdh mHtt cells a MTT tetrazolium reduction assay was used on the DUB inhibitors provided by Leiden University Medical Center (LUMC). These compounds are diluted in dimethylsulfoxide (DMSO) which acts as an control. Figure (6.1) shows the metabolic activity of the living cells measured by the absorbance of 570nm for a range of compound concentrations. The absorbance is normalized to 0 µM compound. The

metabolic activity of the living cells decreases overall with higher concentrations of the control and compounds. This decrease sets in at around 8 µM for the control DMSO. The DUB inhibitors are a bit more toxic with a gradual decrease from around 5 µM to 16 µM shown in Figure 6.1. No toxicity was found around 2 µM and along with recommendation of the providers of the compounds this was used for further experiments on the STHdh Htt cells.

Figure 6.1. Toxicity Mtt assay DUB inhibitors 6RK61-64, 73. The absorbance of 570 nm over different DUB inhibitor concentrations in µM. The data is normalized to absorbance levels with no compound added. The absorbance begins decreasing around 5 µM for most compounds, therefore a concentration of 2 µM was used for further experiments.

(29)

29

Appendix 7. Microscopy data analysis

Table 7.1. Results microscopy analysis with ImageJ, Percentage aggregates green cells for six experiments sorted on date of initiation.

Exp Date of

initiation

DMSO DUB 61 DUB 62 DUB 63 DUB 64 DUB 73

Exp. 1 20-03 30.0 13.8 18.6 12.3 24.4 42.7 Exp. 2 21-03 37.1 20.0 25.7 18.8 20.6 34.1 Exp. 3 03-04 23.4 23.7 18.4 16.4 15.5 23.9 Exp. 4 03-04 26.1 25.0 25.9 15.3 27.2 23.9 Exp. 5 09-04 20.5 15.6 25.3 13.5 22.0 16.3 Exp. 6 10-04 16.6 10.2 11.9 5.9 13.0 17.6 Mean: 25.6 18.1 21.0 13.7 20.5 26.4

Table 7.2. Results microscopy analysis with MATLAB. Percentage aggregates, mean GFP fluorescence of all cells and percentage of GFP positive cells for 13 experiments sorted on date of initiation in the mHttQ97-IRES-GFPQ16. Date of Initiation Condition Percentage of Aggregates Mean GFP Fluorescence Percentage GFP Positive cells 03-04 DMSO DUB63 26.7 13.7 25.9 28.1 60.7 67.0 09-04 DMSO DUB6 22.8 11.5 28.9 26.8 65.6 60.8 10-04 DMSO DUB63 12.8 6.0 22.7 25.2 70.2 81.6 16-05 DMSO DUB63 18.7 12.2 14.5 14.5 46.6 50.0 17-05 DMSO DUB63 13.6 9.3 12.0 11.6 65.2 62.8 23-05 DMSO DUB63 17.6 14.1 21.7 20.8 45.1 43.8 28-05 DMSO DUB63 15.1 13.9 16.8 16.0 70.7 62.3 04-06 DMSO DUB63 11.9 9.8 12.7 13.7 54.5 60.5 04-06 DMSO DUB63 19.4 19.1 21.9 20.8 68.8 60.5 09-04 DMSO DUB63 20.9 15.0 7.9 9.1 78.7 84.6 09-04 DMSO DUB63 25.9 18.2 11.5 10.9 74.3 72.5 09-04 DMSO DUB63 19.3 13.6 11.2 10.8 63.7 62.9

(30)

30 09-04 DMSO DUB63 16.5 12.3 11.8 11.3 61.4 60.1

Table 7.2. Results microscopy analysis with MATLAB. Percentage aggregates, mean GFP fluorescence of all cells and percentage of GFP positive cells for 13 experiments sorted on date of initiation in the mHttQ97(3xR)-IRES-GFPQ16. Date of Initiation Condition Percentage of Aggregates Mean GFP Fluorescence Percentage GFP Positive cells 16-05 DMSO DUB63 7.2 6.1 12.4 12.8 73.8 82.3 17-05 DMSO DUB63 10.1 9.8 12.2 11.8 71.2 67.5 23-05 DMSO DUB63 11.4 10.4 19.1 18.0 62.2 51.5 28-05 DMSO DUB63 15.3 17.2 14.0 13.5 76.6 69.9 04-06 DMSO DUB63 10.2 13.1 12.2 11.5 68.0 62.9 09-04 DMSO DUB63 21.0 21.7 8.0 7.5 77.8 73.2

(31)

31

Appendix 8 Filter Trap concentration range

Different concentrations of protein were tested on the Filter Trap system to determine which concentration should be used in the further experiments. According to figure 5.1 it was decided to use 25 µg.

Figure 8.1 Filtertrap Concentration Range.

The Triton soluble sample was used to perform a Bradford assay and 20 µg was loaded on a 12% SDS gel. This was done to be able to control for the amount of protein loaded on the Filter Trap. Furthermore the soluble levels were quantified and did not differ for the control DMSO and DUB63.

(32)

32

Appendix 9. Agarose Gel Electrophoresis for Resolving Aggregates (AGERA)

The three cell lines, Q97, 3xR Q97 and Q25, were plated to run an agarose gel electrophoresis for resolving aggregates (Agera). Two or three wells from an 6-wells plate were used. The cells were lysed in Triton lysis buffer and sonicated. A bradford assay was performed to measure the protein concentration. The samples were diluted 1:1 into Leammli loading buffer (150mM Trish/HCl pH6.8, 33% glycerol, 1.2% SDS, and bromphenol blue) and incubated for 5 min at 95 °C. The Agera gel was prepared by adding 3.75g Agarose to 250ml TAE buffer (375mmol/L, Tris-HCL, pH 8,8) and microwaved untill the agarose had dissolved. The solution was cooled to 60 °C and 1250 µl 20% SDS was added to an end concentration of 0,1% and cooled in an Biorad DNA Sub Cell tray for 50 minutes with a camb. After sample loading the gel was run in Leammli running buffer (192 mmol/L glycine, 25 mmol/L Tris-base, 0,1% SDS) at 100 V till the bromophenol blue reached ¾ of the gel. The gel was blot on PDVF membranes (Millipore, Zug, Switzerland; Immobilon-P) by capillarity for more than 24h in transfer buffer (192mM glycine, 25mM Tris-base, 0.1% SDS, 15% methanol). Afterwards the membrane was blocked in 5% milk and stained with antibody anti-Htt (N17, Abcam). The results are shown in figure 9.1.

Figure 9.1. Agera gel

For the three cell lines Q97, 3xR Q97 and Q25 48h dox induction and no dox condition was run. The blots were stained with anti-Htt (Abcam) and anti-3B5H10.

Screen for Huntington’s disease; The development and validation of a robust method for compound library screening in Huntington’s disease