Protein Surface Scanning. A crystal structure of the surface of an epitope on mHTT (PDB entry: 2otu) was described by means of surface feature points. The selected surface was then compared to a database of 240,013 interacting pairs of chemical fragments and protein environments extracted from the PDBbind database,9 using the ProSurfScan methodology based on the identification of chemoisosteric protein environments10. When a fragment-interacting surface environment present in PDBbind was found to be similar to a surface environment of the polyQ peptide, the chemical fragment of the former was directly mapped onto the latter. This resulted in a selection of 82 chemical fragments that covered a reasonable portion of the polyQ peptide interface to the antibody. These fragments were then used to filter a catalogue of around 12 million compounds from chemical vendors based on the assumption that compounds containing those chemical fragments will themselves favorably interact with the polyQ peptide. Accordingly, for each molecule in the catalogue, a substructure search was performed against the set of 82 chemical fragments. In the event that multiple fragments are detected as substructures of the catalogue molecule, the matching fragments were ordered by size and iteratively selected if compatible with all the already selected matches. Two chemical fragments were regarded as compatible if their volumes after the mapping were not overlapping. These criteria avoid selecting large compounds containing chemical fragments that were mapped onto the same region of the target surface. For a molecule to be selected, at least two thirds of their atoms are required to match any of the 82 fragments. A total of 2,937 compounds passed the filter at this stage. This initial set of 2,937 compounds was ranked by its predicted binding energy to the target structure using AutoDock11. To ensure a wide and diverse exploration of the polyQ peptide surface, compounds showing a surface overlap larger than 66% were filtered out and only the compound with the best predicted binding energy to a given surface region was kept. A list of 67 compounds predicted to interact favorably with the mHTT was generated. A final set of 49 compounds was purchased from chemical vendors.
The drug target was expressed in E. coli. Exon 1 of the HTT huntingtin (NCBI Gene ID 3064) was expressed as a fusion protein with (i) TRX only (MurTRX); (ii) 16Q (Mur16); and (iii) 46Q (Mur46). MurTRX, Mur16, and Mur46 were produced and purified by HIS-tag and size exclusion chromatography (SEC). Colonies were picked and 5 mL cultures were grown. Cells were induced at Optical Density (OD) 0.7-0.8 for four hours at 37°C. 1mL of the post-induction culture was spun down and the pellet resuspended in lysis buffer. This was spun down at 10,000 RPM for 10 minutes and the soluble fraction loaded on a 4-20% polyacrylamide gel (NuSep). An anti-His western blot was performed to detect protein expression. An ELISA assay was performed to confirm the expression and detection of the proteins. MurTRX, Mur16 and Mur46 are recognized by anti-His antibodies. Mur16 and Mur46 are recognized by an mHTT antibody.
Protein Purification. Proteins were loaded onto IMAC resin (ThermoFisher Scientific HisPur) and eluted in 200 mM imidazole, purified by gel filtration FPLC (HiLoad superdex-200, 26/60; GE Life sciences) and concentrated using 3 kDa cut off Vivaspin 20 PES centrifugal concentrators (Sartorius AG). Proteins were biotinylated using a 1:0.5 molar ratio of EZ-link™ Sulfo-NHS-LC-LC-Biotin (ThermoFisher Scientific) and thoroughly dialysed against PBS prior to Biacore coupling.
Neutravidin CM5 amine coupling immobilisation. It was determined that the ligand density of streptavidin found on a standard SA Chip was not high enough for this project (~3,000 RU). It was decided to make a customized biotin capture chip using NeutrAvidin as the capture ligand as opposed to streptavidin. To obtain high enough ligand density on the sensor chip surface, NeutrAvidin (Thermoscientific Pierce, Waltham, MA USA) was immobilised onto all four sensor chip surfaces to a level of approximately 20,000 RU.
Capture of the Mur ligands to NeutrAvidin. Biotinylated proteins were then diluted in HBS EP+ running buffer and captured on the sensor chip surface. Mur46 was coupled to the sensor chip first as this is the major ligand of interest. After saturating FC3 with Mur46, target response levels for Mur16 and MurTRX were calculated based on their respective molecular weights.
Biacore Screening. Briefly, compounds were diluted in 100% DMSO to a final concentration of 100 mM. These were diluted wherever possible to 1 mM, 0.1 mM and 0.01 mM in 10% DMSO in PBS+ running buffer. The entire assay was run in 10% DMSO in the Biacore T200 (GE, Little Chalfont in the United Kingdom) and an appropriate solvent correction window was selected for the assay. Samples were run by identity, from the lowest to the highest concentration with one concentration per cycle. The assay was run with a 10 Hz data collection frequency in a multi (4-1, 3-1, 2-1) configuration and compounds were in contact with the ligands for 60 seconds. Blank samples (required for double referencing) were passed over the sensor surface every 21 cycles in triplicates, as was the positive MAb control and solvent correction.
PC12 cell assay. The Htt14A2.6 PC12 line was generated and propagated as described previously42. In brief, Htt14A2.6 was maintained in Dulbecco's modified eagle medium-GlutaMAX (Life Technologies, Grand Island, NY, USA) supplemented with horse serum (10%), heat-inactivated fetal bovine serum (5%), Pen Strip (1%), Zeocin (200 μg/ml), and G418 (50 μg/ml). For cell assays, cells in the log phase of growth were replated in 24-well culture plates around 50–60% confluency, grown overnight and induced for the expression of mHTTex1-GFP for 8 h with 2 μM ponasterone A (Santa Cruz Biotechnologies). Filter retardation was performed according to published methods43.
R6/2 Study. All animal experiments were carried out according to the National Institute of Health (NIH) guidelines for the care and use of laboratory animals, and approved by the National Animal Experiment Board, Finland. The animal facility at site is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC), International.
Husbandry. All mice were housed in groups of up to 5 per cage, in a temperature (22 ±1°C) and humidity (30-70%) controlled environment with a normal light-dark cycle (7:00 - 20:00 h light). All mice were housed in cages with clean bedding covering the ground that was changed as frequently as needed, at least once a week to provide the animals with dry bedding. This basic environment was enriched with the addition of a red mouse igloo (K3327), shredded paper and a wooden chewing stick. Food and water were available ad libitum to the mice in their home cages. The water spouts were fitted with extensions to allow mice to easily access from floor level. Each cage contained mice of only one treatment group. In each cage was included also a wildtype mouse to provide normal social stimulation to R6/2 mice.
Breeding and Weaning. 10 female R6/2 mice and 5 female wildtype littermate control mice (F1 generation) were bred by Charles River Laboratories, Sulzfeld, Germany by mating (F0 generation) WT males (C57BL/6J; systematically re-infused with pedigreed JAX mice, stock 000664) with ovarian transferred (OT) TG females (JAX, stock 006494). After weaning mice were sent from Germany to Charles River, Kuopio, Finland at an age of 3 weeks. Following genotyping and acclimation, the mice were enrolled in the study.
Genotyping. Mice were ear marked at the age of 15 - 21 days and tail samples were collected at the same time for genotyping with PCR. Genotyping was performed at Charles River Discovery Services, Kuopio. DNA was isolated from tail or ear samples with Phire Animal Tissue Direct PCR-kit (Thermo Scientific, ref. F140WH) according to the kit’s instructions. Then 1 µl of DNA was multiplied in the PCR reaction with mouse specific (Gapdh) primers and human specific (Htt) primers. Primers sequences and final working concentrations are listed below. After the PCR, multiplied DNA was separated by agarose gel electrophoresis. The expected products were 272bp (human specific product) and 372bp (mouse specific product). Thus wildtype (WT) mouse has one 372 bp band while transgenic (TG) mouse has both the 272bp and 372bp bands.
Human specific 25 pmol/µl, (5'-3'): TCATCAGCTTTTCCAGGGTCGCCAT (SEQ ID NO: 8)
Human specific 25 pmol/µl, (5'-3'): CGCAGGCTAGGGCTGTCAATCATGCT (SEQ ID NO: 9)
Mouse specific 5 pmol/µl, (5’-3’): ACTCCACTCACGGCAAATTCAACGGCAC(SEQ ID NO: 10)
Mouse specific 5 pmol/µl, (5’-3’): GGTCATGAGCCCTTCCACAATGCCAAAG (SEQ ID NO: 11)
Study design. 10 female R6/2 mice and 10 female wildtype littermate control mice were bred at Charles River, Germany. Total of 20 female and male R6/2 mice and 10 female and male WT littermate control mice were used in the study. The mice were genotyped, and the R6/2 mice were divided into different treatment groups based on their litter and baseline body weight. The treatment with Vehicle or GLYN122 (33 mg/kg; 5 ml/kg, i.p. QD) was started at four weeks of age after the baseline behavioral tests. The experimental groups were: (i) Five wildtype mice treated with Vehicle (5 ml/kg, i.p., QD) starting at four weeks of age; (ii) Five R6/2 mice treated with Vehicle (5 ml/kg, i.p., QD) starting at four weeks of age; and (iii) Five R6/2 mice treated with GLYN122 (33 mg/kg; 5 ml/kg, i.p. QD) starting at four weeks of age.
Body weight. Body weights were measured at three weeks of age and twice a week until the end of the study. Motor function testing using rotarod were commenced at four weeks (pre-treatment baseline) and continued at 6, 8 and 10 weeks of age, accompanied with grip strength at 4 (pre-treatment baseline), 10 and 12 weeks of age. At the end point of 12 weeks of age the mice were subjected to tissue collection.
Rotarod test. Rotarod was performed at 4 (pre-treatment baseline), 6, 8 and 10 weeks of age. Each testing day included a training trial of 5 min at 4 RPM on the Rotarod apparatus (AccuScan Instruments, Columbus, USA). 30 min later, the animals were tested for 3 consecutive accelerating trials of 6 min with the speed changing from 0 to 40 RPM over 360 s and with an inter-trial interval of at least 30 min. The latency to fall from the rod was recorded. Mice remaining on the rod for more than 360 s were removed and their time scored as 360 s.
Grip Strength. Mice were tested at 4 (pre-treatment baseline), 10 and 12 weeks of age. Mice were taken to the experimental room and, one at a time, were placed on the grip strength apparatus (San Diego Instruments, San Diego, USA) in such a way that the animal grabbed a small mesh grip with its forepaws. The entire apparatus was placed on a table top for testing. Animals were lowered to the platform and then slowly pulled away from the handle by the tail until the animal released the handle. The equipment automatically measures the strength of the animal’s grip in grams. Five scores were recorded per animal in consecutive sequence, and the average of three best scores for each animal was used for the results. Mice were returned to their home cage after testing.
End-Point and Tissue Processing. Approximately one hour after the last dose the mice were terminally anesthetized with pentobarbital. Thereafter the mice were transcardially perfused with ice cold heparinized saline (Heparin 2.5 IU/ml) 25 ml)), followed by perfusion with ice cold 4 % PFA (80 ml). The brains were fixed by immersion in 4 % paraformaldehyde for minimum of 24 h after which brain samples were cryoprotected by 30 % sucrose solution for 72 h after which the brain samples were frozen in liquid nitrogen. Frozen brain specimens were stored at -80 ºC.
Preparation of Immunohistology slides. Brain tissue collected within the animal study was prepared for IHC studies by Charles River staff. The brains were fixed by immersion in 4% paraformaldehyde for at least 24 h after which brain samples were cryoprotected by 30% sucrose solution for 72 h. Finally, the brain samples were flash-frozen in liquid nitrogen and stored at -80 ºC. The brain samples were cut using a microtome cryostat system, producing coronal brain tissue sections of 40 µm thickness. Those were mounted on individual adhesive-coated microscope glass slides with frosted ends.
Staining. Goat anti-Rabbit IgG (H+L) Alexa Fluor 647 (1:500, ThermoFisher Invitrogen) was used as a secondary antibody for binding to CBP antibody (Supplementary Table S1). DAPI (Sigma-Aldrich) was used to identify the nuclei. Goat anti-Mouse IgG (H+L), Alexa Fluor 488 (1:500, A-11034, ThermoFisher Invitrogen) was used as a secondary antibody for binding to EM48 antibody. Mouse anti-human-mHTT (EM48, 1:500, Sigma-Aldrich), was used to stain huntingtin. Rabbit anti-CBP (1:100, Sigma-Aldrich) was used a primary antibody to stain CBP. The blocking buffer was freshly prepared and consisted of PBS (Sigma-Aldrich) with 5% normal goat serum (NGS), 0.2% BSA, 0.2% lysine and 0.2% glycine. Samples were covered with 750 µL of blocking buffer per sealing chamber and incubated at 4 °C for 24 h. Subsequently, on day two samples were washed three times 10 min each in PBS, before working dilutions of primary antibodies were applied in 750 µL primary antibody buffer per chamber. The primary buffer consisted of PBS with 2% BSA/0.3% Triton X-100 (Sigma-Aldrich) and 0.02% NaN3 as preservative agent. The samples were incubated with primary antibodies at 4 °C for 73 h. On day five, samples were washed like described previously and then incubated at 4 °C for 24 h with secondary antibody in 750 µL secondary antibody buffer at 1:500 working dilutions per chamber. The secondary buffer consisted of PBS with 3% NGS/0.3% Triton X-100/0.02% NaN3. On day six, samples were washed and then incubated with DAPI containing mounting medium Fluoroshield (Sigma-Aldrich), in order to counterstain the nuclei and preserve the fluorescence. Therefore, one drop of mounting medium (Dako) was added per tissue section and the sample carefully coverslipped avoiding introduction of air bubbles. The samples were stored for 24 hours at room temperature shielded from light before being stored at 4 °C until imaging.
Fluorescence Imaging. Imaging sessions were performed on a confocal microscope system (Carl Zeiss Microscopy) equipped with a dual spinning disk unit (Yokogawa). All components of the imaging system were controlled via the ZEN 2 software suite (Carl Zeiss Microscopy). The laser lines used were 405 nm, 488 nm and 639 nm to excite DAPI or the respective fluorophores. The fluorescence images obtained of the immunofluorescence labelled tissue sections were quantified with the help of the “Image Processing and Analysis in Java” or short ImageJ software distributed under the GNU General Public License by the NIH44, i.e. the edition used was the Fiji distribution45. Nested t test was performed to calculate the significance amongst repeated measurements using GraphPad Prism software.
Explorative Mode of Action Study. STHdh 111/111 primary mouse neurons treated with GLYN122 at the indicated concentration for 48 h (Supplementary Table S2). Cells are washed with PBS, trypsinised, and gathered in a 15 ml falcon. After a short spin down for 3 min at 300 g, the cell pellet was washed with PBS once, before it was frozen down in liquid nitrogen. RNA was extracted with a standard protocol. The nCounter® Mouse Neuropathology Panel (Nanostring, Seattle) was used for gene expression analysis according to the manufacturer’s protocols.
Statistical Analysis. Statistical analyses were performed using GraphPad Prism statistical software version 8.1.2 from 2019. Where not noted differently, significance levels were determined using unpaired Student’s t-test for all analyses, without assuming a consistent SD between sample populations. Where suitable, significance was reported as multiplicity adjusted p-value to account for multiple comparisons. The level for significance was set to alpha = 0.05 (95% confidence interval).