Investigation in the cannabigerol derivative VCE-003.2 as a disease-modifying agent in a mouse model of experimental synucleinopathy

Viral vector production

A recombinant AAV vector serotype 2/9 expressing the SNCA gene with A53T mutation driven by the human synapsin promoter was produced by Vector Builder (www.vectorbuilder.com; catalog No. P211024-1010fpa; lot No. 211221AAVJ03). The virus was formulated in PBS buffer (pH 7.4) supplemented with 200 nM NaCl and 0.001% pluronic F-68. Obtained vector concentration was 9.62 × 10¹³ GC/ml. Virus purity was determined by SDS-PAGE followed by silver staining, resulting in > 80% pure. Plasmid map for pAAV-synapsin-SynA53T and sequence are provided in Supplementary File 1.

Animals and surgical lesions

Male C57BL/6 mice were housed in our animal facilities (CAI-Animalario, Faculty of Medicine, Complutense University, ref. ES280790000086) in a room with a controlled photoperiod (08:00–20:00 light) and temperature (22 ± 1ºC). They had free access to standard food and water and were used at adult age (6–8-month-old; weight = 28–33 g). All animal experiments were conducted by researchers having the necessary accreditation and following local, national and European guidelines (directive 2010/63/EU), as well as conformed to ARRIVE guidelines and were approved by our local Ethical Committee for Animal Experimentation (ref: PROEX 201.8/22).

Mice were used for the in vivo overexpression of human α-synuclein with the A53T mutation using recombinant AAV9 encoding for this gene under the control of human synapsin 1 promoter. To this purpose, mice were anaesthetized (ketamine 40 mg/kg + xylazine 4 mg/kg, i.p.) and subjected to unilateral injections of viral particles (AAV9-SynA53T) into the SNpc of the right hemisphere following the procedure described in Lastres-Becker et al. [52]. In brief, 1 µl of a viral suspension containing 3.33 × 10¹² GC/ml (selected from a preliminary concentration-response experiment using a range for the viral suspension of 3.33 × 10¹² − 3.33 × 10¹³ GC/ml) was unilaterally injected at predefined stereotaxic coordinates: -2.5 mm posterior to bregma, − 1.4 mm lateral to midline and − 4.5 mm ventral to dura. The volume of 1 µl was pressure-injected in pulses of 0.2 µl/min and, once completed, the needle was left in place for 5 min before withdrawal to minimize reflux of viral suspension through the injection tract. Control animals were sham-operated and injected with 1 µl of a viral suspension of an empty AAV9 (AAV9-null) containing 1.07 × 10¹² GC/ml using the same coordinates. The lesions were generated using unilateral inoculation, so that contralateral structures serve as controls for the different analyses (see Fig. 1 for a diagram showing the schedule used).

Pharmacological treatments and sampling

After the application of AAV9-SynA53T or control AAV9-null vectors, animals were distributed into 3 groups in each experiment, and were subjected to a daily treatment with VCE-003.2 (20 mg/kg), or vehicle (sesame oil) by oral administration (see Fig. 1 for a diagram showing the schedule used). The number of animals in each experimental group was: (i) vehicle-treated sham-operated mice: 4–6; (ii) vehicle-treated AAV9-SynA53T mice: 8–10; and (iii) VCE-003.2-treated AAV9-SynA53T mice: 8–10 (ranges are due to elimination of potential outliers in those cases that the statistical analysis recommended to do it). The first administration, in all cases, was done approximately 24 h after the lesion and the treatment was prolonged for two weeks. At the end of the treatment (24 h after the last injection), they were analysed in different behavioural tests just before being killed by rapid and careful decapitation, and their brains rapidly removed and divided coronally in two parts. The anterior halves were used to dissect the striatum. Tissues were rapidly frozen by immersion in cold 2-methylbutane and stored at -80ºC for RNA sequencing analysis (RNAseq). The posterior halves containing the midbrains were fixed for one day at 4 °C in fresh 4% paraformaldehyde prepared in 0.1 M PBS, pH 7.4. Samples were cryoprotected by immersion in a 30% sucrose solution for a further day, and finally stored at -80 °C for immunohistochemical analysis in the SN.

Pharmacokinetic and ADMET analysis

Pharmacokinetic for oral VCE-003.2 in mice and in vitro off-target binding, transporter interactions, micronucleus assay, and comparative metabolism assays for VCE-003.2 in hepatocytes from different species were performed by Charles River (Edinburgh, UK) and Eurofins Panlabs Discovery Services (New Taipei City, TW), using protocols summarized in Supplementary File 2.

Behavioural tests

Cylinder rearing test (CRT) Given that the lesions were unilateral, this test attempts to quantify the degree of forepaw (ipsilateral, contralateral, or both) preference for wall contacts in rearing movements after placing the mouse in a methacrylate transparent cylinder (diameter: 15.5 cm; height: 12.7 cm; Fleming et al., [33]. Each score was calculated from a 3 min trial with a minimum of 4 wall contacts. See Supplementary File 3 for an example of video recording in the three experimental groups.

Elevated-body swing test (EBST) Mice were placed head-downward hanging by their tail in a vertical axis (about 15 cm from the surface) and recorded for 60 s [8, 11]. A swing was recorded whenever the animal moved its head out of the vertical axis to either side, whereas for the next swing to be counted, the animal must have returned to the vertical position first. See Supplementary File 4 for an example of video recording in the three experimental groups.

Computer-aided Actimeter For the analysis of motor activity, we used a computer-aided actimeter (Actitrack, Panlab, Barcelona, Spain). This apparatus consisted of a 45 × 45 cm area, with infra-red beams all around, spaced 2.5 cm, coupled to a computerized control unit that analyzes the following parameters: (i) distance run in the actimeter (ambulation); (ii) time spent in inactivity; (iii) frequency of vertical activity (rearing); (iv) mean and maximal velocity developed during the running; and (v) time spent in fast (> 5 cm/s) and slow (< 5 cm/s) movements. Animals remained for a period of 10 min in the actimeter, but measurements were only recorded during the final 5 min (first 5 min was used only for animal acclimation). These data are presented as Supplementary File 5.

Immunohistochemical procedures

Brains were sliced in coronal sections (containing the SN) in a cryostat (30 µm thick) and collected in antifreeze solution (glycerol/ethylene glycol/PBS, 2:3:5) and stored at -20ºC until being used. Brain sections were mounted on gelatine-coated slides, and once adhered, washed in 0.1 M potassium PBS (KPBS) at pH 7.4. Then, endogenous peroxidase was blocked by 30 min incubation at room temperature in peroxidase blocking solution (Dako Cytomation, Glostrup, Denmark). After several washes with KPBS, sections were incubated overnight at 4ºC with the following antibodies: (i) monoclonal mouse anti-human α-synuclein (sc-53955; Santa Cruz Biotechnology, Dallas, TX, USA) used at 1:2000; (ii) polyclonal rabbit anti-mouse tyrosine hydroxylase (TH) (AB152; Chemicon-Millipore, Temecula, CA, USA) used at 1:200; (iii) polyclonal rat anti-mouse CD68 antibody (MCA-1957; AbD Serotec, Oxford, UK) used at 1:200; or (iv) polyclonal rabbit anti-mouse GFAP antibody (Z0334; Dako Cytomation, Glostrup, Denmark) used at 1:200. Dilutions were carried out in KPBS containing 2% bovine serum albumin and 0.1% Triton X-100 (Sigma Chem., Madrid, Spain). After incubation, sections were washed in KPBS, followed by 1 hour incubation at room temperature with the corresponding biotinylated secondary antibody (1:1000 for α-synuclein; 1:200 for the rest) (Vector Laboratories, Burlingame, CA, USA) for 1 hour at room temperature. Avidin-biotin complex (Vector Laboratories, Burlingame, CA, USA) and 3,3’-diaminobenzidine substrate–chromogen system (Dako Cytomation, Glostrup, Denmark) were used to obtain a visible reaction product. Negative control sections were obtained using the same protocol with omission of the primary antibody. A Leica DMRB microscope and a DFC300FX camera (Leica, Wetzlar, Germany) were used for the observation and photography of the slides, respectively. For quantification of the intensity of α-synuclein, TH, CD68 or GFAP immunostaining in the SN (at both ipsilateral and contralateral sides), we used the NIH Image Processing and Analysis software (ImageJ; NIH, Bethesda, MD, USA) using 4–5 sections, separated approximately by 200 μm, and observed with 5x-20x objectives depending on the immunostaining under quantification. In all sections, the same area of the SNpc was analysed. Analyses were always conducted by experimenters who were blinded to all animal characteristics. Data were expressed as percentage of immunostaining intensity in the ipsilateral (lesioned) side over the contralateral (non-lesioned) side. In addition, in the case of CD68 immunostaining, images at higher magnification were also used for a qualitative analysis of certain morphological characteristics (cell body volume; number, thickness and length of cell processes) that may reflect the state of quiescence (low cell body volume and tiny and long processes) or activation (amoeboid-like morphology: high cell body volume and thick and short processes) of these cells.

In a further analysis, a deep-learning dedicated algorithm was prepared with Aiforia^® [61], validated and released (resulting in an error of 4.82%) for quantifying the number of TH-positive neurons in TH-stained, equally-spaced coronal sections covering the whole rostrocaudal extent of the SNpc. The dedicated algorithm was customized to count all positive neurons, independently of the intensity of staining. Sections were scanned at 20x in a slide scanner (Aperio CS2; Leica, Wetzlar, Germany) and uploaded to the Aiforia^® cloud. Next, the boundaries of each selected ROI were outlined at low magnification and the dedicated algorithm was then used as a template quantifying the TH-positive neurons. Details on this analysis have been previously published [21, 63].

Transcriptomic analysis

RNA isolation and sequencing RNA was isolated from the striatum samples using the RNeasy Lipid Tissue Mini kit (Qiagen, Barcelona, Spain) following the manufacturer’s instructions, including the optional DNase I step. Once RNA was isolated, samples were stored at − 80 °C following library preparation. RNA concentration and RNA integrity number (RIN) was assessed using Nanodrop and Bioanalyzer (Centre for Genomic Regulation, Barcelona, Spain). Library preparation and sequencing was outsourced to the Centre for Genomic Regulation (CRG, Barcelona, Spain). Briefly, libraries were prepared using the TruSeq mRNA Library prep following the manufacturer’s instructions. mRNA selection was performed using poly(A)-mRNA selection strategies. Libraries were prepared and sequenced by 50 bp sequencing using Illumina NextSeq 2000 run system (Illumina, Inc., San Diego, CA, USA).

RNA-seq data processing and normalization Raw reads were aligned against the GRCm38 genome (v84) from ENSEMBL with HISAT2 (v2.1.0) [47, 59]. This resulted in a ~ 99% alignment rate with 60% of reads aligned to exonic regions. The count matrix was obtained from the alignments using the corresponding ENSEMBL annotation file with featureCounts (v2.0.0) [54]. Then, counts per million (CPM) were calculated using CPM function from the EdgeR packing in R [53]. Lowly expressed genes were filtered out, which were defined as having less than one CPM in at least 25% of the samples leading to a total of 13,980 genes. Principal Component Analysis (PCA) was used to identify sources of variation and outliers. As samples were sequenced in two independent batches, we regressed out sequencing batch to control variation. The final dataset was composed by 17 samples (6 from vehicle-treated sham-operated mice, 5 vehicle-treated AAV9-SynA53T-injected mice, and 6 VCE-003.2-treated AAV9-SynA53T-injected mice).

Differential expression analysis Differential expression analysis was performed using the R package limma [64]. EdgeR was used for data normalization using TMM values and followed by voom transformation. P-values were corrected using the Benjamini-Hochberg false discovery rate (FDR).

Pathway enrichment analysis. We used Gene Set Enrichment Analysis (GSEA) focusing on Hallmark gene sets [72]. To perform pathway analysis, ranked DEGs in terms of t-value were used as input. We used the fGSEA R-package [49]. Only pathways with adjusted P-value < 0.05 were considered.

Transcription factor enrichment analysis For exploring if genes were enriched for specific transcription factors, we used DoRothEA [37, 41]. Regulons were filtered at confidence levels “A” and “B”. Regulon enrichment was runned using fgsea with preranked data in terms of t value. Only regulons with adjusted P-value < 0.05 were considered.

Statistics

Data were assessed using one-way ANOVA, followed by the Tukey test using GraphPad Prism, version 8.00 for Windows (GraphPad Software, San Diego, CA, USA). A p value lower than 0.05 was used as the limit for statistical significance. The sample sizes in the different experimental groups were always ≥ 5.

Pharmacokinetic, metabolic, and toxicological properties of VCE-003.2

We have previously shown that VCE-003.2 crossed the BBB after oral administration in rats [1]. Now, we extended our PK studies in mice, found that plasma VCE-003.2 levels peaked at 4 h (Tmax) and then slowly declined to basal levels at 24 h. Oral VCE-003.2 resulted in 69% bioavailability (F) in mice (Supplementary Table 1), which was significantly higher than the one observed in rats (F = 13,77%) [1]. Moreover, VCE-003.2 did not significantly inhibit the activity of relevant cytochrome P450 isoforms, did not inhibit hERG channel activity and it was not genotoxic as assessed by AMES assays [1]. To complement these studies, we interrogated for potential off-targets and genotoxicity of VCE-003.2. Using radioligand binding assays, we explored the interaction with a panel of 172 receptors and channels, and we found that VCE-003.2 did not show a significant binging activity to these targets (Supplementary File 2). The potential genotoxicity of VCE-003.2 was investigated using micronucleus assays either in the presence or absence of a metabolic activation system (S9-mix), in cultured human lymphocytes (in vitro) and in rat bone marrow cells (in vivo). We observed that VCE-003.2 is not clastogenic or aneugenic either in human lymphocytes or in the bone marrow micronucleus test of male rats up to a dose of 2000 mg/kg (Supplementary File 2).

We also investigated the metabolic stability and metabolite profiles of VCE-003.2 using suspensions of cryopreserved hepatocytes from CD-1 mouse, Wistar rat, Beagle dog, Göttingen minipig, cynomolgus monkey and human. The rates of metabolism and the metabolite profiles were compared across the species tested. Twenty-six metabolites of VCE-003.2 were found in the hepatocyte incubations of the six species investigated. Based on the MS peak intensities, no human specific metabolites were observed and five major metabolites of VCE-003.2 were detected in at least four of the six species tested. Metabolic reactions observed included oxidations, the loss of C₃H₆ in combination with two oxidations, and a desaturation in combination with one or two oxidations. The postulated major metabolic pathway VCE-003.2 was also predicted (Supplementary File 2).

Finally, we explored the in vitro interaction of VCE-003.2 with human BCRP, BSEP and MDR1 efflux (ABC) transporters, and with human solute carrier (SLC) transporters MATE1, MATE2-K, OAT1, OAT3, OATP1B1, OATP1B3, OCT1 and OCT2. VCE-003.2 did not inhibit SLC transported and up to 22.5 µM showed inhibitory activity of the ABC transporters (BCR = 89% inhibition, BSEP = less than 50% inhibition; MDR1 = 81% inhibition) (Supplementary File 2). Altogether, VCE-003.2 represents an excellent ligand for studying the relevance of PPAR-γ activation in experimental models of PD, as has been demonstrated in previous studies [12, 15, 36] and situates this compound in a clear advantage for future translational development.

Behavioural readouts of treatment with VCE-003.2 in mice injected with AAV9-SynA53T

To explore the neuroprotective potential of an oral administration of VCE-003.2 against the overexpression of neuronal (A53T)α-synuclein, we first analysed the motor performance of these animals using CRT and EBTS. These motor tests discriminate the unilateral hemiparesis and were assessed 24 h after the last VCE-003.2 administration, avoiding any acute effect of the drug. CRT showed a preference for the ipsilateral right paw (controlled by the non-lesioned contralateral SN) in mice-injected with AAV9-SynA53T, which was prevented by the treatment with VCE-003.2 (F(2,17) = 11.50, p < 0.001; Fig. 2 and Supplementary File 3). Also, we noticed a general mobility increase, reflected in the total number of rearings, in the VCE-003.2-treated animals, which tended to be elevated with respect to both vehicle-treated sham-operated and AAV9-SynA53T-injected mice (F(2,18) = 2.99, p = 0.075; Fig. 2).

Next, we performed EBST, which is useful to detect the side more frequent in the torsion movements depending on the lesioned side. We found that vehicle-treated AAV9-SynA53T-injected mice tended to show a preference for contralateral turns, which was not attenuated by the treatment with VCE-003.2, even it tended to be enhanced (F(2,19) = 1.49, ns; Fig. 2 and Supplementary File 4). However, as in CRT, we observed a higher number of total turns (irrespective of the side) shown during the test performance (F(2,19) = 5.13, p < 0.05; Fig. 2), demonstrating a greater activity in the treated animals.

Such elevated motor activity elicited by VCE-003.2 in AAV9-SynA53T-injected mice in comparison with the other two groups was also visible in the data obtained in a computer-aided actimeter (actitrack), with statistically significant increases in locomotor parameters such as general activity (F(2,19) = 4.39, p < 0.05), ambulation (F(2,19) = 4.55, p < 0.05), mean and maximal velocities (F(2,19) = 4.14, p < 0.05 and F(2,19) = 5.53, p < 0.05, respectively), and frequency of slow movements (F(2,19) = 5.82, p < 0.01). This increase remained as a mere numerical trend in parameters as vertical activity (F(2,19) = 1.56, p = 0.24) and frequency of fast movements (F(2,19) = 2.65, p = 0.096), whereas the opposite (lower values) was detected, as expected, for resting time (F(2,19) = 5.13, p < 0.05), then confirming that VCE-003.2-treated AAV9-SynA53T-injected mice frequently exhibited a greater motor activity (see these data in the Supplementary File 5).

Histopathological readouts of treatment with VCE-003.2 in mice injected with AAV9-SynA53T

Immunohistochemical analysis demonstrated an AAV9-driven overexpression of human α-synuclein all over the midbrain in the lesioned mice, in particular in the SNpc (F(2,18) = 19.42, p < 0.0001; see Supplementary File 6) [18, 66]. Furthermore, the enhanced neuronal expression of mutated α-synuclein was associated with a significant reduction (around 35%) of TH immunoreactivity in the SNpc (F(2,21) = 20.67, p < 0.0001; Fig. 3). Treatment with VCE-003.2 of mice injected with AAV9-SynA53T did not alter the immunoreactivity levels of α-synuclein (Supplementary File 6). However, it prevented the reduction in TH immunoreactivity to the same value detected in sham-operated mice (Fig. 3). This observation supports the idea that VCE-003.2 treatment preserves the integrity of TH-positive neurons in the SNpc against the enhanced neuronal expression of mutated α-synuclein, despite our analysis only served to measure immunoreactivity levels. In a further analysis, using a stereological-like analysis based on a deep-learning dedicated algorithm (Aiforia^®), we could confirm that the reduction in TH measured in the SNpc of mice injected with AAV9-SynA53T, as well as the recovery detected after the treatment with VCE-003.2, reflected both a loss and a preservation in the number of TH-positive neurons, respectively (F(2,21) = 6.72, p < 0.01; see Supplementary File 7). This preservation may contribute to the behavioral improvements indicated in the previous subsection.

The neuronal overexpression of mutated α-synuclein in the SN also induced robust glial reactivity in both microglia (CD68: F(2,17) = 27.58, p < 0.0001; Fig. 4), and astrocytes (GFAP: F(2,21) = 26.17, p < 0.0001; Fig. 4 and representative images in Supplementary File 8). In addition, a simple morphological analysis of CD68-labelled cells revealed a higher presence of cells with apparently amoeboid aspect (greater cell body volume and thick and short cells branches) in the SN of vehicle-treated AAV9-SynA53T mice compared to vehicle-treated sham-operated mice (Supplementary File 9), which is compatible with an activated state of these cells. VCE-003.2 treatment attenuated both responses, with a statistically significant effect in CD68 immunostaining (Fig. 4) and a numerical trend in GFAP immunostaining (Fig. 4 and representative images in Supplementary File 8). VCE-003.2 treatment of AAV9-SynA53T mice also reduced the number of CD68-labelled cells with apparently activated phenotype to a more quiescent state (smaller cell bodies and more, tiny and longer processes) (Supplementary File 9).

Transcriptomic readouts of treatment with VCE-003.2 in mice injected with AAV9-SynA53T

Having established the neuroprotective efficacy of VCE-003.2 in this α-synuclein model of PD, and recognizing that functional recovery likely involves the survival of nigrostriatal terminals arising the striatum, we aimed sought to delve into downstream cellular pathways through transcriptomic analysis of this specific brain region. To this end, we carried out RNA-Seq analysis in the right (lesioned) striatum of mice from the three different experimental groups (Supplementary Table 2). We first examined the impact of α-synuclein lesion on gene expression, comparing the AAV9-SynA53T-injected mice versus sham-operated animals, all treated with vehicle. We did observe a strong transcriptomic dysregulation with 1055 differentially expressed genes (DEGs) at FDR < 0.05 (50.8% up-regulated and 49.2% down-regulated) (Fig. 5, left panel; Supplementary Table 3). The oral administration of VCE-003.2 in the AAV9-SynA53T-injected mice produced a much more modest effect, with only 14 genes (Irf2, Ifitm3, Ifit1, Ifit3, Oasl2, H2-Q6, Mir6236, Lgals9, Samd9, B2m, Oas2, H2-K1, H2-D1 and ligp1) becoming significantly altered from their expression in vehicle-treated AAV9-SynA53T-injected mice, most of them up-regulated, with 9 additional genes being close to be statistically significant (Fig. 5, right panel; Supplementary Table 4). Interestingly, these 23 genes appear to be part of a network of proteins related to interferon biology, as confirmed through Local Network Cluster Analysis using STRING (data not shown). The treatment with VCE-003.2 in AAV9-SynA53T-injected mice did not alter the expression of genes encoding for cannabinoid receptors or for those enzymes involved in the endocannabinoid synthesis and degradation in comparison with those animals treated with vehicle, despite some of them did result altered by the local injection of AAV9-SynA53T into the SN. This was the case of the Cnr1 (down-regulated: logFC=-1.068; FDR = 0.007) and the Gpr55 (up-regulated: logFC = + 1.14; FDR = 0.006) encoding the CB₁ and GPR55 receptors, respectively, as well as the Faah gene (down-regulated: logFC=-0.47; FDR = 0.029) encoding the FAAH enzyme (Supplementary Tables 3 and 4). As regards to some PD-related proteins (e.g., α-synuclein, parkin, leucine-rich repeat kinase-2 (LRRK2)), our data indicated an important and expected up-regulation of Snca gene (logFC = 0.48; FDR = 0.0058) by the local injection of AAV9-SynA53T into the SN, but not in Prkn or Lrrk2, and no effect of the VCE-003.2 in any of these three genes (Supplementary Tables 3 and 4).

To understand the biological implications of the transcriptomic dysregulation, we performed GSEA using Hallmark Pathways from Gene Ontology. Pathway analysis revealed that DEGs affected by mutated α-synuclein overexpression were enriched for pathological hallmarks (see Supplementary Table 5) such as an impaired mitochondrial function affecting the oxygen species pathway and the oxidative phosphorylation (Fig. 6 and Supplementary Table 5). We also observed down-regulation in pathways associated with autophagy and lysosomal function (e.g., mTORC1 signaling), as well as lipid metabolism (e.g., fatty acid metabolism, adipogenesis) and the interferon response (Fig. 6 and Supplementary Table 5). In contrast, treatment with VCE-003.2 led to a significant increase in the expression of genes related to the interferon signaling, which was the opposite response when compared with the one detected after the delivery of AAV9-SynA53T (Fig. 6 and Supplementary Table5).

To determine whether the observed transcriptomic dysregulation was orchestrated by specific transcription factors, we conducted regulon enrichment analysis using DoRothEA [37, 41]. In a first set of analysis, we detected a significant enrichment in several regulons, which were down-regulated in vehicle-treated AAV9-SynA53T-injected mice with respect to vehicle-treated sham-operated animals (Fig. 7; see Supplementary Table 6). Among them was the finger family transcription factor Specificity protein-1 (Sp1), the tumorigenic factor Forkhead box I2 (FoxI2), and others such as Erythroblast transformation specific-1 (Ets1) and Hypoxia-inducible factor-1a (Hif1a) (Fig. 7 and Supplementary Table 6). Other down-regulated regulon of interest is Interferon regulatory factor-1 (Irf1), along with other transcription factors such as Signal transducer and activator of transcription-1 (Stat1) and − 2 (Stat2), the latter also down-regulated (Fig. 7 and Supplementary Table 6). When we compared VCE-003.2-treated AAV9-SynA53T-injected mice with the same animals treated with vehicle, we observed up-regulation for Stat1, Stat2, Irf1 and Irf9 (Fig. 7 and Supplementary Table 6). This suggests that VCE-003.2 may reverse the transcriptomic deficiencies in these factors induced by the neuronal overexpression of mutated α-synuclein (Fig. 7). Whether this mechanism of action is a key event in the ability of VCE-003.2 to preserve the structure and functionality of the nigrostriatal projection in the basal ganglia, or whether a mere consequence of such neuroprotective effect exerted by VCE-003.2 will require further research. Nevertheless, this finding is a novel and significant result of our transcriptomic analysis. All raw RNA-Seq data have been uploaded to ArrayExpress repository to make them available for further analysis (accession number: E-MTAB-13768).