Proteomics of Patient-Derived Striatal Medium Spiny Neurons in Multiple System Atrophy

Abstract

The rare and rapidly progressive neurodegenerative disease multiple system atrophy (MSA) mainly affects the striatum and other subcortical brain regions. In this atypical Parkinsonian syndrome, the protein alpha-synuclein aggregates and misfolds in neurons as well as glial cells and is released in elevated amounts by hypoexcitable neurons. Mitochondrial dysregulation affects the biosynthesis of coenzyme Q10 and the activity of the respiratory chain, as shown in an induced pluripotent stem cell (iPSC) model. Proteome studies of cerebrospinal fluid and brain tissue from MSA patients yielded inconsistent results regarding possible protein changes due to small and combined groups of atypical Parkinsonian syndromes. In this study, we analysed the cellular proteome of MSA patient-derived striatal GABAergic medium spiny neurons. We observed 25 significantly upregulated and 16 significantly downregulated proteins in MSA cell lines compared to matched healthy controls. Various protein types involved in diverse molecular functions and cellular processes emphasise the multifaceted pathomechanisms of MSA. These data could contribute to the development of novel disease-modifying treatment strategies for MSA patients.

1. Introduction

In the rare neurological disease multiple system atrophy (MSA), mainly subcortical areas of the brain can be affected by cell degeneration. A more pronounced degeneration in the striatum, Parkinsonian-type (MSA-P), or in the cerebellum, cerebellar type (MSA-C), can be described in this movement disorder. Accordingly, predominantly autonomic, Parkinsonian and cerebellar symptoms may occur, for which there are only limited treatment options [1,2,3,4,5].

Although not identical to Parkinson’s disease (PD), alpha-synuclein pathology also occurs in MSA, resulting in misfolding and aggregation of the chaperone protein in the cytoplasm of oligodendrocytes and neurons [2,4,6]. An increase in neuronal release and distribution of alpha-synuclein was also shown in induced pluripotent stem cell-derived hypoexcitable striatal GABAergic medium spiny neurons from MSA-P patients compared to matched healthy controls [7]. The pathophysiology also includes neuroinflammation with activation of microglial cells and inflammatory signalling pathways [1,2,4]. Genetic studies could not identify any specific mutations leading to MSA [1,2,4].

In neurodegenerative diseases, mitochondrial processes and autophagy are often disrupted [8,9,10,11]. MSA is associated with reduced activity of the mitochondrial respiratory chain and increased biosynthesis of coenzyme Q10 [12,13,14].

Proteome studies in cerebrospinal fluids (CSF) and post-mortem brains of patients with MSA, PD and other atypical Parkinsonian syndromes such as the tauopathy progressive supranuclear palsy (PSP), corticobasal syndrome (CBS) and the synucleopathy dementia with Lewy bodies (DLB) were analysed in comparison to healthy control subjects or among themselves [6,15,16,17,18,19,20,21]. In various studies, a large number of altered proteins were detected, primarily using mass spectrometry. However, the proteins reported differed from study to study. Nevertheless, the proteins amyloid-beta precursor protein (APP), neuronal cell adhesion molecule (NRCAM) and secretogranin-2 (SCG2) were repeatedly reduced in CSF compared to PD and healthy control subjects [16,17]. In both studies, atypical Parkinsonian syndromes were pooled, so these observations apply to the MSA, PSP and CBS groups.

Studies of the prefrontal cortex and cingulate gyrus, in which the individual disorders were not grouped as atypical Parkinsonian syndromes, also revealed protein differences [6,18,19]. A Danish [18] and a Norwegian [19] research group analysed the proteome of the prefrontal cortex and compared the results of MSA patients with those of healthy control subjects. However, no common protein alteration was found in either study, which could be explained by differences in methodological implementation and equipment. In another recently published study, the proximal proteome of alpha-synuclein in the forebrain and midbrain of MSA patients was compared with that of PD and DLB patients. The proteins NADPH-dependent carbonyl reductase 1 (CBR1), alpha-crystallin B chain (CRYAB) and glial fibrillary acidic protein (GFAP) were enriched in MSA [20].

Current proteome studies provide initial approaches for the discovery of biomarkers and insights into the altered proteome of the brain in MSA patients. Nevertheless, this research still faces many challenges, as misdiagnosis, the grouping of several diseases, small patient cohorts, especially in rare, atypical Parkinsonian syndromes, and different methodologies lead to difficult interpretations. In addition to biomarkers, required for the early and reliable differentiation of the neurodegenerative diseases MSA, PD, PSP and DLB [3,4,14,22], basic research is also essential to gain further pathophysiological insights for the development of novel disease-modifying therapies.

This study focuses on the proteome analysis of cell lines from MSA-P patients (n = 3) compared to matched healthy controls (n = 3). Human induced pluripotent stem cells were differentiated into striatal GABAergic medium spiny neurons, and the proteomes were analysed using liquid chromatography-mass spectrometry. A total of 151 proteins differed in their expression between the cell lines from MSA-P patients and the controls.

2. Materials and Methods

2.1. Cell Culture and Striatal GABAergic Medium Spiny Neuron (MSN) Differentiation

In this study, human induced pluripotent stem cell (hiPSC) lines from three MSA-P patients, and three matched healthy controls were used as a cell model (

Table 1. Human induced pluripotent stem cell lines from three healthy control subjects and three MSA-P patients. The cell lines CTR1, CTR2 and CTR3 originate from healthy control subjects, while P1, P2 and P3 originate from MSA-P patients.

ID Code	Sex	Age at Biopsy	Origin/Reference
CTR1	M	59	Henkel et al. [7]
CTR2	F	62	StemBANCC consortium SFC084-03-02-01A
CTR3	F	62	Henkel et al. [7]
P1	F	78	Monzio Compagnoni et al. [13]
P2	M	52	Henkel et al. [7]
P3	F	56	Henkel et al. [7]

), which had been characterised in previous works [7,13]. Skin punch biopsies from MSA-P patients (cell lines P2 and P3) and healthy subjects (cell lines CTR1 and CTR3) were taken after written informed consent. Ethical approval for the generation of these cell lines was granted by the Ethics Committee of Hannover Medical School (No. 8666_BO_K_2019, date of approval: 13 September 2019).

The cultivation of hiPSCs was carried out in mTeSR Plus medium (STEMCELL Technologies, Vancouver, BC, Canada, 100-0276) with 1% Penicillin-Streptomycin (Thermo Fisher Scientific Inc., Waltham, MA USA, 15140122) on hESC-qualified Matrigel matrix (Corning, Now York, NY, USA, 354277). Cells were passaged as clumps using 0.5 mM EDTA (Thermo Fisher Scientific Inc., AM9260G) and mTeSR Plus with 10 µM ROCK inhibitor (Y-27632, STEMCELL Technologies, 72304).

Striatal GABAergic medium spiny neuron differentiation was performed after adaption from Henkel et al. [7] and Staege et al. [23]. Starting on day 0 as embryoid bodies (EBs), hiPSCs were passaged in mTeSR Plus with 1 µM dorsomorphin dihydrochloride (D, Abcam, Cambridge, UK, ab144821), 10 µM SB 431542 (SB, Abcam, ab291112), 1 µM Wnt Antagonist II (IWP-2, Merck KGaA, Darmstadt, Germany, 681671) and 10 µM Y-27632. The basal medium from day 2 to 11 (N2 medium) consisted of KnockOut DMEM/F-12 (Thermo Fisher Scientific Inc., 12660012), N-2 Supplement (Thermo Fisher Scientific Inc., 17502048), GlutaMAX Supplement (Thermo Fisher Scientific Inc., 35050038), 1% MEM Non-Essential Amino Acids Solution (Thermo Fisher Scientific Inc., 11140035), 1% Penicillin-Streptomycin and 15 mM HEPES, (Merck KGaA, H0887). On day 2, 1 µM D, 10 µM SB and 1 µM IWP-2 were added to the N2 medium. The same concentrations and 1 µM purmorphamine (PMA, Abcam, ab120933) were supplemented on day 4. On day 6 and 8, the same two molecules and concentrations were added (1 µM IWP-2 and 1 µM PMA). No additives were required on day 10; only the N2 medium was used.

As maturation medium, DMEM/F-12 with GlutaMAX Supplement (Thermo Fisher Scientific Inc., 31331093), Neurobasal medium (Thermo Fisher Scientific Inc., 21103049), N-2 Supplement, B-27 Supplement without Vitamin A (Thermo Fisher Scientific Inc., 12587010), 1% Penicillin-Streptomycin and 1% GlutaMAX Supplement were mixed. 20 ng/mL brain-derived neurotrophic factor (BDNF, Thermo Fisher Scientific Inc., 450-02), 10 ng/mL glial cell-derived neurotrophic factor (GDNF, Thermo Fisher Scientific Inc., 450-10) and 0.05 mM dibutyryl cyclic-AMP (dbcAMP, Merck KGaA, D0260) were added fresh due to their short-term stability.

EBs were seeded in maturation medium on hESC-qualified Matrigel matrix-coated plates from 12 to 16. The cells were passaged with Accutase (Thermo Fisher Scientific Inc., A1110501) and in maturation medium containing 10 µM Y-27632 on poly-DL-ornithine hydrobromide (80 µg/mL, PORN, Merck KGaA, P8638) and laminin-coated (10 µg/mL, Thermo Fisher Scientific Inc., 23017015) dishes on days 16 to 18 and again on PORN and laminin-coated well plates on days 24 to 30. From then on, the cells matured for up to 70 ± 7 days.

2.2. Immunocytochemistry (ICC)

The differentiated cells were fixed on day 70 (±7) with 4% paraformaldehyde (PFA, Merck KGaA, 8.18715). After fixation for 20 min at room temperature, the cells were washed three times with phosphate-buffered saline (PBS, Thermo Fisher Scientific Inc., 14190169) and treated with a blocking solution for 60 min (1% bovine serum albumin, BSA, Merck KGaA, A7906; 5% goat serum, Thermo Fisher Scientific Inc., 16210072; 0.3% Triton X-100, Merck KGaA, T8787; in PBS). The primary antibodies prepared in the blocking solution mouse anti-beta 3 tubulin (TUJ1, 1:1000; Abcam, ab78078, RRID:AB_2256751), rabbit anti-gamma-aminobutyric acid (GABA, 1:1000; Merck KGaA, A2052, RRID:AB_477652) and rat anti-COUP TF-1 interacting protein 2 (CTIP2, 1:300; Abcam, ab18465, RRID:AB_2064130) were incubated overnight at 4 °C. Afterwards, the cells were washed again before incubating the secondary antibodies for 2 h at room temperature. Goat anti-mouse Alexa Fluor 488 (1:1000; A-21131, RRID:AB_2535771), goat anti-rabbit Alexa Fluor 555 (1:1000; A-21428, RRID:AB_2535849), goat anti-rabbit Alexa Fluor 488 (1:500; A-11034, RRID:AB_2576217) and goat anti-rat (1:500; A-21434, RRID:AB_2535855) (all Thermo Fisher Scientific Inc.) were prepared in PBS. Mowiol 4-88 (Merck KGaA, 475904) with 4′,6-diamidino-2-phenylindole (DAPI, 10 µg/mL, Thermo Fisher Scientific Inc., D1306) was applied as the embedding medium.

For the fluorescence imaging, the Olympus BX61 fluorescence microscope, the Olympus BX-UCB control box, the Olympus DP72 camera (Olympus, Hamburg, Germany), the X-Cite 120Q fluorescence lamp (Micrasys, Herborn, Germany) and the Olympus cellSens Dimension 1.18 analysis software were used. Positively stained cells were counted in four randomly selected visual areas. For each cell line and differentiation, both DAPI, TUJ1 and GABA and DAPI, GABA and CTIP2 stainings were counted separately and calculated as a percentage ratio to DAPI, TUJ1 or GABA. The percentages were summarised for the control group of three healthy control cell lines and the MSA group of three MSA-P patient cell lines.

2.3. Reverse Transcription Quantitative Real-Time Polymerase Chain Reaction (RT-qPCR)

The RNAs of hiPSCs and MSNs (70 ± 7 days) were isolated with RNeasy Mini Kit (QIAGEN, Venlo, The Netherlands, 74104) and transcribed into cDNA via QuantiTect Reverse Transcription Kit (QIAGEN, 205311). As recommended, the RNAs were purified with a RNase-free DNAse Set (QIAGEN, 79254). All kits were performed according to the manufacturer’s protocol. For the determination of the RNA concentrations, Nanodrop 2000c (Thermo Fisher Scientific Inc.) was used. In a StepOnePlus cycler (Applied Biosystems StepOnePlus Real-Time PCR System, StepOne Software v2.3, Thermo Fisher Scientific Inc.), real-time PCR was performed as triplicates with SYBR Green PCR Master Mix (Thermo Fisher Scientific Inc., 4367659), 7 ng cDNA, 1.75 µM forward and reverse primers (see Table S1 for the primer sequences) [7,23]. Neuronal (TUJ1; MAP2, microtubule-associated protein 2; FOXP1, forkhead box protein P1), GABAergic (GAD67, glutamic acid decarboxylase; FOXG1, forkhead box protein G1), MSN-specific (CTIP2) and three reference genes (beta-2 microglobulin, B2M; glyceraldehyde-3-phosphate dehydrogenase, GAPDH; and beta-actin, ACTB) were examined in this study. The PCR reaction was performed as follows: Start at 95 °C for 10 min, 40 cycles of 95 °C for 15 s and 60 °C for 1 min.

The gene expressions were analysed for each cell line, comparing MSNs with their corresponding hiPSCs. ΔCt Reference was calculated as the average of the Ct values of the three reference genes. ΔCt Reference was subtracted from the Ct values of the mean target genes to determine ΔCt Sample. As ΔCt Calibrator, the averaged Ct values of the hiPSCs were calculated for each cell line. The mean ΔCt Calibrator was subtracted from ΔCt Sample to calculate ΔΔCt and, accordingly, the fold expressions (2^−ΔΔCt).

2.4. Liquid Chromatography-Mass Spectrometry (LC-MS) Analysis

On day 70 (±7), the cells were cooled on ice and washed at least once with pre-cooled PBS. Afterwards, MSNs were treated with 80% methanol solution (MtOH), scraped, vortexed and stored at −80 °C. After thawing, the samples were dried and dissolved in lysis buffer (8 M urea and 50 mM ammonium bicarbonate, pH 8.0).

MS-based proteome analyses were performed as described recently [24]. Briefly, proteins were mixed and alkylated by iodoacetamide, digested with trypsin, and further seperated into six fractions by alkaline reversed-phase (RP) chromatography as described. Peptide samples were analysed with data-dependent acquisition in an LC-MS system (RSLC, Orbitrap Exploris 240, both Thermo Fisher Scientific). Raw MS data were processed using MaxQuant (version 2.0, [25]) and Perseus software (version 2.0.11.0, [26]). Human entries from Universal Protein Knowledgebase (UniProtKB) were used for identification, and proteins were stated as identified with a false discovery rate (FDR) < 0.01 at the protein and peptide level. For each sample, the intensities were normalised by subtracting the corresponding median. Only proteins whose intensity was determined in at least all three biological replicates (independent differentiations) of one cell line were considered, and missing values were imputed by using normal distribution (downshift 1.8, width 0.3).

2.5. Proteomics

The proteome of three matched healthy control cell lines and three MSA-P patient cell lines, each with three independent differentiations, was analysed. For each cell line and protein, the log₂ intensities were averaged. The averaged log₂ intensities were assigned to the CTR or MSA group and statistically tested using an unpaired two-tailed t-test. A Benjamini–Hochberg correction was not applied. The fold change (FC) or the difference between the two groups was calculated from the total averaged log₂ intensities of the CTR group subtracted from the total averaged log₂ intensities of the MSA group. Data with −log₁₀ p-values (p) and log₂ FC are shown in the volcano plot. In addition, the averaged log₂ intensities of CTR and MSA groups were visualised as heat maps for the proteins with significant downregulation (−log₁₀ p > 1.3 and log₂ FC < −1) and significant upregulation (−log₁₀ p > 1.3 and log₂ FC > 1). For seven proteins, a scatter plot was selected to visualise the averaged log₂ intensities of the cell lines as data points. The proteome data were analysed using various bioinformatics tools.

2.5.1. PANTHER Classification System

Using the PANTHER classification system (version 19.0, 26 June 2025, [27]), a statistical over-representation analysis was performed using Fisher’s exact test and Benjamini–Hochberg false discovery rate (FDR), with the whole genome of Homo sapiens as reference for all genes that were significant according to the unpaired two-tailed t-test (p < 0.05 or −log₁₀ p > 1.3). Considering the significance criteria p < 0.05 and FDR < 0.05, the analysis revealed entries in biological processes, cellular components and molecular functions according to the Gene Ontology (GO) classification and the PANTHER GO-Slim classification.

For the significantly downregulated (p < 0.05 and log₂ FC < −1) and upregulated proteins (p < 0.05 and log₂ FC > 1), no statistically over-represented results could be obtained with the over-representation test. A functional classification was carried out for these corresponding genes, which were also classified according to GO, PANTHER pathways and protein classes.

2.5.2. Reactome Database

For the reaction analysis, an over-representation test with hypergeometric distribution and FDR according to Benjamini–Hochberg was performed using the Reactome database (version 93, 1 July 2025, [28]). For the proteins that were significant in the unpaired two-tailed t-test, Reactome pathways were determined that were statistically significant with p < 0.05 and FDR < 0.05. An over-representation test with the significantly downregulated (p < 0.05 and log₂ FC < −1) and upregulated proteins (p < 0.05 and log₂ FC > 1) revealed no statistically over-represented Reactome pathways.

2.5.3. Ingenuity Pathway Analysis (IPA)

Using Ingenuity Pathway Analysis (IPA, QIAGEN, 14 July 2025, [29]) to identify potential upstream regulators, the data for the normalised proteins, including p-values (unpaired two-tailed t-test) and log₂ fold changes (FC), were uploaded. Of these, 23 genes were unmapped and were excluded. The Core Analysis, which provided results for the upstream regulator, the causal and mechanistic network, and the downstream effect, was performed based on the expression analysis and the experimental log ratio (log₂ FC). The cut-off value for the p-value was set at 0.05, resulting in 150 analysable molecules. For statistical analysis, overlap p-values were calculated using one-sided Fisher’s exact test. Activation z-scores were calculated according to Krämer et al. [29], which predict the directionality of genes as activated or inhibited. For the statistical significance, results with p < 0.05 and z-score < −2 were considered as predicted inhibited and with p < 0.05 and z-score > 2 as predicted activated [29]. Only genes, RNAs and proteins were defined as molecule types for upstream molecules.

The Core Analysis was repeated with additional cut-offs for FC (−1 and 1), resulting in 40 analysable molecules. A statistical significance could not be achieved for these.

2.6. Statistical Analysis

Unless otherwise stated, statistical analyses were performed with GraphPad Prism 10.4.2. Cell lines were grouped according to CTR or MSA group and tested statistically using unpaired two-tailed t-tests. p-values (p) with p < 0.05 were defined as statistically significant. Graphs and scatter plots show mean values and standard error of the mean (SEM). Further statistical analyses used in the bioinformatic tools can be found in the corresponding methodology or in the figure captions.

3. Results

3.1. Striatal GABAergic Medium Spiny Neuron (MSN) Differentiation

To verify the 70-day differentiation of human induced pluripotent stem cells (hiPSCs) into striatal GABAergic medium spiny neurons (MSNs), the neuronal, GABAergic and striatal marker expressions were analysed in the differentiated cells of three control and three MSA-P cell lines. For this, three independent differentiations of each cell line were evaluated by immunocytochemistry (ICC) using 4′,6-diamidino-2-phenylindole (DAPI) as well as antibodies against neuron-specific beta 3 tubulin (TUJ1), gamma-aminobutyric acid (GABA) and striatal COUP TF-1 interacting protein 2 (CTIP2) (Figure 1A–C). Furthermore, we performed comparative gene expression analysis of TUJ1, microtubule-associated protein 2 (MAP2), forkhead box protein P1 (FOXP1), glutamic acid decarboxylase (GAD67), forkhead box protein G1 (FOXG1), and striatal marker gene CTIP2 (Figure 1D).

Based on the DAPI-stained cells, over 93% of the differentiated cultures expressed TUJ1 and GABA. Approximately 95% of CTR and 97% of MSA cells were TUJ1-positive, 93% of CTR and 95% of MSA cells were GABA-positive. In both groups, 98% of the TUJ1-stained neurons were also GABAergic. Comparing the staining of the striatal marker CTIP2/DAPI, fewer stained cells were observed in the MSA (84%) than in the CTR group (93%), although the difference was not significant (unpaired two-tailed t-test, p = 0.1061). Of the GABAergic-positive cells, 98% of the CTR and 91% of the MSA group expressed CTIP2 (Figure 1C).

The fold expressions (2^−ΔΔCt) of neuronal (TUJ1, MAP2), FOXP1, GABAergic (GAD67), FOXG1 and striatal (CTIP2) genes were significantly more expressed in the differentiated MSNs than in the corresponding hiPSCs (Figure 1D).

These results confirm a homogenous cell culture of mainly hiPSC-derived striatal GABAergic medium spiny neurons from which the proteomes were investigated.

3.2. Proteome Analysis

For the proteome analysis, three independent differentiations of three control (CTR) and three MSA-P patient cell lines (MSA) were analysed, in which a total of 6959 proteins were identified. Only proteins that were quantified in at least three biological replicates (independent differentiations) of one cell line were considered, resulting in a reduction in the number of proteins to 5817. After t-testing, the −log₁₀ p-values and the log₂ fold changes (FC) of these proteins were plotted as volcano plot (Figure 2A).

Comparing the protein expressions of MSA-P cell lines (MSA) with those of controls (CTR), 151 proteins with significantly different protein intensities were identified (p < 0.05, unpaired two-tailed t-test). A list of these proteins with gene names, p-values and log₂ fold changes is provided in the Supplement (see Table S2). Of these, 25 proteins were significantly upregulated in MSA (p < 0.05 and log₂ fold change > 1) (Figure 2A, purple dots). These proteins were gastrotropin (FABP6), centromere protein V (CENPV), microtubule cross-linking factor 1 (MTCL1), histone-lysine N-methyltransferase PR-domain containing 16 (PRDM16), protein zer-1 homologue (ZER1), glycylpeptide N-tetradecanoyltransferase 2 (NMT2), cytohesin-3 (CYTH3), neuroendocrine protein 7B2 (SCG5), zinc finger C-x8-C-x5-C-x3-H (CCCH) domain-containing protein 4 (ZC3H4), selenoprotein M (SELM), histone cell cycle regulator (HIRA)-interacting protein 3 (HIRIP3), Ras-related GTP-binding protein B (RRAGB), ubiquinol-cytochrome-c reductase complex assembly factor 1 (UQCC1), proline-rich coiled coil 2A (PRRC2A), beta-enolase (ENO3), E3 ubiquitin-protein ligase RNF31 (RNF31), phosphatidylinositol-specific phospholipase C X domain containing 3 (PLCXD3), collagen alpha-1(XXVI) chain (COL26A1), histone H1.0 (H1F0), 2-methoxy-6-polyprenyl-1.4-benzoquinol methylase (mitochondrial; COQ5), large ribosomal subunit protein uL29m (MRPL47), netrin receptor DCC (Deleted in Colorectal Carcinoma, DCC), polyglutamine-binding protein 1 (PQBP1), neurogranin (NRGN) and nucleosome assembly protein 1-like 5 (NAP1L5).

In addition, 16 proteins were significantly downregulated in MSA (p < 0.05 and log₂ fold change < −1) (Figure 2A, turquoise dots). These included Orosomucoid 1 (ORM1)-like protein 1 (ORMDL1), ORM1-like protein 2 (ORMDL2), voltage-dependent calcium channel subunit alpha-2/delta-1 (CACNA2D1), caspase-7 (CASP7), torsin-1A-interacting protein 1 (TOR1AIP1), HAUS (homologous to augmin subunit) augmin-like complex subunit 4 (HAUS4), organic solute carrier partner 1 (OSCP1), voltage-gated potassium channel subunit Kv7.2 (KCNQ2), BEN (BANP, E5R and Nac1) domain-containing protein 5 (BEND5), collectin-12 (COLEC12), epidermal growth factor receptor (EGFR), DNA damage-binding protein 2 (DDB2), DENN (differentially expressed in normal and neoplastic cells) domain-containing protein 5B (DENND5B), queuosine 5′-phosphate N-glycosylase/hydrolase (C9orf64), Yip1 domain family member 2 (YIPF2) and plakophilin-2 (PKP2).

For the 16 downregulated (Figure 2B) and the 25 upregulated (Figure 2C) proteins, the normalised log₂ intensities for the controls (CTR) and MSA cell lines (MSA) were averaged and displayed as heat maps. The colours of the heat maps depict the mean log₂ protein intensities, with a brighter colour representing higher protein intensities and a darker colour representing lower protein intensities.

As examples of the calculations of the log₂ fold changes and p-values by unpaired two-tailed t-tests, the normalised and averaged log₂ intensities were visualised as scatter plots, with the averaged log₂ intensities of one cell line represented as one data point. The two voltage-gated channel subunits CACNA2D1 (Figure 2D) and KCNQ2 (Figure 2E) were chosen as examples of functional proteins, EGFR (Figure 2F) as a protein with many functions, and UQCC1 (Figure 2G), COQ5 (Figure 2H) and MRPL47 (Figure 2I) as mitochondrial proteins. The protein alpha-synuclein (SNCA) (p = 0.1415) was additionally displayed as a reference due to the known alterations in MSA (Figure 2J). The log₂ fold changes were calculated by subtracting the mean log₂ intensities of the MSA group from the mean log₂ intensities of the CTR group.

3.3. Gene Ontology Analysis

For the investigation and categorisation of the different functions of the 151 significant proteins (p < 0.05, unpaired two-tailed t-test), the corresponding gene names were entered into the PANTHER classification system [27]. They were categorised according to the Gene Ontology (GO) and PANTHER’s own classification. Firstly, the 151 genes were analysed for over-representation using Fisher’s exact test (significance test, p < 0.05), which indicates the probability that the number of genes in this category occurred randomly. In addition, the false discovery rate (FDR) was calculated according to Benjamini–Hochberg, which indicates whether the classifications are false positives. With an FDR < 0.05, it can be assumed that up to 95% of the classification is a true assignment [27]. For the 151 genes, seven GO biological processes (Figure 3A), 21 GO cellular components (Figure 3B), three GO molecular functions (Figure 3C), 24 PANTHER GO-Slim biological processes (Figure 3D) and five PANTHER GO-Slim cellular components could be assigned with p < 0.05 and FDR < 0.05, whereby four genes (H1F0, C11orf30, EEF1E1-BLOC1S5, EIF2S3L) were not mapped in the over-representation analysis. The different number of annotations (e.g., 24 or 7 for the biological processes) resulted from the fact that the PANTHER GO-Slim uses the phylogenetic approach with experimental and extrapolated data, while GO considers manual and electronic data [27]. However, common annotations could be found for the biological processes (biosynthetic process (GO:0009058), gene expression (GO:0010467), macromolecule biosynthetic process (GO:0009059), metabolic process (GO:0008152), nucleobase-containing compound metabolic process (GO:0006139)) (Figure 3A,D) and for the cellular components (cytosol (GO:0005829), intracellular anatomical structure (GO:0005622), peptidase complex (GO:1905368) and proteasome complex (GO:0000502)) (Figure 3B,E). Binding was generally specified as a molecular function obtained from the GO annotation (Figure 3C). The over-representation analysis of the 41 significantly upregulated and downregulated genes revealed no significant association (p < 0.05 and FDR < 0.05).

In addition to the over-representation analysis, a functional classification can be carried out in the PANTHER classification system. For this purpose, the 41 significantly upregulated and downregulated proteins were classified into molecular function (Figure 4A), biological process (Figure 4B) and cellular component (Figure 4C). Not all proteins could be categorised and some proteins fitted more than one term of one ontology [27].

Seven molecular functions, eight biological processes and two cellular components were found as categories. In molecular functions (Figure 4A), most proteins (RRAGB, KCNQ2, FABP6, PRDM16, PQBP1, DDB2, RNF31, EGFR, DENND5B and HAUS4) were assigned to binding (GO:0005488). The second most common function was the catalytic activity (GO:0003824), which was attributed to nine proteins (RRAGB, CASP7, NMT2, SELM, RNF31, EGFR, ENO3, PLCXD3 and COQ5). Other molecular functions with one to two assigned proteins were structural molecule activity (GO:0005198; MRPL47 and COLEC12), molecular function regulator activity (GO:0098772; SCG5 and DENND5B), transcription regulator activity (GO:0140110; PRDM16), transporter activity (GO:0005215; KCNQ2 and CACNA2D1) and molecular transducer activity (GO:0060089; EGFR). Notably, the last three ones contained either the upregulated protein PRDM16 (Figure 4A, red bar) or the downregulated proteins KCNQ2, CACNA2D1 and EGFR (Figure 4A, blue bars).

In the category biological process (Figure 4B), 19 proteins (DCC, RRAGB, TOR1AIP1, CASP7, KCNQ2, ORMDL1, NMT2, PQBP1, ORMDL2, MRPL47, CENPV, DDB2, COLEC12, EGFR, PKP2, ENO3, HAUS4, PRRC2A and C9orf64) were classified as cellular process (GO:0009987). Nine proteins (ORMDL1, NMT2, PQBP1, ORMDL2, MRPL47, DDB2, RNF31, ENO3 and C9orf64) were involved in metabolic processes (GO:0008152). For the biological regulation (GO:0065007), seven proteins (SCG5, RRAGB, CASP7, KCNQ2, PRDM16, CENPV and EGFR) were assigned. Other biological processes were localisation (GO:0051179; KCNQ2, FABP6, PKP2 and OSCP1), response to stimulus (GO:0050896; RRAGB, DDB2 and EGFR) and developmental process (GO:0032502; EGFR, PKP2 and PRRC2A). For homeostatic processes (GO:0042592; ORMDL1 and ORMDL2) and multicellular organismal processes (GO:0032501; EGFR and PKP2), only downregulated proteins were classified.

Only two types of cellular components were found (Figure 4C): cellular anatomical entity (GO:0110165) with 25 proteins (RRAGB, TOR1AIP1, CASP7, KCNQ2, ORMDL1, FABP6, NMT2, PRDM16, PQBP1, ORMDL2, MRPL47, CENPV, SELM, DDB2, ZC3H4, COLEC12, EGFR, PKP2, OSCP1, ENO3, CACNA2D1, HAUS4, PLCXD3, YIPF2 and HIRIP3) and protein-containing complex (GO:0032991) with 9 proteins (RRAGB, KCNQ2, ORMDL1, ORMDL2, MRPL47, EGFR, ENO3, CACNA2D1 and HAUS4). For the corresponding protein names, please refer to the protein list in the Supplement (see Table S2).

In addition to the GO terminology, PANTHER also categorises the proteins according to protein classes (PC) (Figure 4D) and PANTHER pathways (P) (Figure 4E). A total of 15 protein classes were identified for 23 proteins (Figure 4D). The proteins CASP7 and RNF31 were classified as protein modifying enzymes (PC00260). For the upregulated proteins, nine classes were found, namely scaffold/adaptor protein (PC00226; COL26A1 and PQBP1), chaperone (PC00072; SCG5 and UQCC1), cell adhesion molecule (PC00069; DCC), protein-binding activity modulator (PC00095; RRAGB and CYTH3), transfer/carrier protein (PC00219; FABP6), RNA metabolism protein (PC00031; PRRC2A), gene-specific transcriptional regulator (PC00264; PRDM16 and ZC3H4), translational protein (PC00263; MRPL47) and metabolite interconversion enzyme (PC00262; NMT2, ENO3 and COQ5). Six downregulated proteins were assigned to transmembrane signal receptor (PC00197; EGFR), transporter (PC00227; KCNQ2 and CACNA2D1), DNA metabolism protein (PC00009; DDB2), cytoskeletal protein (PC00085; PKP2) and extracellular matrix protein (PC00102; COLEC12).

Only six proteins could be assigned to ten PANTHER pathways (Figure 4E). Only the upregulated protein DCC was matched to two pathways, axon guidance mediated by Slit/Robo (P00008) and axon guidance mediated by netrin (P00009). Most of the PANTHER pathways were assigned to the downregulated protein EGFR: Gonadotropin-releasing hormone receptor pathway (P06664), cadherin signalling pathway (P00012) and EGF receptor signalling pathway (P00018). CASP7 was also found in two PANTHER pathways, apoptosis signalling pathway (P00006) and FAS signalling pathway (P00020). CACNA2D1 was associated with the muscarinic acetylcholine receptor 2- and 4-signalling pathway (P00043), KCNQ2 with muscarinic acetylcholine receptor 1 and 3 signalling pathway (P00042) and DDB2 with p53 pathway (P00059).

3.4. Reactome Analysis

Reactome analysis focuses on the investigation of reactions in biological pathways and processes using the Reactome database. An over-representation analysis with hypergeometric distribution and a Benjamini–Hochberg FDR was performed to determine a probability indicating whether the number of genes in certain Reactome pathways is higher than would be expected by chance [28]. The 151 genes of the respective significant proteins were analysed (p < 0.05, unpaired two-tailed t-test), whereby 49 were not mapped in the over-representation test. A total of 771 Reactome pathways were found, of which 88 Reactome pathways met the significance criteria (p < 0.05 and FDR < 0.05) (Figure 3F). For these, Reactome pathways were identified in the areas of cell cycle, cell death, degradation, axon guidance, neuronal development, cellular response, signalling, translation, viral infection and regulation, which are listed in more detail in the Supplement (see Table S3).

An over-representation analysis was also carried out for the 41 significantly upregulated and downregulated proteins. A total of 237 different Reactome pathways were found for 29 genes and none for 12 genes (upregulated: OSCP1, BEND5, YIPF2; downregulated: CENPV, MTCL1, ZER1, SCG5, SELM, HIRIP3, PRRC2A, PLCXD3, NAP1L5; see Table S2 for the corresponding protein names). For the reduced number of genes, no Reactome pathway met both significance criteria (p < 0.05 and FDR < 0.05).

3.5. Upstream Regulator Analysis

To identify candidates for the possible origin (upstream regulator) of the normalised and significant proteins (p < 0.05, unpaired two-tailed t-test), an upstream regulator analysis was performed in Ingenuity Pathway Analysis (IPA). The statistically calculated overlap p-value (overlap p, one-sided Fisher’s exact test) and the z-score were used to evaluate the results of the upstream regulator analysis. The probability that the experimental dataset did not randomly overlap with the dataset of the Ingenuity Knowledge Base can be specified with an overlap p < 0.05. Activation or inhibition, represented by a z-score, was predicted by comparing the examined gene expression with the expected gene expression derived from the literature [29].

The upstream regulator analysis of the 151 significant genes, whereby SELM was excluded by IPA, revealed 325 potential candidates that could lead to the observed downregulation or upregulation of proteins. Of these, 320 candidates met the overlap p < 0.05, but only 47 received a z-score.

The molecular chaperone MKKS (Figure 5A), the histone H1.1 (Figure 5B), lysine-specific demethylase 5A (KDM5A) (Figure 5C) and cellular tumour antigen p53 (TP53) (Figure 5D) were proposed as upstream molecules, which were statistically predicted to be activated (MKKS, overlap p < 0.05 and z-score > 2) or inhibited (H1-1, KDM5A and TP53, overlap p < 0.05 and z-score < −2). For these predictions, 28 of 150 significant genes, shown as downregulated molecules (Figure 5A–D), were linked to the potential upstream molecules, with relationships indicated directly with solid lines or indirectly with dashed lines. Four downstream molecules were observed in more than one prediction. The decreased CD81 antigen (CD81) was predicted to be indirectly inhibited by MKKS, indirectly activated by H1-1 and directly linked to TP53 without directionality, for which the Ingenuity Knowledge Base could not provide any information. TP53 and the histone H1-1 could participate in an activation of glutamine synthetase (GLUL) leading to the decreased log₂ FC in MSA cell lines compared to CTR cell lines. KDM5A as well as TP53 were predicted to inhibit 2-methoxy-6-polyprenyl-1,4-benzoquinol methylase (mitochondrial; COQ5) and succinate dehydrogenase [ubiquinone] iron-sulphur subunit (mitochondrial; SDHB), leading to an increase in both downstream molecules. Yellow relationships of TP53 to the multifunctional protein CAD (CAD), cysteine and glycine-rich protein 1 (CSRP1) and cyclin-K (CCNK) demonstrate contradictory results in the experimental data and in the literature-based data. In addition to CD81, the high mobility group protein A1 (HMGA1), CLIP-associating protein 1 (CLASP1), tyrosine-protein kinase CSK (c-terminal Src kinase) and polyadenylate-binding protein 4 (PABPC4) were also associated with TP53, but no activation or inhibition could be predicted.

A repeated upstream regulator analysis was performed with additional cut-offs for log₂ FC < −1 and log₂ FC > 1, to predict candidates for the significantly upregulated and downregulated proteins. This filtered the number of genes to 40, resulting in 294 upstream molecules, of which 293 were significant with an overlap p < 0.05. For these, no significant activation or inhibition was observed in the z-score calculation. Interestingly, TP53 was also predicted as an upstream candidate with COQ5, epidermal growth factor receptor (EGFR), DNA damage-binding protein 2 (DDB2) and beta-enolase (ENO3) as downstream molecules, but it achieved only an overlap p of 0.519 and a z-score of −1.931 (highest score out of four calculated scores). H1-1 was also found, whereas MKKS and KDM5A were not.

3.6. Potential Biomarker Candidates

Proteome analysis revealed various proteins with different functions that were found in the striatal GABAergic medium spiny neurons of MSA-P cell lines. Of the 41 significantly upregulated and downregulated proteins, 13 with a neuro-specific or neuro-relevant function could be used as potential biomarker candidates (

Table 2. Neuro-specific proteins as potential biomarker candidates for multiple system atrophy. Proteins were obtained in the proteome analysis of striatal GABAergic medium spiny neurons (MSNs), which were differentiated from three MSA-P cell lines (n = 3) and three CTR cell lines (n = 3). The corresponding results of the protein expression are presented in Figure 2. A protein list of the significantly altered protein expressions with p-values (unpaired two-tailed t-test) and fold changes can be found in the Supplement (Table S2). For the selected proteins proposed here as potential biomarkers, purple arrows indicate those that were upregulated in MSA (log₂ fold change > 1), and turquoise arrows indicate those that were downregulated in MSA (log₂ fold change < −1).

Gene Names	Protein Names	MSA vs. CTR	p-Value	Fold Change (log2)
MTCL1	Microtubule cross-linking factor 1		0.0037	2.6503
NMT2	Glycylpeptide N-tetradecanoyltransferase 2		0.0294	1.9344
SCG5	Neuroendocrine protein 7B2		0.0377	1.5578
SELM	Selenoprotein M		0.0119	1.3475
RRAGB	Ras-related GTP-binding protein B		0.0463	1.2842
PRRC2A	Proline-rich coiled coil 2A		0.0259	1.2435
DCC	Netrin receptor DCC (Deleted in Colorectal Carcinoma)		0.0193	1.0619
PQBP1	Polyglutamine-binding protein 1		0.0388	1.0494
NRGN	Neurogranin		0.0401	1.017
ORMDL1	Orosomucoid 1 (ORM1)-like protein 1		0.0068	−1.071
ORMDL2	ORM1-like protein 2		0.0068	−1.071
CACNA2D1	Voltage-dependent calcium channel subunit alpha-2/delta-1		0.0405	−1.1958
KCNQ2	Voltage-gated potassium channel subunit Kv7.2		0.0144	−1.4919

).

3.6.1. Proteins Associated with Axons and Neurites

Microtubule cross-linking factor 1 is involved in maintaining the initial segment of the neuronal axon [30]. Netrin receptor DCC (Deleted in Colorectal Carcinoma) plays a role in axon outgrowth and guidance [31], which was also demonstrated in the PANTHER pathway result (Figure 4E, axon guidance mediated by Slit/Robo (P00008) and axon guidance mediated by netrin (P00009)). Polyglutamine-binding protein 1 regulates neurite outgrowth [32].

3.6.2. Neuronal Functional Proteins

The postsynaptic protein neurogranin fulfils several synaptic functions, e.g., long-term potentiation [33,34]. Regulation and trafficking of calcium channels depend on the voltage-gated calcium channel subunit alpha-2/delta-1 [35,36]. The voltage-gated potassium channel subunit Kv7.2 is essential for the postsynaptic excitability [37,38].

3.6.3. Neuronal Protection

Glycylpeptide N-tetradecanoyltransferase 2 plays a central role in suppressing ferroptosis [39]. Selenoprotein M, which is enriched in the brain, protects the brain from oxidative damage and may also have other functions in calcium regulation [40]. The neuronal Ras-related GTP-binding protein B regulates the mechanistic target of rapamycin complex 1 (mTORC1), which increases resistance to nutrient deprivation [41]. The chaperone protein neuroendocrine protein 7B2 demonstrated an anti-aggregative function by preventing the formation of fibrils from amyloid plaques and alpha-synuclein, which are associated with Alzheimer’s disease and Parkinson’s disease [42].

3.6.4. Myelination

Orosomucoid 1 (ORM1)-like protein 1 and ORM1-like protein 2 are involved in the sphingolipid synthesis, with sphingolipids forming the basis of myelin [43]. The proline-rich coiled coil 2A controls the myelination process and other oligodendrocyte specifications [44].

4. Discussion

Biological mechanisms causing or promoting neurodegeneration such as multiple system atrophy (MSA) are not yet fully understood. However, human pluripotent stem cell (hiPSC) models from MSA-P patients can provide new insights into the pathomechanisms. In this study, hiPSCs were differentiated into striatal GABAergic medium spiny neurons, which are the most common cell type in the striatum [45] and are affected by the degeneration that occurs in MSA [4,5].

The proteomes of striatal GABAergic medium spiny neurons derived from cell lines of MSA-P patients obtained a variety of protein types with diverse functions, processes and reactions. In comparison to matched healthy controls, 151 proteins were significantly altered in the protein expressions of MSA cell lines that were tested in over-representation. The results showed that the proteins were over-represented in gene expression, biosynthetic and metabolic processes, were components of various enzyme complexes and located intracellularly or bound to the membrane. Furthermore, the proteins were found in various Reactome pathways including cell cycle, cell death, translation, regulation, degradation, cellular response, signalling, axon guidance, neuronal development and viral infection.

The main neuropathological feature of MSA is the accumulation of aggregated alpha-synuclein, forming cytoplasmic inclusions [5]. As shown in our previous study [7], the neuronal distribution and release of the protein alpha-synuclein was significantly elevated in MSA-P MSNs. In this proteomic study, we found a tendency towards upregulation of alpha-synuclein expression in MSA-P neurons compared to controls (p = 0.1415 and log₂ FC = 1.0668).

In cerebrospinal fluid (CSF) studies of patients with MSA, Parkinson’s disease (PD), progressive supranuclear palsy (PSP), corticobasal syndrome (CBS), dementia with Lewy bodies (DLB) and Alzheimer’s disease (AD), different levels of neurogranin were detected, which were reduced in MSA, PD and PSP and increased in AD compared to controls [21,46]. As a postsynaptic calmodulin-binding protein located in dendritic spines, neurogranin (NRGN) fulfils several synaptic functions such as plasticity, regeneration and long-term potentiation [33,34]. Neurogranin is bound to calmodulin at rest. When the calcium concentration changes by influx, calcium ions bind to calmodulin due to changes in affinity, releasing neurogranin and activating the signalling [47,48]. In a study with hippocampal CA1 pyramidal neurons, an overexpression of neurogranin led to increased amplitudes of miniature excitatory postsynaptic currents (mEPSCs) through AMPAR-mediated responses and increased activation of calcium/calmodulin-dependent protein kinase II (CaMKII), resulting in neurogranin-mediated synaptic insertion of glutamate receptor 1 (GluR1) [47]. In contrast, our previous study showed decreased amplitudes in miniature postsynaptic currents of MSA MSNs [7], which could be regulated by neurogranin in an inhibitory, GABA-mediated manner. Calmodulin was also found in both groups, but it did not differ significantly (p = 0.6692 and log₂ fold change = 0.2854).

The two downregulated proteins in MSA cell lines assigned to transporter activity (GO:0005215), the voltage-gated calcium channel subunit alpha-2/delta-1 (CACNA2D1) and the voltage-gated potassium channel subunit Kv7.2 (KCNQ2), may affect the functional properties of striatal GABAergic medium spiny neurons. The voltage-gated calcium channel subunit alpha-2/delta-1 is one of four other alpha-2/delta subunits that combine to form the auxiliary alpha-2/delta subunit. As an auxiliary alpha-2/delta subunit, it is involved in the regulation of activation and inhibition of the calcium channels Cav1 and Cav2, as well as in the trafficking of Cav2 [35,36]. A reduction in the voltage-gated calcium channel subunit alpha-2/delta-1 could lead to reduced regulation and Cav2 trafficking, which could further explain a reduced frequency of spontaneous calcium transients and a tendency towards decreased gene expression of the Cav2.1 subunit, shown in our previous work [7].

The voltage-gated potassium channel contains the subunits Kv7.2 (KCNQ2) and Kv7.3 (KCNQ3), which form the neuronal M-current [49], which plays an important role in neuronal excitability, especially in the brain. Under activated condition, the elevated depolarisation threshold is held by a continuous potassium efflux, resulting in hypoexcitability. In contrast, an inhibited or knock-out M-current leads to hyperexcitability [37,38]. A reduction in the voltage-gated potassium channel subunit Kv7.2 could therefore also lead to hyperexcitability due to potentially reduced potassium efflux mechanism. After potassium chloride application, we observed an elevated percentage of responding cells and response amplitude in the calcium imaging of MSA-P MSNs compared to healthy controls; however, the frequencies of spontaneous action potentials and miniature postsynaptic currents were decreased in whole-cell patch-clamp recordings of MSA-P neurons [7].

Three mitochondrial proteins were significantly upregulated in MSA-P neurons, with two proteins having a central role in the mitochondrial respiratory chain. 2-Methoxy-6-polyprenyl-1.4-benzoquinol methylase (COQ5) is a precursor in the biosynthesis of the coenzyme Q10 (CoQ10), which supplies electrons to complex III of the mitochondrial respiratory chain. Other proteins involved in the biosynthesis of CoQ10 were found in elevated concentrations in MSA, including decaprenyl diphosphate synthase subunit 2 (PDSS2), 4-hydroxy-3-methoxy-5-polyprenylbenzoate decarboxylase (COQ4) and aarF domain-containing kinase 3 (COQ8A) [13]. The ubiquinol-cytochrome-c reductase complex assembly factor 1 (UQCC1) is required for the assembly of the complex III, which was also significantly upregulated in MSA cell lines. Decreased activity of complex II and III was detected in dopaminergic neurons of MSA cell lines [13]. Together with the increased protein expressions, the elevated mitochondrial mass and the reduced activity of the respiratory chain [13], these findings could indicate compensatory mechanisms. As a component of the mitochondrial protein biosynthesis, the large ribosomal subunit protein uL29m (MRPL47) was increased not only in striatal GABAergic MSNs but also in the proteome of the cingulate gyrus of sporadic Parkinson’s disease patients (n = 3) compared to MSA-P patients (n = 3) [6], indicating a higher demand for mitochondrial protein synthesis in both diseases.

As the most frequently occurring protein in PANTHER classification and Reactome pathways, the epidermal growth factor receptor (EGFR) was assigned not only to its own pathway (EGF receptor signalling pathway (P00018)), but also to Gonadotropin-releasing hormone receptor pathway (P06664), cadherin signalling pathway (P00012) as well as in the Reactome pathways axon guidance (R-HSA-422475) and nervous system development (R-HSA-9675108), with many non-significant entries also reported. The transmembrane signal protein (PC00197) EGFR was also associated with molecular functions such as molecular transducer activity (GO:0060089), binding (GO:0005488), catalytic activity (GO:0003824) and biological processes such as cellular process (GO:0009987), biological regulation (GO:0065007), response to stimulus (GO:0050896), developmental process (GO:0032502) and multicellular organismal process (GO:0032501). This list reflects only a part of the EGFR-mediated processes. In a study with Chinese patients, two polymorphisms of the EGFR gene were associated with susceptibility to Parkinson’s disease [50]. A recently published study [51] analysed EGFR in the context of alpha-synuclein aggregation in human neuroblastoma cells (SH-SY5Y). It was demonstrated that the phosphorylation of the DnaJ protein homologue 1 (DNAJB1) at tyrosine 5, which is carried out by EGFR, is essential for the resolution of alpha-synuclein aggregation. Downregulation of EGFR resulted in an increase in alpha-synuclein aggregation, which was not reduced by an additional overexpression of DNAJB1, indicating the importance of EGFR phosphorylation. Furthermore, reduced EGFR and DNAJB1 expression was detected in the brains of PD patients, whereby phosphorylated DNAJB1 was increased [51]. Accordingly, the reduced expression of EGFR in the MSNs of MSA-P cell lines could result in similarly impaired resolution of the aggregation of alpha-synuclein.

To determine whether certain upstream molecules could explain the altered proteome expression, an upstream regulator analysis was performed. The prediction indicated the molecular chaperone MKKS, the histone H1.1, the lysine-specific demethylase 5A (KDM5A) and cellular tumour antigen p53 (TP53) with significant assignment of activation or inhibition.

The molecular chaperone MKKS (MKKS), also known as chaperonin-like BBS6 protein, can combine with two chaperonin-like BBS proteins (BBS10 and BBS12) and six CCT chaperonin proteins that are required for the assembly of the BBSome complex. The BBSome complex regulates the transport of vesicles to cilia and ciliary signalling [52,53]. An activated BBS6 could probably induce the BBSome machinery, potentially leading to an increase in vesicle trafficking.

Two subtypes of the histone linker H1, the predicted inhibited histone H1.1, localised on chromosome 6, and the significantly upregulated histone H1.0 (H1F0 or H1-0), localised on chromosome 22, are part of chromatin structure and dynamics, whereby H1.0 is predominantly expressed in terminally differentiated cells [54,55]. During the cell cycle and cell differentiation, the different histone 1 subtypes on chromosome 6 co-regulate each other in a redundant manner, so that compensatory mechanisms could counteract the inhibition of H1.1 [56]. An increase in H1.0 could be a result of an interaction with alpha-synuclein, whereby higher concentrations could induce accumulation and accelerated fibrillation of alpha-synuclein in neurons, which was shown in a study on alpha-synuclein with bovine histone 1 [57].

As a histone demethylase, the lysine-specific demethylase 5A (KDM5A, H3K4 demethylase RBP2) is primarily involved in the demethylation of lysine 4 of histone H3 (H3K4me3/me2), thereby deactivating the active state of H3K4me3/2 and suppressing gene expression associated with (neuronal) cell differentiation [58,59]. An inhibition of lysine-specific demethylase 5A might therefore leave the methylated H3K4 in an active state, which could cause overexpression (see Figure 5C). In Parkinson’s disease, the SNCA promoter was transcriptionally activated via H3K4me3 and was elevated in the substantia nigra of post-mortem PD brains, resulting in elevated alpha-synuclein expression, which was further investigated in cell models of the neuronal cell line SH-SY5Y and idiopathic PD-iPSC derived dopaminergic neurons. By incorporating the KDM5A sequence (also known as JARID1A) into the demethylating system targeting the SNCA promoter, a reduction in H3K4me3 as well as a decrease in alpha-synuclein expression were achieved in both cell lines [60].

The cellular tumour antigen p53 (TP53), also known as phosphoprotein p53 (p53), is a transcription protein regulating many biological processes such as DNA repair, senescence, cell differentiation, stress, apoptosis, cell metabolism and autophagy. Activation and inhibition of p53 functionally regulates many downstream proteins (see Figure 5D) [61,62]. In post-mortem brains, cell and animal models of Parkinson’s disease, elevated p53 values were associated with increased alpha-synuclein expression in the substantia nigra, which is due to regulation of the SNCA promoter, but not in other brain regions or related diseases such as MSA [63,64,65].

Limitations

This proteomic study only reflects the proteome results of striatal GABAergic medium spiny neurons, which limits the observed alterations in protein intensities due to homogenous cell cultures. The inclusion of other relevant cell types, e.g., dopaminergic neurons, oligodendrocytes and astrocytes, in the stem cell model of MSA may provide further insights into the pathomechanisms of the affected cell populations in future studies. In addition, six individual pluripotent stem cell lines derived from three MSA-P patients and three matched healthy controls were analysed, which could also contribute to statistical artefacts in the data. Incorporating more cell lines could minimise individual and statistical effects, but disease-specific cell lines from MSA-patients are hardly available. Apart from the biological limitations, methodological artefacts of the imputed values may also distort the observed changes in protein intensities.

5. Conclusions

This study presents the first proteomic analyses of an iPSC model of MSA. Proteome alterations were detected in striatal GABAergic medium spiny neurons from three MSA-P patient-derived cell lines and three matched healthy controls. A comparison of the normalised and averaged protein intensities of the MSA-P group with those of the control group showed a significant change in expression for 151 proteins. By analysing the over-representation, a variety of protein types, molecular functions, biological processes, reactions and cellular components were identified, indicating a multifaceted characteristic of the disease. 25 proteins with at least 2.0-fold upregulation and 16 proteins with less than 0.5-fold downregulation in MSA-P cell lines revealed alterations in neuronal functions such as amplification and attenuation of electrical stimulus transmission, postsynaptic processes, mitochondrial processes, gene regulatory mechanisms, enzymatic processes, apoptosis, signal pathways, etc. As potential upstream regulators, chaperone, histone, histone demethylase and p53 protein were predicted. Since there is currently no disease-modifying therapy for MSA, it is important to gain new insights into the pathomechanisms that may contribute to obtaining possible treatment options.