SNCA and TPPP transcripts increase in oligodendroglial cytoplasmic inclusions in multiple system atrophy

Multiple system atrophy (MSA) is characterized by glial cytoplasmic inclusions (GCIs) containing aggregated α -synuclein ( α -syn) in oligodendrocytes. The origin of α -syn accumulation in GCIs is unclear, in particular whether abnormal α -syn aggregates result from the abnormal elevation of endogenous α -syn expression in MSA or ingested from the neuronal source. Tubulin polymerization promoting protein (TPPP) has been reported to play a crucial role in developing GCI pathology. Here, the total cell body, nucleus, and cytoplasmic area density of SNCA and TPPP transcripts in neurons and oligodendrocytes with and without various α -syn pathologies in the pontine base in autopsy cases of MSA ( n = 4) and controls ( n = 2) were evaluated using RNAscope with immunofluorescence. Single-nucleus RNA-sequencing data for TPPP was evaluated using control frontal cortex ( n = 3). SNCA and TPPP transcripts were present in the nucleus and cytoplasm of oligodendrocytes in both controls and diseased, with higher area density in GCIs and glial nuclear inclusions in MSA. Area densities of SNCA and TPPP transcripts were lower in neurons showing cytoplasmic inclusions in MSA. Indeed, TPPP transcripts were unexpectedly found in neurons, while the anti-TPPP antibody failed to detect immunoreactivity. Single-nucleus RNA-sequencing revealed significant TPPP transcript expression predominantly in oligodendrocytes, but also in excitatory and inhibitory neurons. This study addressed the unclear origin of accumulated α -syn in GCIs, proposing that the elevation of SNCA transcripts may supply templates for misfolded α -syn. In addition, the parallel behavior of TPPP and SNCA transcripts in GCI development highlights their potential synergistic contribution to inclusion formation. In conclusion, this study advances our understanding of MSA pathogenesis, offers insights into the dynamics of SNCA and TPPP transcripts in inclusion formation, and proposes regulating their transcripts for future molecular therapy to MSA.

The origin of α-syn accumulation in GCIs and its relationship to the dynamics of α-syn expression in MSA remain unclear.Previous studies have indicated that mature oligodendrocytes either do not express α-syn or do so only in limited quantities (Miller et al., 2005;Solano et al., 2000).Additionally, investigations into the mRNA levels of the α-syn encoding gene, SNCA, in cerebral tissue from MSA patients have shown the same or lower expression compared to non-affected individuals (Jin et al., 2008;Langerveld et al., 2007;Miller et al., 2005;Ozawa et al., 2001).In vitro studies (Ettle et al., 2014;Kaji et al., 2018;Kisos et al., 2012;Konno et al., 2012;Pukass and Richter-Landsberg, 2014) and in vivo experiments (Reyes et al., 2014;Rockenstein et al., 2012) have demonstrated the ability of exogenous recombinant or neuronally derived α-syn to be taken up by oligodendroglial cell lines, suggesting a "prion-like" transmission of α-syn from neurons to glia.
Contrary to previous assumptions, our recent findings (Kon et al., 2023) using RNAscope and single-nucleus RNA-sequencing (snRNA-seq) revealed the presence of a small amount of SNCA transcripts in oligodendrocytes in postmortem non-diseased human brains.Moreover, Asi et al. (Asi et al., 2014) employed laser capture microdissection and qRT-PCR to evaluate the presence of SNCA transcripts in oligodendrocytes from postmortem MSA brains.They also reported increased SNCA transcripts in MSA oligodendrocytes compared to controls, whereas neurons in MSA expressed lower levels of SNCA, although these findings did not reach statistical significance (Asi et al., 2014).Furthermore, both SNCA transcripts and protein were found to be expressed in cultured oligodendrocytes from humans and rodents (Djelloul et al., 2015;Kaji et al., 2018;Mavroeidi et al., 2019;Richter-Landsberg et al., 2000).Mavroeidi et al. specifically reported that the accumulation of endogenous oligodendroglial α-syn plays a crucial role in the formation of pathological GCI-like α-syn structures within oligodendrocytes in wildtype mice, triggered by exogenously added human fibrillar α-syn seeds (Mavroeidi et al., 2019).
Tubulin polymerization promoting protein (TPPP), also known as p25α, was initially identified in a fraction of tau protein kinase (Takahashi et al., 1991).TPPP plays a crucial role in myelin maturation (Lehotzky et al., 2010) and is essential for the reorganization and stabilization of microtubules (Hlavanda et al., 2002;Tirian et al., 2003).In 2004, Kovacs et al. (Kovacs et al., 2004) reported TPPPimmunoreactivity (IR) in Lewy bodies (LBs) in Lewy body disease (LBD), GCIs in MSA, and granulovacuolar degeneration in Alzheimer's disease (AD).In a follow-up study, Kovacs et al. demonstrated that the immunoreactivity for Lewy bodies or GCIs is dependent on the TPPP antibody utilized (Kovacs et al., 2007).Physiologically, TPPP protein is expressed in the oligodendroglial nucleus, cytoplasm, and myelin (Acevedo et al., 2007;Hoftberger et al., 2010;Kovacs et al., 2004;Lindersson et al., 2005;Olah et al., 2006;Skjoerringe et al., 2006).However, in MSA, TPPP protein redistributes from the nucleus and myelin sheath to the abnormally expanded perinuclear cytoplasm, potentially preceding the accumulation of α-syn in the cytoplasm (Ota et al., 2014;Song et al., 2007).Furthermore, TPPP has been found to modulate the rate and extent of α-syn seeding (Mavroeidi et al., 2019), and its interaction with α-syn inhibits the proteolytic degradation (Lehotzky et al., 2021).These findings suggest that TPPP contributes to the pathogenesis of MSA by promoting the aggregation and accumulation of α-syn in oligodendrocytes.However, much remains unknown regarding the expressional level of TPPP transcripts in inclusions of MSA, such as GCIs and neuronal cytoplasmic inclusions (NCIs).Therefore, we investigated the dynamics of SNCA and TPPP transcripts within specific cell morphologies in MSA brains using a combination of RNAscope and immunofluorescence techniques, complemented by snRNAseq.

Specificity of RNAscope probes
To validate the specificity of RNAscope probes, DNase or RNase treatment was added to the RNAscope with immunofluorescence according to the manufacturer's protocols.Briefly, for DNase treatment, DNase-I solution (Bovine, 1:50 w/v dilution with 1× DNase Buffer; Sigma-Aldrich, St. Louis, MO, USA) was added to the section slides after target retrieval, and then the slides were incubated for 30 min at 37 • C. For RNase treatment, RNase A (5 mg/mL; Qiagen, Germantown, MD, USA) was added to the section slides after RNAscope Protease Plus treatment, and then the slides were incubated for 20 min at 37 • C. All other RNAscope assay steps were conducted as previously described.

Acquisition of images
Image acquisition was performed using Nikon C2Si + confocal on a Nikon Ti2-E inverted microscope equipped with a 40× objective lens for a single stack (NA: 0.95) in a random fashion.Filter settings for DAPI, Alexa 488, Opal 590, and Opal 690 were appropriately adjusted.Images were captured using NIS-Elements AR software (version 5.30.04) with calibration set at 0.31 μm/pixel.Selected sections were captured as zstacks with an inter-slice distance of 0.1 μm, and three-dimensional images were generated using NIS-Elements AR software.Parameters for image acquisition were standardized and maintained throughout the experiment.

Morphometry
RNAscope fluorescent sections (Fig. 1A, B) were subjected to analysis using algorithms developed in the General Analysis 3 module within NIS Elements software (version 5.30.04), as previously outlined (Forrest et al., 2023;Kon et al., 2023).Identification of cells expressing cellspecific markers for neurons and oligodendrocytes was achieved using corresponding probes.Given the challenges in identifying the cytoplasmic area of oligodendrocytes, we defined the oligodendrocytic cytoplasmic area as a specific distance from the nucleus, with a predetermined value of 1.5 μm based on our pilot study (details are provided in Supplementary method and result in Supplementary file).Nuclear areas were discriminated by capturing DAPI signal intensity by the algorithm (Fig. 1C, D).Subsequently, the oligodendroglial cell body region was automatically extended in radius length by 1.5 μm in oligodendrocytes (Fig. 1C).Manual tracing of neuronal cell body borders followed, conducted after the Lookup Table (LUT) strength of DAPI was sufficiently enhanced to visualize autofluorescence from lipofuscin pigments (Fig. 1D).The LUT parameters were then restored to normal settings, and areas positive for phosphorylated-α-syn or TPPP were automatically captured above the programmed threshold (Fig. 1E, F).Subsequently, the NIS-Elements software identified the area positive for SNCA and TPPP transcripts above the threshold within the annotated regions of interest (Fig. 1G, H).Consistent with our prior studies (Forrest et al., 2023;Kon et al., 2023), fluorescence intensity thresholds for SNCA and TPPP transcripts were established, surpassing autofluorescence levels, ensuring independence from potential autofluorescence impacts during analysis.Transcript signals were recognized as individual entities or small confluent clusters (Fig. 1), posing challenges for accurate individual transcript counting.Consequently, this study relied on the area density values of the transcripts.SNCA and TPPP transcripts area densities were determined by dividing the area occupied by the transcripts by the annotated regions of interest and expressed as a percentage.
We assessed RNA positive area density in distinct cellular areas: (I) the total cell body, (II) the nucleus, and (III) the cytoplasm (calculated by subtracting II from I).Additionally, we evaluated the area of p-α-synand TPPP-immunoreactive inclusions.When we found RNA signals both in the nucleus and p-α-syn-immunoreactive area (i.e.inclusion), the signals were counted separately for each cell component.We compared SNCA and TPPP area densities between the cytoplasm and the p-α-syn-IR inclusion area only.We classified GCIs and NCIs when both cytoplasmic inclusion and cell-specific markers were identified in the absence of T. Kon et al. nuclear inclusions.Cells that exhibited p-α-syn-immunoreactive inclusions within the nucleus with cell-specific markers were categorized as GNI or neuronal nuclear inclusion (NNI).In the context of TPPP-IR, we categorized oligodendrocytes with TPPP-IR in the nucleus, accompanied by cell-specific markers but lacking cytoplasmic inclusions, as "no inclusion cells."Conversely, oligodendrocytes with cytoplasmic TPPP-IR inclusions, in conjunction with cell-specific markers but devoid of TPPP-IR inclusion in the nucleus, were designated as TPPPimmunoreactive GCIs.Individual measurements of SNCA and TPPP area density from each cell were pooled into distinct morphological types.The correlation analysis was conducted between SNCA and TPPP transcripts within the same cells.The parameters for image analysis were standardized and consistently applied throughout the entire experiment.

snRNA-seq and data processing
We employed snRNA-seq data retrieved from our prior publication (Forrest et al., 2023), with a specific focus on TPPP, utilizing frontal cortex samples from three control cases.Cell clustering was conducted through the k-means algorithm, and the outcomes were visually represented using UMAP (Uniform Manifold Approximation and Projection).To categorize cell types, established brain cell type markers (Hu et al., 2023) were used and cell types were annotated based on the differentially expressed genes within each cluster (Nassir et al., 2021).Heatmaps and graphs were generated using the plotly package in Python.

Statistical analysis
For categorical variables, Fisher's exact test was employed, while the Mann-Whitney U test was used for continuous variables to compare demographic characteristics across cases.To compare area density of SNCA and TPPP transcripts, a linear mixed model was applied to take into account variability within and across cases.Fixed effects included disease group (control and MSA), cell types (neurons and oligodendrocytes), and morphologies (neurons and oligodendrocytes without α-syn-IR, GCIs, GNIs, NCIs, NNIs, α-syn-immunoreactive inclusions in the nucleus and cytoplasm, and TPPP-immunoreactive inclusions in the cytoplasm).Cases were designated as a random effect.Fisher's least significant difference test served as a post hoc test to identify statistically significant mean differences and adjust for multiple comparisons.Analysis for covariance (ANCOVA) was also performed in MSA to show and adjust the variability across cases.Spearman's rho analysis was applied for the correlation analysis between SNCA and TPPP transcripts.The expression of TPPP transcripts using snRNA-seq data was compared through t-test with Bonferroni correction.A two-sided p-value <0.05 was considered statistically significant.

Demographics and neuropathological data of cases
Demographic and neuropathological information for the patients included in this study is summarized in Table 1.In MSA cases, the age at death ranged from 61 to 64 years (median 62), while controls ranged from 52 to 77 years (median 74).The postmortem interval for MSA cases varied from 4 to 45 h, and for controls, it ranged from 4 to 38 h.No α-syn pathology was observed in the controls.

DNase and RNase treatment for RNAscope
DNase treatment digests DAPI signals, but not for the RNAscope signals or immunofluorescence (Fig. S2A in Supplementary file).RNase digests most of the RNAscope signals, but not for immunofluorescence (Fig. S2B in Supplementary file).These results validate that RNAscope assay detects RNA signals.

Cell-type specific markers between control and MSA
Before analyzing SNCA and TPPP transcripts, we compared the area density of cell-type-specific markers.We used Olig2 for oligodendrocytes and RBFOX3 for neurons, comparing controls to MSA cases.Olig2 area density in pooled oligodendrocytes without α-syn-IR inclusions (n = 224 in control vs n = 210 in MSA) showed no significant difference between the two groups (p = 0.964, Supplementary Fig. S3A).Similarly, RBFOX3 area density in pooled neurons without α-syn-IR inclusions (n = 100 in control vs n = 62 in MSA) did not differ significantly (p = 0.204, Supplementary Fig. S3B).

SNCA transcripts in oligodendrocytes
The signals from SNCA transcripts exhibited clear distinction from autofluorescence (Fig. 2A, B).In agreement with our prior findings (Kon et al., 2023), only a limited number of SNCA transcripts were detected in the nucleus and cytoplasm of the oligodendrocytes in the pons in control (Fig. 2A).SNCA transcripts were also observed in the oligodendrocytes without p-α-syn-IR in MSA (Fig. 2B).Furthermore, SNCA transcripts were present in the nucleus, cytoplasm, and α-syn-IR inclusions within GCI and glial nuclear inclusion (GNI)-bearing oligodendrocytes in MSA (Fig. 2B, Supplementary Video 1).Quantitative SNCA area densities of oligodendrocytes without p-α-syn-IR across the total cell body, nucleus, and cytoplasm were not statistically different between controls and MSA cases (Fig. 2C, Supplementary Fig. S4).However, significantly higher SNCA area densities were observed in GCI-and GNI-bearing oligodendrocytes compared to oligodendrocytes without p-α-syn-IR across the total cell body (all p < 0.0001, Supplementary (caption on next page) T. Kon et al. 0.01), and cytoplasm (all p < 0.0001) in MSA (Fig. 2D).GNI-bearing oligodendrocytes had significantly higher SNCA area densities than GCI-bearing oligodendrocytes in the total cell body (p < 0.05) and cytoplasm (p < 0.01).SNCA area density was increased within cytoplasmic p-α-syn-IR inclusions compared with the cytoplasmic area defined by a 1.5 μm radius enlarged from the nucleus (refer to methods Section 2.4 Morphometry) in oligodendrocytes with GCIs (Fig. 2E).No SNCA area density change was observed within nuclear p-α-syn-IR inclusion compared to the nucleus in GNI-bearing oligodendrocytes (Supplementary Fig. S5A).In the area analyzed, oligodendrocytes without p-α-syn-IR inclusion exhibited a 1.1-fold larger area (p < 0.0001) in MSA cases compared to controls.In MSA, the total cell body area was 1.05-fold larger in GCIs-bearing oligodendrocytes than in oligodendrocytes without p-α-syn-IR inclusion, although this difference did not reach statistical significance (p = 0.053).No p-α-syn-IR was observed in the control cases.

SNCA transcripts in neurons
In line with our previous findings (Kon et al., 2023), numerous SNCA transcripts were expressed in the nucleus and cytoplasm of neurons without p-α-syn-IR in the pontine base both in controls (Fig. 3A) and MSA cases (Fig. 3B).SNCA transcripts were also present in the nucleus, cytoplasm, and inclusions in NCI-and NNI-bearing neurons in MSA (Fig. 3B).Quantitative SNCA area density of neurons without p-α-syn-IR did not exhibit statistical difference between controls and MSA cases (Fig. 3C, Supplementary Fig. S6).Notably, the total cell body and cytoplasmic SNCA area densities were significantly lower in NCI-bearing neurons than in neurons without p-α-syn-IR (all p < 0.001, Supplementary Table S2), while nuclear SNCA area density did not differ between NCI-bearing neurons and neurons without p-α-syn-IR (Fig. 3D) Area densities of SNCA transcripts within inclusions were comparable both in the cytoplasmic inclusion and cytoplasm in NCI-bearing neurons (Fig. 3E) as well as in nuclear inclusion and nucleus in NNI-bearing neurons (Supplemental Fig. S5B).
Subsequently, we compared the SNCA area densities across the total cell body, nucleus, and cytoplasm between pooled neurons and oligodendrocytes.Our analysis revealed that mean SNCA area densities were 5.3-fold higher in neurons without p-α-syn-IR than in oligodendrocytes lacking α-syn-IR both in control and MSA cases (all p < 0.0001).Within MSA inclusions, the mean SNCA area density was 2.8-fold higher in NCIbearing neurons than in GCI-bearing oligodendrocytes (p < 0.0001).Regarding the area in neurons without p-α-syn-IR inclusion, the total cell body area was 0.7-fold smaller (p < 0.0001) in MSA cases than in controls.In MSA, the total cell body area in NCI-bearing neurons was not different from neurons without p-α-syn-IR inclusion (p = 0.167).No pα-syn inclusions were observed in the control cases.

p-α-syn antibody
Initially, we investigated TPPP transcripts using the p-α-syn antibody.TPPP transcripts were expressed in the nucleus and cytoplasm of the oligodendrocytes without p-α-syn-IR in the pons both in controls and MSA cases (Fig. 4A, B).TPPP transcripts were also observed in the nucleus, cytoplasm, and inclusions in GCI-and GNI-bearing oligodendrocytes in MSA (Fig. 4B).Quantitative TPPP area densities of oligodendrocytes without p-α-syn-IR were similar between controls and MSA cases across total cell body, nucleus, and cytoplasm (Fig. 4C, Supplementary Fig. S7).However, TPPP area density was significantly higher in the GCI-and GNI-bearing oligodendrocytes than in oligodendrocytes without p-α-syn-IR in the total cell body (p < 0.001 in GCIbearing oligodendrocytes, p < 0.0001 in GNI-bearing oligodendrocytes, Supplementary Table 2) and cytoplasm (all p < 0.0001) in MSA but not in the nucleus in GCI-bearing oligodendrocytes (Fig. 4D).GNIs had higher nuclear TPPP area density than oligodendrocytes without pα-syn-IR (Fig. 4D).GNI-bearing oligodendrocytes had higher TPPP area densities in the total cell body (p < 0.001) and nucleus (p < 0.001) than GCI-bearing oligodendrocytes (Fig. 4D).Area density of TPPP transcripts was greater in cytoplasmic p-α-syn inclusion compared with the cytoplasm in GCI-bearing oligodendrocytes (p < 0.0001) (Fig. 4E).No TPPP transcripts area density differences were found between nuclear p-α-syn inclusion and nucleus in GNI-bearing oligodendrocytes (Supplementary Fig. S8A).
The area densities of SNCA transcripts were correlated with TPPP transcripts in oligodendrocytes without α-syn-immunoreactive inclusion in control and MSA cases in the total cell body (Supplementary Fig. S9), nucleus, and cytoplasm.Notably, the same correlation was also evident in GCI-bearing oligodendrocytes in MSA.All the calculated r-values reached statistical significance.

TPPP-antibody
Subsequently, we explored TPPP transcripts using the TPPP antibody.While TPPP-IR was predominantly localized in the nucleus and myelin of oligodendrocytes without inclusions in the control (Fig. 5A) and MSA (Fig. 5B), it was prominent in the region of cytoplasmic inclusions in MSA (Fig. 5B).TPPP transcripts were also expressed in the nucleus and cytoplasm of oligodendrocytes without cytoplasmic inclusion in the pons, both in controls and MSA cases (Fig. 5A, B).TPPP transcripts were also observed in the nucleus, cytoplasm, and inclusions in GCI-and GNI-bearing oligodendrocytes in MSA (Fig. 5B, Supplementary Video 2).Quantitative TPPP area density of oligodendrocytes without TPPP-immunoreactive cytoplasmic inclusions exhibited similarity between control and MSA cases (Fig. 5C, Supplementary Fig. S10).However, TPPP area densities of GCI-bearing oligodendrocytes were significantly higher than oligodendrocytes without inclusions in the total cell body and cytoplasm in MSA (all p < 0.0001), but not in the SNCA transcripts area densities in pooled oligodendrocytes without inclusions are similar between control (n = 224 in total, 111 from control 1 and 113 from control 2) and MSA (n = 210 in total, 53 from MSA 1, 53 from MS A2, 51 from MSA 3, and 53 from MSA 4) throughout the total cell body, nucleus, and cytoplasmic area.(D) SNCA transcripts area densities in pooled oligodendrocytes with and without inclusions in MSA.SNCA transcript area densities are significantly greater in GCIbearing oligodendrocytes (n = 173 in total, 45 from MSA 1, 45 from MSA 2, 42 from MSA 3, and 41 from MSA 4) and in GNI-bearing oligodendrocytes (n = 41 in total, 7 from MSA 1, 7 from MSA 2, 13 from MSA 3, and 14 from MSA 4) compared with oligodendrocytes without inclusion (n = 210) across the total cell body, nucleus, and cytoplasmic area.SNCA transcript area density is greater in GNI-bearing oligodendrocytes than in GCI-bearing oligodendrocytes in the total cell body and cytoplasmic area compared with oligodendrocytes lacking inclusions.(E) The SNCA transcript area density is greater in the inclusion compared to the cytoplasm in GCI-bearing oligodendrocytes.Bar graphs represent mean and error bars demonstrate standard deviation.Linear mixed model and Fisher's least significant difference test are applied to statistics.*, p < 0.05; **, p < 0.01; ****, p < 0.0001.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) T. Kon et al. (caption on next page) T. Kon et al. nucleus (Fig. 5D).Area density of TPPP was greater within cytoplasmic TPPP-immunoreactive inclusion compared with cytoplasm in GCIbearing oligodendrocytes (p < 0.0001) (Fig. 5E).No TPPPimmunoreactive glial nuclear inclusions were observed in our study.

p-α-syn antibody
Initially, when p-α-syn antibody was utilized, unexpectedly TPPP transcript expression was detected in the nucleus and cytoplasm of neurons without p-α-syn IR in the pontine base, both in control (Fig. 6A) and MSA cases (Fig. 6B).TPPP transcripts were also observed in the nucleus, cytoplasm, and inclusions in NCI-and NNI-bearing neurons in MSA (Fig. 6B).Quantitative TPPP area density of neurons without pα-syn IR showed similarity between controls and MSA cases (Fig. 6C, Supplementary Fig. S11).TPPP area density of NCI-bearing neurons was significantly lower than neurons without p-α-syn IR in the total cell body (p < 0.01), but not in other areas (Fig. 6D).Area densities of TPPP transcripts within inclusions were comparable both in the cytoplasmic inclusion and cytoplasm in NCI-bearing neurons (Fig. 6E) as well as in nuclear inclusion and nucleus in NNI-bearing neurons (Supplemental Fig. S8B).

TPPP antibody
Given the unexpected detection of TPPP transcripts in neurons, the TPPP antibody was employed to confirm the presence of TPPP protein in neurons.However, consistent with a previous report (Lindersson et al., 2005), the anti-TPPP antibody could not detect a convincing immunoreactivity in neurons in both controls (Fig. 7A) and MSA (Fig. 7B) including NCI-and NNI-bearing neurons.

SNCA and TPPP transcripts with SYN-1 immunostaining
Combination of SNCA and TPPP RNAscope with SYN-1 α-syn immunostaining revealed oligodendrocytes with both SNCA and TPPP transcript expression and fine dot-like SYN-1 cytoplasmic immunoreactivity.This pattern was not seen with when p-α-syn antibody.SYN-1 immunoreactivity in normal-looking oligodendrocytes, both in controls and MSA, mostly fully co-localized with the SNCA RNA signal, but not with TPPP (Supplementary Fig. S12A).In a smaller proportion of normal-looking oligodendrocytes (approximately 10-15%), incomplete co-localization of SNCA transcripts with SYN-1 protein signal was also observed (Supplementary Fig. S12B).GCIs in MSA were labeled by SYN-1 and also showed SNCA and TPPP transcripts (Supplementary Fig. S12C); in these cells, the area of SYN-1 immunoreactivity exceeds those of the SNCA transcripts.

TPPP transcripts in snRNA-seq
Based on our RNAscope analysis, we unexpectedly detected TPPP transcripts in neurons in control and MSA.This finding challenges the currently incomplete understanding of TPPP transcript expression in neurons, as TPPP protein has traditionally been considered as primarily an oligodendroglial protein (Goldbaum et al., 2008;Lehotzky et al., 2010;Skjoerringe et al., 2006;Takahashi et al., 1991).Consequently, we further extracted TPPP mRNA expression from the frontal cortex snRNAseq data (Forrest et al., 2023).The cluster annotation based on the expression of established marker genes associated with known cell types uncovered a distinct TPPP transcript expressional pattern specific to each cell type (Fig. 8A).Mature oligodendrocytes exhibited the most abundant TPPP transcript compared to other cell types (all p < 0.0001, Fig. 8B).Notably, a substantial expression of TPPP transcripts was observed in excitatory and inhibitory neurons.In contrast, TPPP expression was low to absent in oligodendrocyte progenitor cells (OPCs), astrocytes, microglia, and endothelial cells.

Summary of the study
This study revealed greater density of SNCA and TPPP transcripts in GCI-and GNI-bearing oligodendrocytes compared to oligodendrocytes lacking inclusions in the pontine base in MSA (summarized in Supplemental Table S3).Our findings suggest that an increase in RNA expression contributes to the expression of endogenous α-syn may serve as a source for misfolded α-syn templates, contributing to the development of α-syn-immunoreactive oligodendroglial inclusions in MSA.Furthermore, TPPP transcripts exhibit similar behavior to SNCA transcripts, supporting the notion that TPPP contributes to inclusion formation in oligodendrocytes.

SNCA transcripts in oligodendrocytes
The origin of accumulated misfolded α-syn in GCIs has remained unclear.Some researchers have reported the absence of oligodendroglial physiological α-syn and neuronal α-syn transmission to oligodendrocytes (Ettle et al., 2014;Kaji et al., 2018;Kisos et al., 2012;Konno et al., Fig. 3. Representative RNAscope images and SNCA area density in neurons.(A) Neurons exhibit numerous SNCA transcripts both in the nucleus and cytoplasm in control.(B) A neuronal cytoplasmic inclusion (NCI, double-arrow)-bearing neurons shows fewer SNCA transcripts than neurons without inclusion (arrows), while a neuronal nuclear inclusion (NNI, arrowhead)-bearing neurons demonstrates similar SNCA transcripts compared with neurons without p-α-syn inclusions (arrows) in MSA.Each image set represents confocal images taken from the same field of view showing DAPI (blue), phosphorylated-α-syn (p-Syn) immunostaining (green), SNCA transcripts (red), neuron-specific RBFOX3 transcripts (white), and the enlarged merged image.A is obtained from control 2, the left panel of B is obtained from MSA 3, and the right panel of B is obtained from MSA 1. Scale bars represent 10 μm.(C) Quantitative SNCA transcript area densities in pooled neurons without inclusions are similar between control (n = 100 in total, 50 from control 1 and 50 from control 2) and MSA (n = 62 in total, 16 from MSA 1, 10 from MSA 2, 18 from MSA 3, and 18 from MSA 4) across the total cell body, nucleus, and cytoplasmic area.(D) SNCA transcripts area densities in pooled neurons with and without inclusions in MSA.SNCA transcript area density is lower in NCI-bearing neurons (n = 26 in total, 12 from MSA 1, 8 from MSA 2, 1 from MSA 3, and 5 from MSA 4) compared with neurons without inclusion (n = 62) in the total cell body and cytoplasmic area.A significantly lower SNCA transcript area density is observed in NCI-bearing neurons than in NNI-bearing neurons (n = 22 in total, 6 from MSA 1, 5 from MSA 2, 5 from MSA 3, and 6 from MSA 4) across the total cell body, nucleus, and cytoplasmic area.(E) SNCA transcript area density in cytoplasmic inclusion is similar to that in the cytoplasm in NCI-bearing neurons.Bar graphs represent mean and error bars demonstrate standard deviation.Linear mixed model and Fisher's least significant difference test are applied to statistics.**, p < 0.01; ***, p < 0.001.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Our previous study demonstrated that while oligodendrocytes showed significantly lower levels of SNCA transcripts than in neurons, these transcripts were indeed present in oligodendrocytes from control brains using RNAscope and snRNA-seq (Kon et al., 2023).In this study, we describe the presence of a similar level of SNCA transcripts in pα-syn-immunonegative oligodendrocytes in both control and MSA brains, although significantly lower amounts than in neurons.In addition, we demonstrate a significantly greater density of SNCA transcripts in the nucleus and cytoplasm in GCI-and GNI-bearing oligodendrocytes compared to oligodendrocytes without p-α-syn-IR.We expand previous studies on various models showing α-syn-IR in oligodendrocytes including induced pluripotent stem cells produced from fibroblasts of patients diagnosed with MSA (Djelloul et al., 2015).In our study, SYN-1 antibody (detecting the monomer form) showed fine dot-like deposits and co-localization with SNCA but not TPPP RNA transcripts in normallooking oligodendrocytes.Since there were cells that showed SNCA transcripts without the colocalization of SYN-1 immunoreactivity, we interpret this cautiously that this is not a non-specific cross-reaction but, in some oligodendrocytes, indeed there is a protein translation.We propose to confirm this using other methods (Savulescu et al., 2021), but this could suggest that monomeric α-syn could be present in oligodendrocytes allowing a pool of proteins to be misfolded and build inclusions in oligodendrocytes.Despite the absence of a statistically significant difference, Asi et al. (Asi et al., 2014) reported greater SNCA transcripts in MSA oligodendrocytes when compared with controls.However, caution is warranted in interpreting their findings, as they analyzed oligodendrocytes without considering the presence or absence of GCIs.In Parkinson's disease (PD), SNCA multiplications (duplication and triplication) lead to familial PD, suggesting a gene-dosage effect (Farrer et al., 2004;Fujishiro et al., 2013;Miller et al., 2004;Riederer et al., 2019).Although no SNCA multiplications have been reported in MSA, cases carrying SNCA mutations and duplications frequently exhibit pathological α-syn-immunoreactive cytoplasmic inclusions in oligodendrocytes (Fujishiro et al., 2013;Kiely et al., 2013;Kiely et al., 2015;Konno et al., 2016;Pasanen et al., 2014).Garicia-Segura et al. recently investigated the gain of SNCA copy number variants, potentially increasing mRNA, in postmortem MSA brains using DNA fluorescent in situ hybridization (Garcia-Segura et al., 2023).They found that inclusion-harboring oligodendrocytes had significantly more copy number variants than those without inclusions in the substantia nigra from MSA cases.Moreover, mouse models of MSA employing oligodendroglial SNCA overexpression develop GCI-like aggregates in oligodendrocytes (Kahle et al., 2002;Shults et al., 2005;Tanji et al., 2019;Yazawa et al., 2005).The aberrant cellular accumulation of α-syn is linked to an increase in endogenous α-syn expression (Musgrove et al., 2011).Furthermore, misfolded α-syn seeds exhibit the capability to actively recruit endogenous monomeric α-syn (Holec et al., 2022;Martinez-Valbuena et al., 2022a;Martinez-Valbuena et al., 2022b;Rossi et al., 2020).Based on these findings, the observed greater density of SNCA transcripts in GCI-and GNI-bearing oligodendrocytes suggests that an increase in transcripts can be a response of the cell to increase translation into monomer α-syn that then potentially serves as a template for misfolded α-syn, facilitating further progression of the aggregation process.However, we cannot exclude the possibility that the uptake of exogenous α-syn, primarily from neurons, influences the mechanism of GCI development in oligodendrocytes, as this study did not investigate this neuron-to-oligodendrocyte propagation phenomenon.Oligodendrocytes might also participate in pruning axons bearing α-syn-immunoreactive neurites.There is a possibility that, in addition to local production, SNCA transcripts and aggregated α-syn, originally present in neurons, may be taken up by oligodendrocytes.In this scenario, they would come into contact with the oligodendroglial TPPP transcripts and proteins.
The discrepancy between previous reports, which described the absence of SNCA in oligodendrocytes, and our results may stem from methodological differences.Studies (Langerveld et al., 2007;Ozawa et al., 2001) utilizing tissue-digested method cannot directly compare SNCA expression in specific cell types and morphologies such as neurons, oligodendrocytes, GCIs, and NCIs across control and MSA brains, as these approaches provide only averaged expression levels across the diverse array of cell types present in the brain.Furthermore, it is plausible that conventional in situ hybridization studies (Jin et al., 2008;Miller et al., 2005;Solano et al., 2000) failed to detect low levels of SNCA expression due to their low sensitivity.

SNCA transcripts in neurons
In our study, the cytoplasmic SNCA transcripts area density was significantly lower in NCI-bearing neurons compared to neurons without p-α-syn-IR but was the same in NNI-bearing neurons.This observation aligns with our prior investigation (Kon et al., 2023), where we noted a significant decrease in SNCA area density in LBs compared to neurons without disease-associated α-syn-IR and punctate α-syn-IR (preaggregates) in LBD.SNCA transcript decrease in MSA neurons when compared with controls has been reported previously (Asi et al., 2014).However, caution is warranted in interpreting their findings, as they analyzed neurons without considering the presence or absence of NCIs and lacked statistical significance.In prion diseases, the leading model of neurodegenerative proteinopathies linked to misfolded prion protein (PrP), the progressive accumulation of misfolded PrP results in a depletion of the pool of physiological cellular PrP (Kovacs et al., 2002).Consequently, the decrease in cytoplasmic SNCA area density in NCIbearing neurons found in this study likely reflects exhaustion of transcription due to continuous α-syn production, potentially utilized for the   seeding of misfolded α-syn.NNI can manifest independently of NCIs (Wakabayashi et al., 2005), occurring early before evident neuronal loss (Kon et al., 2013;Yoshida, 2007).Additionally, neurons containing NNIs occasionally exhibit diffuse cytoplasmic α-syn-IR, implying pre-NCIs (Yoshida, 2007).These observations suggest that the nuclear translocation of α-syn is likely an early event in neuronal pathology in MSA (Wakabayashi et al., 2024;Yoshida, 2007).Based on these findings, we propose that the preserved SNCA transcript area density in NNIbearing neurons signifies cellular efforts to maintain its physiological function, as indicated by neurons with pre-aggregates in LBD (Kon et al., 2023).It is noteworthy that the discrepancy of SNCA transcript expressional pattern between NCI-bearing neurons and GCI-bearing oligodendrocytes suggests a distinct cell-type-specific vulnerability against α-syn.The evidence that glial cells exhibit a greater capacity for protein degradation may provide a plausible explanation for the observed discrepancy (Jansen et al., 2014;Tydlacka et al., 2008).

TPPP transcripts in oligodendrocytes
TPPP transcripts area density was significantly greater in the cytoplasm of GCI-bearing oligodendrocytes and in both the nucleus and cytoplasm of GNI-bearing oligodendrocytes when compared to oligodendrocytes lacking p-α-syn-IR in this study.This elevation pattern closely resembled that of SNCA transcripts, implying a parallel expression of TPPP protein with α-syn during the development of GCI-bearing oligodendrocytes.Transfecting TPPP significantly enhances SNCA transcripts in cultured oligodendrocytes (Cui et al., 2021).Moreover, TPPP protein in oligodendrocytes transforms their limited cellular α-syn into stable, insoluble, and toxic species (Ferreira et al., 2021;Mavroeidi et al., 2019) and modulates the rate and extent of α-syn seeding (Mavroeidi et al., 2019).In addition, there is a loss of myeline and a decrease in TPPP immunoreactivity in MSA white matter.This neurodegeneration may potentially trigger a response, leading to an increase in TPPP transcripts in oligodendrocytes in MSA.Given these findings, the concurrent greater density, and thus presumed increase in TPPP and SNCA transcripts, may synergistically contribute to the formation and extent of GCIs, as summarized in Fig. 9.

TPPP transcripts in neurons
Despite observing substantial TPPP transcripts in neurons through RNAscope and snRNA-seq studies, the TPPP antibody failed to detect convincing immunopositivity in neurons.Previous studies reported TPPP-IR in NCIs and NNIs in MSA (Baker et al., 2006;Kovacs et al., 2007), with Baker et al. (Baker et al., 2006) also noting TPPP-IR in a small portion of neurons without disease-associated α-syn-IR in the pons and medulla oblongata in MSA.Lindersson et al. (Lindersson et al., 2005) identified TPPP-IR in normal neurons, exclusively in the supraoptic nucleus in wild-type rats.The discrepancy between abundant TPPP transcript expression and the absence of TPPP-IR in neurons may be attributed to differences in the epitopes detected by currently available antibodies.We observed significant lower TPPP transcripts densities in NCIs compared with neurons without p-α-syn-IR.The abundance of NCIbearing neurons are increased in moderate-stage compared to earlystage MSA, and the number is markedly decreased in severe lesions (Nishie et al., 2004;Yokoyama et al., 2001), suggesting that NCI-bearing neurons are dying neurons (Wakabayashi et al., 2024).Lower, presumably decreased, TPPP transcript area density observed in this study implies that exhaustion of cellular function occurs due to misfolded α-syn.The possibility remains that substantial TPPP expression in neurons templates for TPPP protein accumulation in LBs.

Correlation between mRNA and protein expression
The fundamental concept of molecular biology emphasizes that proteins are synthesized from mRNA templates (Buccitelli and Selbach, 2020).Consequently, it is commonly observed that transcript expression levels correlate with protein synthesis (Buccitelli and Selbach, 2020;Farrer et al., 2004;Miller et al., 2004;Riederer et al., 2019;Trabzuni et al., 2012).However, the correlation between mRNA and protein expression levels varies widely and is imperfect (Buccitelli and Selbach, 2020;Gry et al., 2009).In the case of α-syn, alterations in SNCA mRNA levels in cell culture and animal models correspond to changes in α-syn protein expression (Tagliafierro and Chiba-Falek, 2016;Zhang et al., 2023).Notably, SNCA triplication displayed a two-fold over-expression of SNCA mRNA and α-syn protein (Farrer et al., 2004).Interestingly, for another neurodegenerative disease-related protein, tau, regional variations in total tau protein expression levels correlate with similar changes in mRNA expression levels evaluated using the quantitative RT-PCR method (Trabzuni et al., 2012).In our study, however, we show discrepancies between cell types regarding the translation of RNA to protein; this is exemplified by the lack of clear protein signals in neurons expressing TPPP mRNA.

Limitations of the study
It is crucial to acknowledge certain limitations in this study.Detailed morphometric evaluations were conducted on a limited number of cases, warranting validation in larger cohorts.However, the cytopathologies were well-represented, and the clarity of preserved SNCA and TPPP transcripts was consistent across all cases.Additionally, the cytoplasmic area defined in our study is an artificial construct and may not precisely correspond to the true cytoplasmic area.However, recognizing the true cytoplasmic area is challenging in formalin-fixed paraffin sections as oligodendroglial cytoplasm is often invisible.We applied the same method for analyzing both oligodendrocytes with and without inclusions, ensuring that the defined cytoplasm at least included perinuclear cytoplasmic areas.Furthermore, in this study, our focus was exclusively on the pons-a region known to be more severely affected in MSA-C than in MSA-P.However, it is important to note that the pons is typically affected in both MSA subtypes (Ozawa et al., 2004).To validate our findings, future investigations should extend to other brain regions, including the putamen and cortical areas.Unfortunately, due to limited tissue availability, we were unable to directly compare the snRNA-seq data between MSA cases and controls.Therefore, further research is necessary to elucidate the differences in snRNA-seq profiles.We focused on comparing normal-looking cells with those with p-α-syn-IR and the controls were used to show that the method works.Although we did not detect robust differences between normal-looking cells in controls and MSA, this observation merits further confirmation.Finally, this study cannot address whether the change in the morphology of the cells affects the intensity of RNA transcripts.

Conclusions
Our study revealed greater SNCA and TPPP transcripts densities in GCI-and GNI-bearing oligodendrocytes compared with oligodendrocytes without inclusions in the pontine base in MSA.Our findings suggest that an increase in RNA expressions may lead to an increase in endogenous TPPP and α-syn, and the latter may provide templates for the development of misfolded α-syn and the formation of oligodendroglial inclusions in MSA.This study enhances our understanding of MSA pathogenesis by elucidating the upstream regulation of SNCA and TPPP expression during inclusion formation and unveiling the cell-typespecific diversity of TPPP expression.The regulation of their transcripts may hold promise for future molecular therapy in MSA.We underscore novel aspects of disease pathogenesis that will be pertinent for basic researchers investigating cellular mechanisms in synucleinopathies.

Research ethics and patient consent
This study was approved by the University Health Network (UHN) Research Ethics Board (Nr.20-5258) and the University of Toronto (Nr.39459) and was performed per ethical standards established in the 1964 Declaration of Helsinki, updated in 2008.

Fig. 1 .
Fig. 1.Method of outlining the region of interest and capturing the positive signals in RNAscope.Raw data images of the pontine base sections in a case of MSA (A, B).The nuclear area is identified by capturing DAPI signal intensity (C, D).In oligodendrocytes, the region of the total cell body is then automatically extended in a radius length by the NIS-Elements AR software (C).In neurons, the border of the neuronal cell body is manually traced after the Lookup Table strength of DAPI is sufficiently enhanced to visible the autofluorescence from lipofuscin pigments (D).Phosphorylated-α-syn or TPPP-immunoreactive areas are automatically captured above the programmed threshold (E, F).The software captured the area positive for the target transcripts above the threshold within the annotated regions of interest (G, H).Red in A, C, E, and G display SNCA transcripts, Red in B, D, F, and H demonstrate TPPP transcripts, white in A, C, E, and G show Olig2 transcripts, white in B, D, F, and H represents RBFOX3 transcripts, green displays phosphorylated -α-syn, and blue exhibits DAPI.A, C, E, and G are obtained from MSA 4, and B, D, F, and H are obtained from MSA 3. Scale bar represents 5 μm.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2 .
Fig. 2. Representative RNAscope images and SNCA area density in oligodendrocytes.(A) Oligodendrocytes (arrows) exhibit a few SNCA transcripts both in the nucleus and cytoplasm, while neurons (arrowheads) show numerous SNCA transcripts in control.(B) A glial cytoplasmic inclusion (GCI, double-arrow) and a glial nuclear inclusion (GNI, double-arrowheads) show more SNCA transcripts compared with oligodendrocytes without p-α-syn inclusions (arrows) in MSA.Each image set represents confocal images taken from the same field of view showing DAPI (blue), phosphorylated-α-syn (p-Syn) immunostaining (green), SNCA transcripts (red), oligodendrocyte-specific Olig2 transcripts (white), and the enlarged merged image.A is obtained from control 1, the left panel of B is obtained from MSA 4, and the right panel of B is obtained from MSA 3. Scale bars represent 10 μm.(C) Quantitative

Fig. 4 .
Fig. 4. Representative RNAscope images and TPPP area density in oligodendrocytes.(A) Oligodendrocytes (arrows) exhibit TPPP transcripts both in the nucleus and cytoplasm, neurons (arrowheads) also show numerous TPPP transcripts in control.(B) A glial cytoplasmic inclusion (GCI, double-arrow)-bearing oligodendrocytes and a glial nuclear inclusion (GNI, double-arrowheads)-bearing oligodendrocytes show more TPPP transcripts compared with oligodendrocytes without p-α-syn inclusions (arrows) in MSA.Each image set represents confocal images taken from the same field of view showing DAPI (blue), phosphorylated-α-syn (p-Syn) immunostaining (green), TPPP transcripts (red), oligodendrocyte-specific Olig2 transcripts (white), and the enlarged merged image.A is obtained from control 1, the left panel of B is obtained from MSA 3, and the right panel of B is obtained from MSA 2. Scale bars represent 10 μm.(C) Quantitative TPPP transcript area densities in pooled oligodendrocytes without inclusions are similar between control (n = 224) and MSA (n = 210) across the total cell body, nucleus, and cytoplasmic area.(D) TPPP transcript area densities in pooled oligodendrocytes with and without inclusions in MSA.TPPP transcript area density is significantly greater in GCI-bearing oligodendrocytes (n = 173) and in GNI-bearing oligodendrocytes (n = 41) compared with oligodendrocytes without inclusion (n = 210) in the total cell body and cytoplasmic area.A significantly greater nuclear TPPP transcript area density in GCI-bearing oligodendrocytes is observed compared with oligodendrocytes without inclusion.TPPP transcript area density is greater in GNI-bearing oligodendrocytes than in GCIbearing oligodendrocytes in the total cell body and nuclear area.(E) A significantly greater TPPP transcript area density is detected in the inclusion than in the cytoplasm in GCI-bearing oligodendrocytes.Bar graphs represent mean and error bars demonstrate standard deviation.Linear mixed model and Fisher's least significant difference test are applied to statistics.***, p < 0.001; ****, p < 0.0001.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5 .
Fig. 5. Representative RNAscope images and TPPP area density in TPPP-immunoreactive oligodendrocytes.(A) Oligodendrocytes exhibit TPPP transcripts both in the nucleus and cytoplasm in control.(B) TPPP-immunoreactive glial cytoplasmic inclusions (GCIs, arrowheads)-bearing oligodendrocytes show more TPPP transcripts compared with oligodendrocytes without TPPP-immunoreactive inclusions (arrow) in MSA.Each image set represents confocal images taken from the same field of view showing DAPI (blue), TPPP immunostaining (green), TPPP transcripts (red), oligodendrocytespecific Olig2 transcripts (white), and the enlarged merged image.A is obtained from control 2, B is obtained from MSA 4. Scale bars represent 10 μm.(C) Quantitative TPPP transcript area densities in pooled oligodendrocytes without inclusions are similar between control (n = 210) and MSA (n = 196) across the total cell body, nucleus, and cytoplasmic area.(D) TPPP transcript area densities in pooled oligodendrocytes with and without inclusions in MSA.TPPP transcript area density is greater in GCI-bearing oligodendrocytes (n = 162) compared with oligodendrocytes without inclusion (n = 196) in the total cell body and cytoplasmic area.(E) A significantly greater TPPP transcript is detected in the inclusion than in the cytoplasm in GCI-bearing oligodendrocytes.Bar graphs represent mean and error bars demonstrate standard deviation.Linear mixed model and Fisher's least significant difference test are applied to statistics.***, p < 0.0001.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6 .
Fig. 6.Representative RNAscope images and TPPP area density in neurons.(A) Neurons (arrows) exhibit numerous TPPP transcripts both in the nucleus and cytoplasm in control.(B) A neuronal cytoplasmic inclusion (NCI, double-arrow)bearing neurons and a neuronal nuclear inclusion (NNI, arrowhead)-bearing neurons show similar TPPP transcripts expression compared with neurons without inclusion (arrows) in MSA.Each image set represents confocal images taken from the same field of view showing DAPI (blue), phosphorylated-α-syn (p-Syn) immunostaining (green), TPPP transcripts (red), neuron-specific RBFOX3 transcripts (white), and the enlarged merged image.A is obtained from control 1, the left panel of B is obtained from control 2, and the right panel of B is obtained from MSA 4. Scale bars represent 10 μm.(C) Quantitative TPPP transcript area densities in pooled neurons without inclusions are similar between control (n = 100) and MSA (n = 62) across the total cell body, nucleus, and cytoplasmic area.(D) TPPP transcripts area densities in pooled neurons with and without inclusions in MSA.TPPP transcript area density in NCI-bearing neurons (n = 26) is significantly less than neurons without inclusion (n = 62) in the total cell body area.A significantly lower TPPP transcript area density is observed in NCI-bearing neurons than in NNI-bearing neurons in the total cell body area.(E) TPPP transcript area density in inclusion is similar to that in the cytoplasm in NCI-bearing neurons.Bar graphs represent mean and error bars demonstrate standard deviation.Linear mixed model and Fisher's least significant difference test are applied to statistics.**, p < 0.01.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 7 .
Fig. 7. Representative RNAscope images in neurons with TPPP immunostaining.(A, B) TPPP-immunoreactivity is not found in neurons (arrows), while oligodendrocytes without inclusion (arrowheads) and glial cytoplasmic inclusions (double arrows) show strong TPPP-immunoreactivity in control (A) and MSA (B).Each image set represents confocal images taken from the same field of view showing DAPI (blue), TPPP immunostaining (green), TPPP transcripts (red), neuron-specific RBFOX3 transcripts (white), and the enlarged merged image.A is obtained from control 1, B is obtained from MSA 4. Scale bars represent 10 μm.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 8 .
Fig. 8. TPPP expression in single-nucleus RNA sequencing in control cases.(A) Expression profile of TPPP in major cell types of the control frontal cortex uncovering a distinct pattern specific to each cell type.Each point represents a single cell and colour intensity indicates the log 2 normalized expression level of TPPP.(B) Quantification of TPPP expression across cell types.The bar graphs represent mean and error bars exhibit standard deviation.For statistics, t-test with Bonferroni correction is applied.Adjusted p values are listed on the right side of the figure.OPC, oligodendrocyte progenitor cell.

Table 1
Case demographics and regions used in the study.