Abstract
The transcription factor TATA-box-binding protein (TBP) plays a crucial role in neurons, as highlighted by its involvement in spinocerebellar ataxia type 17 (SCA17), where pathological polyglutamine expansions disrupt its function. While the molecular pathomechanisms remain unclear, our previous study showed that calcium-activated calpains cleave TBP in SCA17 cell and animal models. Here, we pinpoint a major calpain cleavage site after residue T106, C-terminal to TBP’s polyglutamine tract. The resulting polyglutamine-containing N-terminal fragments mislocalise to the cytoplasm and exhibit increased aggregation. Both the N-terminal and the nuclear-localised C-terminal fragments contribute to calpain overactivation. SCA17 cell line-based multi-omics analyses revealed that TBP fragments alter pathways linked to neuronal function and neurodegeneration, identifying AMOT as a key player. Further analyses in SCA17 cell lines, patient-derived fibroblasts and induced neurons as well as in rat and zebrafish models confirmed AMOT’s pivotal role in neurons, regulated by TBP and mediated via YAP1-dependent signalling. These findings offer further insights into TBP’s neuronal function and suggest broader implications of AMOT for neurodegenerative and neurodevelopmental disorders.
Subject terms: Molecular neuroscience, Proteolysis, Transcriptomics, Spinocerebellar ataxia, Developmental neurogenesis
TBP fragments activate calpains and disrupt AMOT-YAP1 signalling, revealing a pathway involved in normal neural development and implicated in spinocerebellar ataxia type 17.
Introduction
The TATA-box-binding protein (TBP) is an essential subunit of the transcription factor IID (TFIID) complex. Its highly conserved C-terminal core domain recognises TATA-box motifs, facilitating RNA polymerase II recruitment for transcription initiation. TBP also contributes to RNA polymerase I and III activity, underscoring its critical role in eukaryotic gene expression1–3. Complete loss of TBP is lethal, causing early embryonic arrest4.
TBP is crucial for neurological function, supporting neuronal stem cell renewal and differentiation5. Moreover, several TBP-associated factors participate in neuronal differentiation and nervous system development6. Mutations in TBP have been associated with neurological disorders such as Alzheimer’s disease (AD), Parkinson’s disease (PD), and schizophrenia7–9. Notably, polyglutamine (polyQ) expansion mutations in TBP’s exonic CAG/CAA repeat cause spinocerebellar ataxia type 17 (SCA17), an autosomal dominant neurodegenerative disorder characterised by ataxia, dementia, involuntary movements, and psychiatric symptoms. Normal TBP alleles contain 25-40 repeats, while pathogenic alleles exceed 41 repeats, reaching full penetrance at 50. A hallmark of polyQ disorders is the aggregation propensity of expanded polyQ tracts, forming inclusion bodies10–12. Despite its discovery over two decades ago13, the molecular pathomechanisms driving neuronal loss in SCA17 remain unclear.
SCA17 serves as a valuable model disease for TBP’s role in neuronal processes. PolyQ expansions impair TBP’s ability to dimerise, bind DNA, interact with transcription factors, and regulate transcription14–17. Under the toxic fragment hypothesis, proteolytic cleavage of disease-linked proteins is a key event in neurodegeneration18–20, as seen in AD, PD, and polyQ disorders such as Huntington’s disease (HD), and spinocerebellar ataxia type 3 (SCA3)21–27. Interestingly, an internal deletion in the C-terminal region of TBP caused increased aggregation, disrupted TFIID function, and a worsened neurological phenotype in mice15.
While earlier studies failed to link TBP fragmentation to caspase activity25, we recently demonstrated that calpains generate C-terminal TBP fragments in SCA17 cell and rodent models and that inhibiting calpains mitigates molecular pathology28. Here, we investigated the physiological and pathological consequences of TBP cleavage by calpains, identifying specific cleavage sites and generating TBP fragment constructs. Using a multi-omics approach in cells, combined with analysis of neurons differentiated from an induced pluripotent stem cell (iPSC) line of an SCA17 patient, brain tissue from SCA17 rats, and experiments in zebrafish embryos, we investigated the consequences of TBP fragments and identified a TBP-calpain-AMOT regulatory relationship that contributes to normal neural development.
Results
Calpain cleavage of TBP occurs C-terminally of its polyQ tract
Previously, we demonstrated that calpains are involved in the proteolytic processing of TBP, resulting in the generation of C-terminal breakdown products28. To identify calpain-dependent cleavage sites within TBP, we first extended our in vitro characterisation of the TBP fragmentation. For this, C-terminally myc-tagged non-polyQ-expanded (38Q) or polyQ-expanded (106Q) full-length TBP overexpressed in HEK 293T (293T) cells was immunopurified and cleaved by exogenous calpain-1. To identify N- and C-terminal fragments, samples were analysed by western blotting using a selection of antibodies binding along the myc-TBP protein sequence (Fig. 1a). Immunodetections confirmed successful overexpression and purification of full-length TBP 38Q or 106Q (Fig. 1b–d). Expectedly, the addition of exogenous calpain-1 caused fragmentation of both TBP variants. Immunodetection with antibodies specific for both the myc tag and the C-terminus of TBP (antibody 58C9) revealed three C-terminal fragments (c1, c2, and c3) (Fig. 1b). Detections with antibody D5G7Y, which binds around amino acid residue A110, only detected fragment c1 (Fig. 1c), suggesting that the cleavage site for fragment c1 lies N-terminally to the antibody epitope, while cleavage sites for c2 and c3 are located more C-terminally. Probing with the N-terminus-binding antibody D5C9H, as well as 1TBP18 and 8515, revealed N-terminal, polyQ-containing fragments (n1, n2, and n3) corresponding to the C-terminal TBP breakdown products (Fig. 1d and Supplementary Fig. 1a, b). Immunodetection with the polyQ-specific antibody 1C2 confirmed that these N-terminal TBP fragments indeed contain the polyQ stretch (Supplementary Fig. 1c).
Fig. 1. Antibody-based mapping reveals calpain-dependent N-terminal TBP fragments.

a Schematic overview of myc-tagged non-polyQ-expanded (38Q) and polyQ-expanded (106Q) TBP overexpression constructs, highlighting the polyQ stretch and the C-terminal DNA-binding core domain (core), and binding regions of antibodies used in this study. b–d Western blot analysis of TBP fragmentation after immunoprecipitation (IP) from 293T cell extracts, followed by a calpain-cleavage assay (CCA) with calpain-1. Membranes were probed with a myc-tag-specific antibody combined with TBP-specific antibodies targeting the C-terminus (58C9, in b), an epitope around residue A110 (D5G7Y, in c), or the N-terminus (D5C9H, in d). SYPRO Ruby staining served as a loading control and for validation of IP purity. Full protein extracts from TBP-transfected 293T cells (input) were loaded along with the corresponding IP samples. Light blue- and pink-rimmed arrowheads mark full-length TBP 38Q and 106Q, respectively. Red and blue arrowheads indicate N-terminal and C-terminal (n1–3; c1–3) TBP fragments, respectively. Light pink arrowheads mark signal overlaps of TBP protein species detected with myc-tag-specific and respective TBP-specific antibodies. The approximate binding sites of the antibodies used are illustrated above the corresponding blots. Source data are provided as a Source Data file.
Calpain-derived C-terminal TBP fragments are mainly nuclear
TBP is predominantly localised in the nucleus29,30. However, the subcellular distribution of calpain-dependent TBP fragments remained elusive. To examine this further, we induced calpain-dependent TBP fragmentation in cellulo by co-transfecting 293T cells with myc-tagged full-length TBP 38Q or 106Q and FLAG-tagged calpain-1. Immunofluorescence staining and microscopy of TBP with the N-terminally binding antibody 1TBP18 confirmed the predominantly nuclear localisation of TBP, while calpain-1 was detected exclusively in the cytoplasm (Fig. 2a). Analysis of protein extracts from 293T cells co-expressing full-length TBP 38Q or 106Q with calpain-1 via western blotting confirmed the overexpression of calpain-1 and the proteolytic breakdown of its substrates α-spectrin and calpastatin (Fig. 2b–e). Immunodetection of TBP using the N-terminally binding D5C9H and C-terminally binding 58C9 antibodies revealed a reduction in full-length TBP 38Q or 106Q upon overexpression of calpain-1, along with increasing levels of C-terminal TBP fragments c1 and c2, corroborating the fragmentation of TBP by calpains. However, N-terminal TBP fragments were not detectable (Fig. 2f, g). Similar results were observed in 293T cells transfected with the same TBP 38Q or 106Q constructs and treated with ionomycin—an ionophore that depletes intracellular calcium stores and activates calpains31,32. Correspondingly, ionomycin treatment led to reduced levels of full-length TBP and elevated levels of fragments c1 and c2 (Supplementary Fig. 2a, b). To determine the subcellular localisation of calpain-dependent TBP fragments c1 and c2, we repeated the co-expression of TBP 38Q or 106Q along with FLAG-tagged calpain-1 in 293T cells and performed cytoplasmic-nuclear fractionation. Western blot analysis of the resulting subcellular fractions using the C-terminal antibody 58C9 showed significant TBP cleavage upon co-expression of calpain-1 and accumulation of fragments c1 and c2 primarily in the nuclear fraction (Fig. 2h, i). Interestingly, cytoplasmic levels of full-length TBP 38Q and 106Q dropped significantly, while full-length TBP in the nuclear fraction remained unchanged (Fig. 2i). To rule out that calpain-mediated TBP cleavage might occur in the nucleus, we investigated the subcellular localisation of all calpains expressed at appreciable levels in 293T cells based on our transcriptome analysis (Supplementary Fig. 2c). Immunofluorescence staining and microscopy showed that all endogenously expressed calpains—including calpain-1, -2, -5, -7, -10 and -15—as well as the calpain small subunit CAPNS1 and calpastatin were localised to the cytoplasm (Supplementary Fig. 2d). Moreover, knockdown of CAPN1, but not CAPN2, was sufficient to reduce TBP fragment levels, particularly of fragment c1, at baseline (Supplementary Fig. 2e, f), while calpastatin knockdown increased TBP cleavage (Supplementary Fig. 3). These results indicate that TBP fragmentation by calpain occurs in the cytoplasm, followed by the nuclear translocation of TBP fragments c1 and c2.
Fig. 2. Calpain-dependent C-terminal TBP fragments are enriched in the nucleus of SCA17 cells.

a Immunofluorescence staining and microscopy of 293T cells expressing TBP 38Q or 106Q with and without calpain-1 (CAPN1) overexpression. TBP was stained with N-terminally binding TBP antibody (1TBP18; red) and CAPN1 with a specific antibody (green). DAPI was used to counterstain cell nuclei (blue). Scale bar: 10 µm. b Protein extracts of 293T cells expressing TBP 38Q or 106Q with or without CAPN1 were analysed by western blot. Overexpression of CAPN1 was confirmed using antibodies against CAPN1 and the CAPN1 substrates α-spectrin and calpastatin (CAST). GAPDH served as the loading control. Black-rimmed arrowhead indicates full-length α-spectrin, while the grey arrowhead shows the respective breakdown product. c–e Densitometric analysis of western blots shows CAPN1 levels (levels normalised to GAPDH; in c), α-spectrin cleavage ratio (breakdown product (CL) normalised to full-length (FL) α-spectrin; in d) and CAST levels (levels normalised to GAPDH; in e). Obtained values were normalised to the respective control (293T cells expressing TBP 38Q) of each measurement. n = 3 (c) or 4 (d, e) biological replicates. One-sample t test. f TBP 38Q and 106Q were detected using N-terminally (D5C9H) and C-terminally (58C9) binding TBP antibodies. GAPDH served as the loading control. Light blue- and pink-rimmed arrowheads mark full-length TBP 38Q and 106Q, respectively. Blue arrowheads indicate C-terminal fragments (c1 and c2). g Densitometric analysis of western blot detections with the C-terminal TBP antibody 58C9 shows levels of full-length TBP and TBP fragments c1 and c2. Levels were normalised to GAPDH, and obtained values were normalised to the respective control (293T cells expressing TBP 38Q or 106Q) within each sample set. n = 4 biological replicates. One-sample t test. h 293T cells expressing TBP 38Q or 106Q with and without CAPN1 were subjected to cytoplasmic-nuclear fractionation. Whole cell (W), cytoplasmic fraction (C), and nuclear fraction (N) were analysed by western blotting. Immunodetection was conducted using the C-terminally binding TBP antibody 58C9. Nuclear proteins lamin A and C, and cytoplasmic protein α-tubulin served as fraction-purity and loading control. Light blue- and pink-rimmed arrowheads mark full-length TBP 38Q and 106Q, respectively. Blue arrowheads indicate C-terminal fragments (c1 and c2) (f). i Densitometric analysis shows levels of full-length TBP and TBP fragments c1 and c2. Levels were normalised to the subcellular fraction-specific loading control, followed by normalisation to the respective control condition within the same fraction (293T cells expressing TBP 38Q or 106Q). n = 3 biological replicates. One-sample t test). Bars represent means + s.e.m. All tests were two-tailed. P values are shown in the graphs. Source data are provided as a Source Data file.
Mass spectrometry analysis identifies T106 | S107 in TBP as a main calpain-1 cleavage site
To identify calpain-1 cleavage sites within TBP, we performed in vitro cleavage assays of TBP 38Q purified from 293T cells using calpain-1, followed by mass spectrometry (MS) analysis of excised TBP fragment bands after a chymotrypsin digest (Fig. 3a). Calpain-1 cleavage of TBP 38Q was confirmed via western blotting using the N-terminally binding D5C9H antibody and the C-terminally binding 58C9 antibody (Fig. 3b). Subsequent MS analysis identified 13 distinct peptides derived from myc-tagged TBP, each containing one terminus not attributable to chymotryptic cleavage (Supplementary Table 1). Comparison of these termini with in silico calpain cleavage site predictions for TBP (UniProt ID: P20226) using the GPS-CCD tool revealed that 10 peptides corresponded to predicted calpain cleavage sites within the TBP sequence, whereas three mapped to the C-terminal myc-tag and were excluded (Fig. 3c, d and Supplementary Table 1). Five of the identified peptides mapped to a region C-terminal of the polyQ tract, with their N-terminal sequences starting between residues Q104 and Q109 and extending to residue F122. All N-terminal sequences aligned with a motif of increased calpain cleavage likelihood located between residues A96 and Q117. Importantly, the peptide spanning S107 to F122 was the only one detected in all three experimental runs (Fig. 3d, e), indicating that a primary calpain-1 cleavage event takes place between T106 and S107. This cleavage is likely responsible for generating the observed TBP fragments n1 and c1. To validate this finding, we generated calpain cleavage site-mutant forms of myc-tagged TBP 38Q and 106Q, based on our previously established strategy33,34 and subjected them to in vitro cleavage using calpain-1. Western blotting confirmed that mutating amino acids surrounding the T106 | S107 site to tryptophan residues prevented the formation of the n1 and c1 fragments, while, expectedly, shifting cleavage towards adjacent sites (Fig. 3f, g and Supplementary Fig. 4).
Fig. 3. Mass spectrometry analysis identifies T106 | S107 as a calpain-1 cleavage site in TBP.

a Schematic illustration of the experimental workflow for identifying calpain-1 (CAPN1) cleavage sites in TBP using immunocapture and calpain-cleavage assay (CCA)-based mass spectrometry (MS). exCAPN1, exogenous CAPN1. b Western blot detection of immunopurified (IP) and CAPN1-cleaved myc-tagged TBP 38Q using N-terminally binding (D5C9H) and C-terminally binding (58C9) TBP antibodies. The light-blue-rimmed arrowhead indicates full-length TBP 38Q (38Q), while the black-rimmed arrowhead shows endogenous TBP (eg). Red and blue arrowheads indicate N-terminal (n38Q1-3) and C-terminal (c1–3) TBP fragments, respectively. The dashed rectangle highlights the gel section excised for MS analysis of TBP fragments. c Overview of semi-specific peptides detected by MS. Yellow dashed lines show 13 CAPN1 cleavage sites, while pink dashed lines indicate five chymotrypsin cleavage sites along the myc-TBP sequence. d In silico prediction of calpain cleavage sites in TBP (UniProt ID: P20226) using the GPS-CCD tool, overlaid with expected chymotrypsin cleavage sites. Black lines show normalised calpain cleavage likelihood score (nCCLS) along the protein sequence (left y axis), with the black-dotted line indicating the likelihood threshold (nCCLS = 1). Yellow lines represent the proportion of experimental runs (n = 3 experiments, right y axis) in which CAPN1 cleavage-specific peptides were detected by MS. Pink dashed lines mark chymotrypsin cleavage sites. The polyQ region in TBP (highlighted in light blue) produces a prediction artefact with high CCL scores. e MS spectrum of a calpain-specific peptide identified in all three experimental runs, representing CAPN1 cleavage at T106 | S107 in TBP. f In silico prediction of calpain cleavage likelihood between aa 95 and 115 in TBP without (light blue) and with cleavage site mutation (CSM) by quadruple tryptophan (WWWW; green) substitution around the detected cleavage T106 | S107 (S105-Q108). Black-dotted line marks the likelihood threshold (nCCLS = 1). g, h Western blot-based fragmentation analysis of myc-tagged TBP 38Q (g) and 106Q (h) without and with CSM, after myc tag-specific immunoprecipitation (IP) from 293T cell extracts, followed by a time-dependent CCA with CAPN1. Membranes were detected using N-terminally binding (D5C9H) and C-terminally binding (58C9) TBP antibodies. Light blue- and pink-rimmed arrowheads mark full-length TBP 38Q and 106Q, respectively, and green-rimmed their respective CSM forms. Red and blue arrowheads indicate N-terminal and C-terminal (n1–3; c1–3) TBP fragments, respectively. Green arrowheads indicate alternative TBP fragments caused by the presence of the CSM. The approximate binding sites of the antibodies used are illustrated above the corresponding blots. SE short exposure, LE long exposure. Source data are provided as a Source Data file.
The C-terminal TBP fragment exhibits high turnover, while polyQ-containing N-terminal fragments show increased aggregation propensity
To explore the physiological and pathological roles of calpain-1-dependent TBP cleavage, we generated constructs of N- and C-terminal TBP fragments corresponding to the identified cleavage site T106 | S107 (Fig. 4a). First, we overexpressed full-length TBP 38Q or 106Q, non-polyQ-expanded (TBP1-106 38Q) or polyQ-expanded (TBP1-106 106Q) N-terminal TBP fragments, and their myc-tagged C-terminal (TBP107-339) counterpart in 293T cells. Protein extracts were analysed via western blotting. Using an N-terminally binding TBP antibody (D5C9H), we detected overexpressed full-length TBP as well as TBP1-106 38Q and TBP1-106 106Q. In addition, a C-terminally binding TBP antibody (58C9) confirmed the overexpression of TBP107-339. Notably, all fragment constructs exhibited markedly lower levels compared to full-length TBP (Fig. 4b). To determine whether high proteolytic turnover via cellular clearance mechanisms contributed to the low fragment levels, we treated transfected cells with proteasome inhibitor MG132 or autophagy inhibitor bafilomycin A1 (BafA1). Western blot analysis of K48-linked polyubiquitin chains (K48-pUb), as well as SQSTM1/p62 and MAP1LC3B, confirmed successful inhibition of proteasomal and autophagic pathways (Fig. 4c and Supplementary Fig. 5a). TBP-specific detection revealed a strong increase in levels of full-length TBP 38Q and 106Q upon proteasomal inhibition, whereas autophagy inhibition had only moderate effects (Fig. 4d). N-terminal fragments TBP1-106 38Q and TBP1-106 106Q showed minimal response to either proteasomal or autophagy inhibition (Fig. 4e). In contrast, levels of the C-terminal TBP107-339 fragment markedly increased with proteasomal inhibition, while autophagy inhibition had lesser impact (Fig. 4f). Given the possibility that increased aggregation propensity might explain the low levels on N-terminal TBP fragments, we performed filter retardation assays of protein extracts from transfected 293T cells to analyse SDS-insoluble species. Immunodetection with N-terminal antibodies 1TBP18 and 8515 revealed the presence of SDS-insoluble species of full-length TBP 106Q, as well as of TBP1-106 38Q and TBP1-106 106Q, with TBP1-106 106Q exhibiting the strongest signal, while the C-terminal fragment as detected with antibody 58C9 did not aggregate (Fig. 4g). In summary, TBP fragments generated by calpain cleavage at T106 | S107 display reduced protein levels, driven by rapid proteasomal turnover of the C-terminal fragment and pronounced aggregation of N-terminal polyQ-containing fragments, particularly TBP1-106 106Q.
Fig. 4. The C-terminal TBP fragment undergoes increased proteolytic turnover, while N-terminal polyQ-containing fragments are highly aggregation-prone.

a Schematic representation of TBP fragment constructs derived from the calpain cleavage site T106 | S107. Two N-terminal fragments, non-polyQ-expanded (TBP1-106 38Q) and polyQ-expanded (TBP1-106 106Q), along with the corresponding myc-tagged C-terminal counterpart (TBP107-339), are depicted. b Protein extracts from 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 were analysed by western blotting. Full-length TBP, as well as the N- and C-terminal fragment constructs, were detected using N-terminally binding (D5C9H) or C-terminally binding (58C9) TBP antibodies. GAPDH served as a loading control. Light blue- and pink-rimmed arrowheads mark full-length TBP 38Q and 106Q, respectively. Red arrowheads show N-terminal fragments (TBP1-106 38Q or TBP1-106 106Q), while blue arrowheads indicate C-terminal fragment TBP107-339. c–f 293T cells expressing full-length TBP or TBP fragment constructs were treated with the proteasome inhibitor MG132 or the autophagy inhibitor bafilomycin A1 (BafA1). Protein extracts were analysed by western blotting. Proteasome and autophagy inhibition were validated using antibodies against K48-linked polyubiquitin chains (K48-pUb), SQSTM/p62 (p62), and MAP1LC3B (LC3B) (in c). Black-rimmed arrowheads mark LC3B-I and -II. TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 were detected using N-terminally binding (D5C9H) or C-terminally binding (58C9) antibodies (d–f). GAPDH served as a loading control. Densitometric analysis of the blots shows levels of full-length TBP (d) or TBP fragments (e, f). TBP levels were normalised first to GAPDH and then to the vehicle control. The vehicle control is indicated by a dotted line. n = 5 (d, f; e for TBP1-106 38Q), or 6 (e for TBP1-106 106) biological replicates. Unpaired Student’s t test. g Filter retardation assays of 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q or TBP107-339. SDS-insoluble TBP species were detected using N-terminally binding (1TBP18 or 8515) TBP antibodies or the C-terminally binding (58C9) TBP antibody. Densitometric analysis of SDS-insoluble full-length TBP and TBP fragments, normalised to full-length TBP 106Q signals within each sample set. n = 7 (1TBP18; 8515, except for TBP1-106 106Q with n = 6 biological replicates), or 3 (58C9) biological replicates. One-sample t test and one-way ANOVA with Tukey’s post hoc test. Bars represent means + s.e.m. All tests were two-tailed. P values are shown in the graphs. Source data are provided as a Source Data file.
N-terminal TBP fragments are cytoplasmic and form perinuclear inclusion bodies
PolyQ-expanded TBP is known to form intranuclear inclusion bodies (IBs) in the brains of SCA17 patients13. As our findings demonstrate that N-terminal TBP fragments form SDS-insoluble species and exhibit enhanced aggregation propensities compared to the full-length protein, we investigated their aggregation behaviour in 293T cells expressing full-length TBP and TBP fragments using fluorescence microscopy. Consistent with our filter retardation analyses, TBP 106Q, TBP1-106 38Q, and TBP1-106 106Q formed nucleus-associated IBs, whereas TBP 38Q and TBP107-339 displayed a diffuse nuclear staining (Fig. 5a). Quantitative analysis supported by confocal microscopy revealed a significantly higher frequency of IB formation by TBP1-106 106Q compared to TBP 106Q and TBP1-106 38Q (Fig. 5b, c). IBs formed by TBP1-106 106Q also exhibited a larger cross-sectional area and diameter. Additionally, structural differences were observed, with IBs formed by TBP1-106 38Q and TBP1-106 106Q showing reduced roundness (Fig. 5b, c). Strikingly, while IBs formed by TBP 106Q were mainly intranuclear, those formed by the N-terminal TBP fragments TBP1-106 38Q and TBP1-106 106Q exhibited a perinuclear localisation (Fig. 5b and Supplementary Fig. 5c, d). To explore whether the intracellular distribution of soluble TBP fragments underlies the observed spatial patterns of IBs, we conducted cytoplasmic-nuclear fractionation of 293T cells expressing full-length TBP and TBP fragments. As expected, full-length TBP was predominantly, while not solely, found in the nuclear fraction (Fig. 5d–f). In contrast, N-terminal TBP fragments were exclusively detected in the cytoplasmic fraction, which may account for the perinuclear localisation of the IBs they form (Fig. 5g, h). Conversely, the C-terminal fragment TBP107-339, which contains the DNA-binding core domain, is enriched in the nuclear fraction, consistent with microscopy data, while lower levels remain detectable in the cytoplasmic fraction (Fig. 5i, j and Supplementary Fig. 5b). In summary, whereas C-terminal TBP predominantly localises to the nucleus, N-terminal polyQ-expanded TBP fragments mislocalise to the cytoplasm, where their aggregation-prone nature promotes the formation of perinuclear IBs.
Fig. 5. N-terminal polyQ-expanded TBP fragments show cytoplasmic localisation and form perinuclear inclusion bodies at a higher frequency.

a Immunofluorescence staining and microscopy of 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 (green). To analyse subcellular localisation and aggregation, an N-terminally binding TBP antibody (8515) or a myc tag-specific antibody (both in green) were used. DAPI was used as a nuclear counterstain (blue). Insets show 3.3× magnified views of selected regions (white dashed boxes). Scale bar: 10 µm. b Immunofluorescence staining and confocal microscopy of 293T cells expressing TBP 106Q, TBP1-106 38Q, or TBP1-106 106Q. The N-terminally binding TBP antibody 1TBP18 was used to analyse subcellular localisation of TBP inclusion bodies (IB; green). DAPI was used as a nuclear counterstain (blue). Maximum intensity projections (MIP) and insets of 3× magnified selected areas (white dashed boxes) are shown alongside three additional MIPs of representative TBP IBs. For TBP 106Q and TBP1−106 38Q, IB-positive cells are highlighted with white arrowheads and individual IBs accented with orange arrowheads. Scale bars: 10 µm. c Fiji-based quantitative analysis of TBP IB frequency, cross-sectional area, equivalent circular area diameter (dECA), and roundness. IB frequency was assessed as per 100 cells. n = 3 experiments. Bars represent means + s.e.m. Violin plots show the distribution of every single IB analysed. Dashed lines indicate median. Dotted lines indicate the 25th and 75th percentiles. One-way ANOVA with Tukey’s post hoc test or Kruskal–Wallis test with Dunn’s post hoc test. d–j 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 were subjected to cellular-nuclear fractionation. Obtained fractions (whole cell (W); cytoplasmic fraction (C); nuclear fraction (N)) were analysed by western blot. Cytoplasmic GAPDH and α-tubulin, as well as nuclear lamin A/C, served as fraction-purity and loading controls (d). Western blots were probed with TBP binding antibodies 58C9 or D5C9H, or myc tag-specific antibody to detect TBP 38Q and TBP 106Q (e, f; blue-rimmed and pink-rimmed arrowheads), TBP1-106 38Q and TBP1-106 106Q (g, h; red arrowheads) or TBP107-339 (i, j; blue arrowheads). Densitometric analysis shows relative levels of full-length TBP or TBP fragments in the cytoplasmic and nuclear fractions. Levels were normalised first to fraction-specific loading controls and then to whole-cell (indicated by a dotted line) within each sample set. n = 3 biological replicates. Bars represent means + s.e.m. Unpaired Student’s t test. All tests were two-tailed. P values are shown in the graphs. Source data are provided as a Source Data file.
TBP fragments trigger calpain overactivation and dysregulation of the cellular proteome
To investigate whether TBP fragments generated by cleavage at T106 | S107 exhibit increased cytotoxicity and further activate calpains, in line with the toxic fragment hypothesis18,19, we assessed cell viability and calpain cleavage markers in 293T cells expressing full-length TBP or TBP fragments. Notably, only TBP1-106 106Q caused a slight but significant decrease in cell viability (Fig. 6a), indicative of its inherent cytotoxicity. Western blot analysis of calpain markers α-spectrin, calpastatin, and calpain-1 revealed increased calpain activation in cells expressing full-length TBP 106Q and the N-terminal fragment TBP1-106 106Q (Fig. 6b–d). Moreover, the DNA-binding C-terminal fragment TBP107-339 led to calpain overactivation, whereas the non-polyQ-expanded N-terminal TBP1-106 38Q did not induce any changes (Fig. 6c, d). To gain broader insight into the downstream effects of TBP fragment expression and, in comparison, full-length TBP, we conducted a mass spectrometry-based label-free quantitative (LFQ) proteome analysis. Compared to TBP 38Q, overexpression of TPB 106Q in 293T cells resulted in a significant alteration of 20 proteins ( ↓ 9, ↑11). Analysis of cells expressing TBP1-106 38Q, TBP1-106 106Q, and TBP107-339 revealed 31 ( ↓ 10, ↑21), 27 ( ↓ 11, ↑16) and 36 ( ↓ 15, ↑21) differentially regulated proteins, respectively (Fig. 6e). A total of 21 proteins were dysregulated in at least two conditions, including ADP-ribosylation factor 1 (ARF1), citrate synthase (CS), lysophospholipase 1 (LYPLA1), and nucleoside diphosphate kinase A (NME1)—all of which have been implicated in the polyQ disease-related molecular pathogenesis35–38. Serine/arginine-rich splicing factor 7 (SRSF7) was the only protein consistently increased across all conditions relative to TBP 38Q (Fig. 6e, Supplementary Fig. 6a and Supplementary Data 1). Most prominently, levels of angiomotin (AMOT), a member of the motin family linked to cell motility, angiogenesis, and regulation of YAP1-dependent Hippo signalling39, were significantly reduced in 293T cells expressing TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 compared to cells expressing full-length TBP 38Q (Fig. 6e, f). The most pronounced reductions were observed in cells expressing the N-terminal fragments TBP1-106 38Q and 106Q. To confirm this reduction, we analysed AMOT—present as a polyQ-stretch-containing p130 (UniProt ID: Q4VCS5-1) and a p80 (UniProt ID: Q4VCS5-2) isoform—by western blotting of protein extracts from 293T cells expressing full-length TBP or TBP fragments, as well as mock controls transfected with an empty vector to assess the effect of TBP 38Q on AMOT baseline expression. Compared to mock-transfected controls, both AMOT p130 and p80 levels increased with TBP 38Q overexpression. When compared to TBP 38Q, AMOT p130 and p80 levels significantly decreased in TBP 106Q-, TBP1-106 38Q-, TBP1-106 106Q-, and TBP107-339-expressing cells, with TBP1-106 106Q lowering AMOT below mock-transfected control levels (Fig. 6g and Supplementary Fig. 6b, c). Notably, subsequent protein network analysis, based on the differential proteome detected for TBP1-106 106Q, revealed a protein interaction cluster linked to AMOT (Supplementary Fig. 6d), with all proteins implicated in neurological contexts (Supplementary Data 2).
Fig. 6. TBP fragments exhibit increased toxicity, induce calpain overactivation, and trigger proteomic alterations.

a Resazurin-based cell viability assay of 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339. Signals were normalised to 293T cells expressing TBP 38Q. n = 6 biological replicates. One-sample t test. b Western blot analysis of 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q or TBP107-339. TBP was detected with D5C9H and myc-specific antibodies. GAPDH served as a loading control. Light blue- and pink-rimmed arrowheads mark full-length TBP 38Q and 106Q, respectively. Red arrowheads indicate N-terminal fragments (TBP1-106 38Q or TBP1-106 106Q), while blue arrowheads indicate C-terminal fragment TBP107-339. c Protein extracts of 293T cells expressing full-length TBP or TBP fragments were analysed for calpain activation by western blot. Immunodetection was conducted using antibodies against α-spectrin, calpastatin (CAST), and calpain-1 (CAPN1). GAPDH served as the loading control. Black-rimmed arrowheads indicate full-length proteins, while grey arrowheads show respective breakdown products. d Densitometric analysis of α-spectrin cleavage ratio (breakdown product (CL) normalised to full-length (FL) α-spectrin), CAST levels (CAST levels normalised to GAPDH) and CAPN1 cleavage ratio (breakdown product (CL) normalised to full-length (FL) CAPN1). Values were normalised to 293T cells expressing TBP 38Q in each measurement. n = 5 biological replicates. One-sample t test. e Label-free quantitative mass spectrometric analysis of protein extracts from 293T cells expressing full-length TBP and TBP fragments. Volcano plots show significance values (−log10p) plotted against the fold change (log2fc) of detected proteins. Proteins significantly altered (P ≤ 0.05; dotted horizontal line) in their levels are indicated in colour. Prominent hits are labelled. f Overlap of significantly dysregulated proteins in different conditions (293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q or TBP107-339) shown by a Venn diagram. Black arrows indicate the direction of dysregulation. Red arrow includes AMOT. The Venn diagram was created using InteractiVenn. g Western blot analysis of AMOT levels in 293T cells expressing full-length TBP or TBP fragments, and in mock-transfected controls. An N-terminally binding AMOT antibody (N-term) and a C-terminally binding AMOT antibody (C-term) were used for immunodetection. GAPDH served as the loading control. Black-rimmed arrowheads indicate the p130 and p80 isoforms of AMOT. Densitometric analysis shows AMOT p130 levels normalised to the loading control. n = 6 biological replicates. One-sample t test and one-way ANOVA with Dunnett’s post hoc test (*). Bars represent means + s.e.m. All tests were two-tailed. P values are shown in the graphs. Source data are provided as a Source Data file.
TBP fragments induce a multifactorial and polyQ-dependent depletion of AMOT
Previous studies have demonstrated that polyQ-expanded TBP exhibits impaired DNA-binding abilities and altered self-dimerisation, leading to disruptions in its transcriptional activity14,15,17,28. To determine whether the observed reduction in AMOT protein levels results from TBP-induced transcriptional changes and to uncover affected pathways, we performed RNA sequencing of 293T cells expressing full-length TBP or TBP fragments. We identified 291 differentially expressed genes (DEGs) in TBP 106Q, 1684 DEGs in TBP1-106 38Q, 1733 DEGs in TBP1-106 106Q, and 1150 DEGs in TBP107-339, compared to TBP 38Q (Fig. 7a). Interestingly, AMOT expression was significantly downregulated in TBP1-106 38Q, TBP1-106 106Q, and TBP107-339 in our transcriptomics data, while TBP 106Q showed only a trend towards AMOT reduction (Fig. 7b). In contrast, the other two motin family members, angiomotin-like 1 (AMOTL1) and 2 (AMOTL2)39, were either significantly downregulated only in TBP 106Q-expressing cells or remained unaltered, respectively (Supplementary Fig. 7). To further evaluate the TBP-dependent modulation of AMOT expression, we performed qPCR analysis in 293T cells transfected with TBP 38Q, TBP 106Q, or an empty vector. Importantly, and corresponding to our western blot data, TBP 38Q increased expression of AMOT, while TBP 106Q showed a deficiency in AMOT upregulation (Fig. 7c).
Fig. 7. TBP fragments trigger transcriptional disturbances with repercussions on AMOT-YAP1 signalling, neuronal pathways, and calpain activation.

a Transcriptomic analysis of 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 by RNA sequencing. Volcano plots show −log10p values plotted against expressional fold changes (log2fc) of identified genes for comparisons against the TBP 38Q control condition. Significantly differentially expressed genes (DEGs) (P ≤ 0.05; dotted horizontal line) are visualised in colour. Red data points mark AMOT. n = 3 biological replicates. b Transcript levels of AMOT in 293T cells expressing TBP 38Q, TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 detected by RNA sequencing and shown as transcripts per million (tpm). n = 3 biological replicates. For statistics, see Methods: RNA sequencing. c Quantitative PCR of TBP and AMOT expression in 293T cells transfected with TBP 38Q, TBP 106Q, or an empty vector. n = 3 biological replicates. One-way ANOVA with Tukey’s post hoc test. d Overlap of significant DEGs in TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339 compared to TBP 38Q visualised by Venn diagram. An asterisk marks the intersection, which includes AMOT. Red-rimmed intersection marks 14 DEGs shared between TBP 106Q, TBP1-106 106Q, and TBP107-339 compared to TBP 38Q. The Venn diagram was created via InteractiVenn. e Heatmap of shared 14 DEGs (highlighted in red in c) showing row-wise calculated z-scores, log2fc, and −log10p of TBP 106Q, TBP1-106 106Q, and TBP107-339 compared to TBP 38Q. n = 3 biological replicates. f Bubble plot of top 5 significantly up- and downregulated pathways in 293T cells expressing TBP 106Q, TBP1-106 38Q, TBP1-106 106Q, or TBP107-339. Data includes z-scores and presents −log10p indicated by bubble colour and number of pathway-related DEGs by bubble size. MD movement disorder, NS nervous system, dysfunct. dysfunction. g, h Western blot analysis of protein extracts from 293T or SH-SY5Y (SY5Y) cells transfected with esiRNA directed against AMOT. Renilla luciferase targeting esiRNA was used as a mock control. AMOT as well as calpain markers α-spectrin, calpastatin (CAST), and calpain-1 (CAPN1) were detected. GAPDH served as the loading control. Densitometric analysis shows AMOT and calpastatin levels normalised to GAPDH, as well as α-spectrin cleavage ratio (α-spectrin breakdown product (CL) indicated by grey arrowhead normalised to full-length (FL) α-spectrin indicated by black-rimmed arrowhead). n = 3 (293T) and n = 7 (SY5Y) biological replicates. One-sample t test. Bars represent means + s.e.m. All tests were two-tailed. P values are shown in the graphs. Source data are provided as a Source Data file.
Given that AMOT levels decreased upon TBP 106Q overexpression, we investigated additional cellular mechanisms responsible for this reduction. Since previous experiments demonstrated calpain overactivation in the presence of polyQ-expanded TBP and TBP fragments, we examined whether AMOT reduction could result from calpain-dependent proteolysis. In silico calpain cleavage site prediction for AMOT p130 (UniProt ID: Q4VCS5-1) using the GPS-CCD tool demonstrated several potential sites along the protein sequence (Supplementary Fig. 8a). To validate the prediction results, we activated calpains in 293T cells using ionomycin. Calpain activation was confirmed by increased cleavage of the calpain substrates α-spectrin and calpastatin. In ionomycin-treated cells, AMOT levels were strongly reduced, and fragment bands increased, while pretreatment with the calpain inhibitors ALLN, calpeptin, or CI-III rescued AMOT levels and reduced cleavage, demonstrating that this reduction was calpain-specific (Supplementary Fig. 8b). Next, we co-expressed AMOT with calpain-1 in 293T cells to confirm its calpain-dependent proteolysis. As expected, calpain-1 overexpression led to increased α-spectrin fragmentation, reduced calpastatin levels, and a significant reduction in AMOT, identifying AMOT as a direct calpain-1 substrate (Supplementary Fig. 8c). Conversely, calpain-1 knockdown reduced substrate cleavage and significantly increased AMOT levels (Supplementary Fig. 8d).
In addition, we explored whether AMOT levels declined due to recruitment into IBs formed by polyQ-expanded TBP or N-terminal TBP fragments, potentially driven by the polyQ stretch at the N-terminus of the p130 isoform. Fluorescence microscopy revealed that AMOT colocalised with IBs formed by N-terminal TBP fragments (Supplementary Fig. 9a), with ~25% of TBP IBs formed by TBP1-106 38Q and TBP1-106 106Q being AMOT-positive. In contrast, fewer AMOT-positive IBs were observed in cells expressing full-length TBP 106Q (Supplementary Fig. 9b). As we observed intracellular AMOT accumulations independent of TBP, we next assessed whether endogenous or overexpressed V5-tagged AMOT is intrinsically aggregation-prone. Using in situ and membrane-based IB fluorescence microscopy, we confirmed that AMOT co-aggregates with polyQ-expanded TBP1-106 106Q, whereas AMOT alone does not form SDS-insoluble or ubiquitin-positive aggregates (Supplementary Fig. 9c–g). Taken together, our findings suggest that AMOT depletion arises from a combination of transcriptional dysregulation, calpain overactivation-dependent proteolysis, and recruitment into TBP IBs.
As both calpain overactivation and IB formation are hallmark features of several other polyQ disorders18,40, we examined whether overexpression of polyQ-expanded ataxin-3, the pathogenic driver of SCA3, would produce similar effects. While non-polyQ-expanded ataxin-3 15Q did not alter AMOT levels at baseline (Supplementary Fig. 10a)—contrasting our observation for TBP 38Q—overexpression of polyQ-expanded ataxin-3 148Q led to reduced levels of both AMOT p130 and p80, alongside increased calpain-dependent α-spectrin cleavage (Supplementary Fig. 10b). Further, we observed AMOT sequestration into ataxin-3-positive aggregates (Supplementary Fig. 10c), confirming a dual mechanism of AMOT depletion in the SCA3 context.
PolyQ-expanded TBP and its fragments cause AMOT-YAP1 signalling disturbances, impacting neuronal pathways and calpain activity
To identify specific pathways disrupted by polyQ-expanded TBP and its fragments, we compared DEGs across different groups (Fig. 7d). A comparative analysis of TBP 106Q, TBP1-106 106Q, and TBP107-339 revealed 14 commonly dysregulated genes, including MT-CYB, MT-CO1, MT-CO2, MT-CO3, MT-ATP6, which encode subunits of the mitochondrial respiratory chain complexes III to V (Supplementary Fig. 11a). In addition, CCN2 was consistently downregulated across all three groups (Fig. 7e). Importantly, CCN2 is a downstream target of the transcriptional co-regulator Yes-associated protein 1 (YAP1), which interacts with AMOT39. Ingenuity Pathway Analysis (IPA) predicted YAP1 inhibition upon TBP1-106 106Q overexpression, with a total of 89 YAP1 target genes showing dysregulation (Supplementary Fig. 11b and Supplementary Data 3). Based on these results, we hypothesised that polyQ-expansion and fragmentation of TBP dysregulated the AMOT-YAP1 pathway. To further evaluate this, we examined the presence of differentially expressed YAP1 target genes amongst DEGs shared by TBP 106Q and TBP1-106 106Q, TBP 106Q and TBP107-339, as well as TBP1-106 106Q and TBP107-339, and identified eight candidates: RPL26, TNIP2, SGK1, IFI6, CCN2, ACTA2, PLK2, and RPL3L. Notably, except for PLK2 and TNIP2, six of these genes mapped to an extended protein network, revealing previously reported or predicted interactions (Supplementary Fig. 11c). Collectively, our findings suggest that polyQ-expanded and fragmented TBP induce transcriptional perturbations that disrupt the AMOT-YAP1 pathway and lead to differential expression of downstream target genes.
To assess the relevance of the identified DEGs to neurological function, we conducted an additional IPA analysis, which linked DEGs to several neurological processes, including neuronal cell proliferation, nervous tissue morphogenesis, and neurological disorders such as motor dysfunction, movement disorders, and abnormal nervous system morphology (Fig. 7f and Supplementary Data 4).
Finally, to further explore AMOT’s role in calpain overactivation, we performed AMOT knockdown experiments in 293T and SH-SY5Y cells—as a neuronal model—and analysed calpain system markers. In both cell lines, AMOT knockdown resulted in increased α-spectrin cleavage (Fig. 7g, h). Moreover, in 293T cells, calpastatin levels significantly decreased, suggesting that AMOT is functionally linked to calpain activation.
SCA17 patient-derived cell models reproduce calpain overactivation, TBP fragmentation, and AMOT reduction
To validate our findings from overexpression cell models, we obtained fibroblasts from an SCA17 patient and an age- and sex-matched unaffected control subject and performed an initial analysis of TBP expression and AMOT levels. The SCA17 patient carries 37 CAG/CAA repeats on the wild-type allele and 52 repeats on the pathogenic expanded allele of TBP, whereas the control shows homozygous repeat lengths of 37 CAG/CAA repeats on both alleles (Supplementary Fig. 12a). Fluorescence microscopy demonstrated nuclear localisation of TBP and cytoplasmic localisation of AMOT (Supplementary Fig. 12b, c). Subsequent western blot analysis confirmed the presence of polyQ-expanded TBP in SCA17 fibroblasts and, importantly, revealed reduced AMOT p130 levels (Supplementary Fig. 12d, e), consistent with observations in 293T cells.
To further validate these effects in a neuronal context, we generated iPSCs from SCA17 patient and control fibroblasts and differentiated them into induced cortical neurons (iCNs). Fluorescence microscopy confirmed the cortical identity of the differentiated neuronal cells via TBR1, CTIP2, βIII-tubulin, and MAP2 staining, as well as the expression of nuclear TBP and cytoplasmic AMOT, the latter also present in the neuronal projections (Fig. 8a and Supplementary Fig. 12f).
Fig. 8. SCA17 patient-derived induced neurons recapitulate calpain overactivation, TBP cleavage, and AMOT reduction.

a Fluorescence microscopy of induced cortical neurons (iCNs) of an unaffected individual and an SCA17 patient, immunostained for TBP (1TBP18, green), AMOT (red), and the pan-neuronal marker MAP2 (white). DAPI was used as a nuclear counterstain (blue). Scale bar: 50 µm. b Western blot analysis of α-spectrin cleavage as a calpain activation marker, TBP levels and fragmentation, and AMOT p130 levels in ctrl and SCA17 iCNs. GAPDH served as the loading control. Black-rimmed arrowheads indicate full-length (FL) proteins, while the grey arrowhead shows respective cleavage (CL) products. Light blue-rimmed arrowhead marks full-length non-polyQ-expanded TBP, pink-rimmed arrowhead polyQ-expanded TBP, and light blue arrowheads C-terminal TBP fragments c1–3. c Densitometric analysis of α-spectrin cleavage ratio (CL to FL α-spectrin), FL-TBP (normalised to GAPDH), cleaved TBP (c1&c2 normalised to FL-TBP), and AMOT p130 (normalised to GAPDH). n = 4 (CL to FL α-spectrin; all ctrl conditions) or 3 (SCA17 conditions of TBP normalised to GAPDH, c1&c2 normalised to FL-TBP, and AMOT p130 normalised to GAPDH) biological replicates. Student’s t test. d Ctrl and SCA17 iCNs were treated with ionomycin (Iono) for calpain activation and analysed by western blotting. Calpain inhibitor III (CI-III) treatment served as a specificity control. DMSO was added as a treatment control. Calpain activation was confirmed by detecting α-spectrin cleavage. Calpain-induced TBP fragmentation was detected using the C-terminally binding TBP antibody 58C9. GAPDH served as a loading control. Light blue-rimmed arrowhead marks full-length non-polyQ-expanded TBP, pink-rimmed arrowhead polyQ-expanded TBP, and light blue arrowheads c1 and c2. e Densitometric analysis of FL-TBP (normalised to GAPDH and DMSO control) and TBP fragmentation (c1&c2 normalised to FL-TBP and DMSO control). n = 3 biological replicates. Bars represent means + s.e.m. Two-tailed one-sample t test. P values are shown in the graphs. Source data are provided as a Source Data file.
Baseline analysis of iCNs by western blotting revealed increased calpain activity in SCA17 cells, as indicated by enhanced α-spectrin cleavage and trends toward increased processing of calpain-1 and CAPNS1 (Fig. 8b, c and Supplementary Fig. 12g, i). Consistent with this, we observed increased TBP fragmentation in SCA17 iCNs, including the well-characterised C-terminal fragments c1, c2, and c3, as well as corresponding N-terminal cleavage products (Fig. 8b, c and Supplementary Fig. 12h, i). Importantly, SCA17 neurons showed reduced levels of AMOT p130 (Fig. 8b, c).
The calpain-dependent origin of the observed C-terminal TBP fragments was confirmed by performing ionomycin-induced calpain activation in control and SCA17 iCNs, followed by western blot analysis (Fig. 8d, e).
Together, the calpain overaction, TBP fragmentation, and AMOT depletion observed in overexpression-based models were recapitulated in patient-derived neurons, reinforcing their pathophysiological relevance in SCA17.
SCA17 rat cerebellum shows AMOT reduction, YAP1 signalling disturbances, and proteomic perturbations with links to oligodendrocytes and neurofunction
To validate the observed AMOT reduction in the context of polyQ-expanded TBP in vivo, we analysed cerebellar tissue from our SCA17 TBPQ64 rats28,41 by fluorescence microscopy and western blot. In this model, we confirmed the previously reported calpain overactivation (Supplementary Fig. 13a) and demonstrated the presence of N-terminal fragments of transgenic myc-tagged TBP in the cerebellum, using a myc-IP-based enrichment and a selection of TBP-specific antibodies (Supplementary Fig. 13b–e). Importantly, immunofluorescence staining and microscopy of rat brain sections showed not only reduced AMOT signals, particularly in the molecular and Purkinje cell layers of the TBPQ64 cerebellum (Fig. 9a, b and Supplementary Fig. 14a–c), but also a decreased YAP1 intensity (Supplementary Fig. 14d). These effects were further validated by western blot analysis of cerebellar protein extracts (Fig. 9c and Supplementary Fig. 15a). Moreover, the YAP1-associated Hippo signalling target genes CCN1/CYR6142 and CCN2/CTGF43 were likewise reduced (Fig. 9d).
Fig. 9. AMOT, YAP1, and its downstream targets are depleted in SCA17 rat cerebellum, causing oligodendrocyte-enriched proteome disturbances with neurofunctional impact.

a Immunofluorescence staining and microscopy of cerebellum (cb) sections from 10-month-old wild-type (WT) and transgenic SCA17 (TBPQ64) rats expressing N-terminally myc-tagged, polyQ-expanded human TBP with 64 glutamines. Myc-TBP signals are shown in green, and AMOT signals in red. DAPI was used as a nuclear counterstain (blue). MOL molecular layer, PCL Purkinje cell layer, GCL granular cell layer. Scale bar: 100 µm. b Quantitative analysis of AMOT staining in the MOL, PCL, and GCL of WT and TBP64Q rat cerebellum. Signals were normalised to WT. n = 3 animals, 5 images per animal. Unpaired Student’s t test. c Western blot analysis of cerebellar protein extracts from WT and TBPQ64 rats. PolyQ-expanded (64Q, pink-rimmed arrowhead) and endogenous rat TBP (reg, black-rimmed arrowhead), as well as AMOT and YAP1, were detected. GAPDH served as the loading control. Densitometric analysis shows AMOT and YAP1 levels in TBPQ64 rats relative to WT rats. Levels were normalised to GAPDH. n = 7 animals. Unpaired Student’s t test. d Western blot analysis of CCN1, CCN2, and MBP in WT and TBPQ64 rat cerebellum. Black-rimmed arrowheads mark long (l) and short (s) MBP isoforms. Vinculin (VCL) served as a loading control. Densitometric analysis shows protein levels normalised to VCL. n = 7 animals. Unpaired Student’s t test. e Label-free quantitative mass spectrometric analysis of cerebellar protein extracts from WT and SCA17 rats. n = 7 animals. Volcano plots show significance values (−log10p) plotted against the fold change (log2fc) of detected proteins. Proteins significantly altered (P ≤ 0.05; dotted horizontal line) in their levels are indicated in colour. PanglaoDB-curated oligodendrocyte (OLG) marker proteins are labelled. f PanglaoDB-based cell type enrichment analysis for the significantly differentially regulated proteome in TBPQ64 rat cerebellum. Bubble plots show ranking by gene ratio, with bubble colour indicating FDR and bubble size gene count. g Heatmap of differentially regulated OLG marker proteins in TBPQ64 rat cerebellum, showing row-wise calculated z-scores, log2fc, and −log10p. n = 7 animals. h, i Bubble plot of top ten enrichments in the TBPQ64 rat cerebellum proteome for Cellular Component (Gene Ontology; in h) and KEGG pathways (i) obtained from STRING (string-db.org), ranked by signal. ALS amyotrophic lateral sclerosis, SCA spinocerebellar ataxia, PD Parkinson’s disease, AD Alzheimer’s disease. Bubble colour indicates FDR and bubble size protein counts. Bars represent means + s.e.m. All tests were two-tailed. P values are shown in the graphs. Source data are provided as a Source Data file.
Given that loss of CCN1 has been associated with impaired myelination44, we assessed myelin basic protein (MBP) levels in cerebellar extracts and found a significant reduction in total MBP (Fig. 9d and Supplementary Fig. 15b). To further investigate SCA17-associated alterations on a broader scale, we performed LFQ proteomics, identifying 144 downregulated and 29 upregulated proteins compared to wild-type animals (Fig. 9e). While insufficiently covering YAP1 signalling-related targets in the significantly dysregulated proteome (Supplementary Fig. 15c), the majority of dysregulated proteins mapped to a large interaction cluster (Supplementary Fig. 16).
Notably, cell-type enrichment analysis revealed a significant overrepresentation of dysregulated proteins associated with oligodendrocytes, indicating a pronounced impact of SCA17 pathology on this cell type and, consequently, on myelination-related processes (Fig. 9e–g and Supplementary Fig. 15d). Moreover, pathway and gene enrichment analyses highlighted strong associations with neuron function and movement disorder-related pathways, while additionally uncovering dysregulations in mRNA processing-related mechanisms (Fig. 9h, i and Supplementary Fig. 15e, f), further supporting the validity of the TBPQ64 rat as a model for SCA17 and the impact of polyQ-expanded TBP on these processes.
Amot is reduced in zebrafish expressing TBP 106Q, while its genetic depletion disrupts YAP1 signalling and induces a motor neuron phenotype
We turned to zebrafish to explore a potential role of Amot in the nervous system. While Amot has been characterised for its function during angiogenesis45, it has not been investigated outside the vascular system in zebrafish. Consistent with our cell model findings, overexpression of human TBP 38Q in zebrafish embryos (Supplementary Fig. 17a) induced strong amot expression. This was not achieved by TBP 106Q, which resulted in lower amot signals compared to TBP 38Q (Fig. 10a and Supplementary Fig. 17b, c), mirroring the reduced AMOT protein levels observed in SCA17 rat cerebellum compared to wild-type. In situ hybridisation revealed that amot is ubiquitously expressed during early developmental stages, later becoming restricted to early neuronal and mesodermal structures during organogenesis. When the body plan has been established, amot transcripts can be detected in the developing vascular system and in the neural tube (Supplementary Fig. 17d), suggesting its involvement not only in angiogenesis, but also in neurogenesis. To analyse this, we designed a loss-of-function approach using transient Crispr/Cas9-mediated gene editing (Fig. 10b, c). As this may result in mosaic gene editing and can hence increase the variability of the phenotype, we genotyped each embryo. Importantly, Crispr/Cas9 against amot resulted in downregulation of typical Yap1-associated Hippo signalling target genes ctgf and cyr61 (the zebrafish homologs of CCN1 and CCN2, respectively)46 consistent with a previous report47 (Fig. 10d, e), while not causing malformations in the embryos (Supplementary Fig. 16e–g). As a measure of ataxia, we performed simple behavioural tests routinely used to assess mechanosensory and motoric functions in zebrafish embryos48,49. Depletion of amot by Crispr/Cas9 manifested in more spontaneous movements in 24 hours post fertilisation (hpf) zebrafish embryos as well as loss of touch response and touch-evoked swimming at 48 hpf, suggesting general motor impairments, which were rescued by overexpression of human AMOT (Fig. 10f–h and Supplementary Fig. 17h). Moreover, these disturbances were accompanied by a small, but robust reduction in the number of motor neurons (Fig. 10i), which was reversed by AMOT overexpression in amot crispants, while overexpression of AMOT in controls had no effect. We confirmed the reduced number of motor neurons with a previously published45, established antisense morpholino oligonucleotide (MO)-mediated knockdown approach (Fig. 10j). In addition, we observed that axons outgrowing from the motor neuron’s cell bodies of amot crispants followed erratic paths, suggesting axon guidance defects (Fig. 10k). As in the context of HD, a related polyQ-dependent condition, changes in the dopaminergic system have been reported50–52, we additionally examined the zebrafish brain. Here, loss of Amot impacted the dopaminergic system by increasing the number of dopaminergic neurons, which also displayed a higher expression of the dopamine transporter (dat), a protein responsible for dopamine reuptake from the synaptic cleft into the presynapse (Fig. 10l). Taken together, these data link Amot to Yap1-associated Hippo signalling and support a role for Amot in neural development.
Fig. 10. Amot loss-of-function in zebrafish affects Yap1-Hippo signalling, interfering with normal motor function and motor neuron development.

a In situ hybridisation for amot in 60% epiboly embryos, injected with a control RNA (ctrl) or RNA encoding human TBP 38Q or TBP 106Q. n = 2 experiments with 69 (ctrl), 61 (TBP 38Q), and 53 (TBP 106Q) embryos in total. Fisher’s exact test. b Two guideRNAs (gR1 and gR2) were designed against exon 2 of the amot gene. PCR and subsequent Sanger sequencing revealed that editing occurred 3 bases upstream of the PAM sequence (yellow box) of gR2. c qPCR analysis of amot expression in 24 hours post fertilisation (hpf) embryos, injected either with Cas9 alone as control (Cas9) or a ribonucleoprotein consisting of both amot-specific guideRNAs and Cas9 (amot Crispr). n = 7 experiments. Wilcoxon test. d, e qPCR analysis of Hippo-Yap1 signalling target genes cyr61 (ccn1) and ctgf (ccn2) in 24 hpf Cas9-injected control and amot crispant embryos. n = 7 experiments. Wilcoxon test. f Amot crispants (amot Crispr) display more spontaneous movements than control-injected (Cas9) embryos or embryos co-injected with RNA encoding human HA-tagged AMOT (AMOT rescue). n = 3 experiments with 54 (Cas9), 53 (amot Crispr), and 61 (AMOT rescue) embryos per condition. Kruskal–Wallis test with Dunn’s post hoc test. g Amot crispants (amot Crispr) show reduced touch response compared to Cas9 controls and AMOT rescue embryos. n = 3 experiments with 38 (Cas9), 40 (amot Crispr), and 58 (AMOT rescue) embryos per condition. Kruskal–Wallis test with Dunn’s post hoc test. h Swimming ability is reduced in amot Crispr embryos compared to Cas9 controls and AMOT rescue embryos. n = 4 experiments with 56 (Cas9), 58 (amot Crispr), and 78 (AMOT rescue) embryos per condition. Kruskal–Wallis test with Dunn’s post hoc test. i Amot crispants of tg(isl1:gfp) embryos have fewer motor neurons at 72 hpf than Cas9 controls, and AMOT rescue. AMOT overexpression (AMOT) does not alter motor neuron numbers. Images show maximal projections of confocal z-stacks. Scale bar: 20 µm. Isl1-GFP-positive (Isl1-GFP+) cells were quantified for all conditions. n = 3 with 43 (Cas9), 28 (amot Crispr), 23 (AMOT rescue), and 39 (AMOT) embryos in total. One-way ANOVA with Šídák’s post hoc test. j Antisense morpholino oligonucleotide (MO)-mediated knockdown of amot (amot MO) decreases motor neuron number in tg(isl1:gfp) embryos at 72 hpf compared to embryos injected with a control MO (ctrl MO). Isl1-GFP-positive (Isl1-GFP+) cells were quantified for all conditions. n = 3 experiments with 26 embryos in total. Welch’s test. Grey horizontal lines represent the mean. k Amot tg(isl1:gfp) crispants display axon guidance defects as shown for axons extending from motor neurons in comparison to Cas9 controls, AMOT rescue, and AMOT overexpression (AMOT) embryos. Scale bar: 20 µm. n = 3 experiments with 43 (Cas9), 28 (amot Crispr), 23 (AMOT rescue), and 39 (AMOT) embryos in total. Fisher’s exact test. l In situ hybridisation of 72 hpf embryos for the dopamine transporter (dat) demonstrates an increase in the number of dopaminergic neurons and higher levels of dat in amot crispants compared to Cas9-injected controls. Scale bar: 100 µm. n = 3 experiments with 54 (Cas9) and 43 (amot Crispr) embryos in total. Mann–Whitney test (dat+ cells) and Fisher’s exact test (dat signal). Grey horizontal lines in graphs represent the median, unless stated otherwise. All tests were two-tailed. P values are shown in the graphs. Source data are provided as a Source Data file.
Discussion
In this study, we confirmed calpain-dependent TBP fragmentation, identifying T106 as a major cleavage site located C-terminal to the polyQ tract. C-terminal TBP fragments were predominantly nuclear and exhibited high proteostatic turnover, whereas N-terminal fragments containing the polyQ tract mislocalised to the cytoplasm, where they formed perinuclear IBs. PolyQ-expanded N-terminal fragments were cytotoxic and further overactivated calpains. Multi-omics analysis revealed TBP fragment-driven dysregulation of neuronal pathways, highlighting AMOT and its downstream YAP1 signalling as key players in SCA17-linked neurodegeneration, which was validated in vivo.
Earlier studies reported TBP fragmentation in SCA17 and AD without linking it to proteolysis8,15,53. We previously demonstrated calpain-mediated TBP cleavage in SCA17 models and showed that calpain inhibition alleviates molecular pathology. However, the precise cleavage sites and the resulting N-terminal fragments remained unidentified28. In the present study, we confirmed our initial findings and demonstrated that calpain-dependent TBP cleavage—given the predominantly cytoplasmic localisation of these proteases—is a non-nuclear process that affects cytoplasmic TBP after translation but prior to its reported nuclear import30. However, further research is required to determine whether this mechanism is conserved across cell types and tissues. By refining our calpain cleavage site analysis33, we confirmed the generation of N-terminal, polyQ-containing fragments, as well as their presence in vivo, and identified T106 | S107 as a major cleavage site. Constructs mimicking cleavage at T106 | S107 allowed us to characterise the functional and pathological properties of TBP fragments.
C-terminal TBP fragments containing the DNA-binding core domain localised to the nucleus—contrary to our earlier findings28, which may be explained by a misidentification of the previously observed fragments in cytoplasmic fractions due to the lack of calpain-specific control conditions. However, the current observation is consistent with predicted C-terminal nuclear localisation signals15 and previously reported karyopherin-mediated nuclear shuttling in yeast, which depends on a recognition motif located near the protein’s C-terminal core domain54,55. In contrast, N-terminal polyQ-containing fragments mislocalised to the cytoplasm, aligning with findings from internally truncated TBP variants15.
We observed rapid proteasomal turnover of full-length and C-terminal TBP, consistent with prior studies of TBP regulation56. However, inhibition of protein clearance pathways did not increase the low levels of N-terminal fragments, likely due to their lack of lysine residues required for canonical ubiquitination. Instead, these fragments—particularly the polyQ-expanded ones—were highly prone to aggregation, forming perinuclear IBs. This differs from the smaller intranuclear IBs seen with full-length TBP in SCA17 patients and models13,14,17,57. Notably, and consistent with our data, a non-polyQ-expanded N-terminal splice variant of TBP has been shown to aggregate into cytoplasmic and perinuclear inclusions in AD brains8,53, suggesting that both TBP fragmentation and aggregation may play a broader role in neurodegenerative diseases.
Transcriptional dysregulation in SCA17 has previously been linked to polyQ expansion-induced disruptions in TBP dimerisation and interactions with transcriptional regulators such as TFIIB, XBP1, and SP1, identifying targets like HSPB1, MANF, and INPP5A14,16,17. Our multi-omics approach revealed that TBP fragments induced widespread proteomic and transcriptomic changes affecting pathways related to neuronal function and neurodegeneration. These included mitochondrial gene dysregulation, consistent with polyQ- or calpain-associated mitochondrial dysfunction observed in SCA3, HD, and AD58–60.
Among the affected factors, we identified AMOT as significantly downregulated in cells expressing polyQ-expanded TBP or its fragments, whereas non-polyQ-expanded TBP appeared to promote AMOT expression, as demonstrated in TBP-overexpressing 293T cells compared to mock-transfected controls at both the mRNA and protein level. AMOT reduction appears to be multifactorial, resulting from transcriptional dysregulation, co-aggregation, and calpain-dependent fragmentation. We reproduced these molecular findings in SCA17 patient-derived neurons, which, in addition to calpain overactivation and increased TBP fragmentation, exhibited a significant reduction in AMOT. In SCA17 rats, the presence of polyQ-expanded TBP similarly led to reduced AMOT levels, particularly in the molecular layer of the cerebellum. A comparable regulatory pattern was observed in zebrafish: overexpression of non-polyQ-expanded TBP increased amot mRNA levels, whereas polyQ-expanded TBP failed to promote amot expression, suggesting an evolutionarily conserved mechanism.
AMOT is known to be essential for dendritic tree complexity and locomotor coordination61, and we found its reduction to be additionally associated with both calpain-mediated cleavage and sequestration into IBs. While we have observed calpain-mediated cleavage of AMOT in our experimental models, N-terminal processing by the protease DDI2 has been shown to alter its role in collective cell migration62. Our findings raise the possibility that calpain-mediated cleavage may also contribute to AMOT regulation in neuronal cells. We speculate that the co-aggregation of AMOT into TBP-positive IBs observed in this study may be promoted by the presence of an N-terminal polyQ stretch in AMOT. Consistent with this, we also observed reduced AMOT levels upon overexpression of polyQ-expanded ataxin-3, likely driven by co-aggregation and calpain overactivation, whereas non-polyQ-expanded ataxin-3 did not affect AMOT expression. Together, these findings suggest that AMOT dysregulation may not be specific to SCA17 but could represent a common feature of polyQ disorders involving calpain overactivity and aggregation-prone proteins18. Moreover, AMOT knockdown in 293T and neuronal SH-SY5Y cells led to increased calpain activity, indicating a regulatory interplay between AMOT and the calpain system.
AMOT depletion resulted in dysregulation of YAP1 signalling, affecting key neuronal targets such as CCN1 and CCN2, both of which are essential for neuronal development and function63. Notably, CCN1 has recently been linked to neural circuit stabilisation and myelination in the adult brain44.
Importantly, we observed reduced levels of YAP1 and CCN1 in the SCA17 rat cerebellum, accompanied by decreased MBP. Proteomic analysis further revealed an overrepresentation of downregulated proteins associated with oligodendrocytes, the myelinating cells of the central nervous system, whose dysfunction has been implicated in spinocerebellar ataxias and HD64,65.
Subsequent pathway and gene enrichment analysis highlighted widespread proteomic alterations in the SCA17 rat cerebellum, with strong associations to neuronal dysfunction and movement disorder-related neurodegeneration but also mRNA metabolism.
To investigate the neuronal role of AMOT in vivo, we genetically depleted amot in zebrafish embryos. This resulted in downregulation of YAP1 target genes, impaired motor neuron development, and altered motor function. These findings align with previously observed motor neuron and locomotor phenotypes in an SCA3 zebrafish model66,67, and with studies implicating motor neuron involvement in certain spinocerebellar ataxias68–70. Interestingly, beyond its known function in angiogenesis45, amot transcripts were detected in developing neuronal structures, indicating a potential role in neurogenesis.
Very little is known about early cellular or molecular changes in the development of SCA17. When the disease has progressed, however, a reduction in dopamine transporter expression or activity has been reported71. Our study shows that amot depletion increases dopamine transporter levels and produces a higher number of dopaminergic neurons—features also described in the context of HD50–52. Importantly, overexpression of the dopamine transporter has been shown to eventually cause dopaminergic neuron loss72. Thus, we speculate that early upregulation of the dopaminergic system may precede the loss of dopaminergic activity observed at later stages. As proposed for HD52, these neurodevelopmental abnormalities may contribute to early disease stages in SCA17 or other neurodegenerative disorders characterised by dysregulation of the AMOT-YAP1 signalling axis.
In summary, our study shows that calpain-mediated TBP fragmentation in SCA17 leads to a multifactorial depletion of AMOT, linking this process to both disrupted neurodevelopment and neurodegeneration. These findings suggest that AMOT deficiency, resulting from calpain activity and TBP dysfunction, may represent a shared mechanism contributing to neuronal pathology.
Methods
Ethics declarations
For experiments with SCA17 patient- and unaffected subject-derived fibroblast lines, informed written consent was required and obtained from all individuals or their legal guardians. All procedures involving human participants were conducted in accordance with the principles of the Declaration of Helsinki, and sampling was performed in accordance with the relevant local guidelines and regulations. The control line originated from a study approved by the Ethics Committee of the University of Tübingen (vote 598/2011BO1). All experiments involving wild-type and SCA17 rats were conducted after approval by the local ethics committee at the Regierungspräsidium Tübingen, in accordance with the German Animal Welfare Act and the guidelines of the Federation of European Laboratory Animal Science Associations, based on European Union legislation (Directive 2010/63/EU). Zebrafish husbandry and experiments in this study were approved by local authorities (Veterinary Care Unit at the University of Tübingen and the animal welfare commissioner of the regional board for scientific animal experiments in Tübingen, Germany). Experiments were conducted in agreement with the European Union Directive 86/609/EEC for the protection of animals used for experimental and other scientific purposes.
Sex as a biological variable
Sex was not considered a biological variable in this study. The SCA17 patient-derived fibroblast and iPSC line, as well as derived neurons, were analysed alongside age- and sex-matched control lines from an unaffected individual. All SCA17 rat tissues were derived from male individuals. Previous characterisation showed comparable phenotypes in both sexes41; therefore, sex was not included in the study design and is not expected to influence the measured parameters. Zebrafish experiments were conducted at embryonic stages prior to gonad formation, at which point biological sex is not yet established.
Plasmids and siRNAs
To overexpress human TBP in cells (UniProt identifier: P20226-1), pcDNA3.1 vectors carrying cDNA for C-terminally myc-tagged full-length TBP with 38 or 106 glutamines (myc-TBP 38Q or 106Q), their calpain cleavage site-mutated (S105-S108 > WWWW) forms, or respective N-terminal or C-terminal fragment constructs of TBP were synthesised (BioCat, Heidelberg, Germany). For the generation of capped RNA for zebrafish injection, untagged human TBP 38Q and 106Q, respectively, were subcloned into pCS2+ plasmids. Additionally, pcDNA3.1 vectors coding for C-terminally myc-tagged ataxin-3 isoform 1c (ATXN3; UniProt identifier: P54252-2) with 15Q or 148Q glutamines (myc-ATXN3 15Q or 148Q) were synthesised (BioCat) to overexpress human ataxin-3. The p3XFlag-CAPN1 construct was a gift from Yi Zhang (Addgene plasmid #60941; http://n2t.net/addgene:60941; RRID: Addgene_60941)73. The pcDNA3.1_nV5-AMOT (p130) was a gift from Michael Wehr (Addgene plasmid #172997; http://n2t.net/addgene:172997; RRID: Addgene_172997)74. Vectors were amplified in transformed E. coli DH5a cells and isolated using the QIAprep Spin Miniprep Kit (QIAGEN, Hilden, Germany).
For gene knockdown in cell models, MISSION® endoribonuclease-prepared small interfering RNAs (esiRNA) were used targeting human AMOT (esiAMOT; EHU129771, Merck, Darmstadt, Germany), human CAPN1 (esiCAPN1; EHU032581; Merck), human CAPN2 (esiCAPN2; EHU025391, Merck), as well as human CAST (esiCAST; EHU096291, Merck). As a control, an esiRNA targeting the Renilla luciferase gene RLUC (esiRLUC; EHURLUC, Merck) was employed. Information on esiRNA cDNA target sequences can be obtained from the corresponding product website.
Cell culture
HEK 293T (293T) (ATCC: CRL-3216) and SH-SY5Y (SY5Y) (ATCC: CRL-2266) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) foetal bovine serum (FBS), 1% (v/v) non-essential amino acids (MEM NEAA) and 1% (v/v) Antibiotic-Antimycotic (A/A) (all Gibco®, Thermo Fisher Scientific, Waltham, MA, USA) in 5% CO2 at 37 °C. 293T and SY5Y cells were transfected for 72 h using TurboFectin 8.0 (OriGene Technologies, Inc., Rockville, MD, USA) or Lipofectamine 3000 Reagent (Thermo Fisher Scientific), according to the manufacturer’s instructions.
Human fibroblast cell lines were established according to routine procedures after taking skin punch biopsies from an SCA17 patient as well as a sex- and age-matched unaffected control subject. Fibroblasts were grown in AmnioFroxx (Type A) culture medium (neoFroxx GmbH, Einhausen, Germany) in 5% CO2 at 37 °C.
Fragment analysis by capillary electrophoresis
Genomic DNA was extracted from fibroblasts derived from an SCA17 patient and a sex- and age-matched unaffected control subject using the High Pure PCR Template Preparation Kit (REF 11796828001; Roche Diagnostics, Mannheim, Germany) according to the manufacturer’s protocol.
SCA17 alleles were PCR-amplified using the HotStarTaq PCR Master Mix (QIAGEN, Hilden, Germany) and fluorescently labelled primers (forward: JOE-5’-TGG AAG AGC AAC AAA GGC AG-3’; reverse: 5’-GGC CTG AGG TTC CCT GTG-3’; Metabion International AG, Planegg, Germany). Amplicon sizes were determined by capillary electrophoresis using a 3500xL Genetic Analyzer with the GeneScan™ 500 ROX™ dye Size Standard and the Pop-7™ polymer (all Applied Biosystems, Thermo Fisher Scientific). Data were analysed using GeneMapper™ Software version 6.1 (Applied Biosystems, Thermo Fisher Scientific). CAG repeat lengths were calculated by subtracting the number of bases outside the repeat region (87 bp) from the product size of the most intense peak and dividing the result by three. Repeat lengths are reported with an accuracy of ± 1 repeat.
iPSC generation and differentiation to cortical neurons
Fibroblasts from an SCA17 patient were cultured in DMEM high glucose medium supplemented with 10% (v/v) FBS (all Gibco®, Thermo Fisher Scientific, Waltham, MA, USA) for up to 10 days in 5% CO2 at 37 °C. Fibroblasts were reprogrammed into iPSCs by nucleofection with the episomal plasmids pCXLE-hUL, pCXLE-hSK, and pCXLE-hOCT4. In brief, 1 × 105 cells were nucleofected with 1 µg of each plasmid. After 2 days, FGF2 (PeproTech) was added to the culture medium (DMEM, high glucose with 10% (v/v) FBS) in a concentration of 2 ng/ml. The following day, the culture medium was changed to Essential 8 (E8) medium (Gibco®, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 100 µM NaB (Sigma-Aldrich). Medium was changed every other day for 3–4 weeks, and afterwards colonies were picked, and iPSCs were differentiated into cortical neurons following an established protocol75,76. Briefly, iPSCs were seeded at 3 × 10⁵ cells/cm² on Matrigel (Corning)-coated 6-well plates in E8 medium supplemented with 10 μM Y-27632 (Selleckchem). Neural differentiation was carried out in 3 N medium throughout the differentiation. Until day 10, the medium was supplemented with 10 μM SB431542 (Sigma-Aldrich) and 500 nM LDN-193189 (Sigma-Aldrich), with the addition of 20 ng/ml FGF2 (PeproTech) on day 9. Cells were passaged at day 10 at a 1:3 ratio and expanded in 3 N medium supplemented with 10 μM Y-27632 on the day of passaging and 20 ng/ml FGF-2 until day 11. Heparin (100 μg/ml; Sigma-Aldrich) was added on days 14 and 16. A second passage at a 1:3 ratio was performed on day 20, with optional cryopreservation of neural precursor cells. At day 27, cells were dissociated using Accutase (Sigma-Aldrich) and replated onto poly-L-ornithine (Sigma-Aldrich) and Matrigel-coated plates at a density of 1.5 × 10⁵ cells/cm² (six-well format, for protein isolation) or 3× 10⁴ cells/cm² (24-well format, for immunocytochemistry). The following day, 10 μM PD0325901 (Tocris) and 10 μM DAPT (Sigma-Aldrich) were added, with a medium change after 2 days. From day 32 onwards, cells were maintained in 3 N medium with medium changes every other day until day 37.
Cell-based calpain assays
Cell-based calpain activation and inhibition assays were performed as previously described28. For activation of endogenous calpains, transiently transfected 293T cells and induced cortical neurons (iCNs) were incubated in Opti-MEM® I Reduced Serum Media (Gibco®, Thermo Fisher Scientific) containing 3 µM (for 293T cells) or 1 µM (for iCNs) of ionomycin (in DMSO) (407950, Calbiochem, Merck) and 5 mM CaCl2 for 3 h (for 293T cells) or 1 h (for iCNs) at standard culture conditions. For calpain-specificity control, 293T cells were pre-incubated with 25 µM N-Acetyl-L-leucyl-L-leucyl-L-norleucinal (ALLN; in DMSO) (A6185; Sigma-Aldrich, St. Louis, MO, USA), 25 µM calpeptin (n-benzyloxycarbonyl-L-leucylnorieucinal; in DMSO) (C8999, Sigma-Aldrich) or 25 µM calpain inhibitor III (CI-III; carbobenzoxy-valinyl-phenylalaninal; in DMSO) (208722, Calbiochem, Merck), pre-diluted in Opti-MEM® I Reduced Serum Media, 1 h prior to ionomycin/CaCl2 administration. For iCNs, 1-h pre-treatment with 20 µM CI-III diluted in Opti-MEM® I Reduced Serum Media served as the calpain-specificity control. For vehicle controls, DMSO was used.
Cell viability analysis
For the analysis of cell viability in cells expressing different TBP variants, 293T cells were seeded in poly-L-lysine (P4707, Sigma-Aldrich) pre-coated 96-well culture plates (ViewPlate-96 Black, Perkin Elmer, Massachusetts, USA) and transfected 24 h later. After an additional 72 h of incubation, cell viability was assessed using the PrestoBlue assay (Thermo Fisher Scientific, Karlsruhe, Germany), according to the manufacturer’s instructions, with the exception of reading the fluorescence signals after 30 min of incubation with PrestoBlue reagent.
Immunocytochemistry (ICC) staining and fluorescence microscopy
For immunocytochemistry (ICC) staining of transfected cells and fibroblasts, cells were seeded on poly-L-lysine (P4707, Sigma-Aldrich) pre-coated eight-well chamber slides (80841, Ibidi, Gräfelfing, Germany), transfected after 24 h with respective constructs and returned to standard culture conditions. After 72 h, cells were pre-fixed via supplementation of 0.4% (w/v) paraformaldehyde in 1× DPBS (Gibco®, Thermo Fisher Scientific) (PFA) to the medium and incubation at 37 °C for 10 min. Pre-fixative was discarded, cells were washed briefly with 1× DPBS and fixed with 4% (w/v) PFA at room temperature (RT) for 15 min. For ICC staining of iCNs, cells were grown on coverslips and fixed in 4% (w/v) PFA at RT for 15 min without prior pre-fixation. After washing cells with 1× DPBS, cells were incubated in blocking/permeabilisation buffer (10% (w/v) BSA, 0.5% (v/v) Triton X-100, 0.02% (w/v) NaN3 in 1× DPBS) for 1 h at RT. Subsequently, cells were probed with primary antibodies diluted in ICC antibody diluent (1% (w/v) BSA, 0.5% (v/v) Triton X-100, 0.02% (w/v) NaN3 in 1× DPBS) at 4 °C overnight. A detailed list of primary antibodies used for ICC stainings can be found in Supplementary Table 2. Afterwards, cells were washed with 1× DPBS and incubated with respective secondary antibodies: goat anti-mouse Alexa Fluor 488, goat anti-rabbit Alexa Fluor 488, goat anti-mouse Alexa Fluor 555, goat anti-mouse Alexa Fluor 594, goat anti-rabbit Alexa Fluor 594, goat anti-rat Alexa Fluor 594, or goat anti-chicken Alexa Fluor 647 (A32723, A27034, A-21424, A32742, A-11037, A-11007, A32933; Thermo Fisher Scientific) diluted 1:1000 for iCN staining, or 1:500 for all other cells, in antibody diluent at RT for 1 h. After washing with 1× DPBS and an additional rinse with ddH2O, stained cells were mounted and counterstained with VECTASHIELD® Antifade Mounting Medium with DAPI (H-1200, Vector Laboratories, Newark, CA, USA).
Cells were analysed with an Axioplan 2 fluorescence microscope, equipped with an ApoTome, Plan-Neofluar 40×/0.75 and Plan-Apochromat 63×/1.4 oil objectives, and an AxioCam MRm camera, using the AxioVision 4.8 imaging software (all Zeiss, Oberkochen, Germany).
To acquire confocal images, two confocal microscopes were used. Confocal images for orthogonal cross sections were acquired at a 1000× magnification on an ECLIPSE Ti2 microscope equipped with a CSU-W1 SoRa confocal system and a CFI SR HP Plan Apochromat Lambda S 100×C Sil objective, using NIS-Elements AR 5.42 (all Nikon, Tokyo, Japan). All other confocal images were acquired at a 200× or 630× magnification on a Stellaris 5 microscope, HC PL APO 20x/0.75 IMM CORR CS2 and HC PL APO 63x/1,30 GLYC CORR CS2 objective, with the LAS X 4.8.2.29567 imaging software (all Leica, Wetzlar, Germany).
Immunohistochemistry staining and fluorescence microscopy of rat tissue
For immunohistochemistry (IHC) staining, paraffin-embedded brains of 10-month-old male wild-type and SCA17 (TBPQ64) rats of the Sprague Dawley strain, obtained from a previous study41, were cut into 7 µm sagittal sections and mounted on slides. Deparaffinization and rehydration were performed in a xylene and ethanol series using the Leica autostainer XL (Leica). Antigens were retrieved with a 15 min microwave treatment in 80 mM sodium citrate and 20 mM citric acid pH 6.0, and then washed with 1× PBS. Sections were blocked for 1 h at RT with blocking solution (5% normal goat serum (S-1000, Vector Laboratories), 3% BSA, 0.1% Triton X-100, 0.02% NaN3 in 1× DPBS). Slides were incubated with primary antibodies diluted in IHC antibody diluent (2% normal goat serum, 3% BSA, 0.1% Triton X-100, 0.02% NaN3) overnight at 4 °C in a humid chamber. A detailed list of used primary antibodies for IHC staining can be found in Supplementary Table 2. Afterwards, sections were washed with 1× PBS and incubated with fluorescence-tagged secondary antibodies goat anti-rabbit Alexa Fluor 555 and goat anti-mouse 647 (A32732, A-21235; Thermo Fisher Scientific) diluted 1:300 in antibody diluent at RT for 1 h. Sections were washed in 1× PBS, rinsed with ddH2O, mounted and counterstained in VECTASHIELD® Antifade Mounting Medium with DAPI (H-1200, Vector Laboratories).
To image sections at ×40 and ×100 magnification, an Axioplan 2 fluorescence microscope, equipped with Achroplan 4×/0.10 and Plan-Neofluar 10×/0.30 objectives, and an AxioCam MRm camera, using the AxioVision 4.8 imaging software (all Zeiss) was used.
Fluorescence intensities of stained animal brain sections were assessed using the Fiji software77 (version 20240605-1417), and an in-house established Fiji script for semi-automated intensity analysis. Briefly, the mean signal intensities of ×40 magnified images of cerebellar layers (molecular layer, Purkinje cell layer, and granular cell layer) were determined and corrected for local mean background intensities. Details on the Fiji script will be shared upon request.
Protein extraction
For cell homogenates and lysates, 293T cells were detached by gentle pipetting and collected by centrifugation at 500×g for 5 min and 4 °C. Cell pellets were washed with cold 1× DPBS followed by brief centrifugation at 500×g for 2.5 min. For cell homogenisation, pellets were resuspended in an appropriate amount of RIPA buffer (50 mM Tris pH 7.5, 150 mM NaCl, 10% glycerol, 0.1% (w/v) SDS, 0.5% sodium deoxycholate and 1% (v/v) Triton X-100) supplemented with cOmplete™ protease inhibitor cocktail (Roche Diagnostics) and PhosSTOP™ phosphatase inhibitor (Roche Diagnostics) and incubated on ice for 30 min. Homogenates were ultrasonicated using a Sonopuls ultrasonic homogeniser (Bandelin electronic, Berlin, Germany) for 3 s and 10% pulse duration at 10% power, whereas cell lysates were obtained by centrifugation of protein extracts for 30 min at 4 °C and 20,000×g.
Protein extracts from the cerebella of 10-month-old male wild-type and SCA17 (TBPQ6441) rats of the Sprague Dawley strain were generated in a previous study, which provides detailed information on animal maintenance and tissue sampling28.
Protein concentrations of homogenates and lysates were determined by the Bradford assay (Bio-Rad Protein Assay; Bio-Rad Laboratories, Basel, Switzerland) and measured spectrophotometrically in a BioTek Synergy HT microtiter plate reader using the BioTek Gen5 software (both Agilent Technologies, Inc., Santa Clara, CA, USA). Samples were stored at −80 °C until further analysis.
Cytoplasmic-nuclear fractionation
To separate cytoplasmic and nuclear proteins, the Rapid, Efficient And Practical (REAP) fractionation method, based on the original protocol by Suzuki et al78., was applied, with following adaptations: Briefly, cell pellets were resuspended in TBS-I lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 0.1% (v/v) IGEPAL CA-630) supplemented with cOmplete™ protease inhibitor cocktail EDTA-free (Roche Diagnostics), PhosSTOP™ phosphatase inhibitor (Roche Diagnostics), and 10 µM calpeptin (C8999, Sigma-Aldrich). To receive a total protein (whole cell) homogenate, an aliquot of one-fourth of the lysis volume was immediately transferred to a new tube and diluted with an equal volume of TBS-I lysis buffer. The remaining material was centrifuged at 20,800×g for 10 s. The supernatant, containing the cytoplasmic fraction, was carefully transferred to a separate tube, while avoiding the disruption of the nuclear pellet. The nuclear fraction was washed by resuspending the pellet with double the initial lysis volume in TBS-I lysis buffer, centrifuged at 20,800×g for 10 s, and discarding the supernatant. This step was repeated once. The washed nuclear fraction was resuspended in half of the initial lysis volume in TBS-I lysis buffer. Total protein homogenate and nuclear fraction were ultrasonicated at a 10% pulse duration and 10% power for 3 s using a Sonopuls ultrasonic homogeniser (Bandelin electronic). The nuclear fraction underwent up to two additional ultrasonication steps to effectively disrupt nuclear membranes. Protein concentrations were determined by Bradford assay and measured spectrophotometrically in a microtiter plate.
Immunoprecipitation and on-bead calpain cleavage assay
Immunoprecipitation (IP) of target proteins was performed using ChromoTek Myc-Trap® Agarose (ChromoTek GmbH, Planegg, Germany) according to the manufacturer’s protocol with the following adaptations. Shortly, cell lysates were prepared by resuspending cell pellets in TRAP lysis buffer (10 mM Tris pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 0.5% IGEPAL CA-630) containing cOmplete™ protease inhibitor cocktail EDTA-free and PhosSTOP™ phosphatase inhibitor (Roche Diagnostics), ultrasonication for 3 s and 10% pulse duration at 10% power using a Sonopuls ultrasonic homogeniser (Bandelin electronic) and centrifugation for 30 min at 4 °C and 20,000×g. Cell and tissue lysates were diluted to 5 µg/µl protein with TRAP lysis buffer after protein concentrations were determined by Bradford assay, and TRAP dilution buffer (10 mM Tris/Cl pH 7.5, 150 mM NaCl, 0.5 mM EDTA, containing cOmplete™ protease inhibitor cocktail EDTA-free and PhosSTOP™ phosphatase inhibitor (Roche Diagnostics)) was added in a ratio of 2:3 to the samples. For IP, 500 µg or 1000 µg protein in a maximum volume of 500 µl was added to the equilibrated beads and incubated overnight at 4 °C using an end-over-end rotator. Beads were washed thrice using TRAP dilution buffer, and proteins were eluted with 2× LDS buffer (2× LDS, 100 mM dithiotreitol (DTT)) under heat-denaturation for 10 min at 70 °C.
In case of an on-bead calpain cleavage assay, beads were resuspended in 100 µl calpain reaction buffer (CRB; 20 mM HEPES/KOH pH 7.6, 10 mM KCl, 1.5 mM MgCl2, 1 mM DTT) after the last washing step. 200 mM CaCl2 and 200 ng of recombinant calpain-1 (208712, Merck) were added and incubated for 10 min on an end-over-end rotator at RT. The time series of calpain-1 cleavage was carried out for 0, 1, and 5 min. The reaction was stopped by adding 4× LDS sample buffer (1 M Tris pH 8.5, 50% (v/v) glycerol, 8% (w/v) LDS, 2 mM EDTA, 0.1% (w/v) Orange G) in a ratio 3:1 and 100 mM DTT, followed by heat-denaturation for 10 min at 70 °C. Beads were sedimented by centrifugation for 5 min at maximum speed and 4 °C. Supernatant was transferred into new tubes and stored at -80 °C until further analysis.
Western blotting
Western blotting was performed according to standard procedures. Briefly, 25 µg of protein from cell homogenates or lysates were supplemented with 4× LDS sample buffer in a ratio of 3:1 and 100 mM DTT, followed by heat-denaturation for 10 min at 70 °C. Electrophoretic separation of protein samples was achieved via SDS-polyacrylamide gel electrophoresis (SDS-PAGE) using self-made 9% Bis-Tris gels and MOPS electrophoresis buffer (50 mM MOPS, 50 mM Tris pH 7.7, 0.1% (w/v) SDS, 1 mM EDTA). Proteins were blotted on Amersham™ Protran™ Premium 0.2 µm nitrocellulose membrane (Cytiva, Freiburg, Germany) using Bicine/Bis-Tris transfer buffer (25 mM Bicine, 25 mM Bis-Tris pH 7.2, 1 mM EDTA, 15% (v/v) methanol) in a TE22 Transfer Tank (Hoefer, Inc., Holliston, MA, US) at 80 V and a maximum of 250 mA for 2 h followed by standard immunodetection protocol.
Filter retardation assay
Detection of SDS-insoluble TBP was performed as previously described28. Briefly, 3 µg total protein of cell homogenates was diluted in 1× DPBS containing 2% (w/v) SDS and 50 mM DTT, and heat-denatured for 5 min at 95 °C. After equilibrating Amersham™ Protran™ 0.45 µm nitrocellulose membrane (GE Healthcare) with 1× DPBS containing 0.1% (w/v) SDS, samples were filtered through the membrane using a Minifold® II Slot Blot System (Schleicher & Schuell). Slots were washed with 1× DPBS containing 0.1% SDS (w/v) and 1× DPBS before the standard immunodetection protocol was applied.
Membrane-based immunodetection
Membranes were blocked with 5% (w/v) skim milk powder (Sigma-Aldrich) in TBS (10 mM Tris pH 7.5, 150 mM NaCl) for 1 h at RT, and incubated with primary antibodies diluted in 1× TBS-T (1× TBS with 0.1% (v/v) Tween-20) overnight at 4 °C. A detailed list of primary antibodies used for membrane-based immunodetection can be found in Supplementary Table 3. Afterwards, membranes were washed with 1× TBS-T and probed with respective secondary IRDye® antibodies goat anti-mouse 680LT (P/N 926-68020), goat anti-mouse 800CW (P/N 926-32210), and goat anti-rabbit 800CW (P/N 926-32211) (all 1:5,000; LI-COR Biosciences) or horseradish peroxidase (HRP)-linked antibodies goat anti-mouse HRP (#A28177; Thermo Fisher Scientific), donkey anti-mouse HRP (715-035-150; Jackson ImmunoResearch Europe Ltd, Cambridgeshire, UK), and goat anti-rabbit HRP (#G-21234; Thermo Fisher Scientific) at RT for 1 h. After secondary antibody incubation, membranes were washed with 1× TBS-T. For chemiluminescence detections, membranes were shortly incubated with 1× TBS followed by incubation with WesternBrightTM Sirius HRP substrates (K-12043; Advansta Inc., San Jose, CA, USA) for 2 min at RT. Fluorescence and chemiluminescence signals were detected using the LI-COR ODYSSEY® FC and quantified with Image Studio 6.1.0.79 software (both LI-COR Biosciences).
Membrane-based inclusion body microscopy
For membrane-based imaging of TBP inclusion bodies (IBs), a filter retardation assay was performed as described above. After primary antibody incubation, secondary antibodies goat anti-mouse Alexa Fluor 488 and rabbit Alexa Fluor 594 (all 1:500; A32723, A-11037, both Thermo Fisher Scientific) were added and incubated at RT for 1 h. Membranes were mounted on glass slides and fluorescent images were taken at a ×40 magnification on an Axioplan 2 imaging microscope equipped with an Achroplan 4×/0.10 objective and an AxioCam MRm camera, using the AxioVision 4.8 imaging software (all Zeiss).
Microscopy-based assessment of inclusion bodies
To quantify the frequency of staining-positive particles imaged by membrane-based inclusion body microscopy, an in-house-developed Fiji macro was used, as previously described79. For each slot, three images were acquired and analysed. To reduce false positive events, particles with intensity densities below the 75% percentile of respective mock conditions were excluded from further analyses.
Using the same Fiji macro, inclusion body (IB) frequency, cross-sectional area, equivalent circular area diameter (dECA), and roundness were determined in cells stained by ICC.
Orthogonal slices and maximum intensity projections of confocal z-stacks were generated using Fiji77, and three-dimensional reconstructions were created using the built-in tools of Leica LAS X software.
Microscopic evaluation of TBP and AMOT co-aggregation was performed manually by assessing the fluorescence signal intensity of both immunostainings.
Mass spectrometry and data processing
Protein extracts from 293T cell or rat cerebellum samples for label-free quantitative (LFQ) proteomics or immunoprecipitated calpain-dependent protein fragments for calpain cleavage site analysis were purified using Bis-Tris-based SDS-PAGE. Coomassie-stained gel pieces were excised, and in-gel digested using trypsin (for protein extracts) or chymotrypsin (for purified protein fragments), as described previously80. After desalting using C18 Stage tips81 extracted peptides were separated on an Easy-nLC 1200 system coupled to a Q Exactive HF mass spectrometer (both Thermo Fisher Scientific) as described elsewhere82 with slight modifications: The peptide mixtures were separated using a 1-h gradient, and the 7 most intense precursor ions were sequentially fragmented in each scan cycle using higher energy collisional dissociation (HCD) fragmentation.
Acquired MS spectra were processed with the MaxQuant software package version 2.2.0.083 with integrated Andromeda search engine84. For 293T cell analysis, a database search was performed against a target-decoy Homo sapiens database obtained from UniProt, containing 96,817 protein entries and 286 commonly observed contaminants. For analysis of the proteome from rat cerebella, a database search was performed against a Rattus norvegicus database obtained from UniProt containing 49,592 protein entries and 286 commonly observed contaminants. In addition, one small database was created containing sequences used for pulldown. For calpain cleavage site identification, endoprotease chymotrypsin was defined as a protease with a maximum of two missed cleavages; the enzyme mode was semi-specific. Oxidation of methionine and N-terminal acetylation were specified as variable modifications, whereas carbamidomethylation on cysteine was set as a fixed modification. Initial maximum allowed mass tolerance was set to 4.5 parts per million (ppm) for precursor ions and 20 ppm for fragment ions. Peptide, protein and modification site identifications were reported at a false discovery rate (FDR) of 0.01, estimated by the target/decoy approach85. The LFQ (Label-Free Quantification) algorithm was enabled, and a match between runs86,87 and LFQ protein intensities was used for relative protein quantification of three replicates. Downstream bioinformatic analysis (two-sample t tests and Volcano plots) was performed using the Perseus software package, version 1.6.2.3. Data were filtered for contaminants, reversed and only identified by site entries.
RNA extraction
RNA from 293T cells was isolated using the RNeasy Mini Kit (QIAGEN) with minor modifications. After disrupting cells in an appropriate volume of RLT buffer, cells were homogenised using QIAshredder spin columns. DNA digestion was conducted using QIAGEN’s RNase-free DNase Set according to the specifications of the manufacturer’s protocol. The RNeasy spin column was dried by centrifugation at full speed for 3 min. RNA was eluted after an incubation of 5 min at RT in 30 µl RNase-free water.
Total RNA of zebrafish embryos was extracted using Zymo’s RNA Microprep kit (Zymo Research Europe GmbH, Freiburg, Germany), which includes a DNaseI step to remove genomic DNA.
RNA sequencing and data processing
The concentration of RNA was measured using the Qubit Fluorometric Quantitation and RNA Broad-Range Assay (Thermo Fisher Scientific). RNA Integrity Number RIN was determined using the Fragment Analyzer 5300 and the Fragment Analyzer RNA kit (Agilent Technologies) and presented a good integrity (RIN > 9.5). For library preparation, the mRNA fraction was enriched using polyA capture from 100 ng of total RNA using the NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs GmbH, Frankfurt am Main, Germany). Subsequently, libraries were prepared using the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina and NEBNext UDI UMI (New England Biolabs GmbH) following the manufacturer’s instructions. To minimise technical batch effects, library preparations were performed using the liquid handler Biomek i7 (Beckman Coulter GmbH, Krefeld, Germany). The library molarity was determined by measuring the library size (approximately 400 bp) using the Fragment Analyzer 5300 and the Fragment Analyzer DNA HS NGS fragment kit (Agilent Technologies), and the library concentration ( > 2 ng/µl) using Qubit Fluorometric Quantitation and dsDNA High sensitivity assay (Thermo Fisher Scientific). The libraries were denatured according to the manufacturer’s instructions, diluted to 150 pM and sequenced as paired-end 100 bp reads on an Illumina NovaSeq 6000 (Illumina, Inc., San Diego, CA, USA). The sequencing aimed to achieve a depth of approximately >25 million clusters per sample.
Read quality of RNA-seq data in fastq files was assessed using ngs-bits (2023_11-197-g1bcb6874) to identify sequencing cycles with low average quality, adaptor contamination, or repetitive sequences from PCR amplification. Reads were aligned using STAR v2.7.10b88 to the GRCh38, and alignment quality was analysed using ngs-bits and visually inspected in the Integrative Genome Viewer (v2.15.4). Normalised read counts for all genes were obtained using Subread (v2.0.4) and edgeR (v3.42.4).
Raw expression values are available for 61,850 genes in 15 samples. Raw gene expression was filtered by demanding a minimum expression value of 1 cpm (counts per million) in at least 2 samples. Filtered data contains expression values for 16,340 genes. The distribution of logarithmised cpm-normalised expression values shows similar characteristics over all samples. Based on the filtered dataset, samples were investigated with respect to their pairwise similarity. Spearman’s rank correlation coefficient was calculated for each pair of samples. A hierarchical clustering was performed on the resulting similarity values. Differential gene expression analysis was conducted based on the filtered gene expression dataset. A statistical model incorporating the group property of samples was tested by fitting a negative binomial distribution using a generalised linear model (GLM) approach. For each gene, gene expression fold changes (log2 fold change) were computed, and a statistical test was performed to assess the significance, which is given as raw P value and adjusted P value (FDR).
Zebrafish maintenance and manipulation
Zebrafish were maintained in a circulating water system (Tecniplast Germany GmbH, Hohenpeißenberg, Germany) with automated monitoring and adjustment of water parameters (conductivity, pH, and temperature). Zebrafish were kept under a dark–light cycle (10 h dark/14 h light) and fed four times a day. In the morning, fish received freshly-hatched artemia and afterwards dry food (Zebrafeed by Sparos, Olhão, Portugal) dispensed by the TRITONE automatic feeding system (Tecniplast Germany GmbH). Lines used included AB and EK as wild-type zebrafish and the reporter line tg(isl1:gfp)89, expressing GFP in motor neurons. To generate amot loss-of-function embryos using Crispr/Cas9, a ribonucleoprotein complex consisting of two guideRNAs and Cas9 HiFi (both Integrated DNA Technologies, Inc., Coralville, IA, USA) was assembled and injected at the 1-cell stage. The following sequences in exon 2 of amot were targeted: Amot gRNA1: 5’-AAGCTCCTCCTTCCATGAGG, Amot gRNA2: 5’-GACTGGAGTCACCAACTCTA. As control, embryos were injected with Cas9 alone or with a complex consisting of Cas9 protein and a negative control gRNA (IDT, Leuven, Belgium). Rescue experiments were achieved by subsequent injections of ribonucleoprotein complex and capped RNA. Gene editing in every individual embryo was verified by PCR amplification of exon 2 before further analysis. In addition, the editing capacity of the gRNAs was assessed by Sanger sequencing (Microsynth AG, Göttingen, Germany) of exon 2 PCRs in edited and control-injected embryos. To knock down amot, a previously published MO was used45. MO injections were controlled against embryos injected with a standard control morpholino oligonucleotide (MO) (Gene Tools, LLC, Philomath, OR, USA). In addition, in all experiments, embryos were compared to non-injected controls to assess clutch quality. Rescue experiments with human AMOT as well as overexpression of TBP variants were achieved by injection of capped RNA at the 1-cell stage. Capped RNA was generated from linearised plasmids, which were then transcribed using the mMESSAGE mMACHINE SP6 transcription kit (Ambion, Thermo Fisher Scientific). As a template for human AMOT a pcDNA3-HA AMOT overexpression plasmid was used, which was a gift from Kunliang Guan (Addgene plasmid # 32821; http://n2t.net/addgene:32821; RRID:Addgene 32821)90. After manipulation, all embryos were allowed to develop at 28.5 °C until the desired stage.
In situ hybridisation
A 1095 bp fragment of zebrafish amot (Ensemble accession number ENSDART00000141096.2) was amplified from cDNA generated from 24 hpf zebrafish embryos and cloned via TOPO TA cloning (Invitrogen, Thermo Fisher Scientific) into pCRII. A DIG-labelled in situ probe was generated after linearization of the plasmid with XhoI (New England Biolabs GmbH) and in vitro transcription using SP6 RNA polymerase and the DIG labelling mix (Roche Diagnostics). To generate a probe against the dopamine transporter (dat, slc6a3, Genbank accession no. NM_131755.1), a 1256 bp fragment was amplified from 24 hpf cDNA. A T3 polymerase binding element was fused to the reverse primer and a probe transcribed using T3 polymerase (Promega GmbH, Walldorf, Germany). In situ hybridisation was done according to standard protocols91 using embryos at the indicated stages.
Movement analyses of zebrafish embryos
To investigate spontaneous coiling and tailflick movements, 5-min movies were recorded of 27 hpf embryos still in their chorion using a Leica S9i stereomicroscope (Leica Microsystems GmbH, Wetzlar, Germany) with an integrated camera. Spontaneous bursts of activity were counted manually.
The swimming ability of zebrafish embryos was assessed at 48 hpf after manual dechorionisation using forceps. Embryos were gently tapped on the back of their head, and their reaction was recorded and scored based on the following: no reaction (0), twitch upon touch (1), swimming behaviour upon touch (2). All embryos were tapped at least twice and up to five times.
Immunofluorescence in zebrafish embryos
To allow for the detection of not only cell bodies, but also the axons extended from motor neurons, the expressed GFP of the transgenic fish line was enhanced by an additional anti-GFP staining. Similarly, human TBP overexpression was detected by immunofluorescence as follows: Embryos at the desired stage were fixed in 4% buffered paraformaldehyde for 1 h. After several washes in PBS-T (1× PBS with 0.1% Tween-20), the embryos were incubated in distilled water for 5 min, permeabilised in ice-cold acetone for 7 min and again incubated for 5 min in water. Following another 5-min incubation step in PBS-T with 1% DMSO (PBD-T), the embryos were blocked for at least 2 h in 10% normal goat serum (Biozol Diagnostica Vertrieb GmbH, Hamburg, Germany) diluted in PBD-T. Primary and secondary antibodies were diluted in blocking buffer. Embryos were first incubated overnight at 4 °C in the primary antibody solution. After several washing steps in PBD-T containing 1% normal goat serum and 0.1 M NaCl, the embryos were incubated for several hours at RT in secondary antibody solution, followed by several final washes. For confocal imaging, the embryo’s tail was cut off and mounted between two coverslips in VECTASHIELD® Antifade Mounting Medium with DAPI (Vector Laboratories). Antibodies used were: mouse anti-GFP (1:250; clone 3E6, A-11120, Thermo Fisher Scientific), mouse anti-TBP (clone 1TBP18, sc-56794, Santa Cruz Biotechnology), goat anti-mouse Alexa Fluor 488 (1:1,000, #4408, Cell Signaling Technology), donkey anti-mouse Alexa Fluor 568 (1:1,000; A10037; Thermo Fisher Scientific).
Imaging of zebrafish embryos
Zebrafish embryos processed by in situ hybridisation were imaged on a Leica M125 stereomicroscope and a Flexacam C1 colour camera (both Leica Microsystems GmbH). Fluorescence images of immunostained embryos were acquired with a M205 FCA stereomicroscope and a K8 sCMOS camera (both Leica Microsystems GmbH). Z-stacks of immunostained zebrafish embryos were acquired with a Leica Stellaris 5 confocal microscope (Leica Microsystems GmbH) equipped with a system-optimised section distance. Cell bodies of motor neurons were counted with the help of the Cell Counter plug-in in Fiji77.
Quantitative real-time PCR
To analyse gene expression in 293T cells, 1 µg of RNA was reverse transcribed using the QuantiTect Reverse Transcription Kit (Qiagen) according to the manufacturer’s protocol. Relative gene expression of TBP and AMOT was analysed by quantitative real-time PCR (qPCR) using the SYBR Green PCR Master Mix (Qiagen). In short, 1× QuantiTect SYBR Green PCR Master Mix, 1 µM of forward and 1 µM of reverse primer (metabion international AG), as well as RNase-free water were added to cDNA. The qPCR was run on the LightCycler 480 II (Roche Diagnostics) using the LightCycler 480 Software 1.5.62 SP3 following the manufacturer’s protocol. For normalisation, reference genes SDHA and YWHAZ were employed. The following primers were used: TBP: forward: 5’-GAAGTTGGGTTTTCCAGCTAA, reverse: 5’-GGAGAACAATTCTGGGTTTGA; AMOT: forward: 5’-GGGCATGTCCGTTCCTTGAG, reverse: 5’-AGGTGGATGGGCCTTAACTC; SDHA: forward: 5’-AGAAGCCCTTTGAGGAGCA, reverse: 5’-CGATTACGGGTCTATATTCCAGA; YWHAZ: forward: 5’-TCCAGGGACAGAGTCTCAGC, reverse: 5’-TCCATGACTGGATGTTCTGTG.
To analyse gene expression in zebrafish embryos, equal amounts of RNA were transcribed into cDNA using the Protoscript II Kit (New England Biolabs GmbH). qPCR was done on a QuantstudioTM 3 Real-Time PCR System (Thermo Fisher Scientific) with Luna® Universal qPCR Master Mix (New England Biolab GmbH) and the Universal Probe system (Roche Diagnostics). All reactions were done in triplicate and normalised to beta 2-microglobulin (b2m). The following primers and probes were used: b2m (UP65): forward: 5’- ACATCACTGTACAGGGGAAAGTC-3’, reverse: 5’-TCCGTTCTTCAGCAGTTCAA-3’; amot (UP23): forward: 5’-AAGAGCCAGGAGAACAGCAAA-3’, reverse: 5’-TACTGCTGTGAAACACTGGCA-3’; ctgf (UP5): forward: 5’-CCACAAGAAGACACCTTCCAG-3’, reverse: 5’-CAGGCTGGAGACCGTGAG-3’; cyr61 (UP68): forward: 5’-TTGCTCTGCTGACGG AG-3’, reverse: 5’-GGGCAGGATGGAGGAGAC-3’.
In silico calpain cleavage site prediction
For in silico prediction of calpain cleavage site in human TBP (UniProt identifier: P20226-1), the Group-based Prediction System–Calpain Cleavage Detector (GPS-CCD v1.0) tool was employed (available at http://ccd.biocuckoo.org)92. Resulting cleavage likelihood scores for every amino acid position were normalised to the maximum threshold value (0.654) and graphically visualised using the GraphPad Prism software version 10.6.1 (GraphPad Software, Dotmatics, Boston, MA, USA).
Pathway analysis
For canonical pathway enrichment, a core analysis for differentially expressed genes (DEGs) obtained from RNA sequencing was performed using QIAGEN Ingenuity Pathway Analysis (IPA) software (version 24.0.2). Suggested default parameters and DEGs with P value ≤ 0.05 were used for analysis. Results were graphically visualised by GraphPad Prism software version 10.6.1 (GraphPad Software).
Additional in silico pathway and protein-protein interaction analyses were performed using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) version 12.0 (https://string-db.org/)93. Networks were adjusted using CorelDRAW Graphics Suite 2024.
Cell-type enrichment analysis using PanglaoDB
To infer the cell type from which the proteomics signatures were obtained, potential cell-type enrichment analysis was performed using the PanglaoDB Augmented 2021 database94. The PanglaoDB dataset was downloaded from EnrichR libraries (https://maayanlab.cloud/Enrichr/#libraries) and processed to obtain cell-type-specific marker genes. Each cell type was represented as a set of curated marker genes.
For the analysis, the decoupleR-based over-representation analysis (ORA)) was performed using the decoupleR (v2.16.0) R package95. The PanglaoDB data were converted to network format, with cell types as the source and marker genes as the targets. The rat gene symbols were converted to uppercase for the correct mapping with PanglaoDB.
The proteomics data were obtained from mass spectrometry-based experiments and consisted of significantly altered proteins (173) and total detected proteins (1830). The total detected proteins were used as the background universe, and the significant proteins were used as the foreground (hits).
To achieve this, a binary matrix was created, where each gene in the background was given a value of 1 if it was found in the significant gene set, and 0 if it was not. The run_ora function in the decoupleR package was used for the over-representation analysis, which performs Fisher’s exact test to determine the enrichment of cell-type marker genes in the input gene list. However, only gene sets with at least 5 genes were considered for the analysis. The adjusted P values were determined using the Benjamini & Hochberg procedure for correcting the false discovery rate.
An independent analysis was also performed using the enricher function in the clusterProfiler package, which uses the hypergeometric test96. Although the analysis was performed on the PanglaoDB gene sets, the background was not specified, making it comparable to the Enrichr platform97. The aforementioned approaches were used in an overlapping manner to identify cell-type signatures in the proteomics data.
InteractiVenn
For comparative analysis, data obtained from multi-omics approaches were graphically visualised using the web-based InteractiVenn platform98.
Bubble plot
Data obtained from QIAGEN IPA was graphically visualised by bubble plots that were generated using the ggplot2 package99 in R 4.2.2100. Full script can be provided upon request.
Statistics and reproducibility
For qualitative analyses, all experiments were independently repeated at least three times and yielded comparable results, ensuring validity and reproducibility. Quantitative data were statistically analysed using GraphPad Prism software version 10.6.1 (GraphPad Software). For in cellulo- and rat tissue sample-based data sets, normal distribution was assessed using the Shapiro–Wilk test. Data Statistical outliers were identified by the ROUT test (Q = 1%). Normally distributed data are presented as means + standard error of the mean (s.e.m.). Non-normally distributed data are shown as median, indicating the 25th and 75th percentiles. Two-sided one-sample t test, Student’s t test, one-way ANOVA, or Kruskal–Wallis test with an adequate post hoc analysis were applied for statistical testing.
Zebrafish data were first assessed for normal distribution using a Shapiro–Wilk test, before the respective test was applied. All t test-based analyses as well as Fisher’s exact tests were two-sided.
Significance was assumed with a P value ≤ 0.05. For further details, see the respective figure and figure legends, and the Source Data file.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Supplementary information
Description of Additional Supplementary Information
Source data
Acknowledgements
We thank Theresa Höchstädter (Institute of Medical Genetics and Applied Genomics, Tübingen, Germany) for providing SCA3-related cell culture samples; Jessica Nowacki-Hansen and the diagnostics team (Department of Human Genetics, Ruhr University Bochum, Bochum, Germany) for TBP repeat length determination in human-derived samples; Libo Yu-Täger (Department of Human Genetics, Ruhr University Bochum, Bochum, Germany) for advice on immunohistochemistry staining; and Martin D. Burkhalter (Department of Experimental and Clinical Pharmacology and Pharmacogenomics, Section of Pharmacogenomics, University of Tübingen, Tübingen, Germany) for support with confocal microscopy. We thank the animal care staff for excellent fish husbandry. We also thank Yi Zhang (Boston Children’s Hospital, Harvard Medical School, MA, USA), Kunliang Guan (Westlake University, Hangzhou, China), and Michael Wehr (Department of Psychiatry and Psychotherapy, LMU University Hospital, LMU Munich, Munich, Germany) for providing CAPN1 and AMOT overexpression vectors via the Addgene plasmid repository. We are grateful to the SCA17-affected donor who provided the skin biopsy used to generate the corresponding fibroblasts and iPSC lines.
Author contributions
R.D.I.E. and F.F. jointly designed and performed most of the experiments, analysed and interpreted the data, generated the figures, and wrote the manuscript. C.H. performed additional cell culture-based assays, qPCR, and fluorescence microscopy, generated figures, and wrote the manuscript. M.D.P performed and analysed zebrafish experiments. T.I.S. and J.H. generated and characterised SCA17 patient fibroblast-derived iPSCs and iCNs. P.P.S. conducted western blot analysis and cell viability assays. N.C. designed and performed the RNA sequencing-based transcriptome analysis. A.V. designed and conducted mass spectrometry-based proteomics and calpain cleavage site analysis. R.P.P. performed cell type enrichment analysis of proteome datasets from TBPQ64 rats. B.M. designed and supervised the mass spectrometry-based proteomics and calpain cleavage site analysis. O.R. supervised the project and secured funding. S.H. supervised the generation and characterisation of SCA17 patient fibroblast-derived iPSCs and iCNs. H.P.N. conceptualised and supervised the project, secured funding, and provided skin biopsies for the generation of SCA17 patient-derived fibroblasts. M.P. designed, performed, and analysed zebrafish experiments and wrote the manuscript. J.J.W. conceptualised and supervised the project, secured funding, analysed and interpreted the data, generated figures, and wrote the manuscript.
Peer review
Peer review information
Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.
Funding
This study was funded by the German Research Foundation (DFG; grants WE 6585/1-1 to J.J.W. and NG 101/6-1 to H.P.N.) and the Sophia und Fritz Heinemann Foundation (grant to J.J.W., 2021). Next-generation sequencing (NGS) was performed with support from the German Research Foundation-funded NGS Competence Center Tübingen (INST 37/1049-1). Confocal imaging was supported by additional funding from the German Research Foundation (INST 37/1171-1 FUGG/467868420 to M.P.). Open Access funding enabled and organized by Projekt DEAL.
Data availability
All data required to support the findings of this study are available within the article, the supplementary information, or are available in publicly available data repositories. The RNA sequencing data generated in this study have been deposited in the NCBI Gene Expression Omnibus (GEO) database under accession code GSE346419. The mass spectrometry proteomics data generated in this study have been deposited in the ProteomeXchange Consortium via the PRIDE101 partner repository under accession code PXD083710. Additional information can be provided upon request. Source data are provided with this paper.
Code availability
Standard codes and software tools, as listed in the “Methods” section, were used in this study. Custom Fiji macros developed for the analysis of microscopy images are provided as Supplementary Software and have been deposited at Zenodo (https://doi.org/10.5281/zenodo.22231266).
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Rana Dilara Incebacak Eltemur, Florian Fath.
Supplementary information
The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-79055-x.
References
- 1.Akhtar, W. & Veenstra, G. J. C. TBP-related factors: a paradigm of diversity in transcription initiation. Cell Biosci.1, 23 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kramm, K., Engel, C. & Grohmann, D. Transcription initiation factor TBP: old friend new questions. Biochem. Soc. Trans.47, 411–423 (2019). [DOI] [PubMed] [Google Scholar]
- 3.Davidenko, A., Bogomazova, A., Illarioshkin, S. & Lagarkova, M. Molecular mechanisms of spinocerebellar ataxia type 17. Mol. Neurobiol.62, 5720–5729 (2025). [DOI] [PubMed] [Google Scholar]
- 4.Martianov, I., Viville, S. & Davidson, I. RNA polymerase II transcription in murine cells lacking the TATA binding protein. Science298, 1036–1039 (2002). [DOI] [PubMed] [Google Scholar]
- 5.Shen, Y. et al. TBP bookmarks and preserves neural stem cell fate memory by orchestrating local chromatin architecture. Mol. Cell85, 413–429.e410 (2025). [DOI] [PubMed] [Google Scholar]
- 6.Luna-Arias, J. P. & Castro-Muñozledo, F. Participation of the TBP-associated factors (TAFs) in cell differentiation. J. Cell. Physiol.239, e31167 (2024). [DOI] [PubMed] [Google Scholar]
- 7.Choubtum, L. et al. Trinucleotide repeat expansion of TATA-binding protein gene associated with Parkinson’s disease: a Thai multicenter study. Parkinsonism Relat. Disord.28, 146–149 (2016). [DOI] [PubMed] [Google Scholar]
- 8.Reid, S. J. et al. TBP, a polyglutamine tract containing protein, accumulates in Alzheimer’s disease. Mol. Brain Res.125, 120–128 (2004). [DOI] [PubMed] [Google Scholar]
- 9.Ohi, K. et al. TATA Box-Binding Protein gene is associated with risk for schizophrenia, age at onset and prefrontal function. Genes, Brain Behav.8, 473–480 (2009). [DOI] [PubMed] [Google Scholar]
- 10.Park, H., Jeon, B. S., Shin, J. H. & Park, S.-H. A patient with 41 CAG repeats in SCA17 presenting with Parkinsonism and chorea. Parkinsonism Relat. Disord.22, 106–107 (2016). [DOI] [PubMed] [Google Scholar]
- 11.Zühlke, C., Dalski, A., Schwinger, E. & Finckh, U. Spinocerebellar ataxia type 17: report of a family with reduced penetrance of an unstable Gln49 TBP allele, haplotype analysis supporting a founder effect for unstable alleles and comparative analysis of SCA17 genotypes. BMC Med. Genet.6, 27 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Toyoshima, Y. & Takahashi, H. Spinocerebellar ataxia type 17 (SCA17). In Polyglutamine Disorders (eds Nóbrega, C. & Pereira de Almeida, L.) 219–231 (Springer International Publishing, Cham, 2018). [DOI] [PubMed]
- 13.Nakamura, K. et al. SCA17, a novel autosomal dominant cerebellar ataxia caused by an expanded polyglutamine in TATA-binding protein. Hum. Mol. Genet.10, 1441–1448 (2001). [DOI] [PubMed] [Google Scholar]
- 14.Friedman, M. J. et al. Polyglutamine domain modulates the TBP-TFIIB interaction: implications for its normal function and neurodegeneration. Nat. Neurosci.10, 1519–1528 (2007). [DOI] [PubMed] [Google Scholar]
- 15.Friedman, M. J., Wang, C.-E., Li, X.-J. & Li, S. Polyglutamine expansion reduces the association of TATA-binding protein with DNA and induces DNA binding-independent neurotoxicity. J. Biol. Chem.283, 8283–8290 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Yang, S., Huang, S., Gaertig, M. A., Li, X. J. & Li, S. Age-dependent decrease in chaperone activity impairs MANF expression, leading to Purkinje cell degeneration in inducible SCA17 mice. Neuron81, 349–365 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Liu, Q. et al. Cerebellum-enriched protein INPP5A contributes to selective neuropathology in mouse model of spinocerebellar ataxias type 17. Nat. Commun.11, 1101 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Incebacak Eltemur, R. D., Nguyen, H. P. & Weber, J. J. Calpain-mediated proteolysis as driver and modulator of polyglutamine toxicity. Front. Mol. Neurosci.15, 1020104 (2022). [DOI] [PMC free article] [PubMed]
- 19.Matos, C. A., Almeida, L. P. & Nobrega, C. Proteolytic cleavage of polyglutamine disease-causing proteins: revisiting the toxic fragment hypothesis. Curr. Pharm. Des.23, 753–775 (2017). [DOI] [PubMed] [Google Scholar]
- 20.Rai, M., Curley, M., Coleman, Z. & Demontis, F. Contribution of proteases to the hallmarks of aging and to age-related neurodegeneration. Aging Cell21, e13603 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Esch, F. S. et al. Cleavage of amyloid beta peptide during constitutive processing of its precursor. Science248, 1122–1124 (1990). [DOI] [PubMed] [Google Scholar]
- 22.Siman, R., Card, J. P. & Davis, L. G. Proteolytic processing of beta-amyloid precursor by calpain I. J. Neurosci.10, 2400–2411 (1990). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Baba, M. et al. Aggregation of alpha-synuclein in Lewy bodies of sporadic Parkinson’s disease and dementia with Lewy bodies. Am. J. Pathol.152, 879–884 (1998). [PMC free article] [PubMed] [Google Scholar]
- 24.Dufty, B. M. et al. Calpain-cleavage of alpha-synuclein: connecting proteolytic processing to disease-linked aggregation. Am. J. Pathol.170, 1725–1738 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wellington, C. L. et al. Caspase cleavage of gene products associated with triplet expansion disorders generates truncated fragments containing the polyglutamine tract. J. Biol. Chem.273, 9158–9167 (1998). [DOI] [PubMed] [Google Scholar]
- 26.Gafni, J. & Ellerby, L. M. Calpain activation in Huntington’s disease. J. Neurosci.22, 4842–4849 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Koch, P. et al. Excitation-induced ataxin-3 aggregation in neurons from patients with Machado-Joseph disease. Nature480, 543–546 (2011). [DOI] [PubMed] [Google Scholar]
- 28.Weber, J. J. et al. Calpains as novel players in the molecular pathogenesis of spinocerebellar ataxia type 17. Cell. Mol. Life Sci.79, 262 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Orphanides, G., Lagrange, T. & Reinberg, D. The general transcription factors of RNA polymerase II. Genes Dev.10, 2657–2683 (1996). [DOI] [PubMed] [Google Scholar]
- 30.Pemberton, L. F., Rosenblum, J. S. & Blobel, G. Nuclear import of the TATA-binding protein: mediation by the karyopherin Kap114p and a possible mechanism for intranuclear targeting. J. Cell Biol.145, 1407–1417 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lam, M., Dubyak, G. & Distelhorst, C. W. Effect of glucocorticosteroid treatment on intracellular calcium homeostasis in mouse lymphoma cells. Mol. Endocrinol.7, 686–693 (1993). [DOI] [PubMed] [Google Scholar]
- 32.Tsujinaka, T. et al. Synthesis of a new cell penetrating calpain inhibitor (calpeptin). Biochem. Biophys. Res. Commun.153, 1201–1208 (1988). [DOI] [PubMed] [Google Scholar]
- 33.Weber, J. J. et al. A combinatorial approach to identify calpain cleavage sites in the Machado-Joseph disease protein ataxin-3. Brain140, 1280–1299 (2017). [DOI] [PubMed] [Google Scholar]
- 34.Menden, B. et al. Loss-of-function variants in the CAPN1 activator CD99L2 cause X-linked spastic ataxia. Nat. Commun.17, 1698 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Liu, Y. et al. Endogenous mutant Huntingtin alters the corticogenesis via lowering Golgi recruiting ARF1 in cortical organoid. Mol. Psychiatry29, 3024–3039 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Askeland, G. et al. Increased nuclear DNA damage precedes mitochondrial dysfunction in peripheral blood mononuclear cells from Huntington’s disease patients. Sci. Rep.8, 9817 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Virlogeux, A. et al. Increasing brain palmitoylation rescues behavior and neuropathology in Huntington disease mice. Sci. Adv.7, eabb0799 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wertz, M. H. et al. Genome-wide in vivo CNS screening identifies genes that modify CNS neuronal survival and mHTT toxicity. Neuron106, 76–89.e78 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Huang, T. et al. The physiological role of Motin family and its dysregulation in tumorigenesis. J. Transl. Med.16, 98 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lieberman, A. P., Shakkottai, V. G. & Albin, R. L. Polyglutamine repeats in neurodegenerative diseases. Annu Rev. Pathol.14, 1–27 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Kelp, A. et al. A novel transgenic rat model for spinocerebellar ataxia type 17 recapitulates neuropathological changes and supplies in vivo imaging biomarkers. J. Neurosci.33, 9068–9081 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Zhang, H., Pasolli, H. A. & Fuchs, E. Yes-associated protein (YAP) transcriptional coactivator functions in balancing growth and differentiation in skin. Proc. Natl. Acad. Sci. USA. 108, 2270–2275 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Urtasun, R. et al. Connective tissue growth factor autocriny in human hepatocellular carcinoma: oncogenic role and regulation by epidermal growth factor receptor/yes-associated protein-mediated activation. Hepatology54, 2149–2158 (2011). [DOI] [PubMed] [Google Scholar]
- 44.Sancho, L. et al. Astrocyte CCN1 stabilizes neural circuits in the adult brain. Nature649, 948–958 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Aase, K. et al. Angiomotin regulates endothelial cell migration during embryonic angiogenesis. Genes Dev.21, 2055–2068 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Burczyk, M. et al. Phenotypic regulation of the sphingosine 1-phosphate receptor miles apart by G protein-coupled receptor kinase 2. Biochemistry54, 765–775 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Lv, M. et al. Angiomotin promotes renal epithelial and carcinoma cell proliferation by retaining the nuclear YAP. Oncotarget7, 12393–12403 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Mahmood, F. et al. Generation and validation of a zebrafish model of EAST (epilepsy, ataxia, sensorineural deafness and tubulopathy) syndrome. Dis. Model Mech.6, 652–660 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Low, S. E. et al. Touch responsiveness in zebrafish requires voltage-gated calcium channel 2.1b. J. Neurophysiol.108, 148–159 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Jahanshahi, A. et al. Hyperdopaminergic status in experimental Huntington disease. J. Neuropathol. Exp. Neurol.69, 910–917 (2010). [DOI] [PubMed] [Google Scholar]
- 51.Jahanshahi, A. et al. Changes in brainstem serotonergic and dopaminergic cell populations in experimental and clinical Huntington’s disease. Neuroscience238, 71–81 (2013). [DOI] [PubMed] [Google Scholar]
- 52.Siebzehnrübl, F. A. et al. Early postnatal behavioral, cellular, and molecular changes in models of Huntington disease are reversible by HDAC inhibition. Proc. Natl. Acad. Sci. USA. 115, E8765–E8774 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Reid, S. J., Whittaker, D. J., Greenwood, D. & Snell, R. G. A splice variant of the TATA-box binding protein encoding the polyglutamine-containing N-terminal domain that accumulates in Alzheimer’s disease. Brain Res.1268, 190–199 (2009). [DOI] [PubMed] [Google Scholar]
- 54.Liao, C. C. et al. Karyopherin Kap114p-mediated trans-repression controls ribosomal gene expression under saline stress. EMBO Rep.21, e48324 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Morehouse, H., Buratowski, R. M., Silver, P. A. & Buratowski, S. The importin/karyopherin Kap114 mediates the nuclear import of TATA-binding protein. Proc. Natl. Acad. Sci. USA96, 12542–12547 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Li, L., Martinez, S. S., Hu, W., Liu, Z. & Tjian, R. A specific E3 ligase/deubiquitinase pair modulates TBP protein levels during muscle differentiation. eLife4, e08536 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Bruni, A. C. et al. Behavioral disorder, dementia, ataxia, and rigidity in a large family with TATA box-binding protein mutation. Arch. Neurol.61, 1314–1320 (2004). [DOI] [PubMed] [Google Scholar]
- 58.Harmuth, T. et al. Mitochondrial morphology, function and homeostasis are impaired by expression of an N-terminal calpain cleavage fragment of ataxin-3. Front. Mol. Neurosci.11, 368 (2018). [DOI] [PMC free article] [PubMed]
- 59.Jiang, S., Shao, C., Tang, F., Wang, W. & Zhu, X. Dynamin-like protein 1 cleavage by calpain in Alzheimer’s disease. Aging Cell18, e12912 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Hu, D. et al. Small-molecule suppression of calpastatin degradation reduces neuropathology in models of Huntington’s disease. Nat. Commun.12, 5305 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Rojek, K. O. et al. Amot and Yap1 regulate neuronal dendritic tree complexity and locomotor coordination in mice. PLoS Biol.17, e3000253 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Wang, Y. et al. Angiomotin cleavage promotes leader formation and collective cell migration. Dev. Cell60, 101–118.e107 (2025). [DOI] [PubMed] [Google Scholar]
- 63.Malik, A. R., Liszewska, E. & Jaworski, J. Matricellular proteins of the Cyr61/CTGF/NOV (CCN) family and the nervous system. Front. Cell Neurosci.9, 237 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Putka, A. F., Mato, J. P. & McLoughlin, H. S. Myelinating glia: potential therapeutic targets in polyglutamine spinocerebellar ataxias. Cells12, 601 (2023). [DOI] [PMC free article] [PubMed]
- 65.Cvetanovic, M. & Gray, M. Contribution of glial cells to polyglutamine diseases: observations from patients and mouse models. Neurotherapeutics20, 48–66 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66.Watchon, M. et al. Calpain inhibition is protective in Machado–Joseph disease zebrafish due to induction of autophagy. J. Neurosci.37, 7782–7794 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67.Pereira Sena, P. et al. Pathophysiological interplay between O-GlcNAc transferase and the Machado-Joseph disease protein ataxin-3. Proc. Natl. Acad. Sci. USA. 118, e2025810118 (2021). [DOI] [PMC free article] [PubMed]
- 68.Coarelli, G. et al. Motor neuron involvement threatens survival in spinocerebellar ataxia type 1. Neuropathol. Appl. Neurobiol.49, e12897 (2023). [DOI] [PubMed] [Google Scholar]
- 69.Ferrari, V. et al. Rare association between spinocerebellar ataxia and amyotrophic lateral sclerosis: a case series. Neurol. Sci.45, 4367–4371 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70.Jaques, C. S. et al. Spinocerebellar ataxia type 3 presenting simultaneously with motor neuron disease and cerebellar ataxia. Arq. Neuropsiquiatr.79, 851–852 (2021). [DOI] [PubMed] [Google Scholar]
- 71.Minnerop, M. et al. Putamen dopamine transporter and glucose metabolism are reduced in SCA17. Ann. Neurol.58, 490–491 (2005). [DOI] [PubMed] [Google Scholar]
- 72.Masoud, S. T. et al. Increased expression of the dopamine transporter leads to loss of dopamine neurons, oxidative stress and l-DOPA reversible motor deficits. Neurobiol. Dis.74, 66–75 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 73.Wang, Y. & Zhang, Y. Regulation of TET protein stability by calpains. Cell Rep.6, 278–284 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74.Ma, X. et al. Comprehensive split TEV based protein-protein interaction screening reveals TAOK2 as a key modulator of Hippo signalling to limit growth. Cell Signal113, 110917 (2024). [DOI] [PubMed] [Google Scholar]
- 75.Hauser, S. et al. Comparative transcriptional profiling of motor neuron disorder-associated genes in various human cell culture models. Front. Cell Dev. Biol.8, 544043 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 76.Rehbach, K. et al. Multiparametric rapid screening of neuronal process pathology for drug target identification in HSP patient-specific neurons. Sci. Rep.9, 9615 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 77.Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods9, 676–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 78.Suzuki, K., Bose, P., Leong-Quong, R. Y., Fujita, D. J. & Riabowol, K. REAP: a two minute cell fractionation method. BMC Res. Notes3, 294 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 79.Weber, J. J. et al. The parkin V380L variant is a genetic modifier of Machado-Joseph disease with impact on mitophagy. Acta Neuropathol.148, 14 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 80.Borchert, N. et al. Proteogenomics of Pristionchus pacificus reveals distinct proteome structure of nematode models. Genome Res.20, 837–846 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 81.Rappsilber, J., Mann, M. & Ishihama, Y. Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nat. Protoc.2, 1896–1906 (2007). [DOI] [PubMed] [Google Scholar]
- 82.Kliza, K. et al. Internally tagged ubiquitin: a tool to identify linear polyubiquitin-modified proteins by mass spectrometry. Nat. Methods14, 504–512 (2017). [DOI] [PubMed] [Google Scholar]
- 83.Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol.26, 1367–1372 (2008). [DOI] [PubMed] [Google Scholar]
- 84.Cox, J. et al. Andromeda: a peptide search engine integrated into the MaxQuant environment. J. Proteome Res10, 1794–1805 (2011). [DOI] [PubMed] [Google Scholar]
- 85.Elias, J. E. & Gygi, S. P. Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nat. Methods4, 207–214 (2007). [DOI] [PubMed] [Google Scholar]
- 86.Schwanhäusser, B. et al. Global quantification of mammalian gene expression control. Nature473, 337–342 (2011). [DOI] [PubMed] [Google Scholar]
- 87.Luber, C. A. et al. Quantitative proteomics reveals subset-specific viral recognition in dendritic cells. Immunity32, 279–289 (2010). [DOI] [PubMed] [Google Scholar]
- 88.Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics29, 15–21 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 89.Higashijima, S., Hotta, Y. & Okamoto, H. Visualization of cranial motor neurons in live transgenic zebrafish expressing green fluorescent protein under the control of the islet-1 promoter/enhancer. J. Neurosci.20, 206–218 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 90.Zhao, B. et al. Angiomotin is a novel Hippo pathway component that inhibits YAP oncoprotein. Genes Dev.25, 51–63 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 91.Thisse, C. & Thisse, B. High-resolution in situ hybridization to whole-mount zebrafish embryos. Nat. Protoc.3, 59–69 (2008). [DOI] [PubMed] [Google Scholar]
- 92.Liu, Z. et al. GPS-CCD: a novel computational program for the prediction of calpain cleavage sites. PLoS ONE6, e19001 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 93.Szklarczyk, D. et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res.51, D638–d646 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 94.Franzén, O., Gan, L. M. & Björkegren, J. L. M. PanglaoDB: a web server for exploration of mouse and human single-cell RNA sequencing data. Database2019, baz046 (2019). [DOI] [PMC free article] [PubMed]
- 95.Badia, I. M. P. et al. decoupleR: ensemble of computational methods to infer biological activities from omics data. Bioinform. Adv.2, vbac016 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 96.Xu, S. et al. Using clusterProfiler to characterize multiomics data. Nat. Protoc.19, 3292–3320 (2024). [DOI] [PubMed] [Google Scholar]
- 97.Xie, Z. et al. Gene set knowledge discovery with EnrichR. Curr. Protoc.1, e90 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 98.Heberle, H., Meirelles, G. V., da Silva, F. R., Telles, G. P. & Minghim, R. InteractiVenn: a web-based tool for the analysis of sets through Venn diagrams. BMC Bioinforma.16, 169 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 99.Wickham, H. ggplot2:Elegant Graphics forData Analysis (Springer-Verlag New York, 2016).
- 100.R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, Vienna, Austria, 2024).
- 101.Perez-Riverol, Y. et al. The PRIDE database at 20 years: 2025 update. Nucleic Acids Res.53, D543–d553 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Description of Additional Supplementary Information
Data Availability Statement
All data required to support the findings of this study are available within the article, the supplementary information, or are available in publicly available data repositories. The RNA sequencing data generated in this study have been deposited in the NCBI Gene Expression Omnibus (GEO) database under accession code GSE346419. The mass spectrometry proteomics data generated in this study have been deposited in the ProteomeXchange Consortium via the PRIDE101 partner repository under accession code PXD083710. Additional information can be provided upon request. Source data are provided with this paper.
Standard codes and software tools, as listed in the “Methods” section, were used in this study. Custom Fiji macros developed for the analysis of microscopy images are provided as Supplementary Software and have been deposited at Zenodo (https://doi.org/10.5281/zenodo.22231266).
