Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: a, Expression and distribution of ATXN2 in oligodendrocytes was assessed using Oli-neu cells under normal and oxidative stress conditions. Normally diffuse ATXN2 (green) was found to relocate into SGs marked by PABP (red) under sodium-arsenite (NaAsO 2 ) treatment. DAPI (blue) marks the nuclei. b, Expression of Atxn2 transcript in differentiated Oli-neu cells was measured under normal growth conditions by qRT-PCR. Upon starvation stress, Atxn2 transcript is significantly upregulated over time (2-6 h). Actb was used as housekeeping gene. Each data point represents a biological replicate. c, Presence of ATXN2 protein in differentiated Oli-neu cells was validated with quantitative immunoblots under normal growth conditions. ATXN2 protein abundance was significantly upregulated upon starvation stress at 6 h. ACTB was used as loading control. Each data point represents a biological replicate. d, Immunohistochemical assessment of ATXN2 protein (red) in OLIG2+ oligodendrocytes (green) in WT and KIN cerebella at terminal stage confirms its expression and diffuse distribution in WT oligodendrocytes, and shows its toxic cytosolic aggregation in KIN oligodendrocytes. e, Immunohistochemical assessment of ATXN2 co-localization within PDGFRα- and CC1-positive cells in cerebellar WM showed its expression in both OPCs and mature oligodendrocytes. f, Quantification of all OPCs in cerebellar WM showed a significant increase in KIN mice, however a reduction in ATXN2-positive portion was observed. g, Quantification of all mature oligodendrocytes in cerebellar WM showed a significant decrease in KIN mice, however the number of ATXN2-positive portion was significantly increased. h, Comparison of ATXN2 signals from oligodendroglia versus other cells in cerebellar WM showed majority of ATXN2 signal to come from oligodendroglial lineage in both WT and KIN mice.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: Expressing, Quantitative RT-PCR, Western Blot, Quantitative Proteomics, Control, Immunohistochemical staining, Comparison

Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: Transcript levels of all neuronal and myelin compartment factors in WT and transgenic (Tg) ATXN2 -Q58 mouse cerebellum at pre-onset (16 wk) and late-symptomatic (46 wk) stages. While no dysregulation of axonal markers was observed in early or late disease stages, only a mild downregulation was observed in several oligodendroglial transcripts, such as Mbp , Mag , Mog and Aspa. The dysregulation pattern of Mag and Mog were found opposite to that of KIN mice at this stage. Tbp was used as housekeeping gene in qRT-PCR experiments. Each data point represents a single animal.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: Transgenic Assay, Quantitative RT-PCR

Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: a, c, Schemes depicting the alternative splicing patterns of Mag and Plp1 transcripts, with indication of which isoform gets upregulated in the case of QKI loss. Expression levels of Mag (b) and Plp1 (d) splice isoforms were measured by semi-qPCR in KIN cerebellum (Cb, 3 mo, 14 mo), KIN spinal cord (SC, 14 mo) and KO cerebellum (Cb, 6 mo) in comparison to age-matched WT controls. Strong splicing disturbances indicated by L-Mag/S-Mag and Dm-20/Plp1 ratios were observed readily at pre-onset stage (3 mo) in KIN cerebellum, that showed further progression until the terminal stage (14 mo) in both cerebellum and spinal cord. Lack of a similar missplicing pattern in KO cerebellum ensured the ATXN2 aggregation-dependent nature of these dysregulations. The mild alteration of Plp1 splicing in KO samples observed here was not confirmed by further qRT-PCR experiments (see Figure S7b ). Actb was used as housekeeping gene and loading control in semi-qPCR experiments. Each data point represents a single animal.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: Alternative Splicing, Expressing, Comparison, Quantitative RT-PCR, Control

Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: Transcript levels of Mag and Plp1 splice isoforms were measured by qRT-PCR in (a) KIN cerebellum (Cb, 3 mo, 14 mo) and spinal cord (SC, 14 mo), (b) KO cerebellum (Cb, 6 mo) and (c) transgenic (Tg) ATXN2 -Q58 mouse cerebellum (Cb, 16 wk, 46 wk) in comparison to age-matched WT controls. The splicing defects previously observed with semi-qPCRs were validated for all KIN tissues. No splice defects were observed in KO or Tg ATXN2 -Q58 cerebellum, suggesting that ATXN2 aggregation toxicity specifically in oligodendroglia impacts the splicing pattern of mature myelin markers. Actb was used as housekeeping gene in qRT-PCR experiments. Each data point represents a single animal.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: Quantitative RT-PCR, Transgenic Assay, Comparison

Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: Expression levels of Cnp transcript isoforms, produced by an unknown splice factor, were measured by qRT-PCR in KIN cerebellum (Cb) at 14 months. No alteration of its splicing pattern was observed even at the terminal disease stage, suggesting that the alterations of Mag and Plp1 levels observed earlier were specific outcomes of ATXN2 pathology. Actb was used as housekeeping gene. Each data point represents a single animal.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: Expressing, Produced, Quantitative RT-PCR

Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: a-c, Subcellular distribution of QKI isoforms (red) and their co-localization with ATXN2 (green) were investigated using Oli-neu cells under normal and oxidative stress conditions. DAPI (blue) marks the nuclei. Under normal growth conditions, QKI5 showed a nuclear localization, while QKI6 and QKI7 were detected in both cytosol and nucleus. Under oxidative stress induced by sodium-arsenite (NaAsO 2 ), QKI6 showed the highest rate of association with the SGs marked by ATXN2, although a portion of QKI7 and a smaller portion of QKI5 also co-localized with SGs. Insets marked with yellow squares in NaAsO 2 images were shown enlarged in lower panels. White dotted lines mark the straightened segments shown below. Normalized plot profiles of each channel in straightened segments depict the co-localization of respective QKI isoform with ATXN2 in SGs. d, Separation of WT and KIN cerebella at the terminal stage into PN (soluble) and Urea (insoluble) fractions showed an increased abundance of QKI6 and QKI7 to co-fractionate with mutant ATXN2 in the insoluble fraction (red arrows). The segregation of both QKI isoforms at different levels than that of their original size (≈35 kDa) in Urea fraction is suspected to be the result of post-translational modifications they encounter at the aggregates. e-h, Immunohistochemical assessment of ATXN2 and QKI6 co-aggregation in PDGFRα- positive and negative cells of the white matter (OPCs and more mature stages of the lineage, respectively). At least two sections were imaged from three WT and KIN animals with similar findings.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: Mutagenesis, Immunohistochemical staining

Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: a, Protein levels of QKI5, QKI6 and QKI7 were measured in KIN cerebellum (Cb) at pre-onset (3 mo) and terminal (14 mo) disease stages with age-matched WT controls. QKI5 remained upregulated and QKI7 remained downregulated in KIN tissue throughout the disease course, but QKI6 showed a pre-onset upregulation that transformed into a strong downregulation at the terminal stage, mimicking a profile usually displayed by proteins sequestrated into aggregation in the disease course. ACTB was used as loading control in quantitative immunoblots. Each data point represents a single animal. b, Immunohistochemical assessment of QKI6 (green) localization in WT and KIN cerebellar WM at the terminal stage (14 mo) showed its diffuse cytosolic distribution in WT samples, and its sequestration into cytosolic aggregates in KIN samples marked by SG marker PABP (red). DAPI (blue) marks the nuclei. Transcript levels of Qki5 , Qki6 and Qki7 were measured with specific primers designed for qRT-PCR in KIN (c) , KO (d) and transgenic (Tg) ATXN2 -Q58 mouse (e) tissues at different disease stages with age-matched WT controls. Only late-onset dysregulations were observed in KIN cerebellum (Cb) and spinal cord (SC), without a change in KO or Tg ATXN2 -Q58 cerebellum. Actb was used as housekeeping gene in qRT-PCR experiments. Each data point represents a single animal.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: Transformation Assay, Control, Western Blot, Immunohistochemical staining, Marker, Quantitative RT-PCR, Transgenic Assay

Journal: bioRxiv

Article Title: ATXN2 polyglutamine expansion impairs QKI-dependent alternative splicing and oligodendrocyte maintenance

doi: 10.1101/2025.08.08.669189

Figure Lengend Snippet: Progressive aggregation of ATXN2 in neurons is a hallmark of its toxicity across multiple neurodegeneration syndromes. These aggregates also sequester other RNA-binding proteins, including PABP, TDP-43, and FUS, disrupting a multitude of subcellular processes ranging from altered autophagy to axonal cargo transport, which ultimately lead to the degeneration of axon terminals and cell somata. Our findings from the Atxn2 -CAG100-KIN mouse that expresses the mutant protein throughout the body at an endogenous level show that Purkinje cell somata are largely preserved even at the terminal stages of the disease course, whereas demyelination, and consequently, axon-myelin disconnection are among the first events of pathogenesis. Damaged myelin stimulates OPC proliferation and differentiation, however with unsuccessful maturation. We show that ATXN2 aggregates also form in cerebellar oligodendrocytes, sequestering the splice factor QKI, which leads to a dramatic loss of key mature myelin proteins, such as L-Mag and Plp1. Additionally, pronounced astrogliosis and microgliosis are also observed in KIN tissue, likely as a consequence of myelin and axonal damage.

Article Snippet: The membranes were blocked in 5% BSA/TBS-T for 1 h at RT, and incubated overnight at 4°C with primary antibodies against ATXN2 (Proteintech #21776-1-AP, 1:500), ACTB (Sigma #A5441, 1:10000), CALB1 (Cell Signaling #13176, 1:2500), CNP (Cell Signaling #5664S, 1:1000), MAG (Cell Signaling #9043S, 1:500), MBP (Merck #05-675, 1:250), MOG (Abcam #ab32760, 1:500), NEFH/M (Proteintech #18934-1-AP, 1:500), NEFL (Cell Signaling #2837S, 1:1000), NPTN (Alomone Labs #ANR-090, 1:500), PLP1 (Abcam #ab28486, 1:1000), TUBA4A (Aviva Systems #ARP40179-P050, 1:500), QKI5 (Merck Millipore, #MABN661, 1:500), QKI6 (Neuromab, #75-190, 1:2500) and QKI7 (Merck Millipore, #AB9908, 1:1000).

Techniques: RNA Binding Assay, Mutagenesis