DTA expression by oligodendrocyte-specific AAVs deletes oligodendrocytes in the mouse cerebellum during the specific developmental period.

A, Diagram illustrating the injection of AAV-hMAG-mCherry and AAV-hMAG-DTA into the mouse cerebellum at early postnatal days (P6-7). B, Fluorescence microscopy image showing mCherry expression (red) in the section of cerebellum at P14 following AAV-hMAG-mCherry injection at P7. Scale bar, 300 µm. C, Specificity of mCherry expression (red) in ASPA-positive oligodendrocytes (green) at P14 by AAV-hMAG-mCherry injection at P7. Scale bar, 100 µm. D, Scatter plot graph depicting the percentage of mCherry and ASPA double-positive cells among mCherry-positive populations (8 regions from 2 mice). E, Low-magnification image showing the distribution of MAG-mCherry expression across sagittal brain sections at P14 after AAV-hMAG-mCherry injection at P7. Expression is confined to the cerebellum and absent in extracerebellar regions. Scale bar, 500 µm. F, Higher-magnification views of regions indicated in panel E. Robust mCherry expression is observed in the cerebellum (1), whereas no detectable expression is found in the midbrain (2) or cerebral cortex (3). Scale bar, 200 µm. G, Quantification of the regional specificity of MAG-mCherry expression. Scatter plots indicate the percentage of mCherry-positive cells detected in the cortex, medulla, pons, and midbrain (4 regions from 2 mice). There was no mCherry-positive cell in these regions. H, Sequential visualization of oligodendrocyte reduction over time, indicated by ASPA staining in cerebellar sections from control (AAV-hMAG-mCherry) and DTA-treated (AAV-hMAG-DTA) mice at P10, P14, P21, and P78. Scale bar, 100 µm. I, Quantification of ASPA-positive cell density at each stage (from 4 mice per group per each stage). Bars and dots indicate mean and data from individual fields of view, respectively. Adulthood data were collected at P60-P94. **** p < 0.0001 (Nested t-test). J, Representative images of MBP immunostaining (green) in the cerebellar white matter of control and DTA-treated mice at P10, P14, P21, and adulthood (P60). Scale bar, 100 µm. K, Quantification of MBP-positive bundle area at each stage (from 4 mice per group per stage). Bars and dots indicate mean and data from individual fields of view, respectively. p < 0.05, **** p < 0.0001 (Nested t-test).

Cerebellar oligodendrocyte ablation reduces the synchrony of PC population activity during mouse cerebellar development.

A, Experimental timeline showing AAV injection at P7 and in vivo calcium imaging at P13– P15 in control and DTA-expressing mice with jGCaMP7f expressed in PCs. B, Schematic of in vivo one photon calcium imaging and a representative frame from a field scan in lobule VII/VIII of a control mouse at P14 with analyzed ROIs. Scale bar, 0.5 mm. C-F, Decrease in correlation coefficients (CCs) indicating reduced synchrony of spontaneous activities among PC population at P13-15 following DTA-mediated oligodendrocyte ablation. C, Representative traces of spontaneous calcium transients (ΔF/F0) extracted from 25 regions of interest (ROIs) in the cerebellum of control and DTA mice at P14. Scale bar, 20 s. D, Correlation coefficient matrices of spontaneous calcium activity across 300 ROIs in control (top) and DTA (bottom) mice at P14. Color scale indicates the strength of pairwise correlations. E, Mean correlation coefficients between calcium traces from pairs of ROIs plotted against the distance of separation between pairs of ROIs in control (n = 5 mice) and AAV-hMAG-DTA-injected (n = 7 mice) mice at P13-15. Error bars indicate SEM. **** p < 0.0001 (two-way ANOVA). F, Scatter plot graph of CCs for ROI pairs, categorized by separation distances of 0–80 mm and 120–200 mm (analyzed from four mice per group). Lines and plots indicate mean and data from individual separation distances, respectively. Significant group differences were observed (* p < 0.05, Nested t-test, 4 mice for each). G-I, Frequency distributions of Frequency (G), Amplitude (H), and Area (I) of calcium transients in CTL (AAV-hMAG-mCherry) and DTA-treated (AAV-hMAG-DTA) mice at P13–15. Circles and squares indicate data from individual ROIs, and bars represent SEM. There was no significant difference between groups (two-way ANOVA).

Developmental oligodendrocyte ablation reduces CF synchrony but does not alter PF excitatory or inhibitory synaptic inputs to PCs.

A, Experimental scheme. AAV-hMAG-RFP or AAV-hMAG-DTA was injected into the cerebellum at P7, and AAV-TRE-jGCaMP7f was injected into the inferior olive at P10 to label CFs. CF activity was imaged in the cerebellar cortex at P13-15. B, Representative calcium traces of CFs from control (CTL) and DTA-treated mice at P14. Scale bars, 60 s, 5 dF/F0. C, Cross-correlation analysis showing reduced CF–CF synchrony in DTA mice compared to controls at P13-15. Mean correlation coefficients between calcium traces from pairs of ROIs plotted against the distance of separation between pairs of ROIs in control (n = 6 mice) and AAV-hMAG-DTA-injected (n = 5 mice) mice. Error bars indicate SEM. * p < 0.05 (two-). D, Frequency distributions of CF calcium transient properties, including amplitude, event frequency, Area, and width in CTL and DTA mice. No significant group differences were observed (two-way ANOVA). E, Representative traces of miniature inhibitory postsynaptic currents (mIPSCs) recorded from PCs in CTL and DTA mice at P14. Scale bars, 100 ms, 50 pA. F, Quantification of mIPSC amplitude and frequency showing no significant differences between groups (9 mice for control, 11 mice for DTA, Mann–Whitney U test). Bars indicate median, dots represent the data for individual animals. G, Representative traces of parallel fiber–evoked EPSCs (PF-EPSCs) recorded from PCs in CTL and DTA mice at P14. Scale bars, 5 ms, 200 pA. H, Input–output curves of PF-EPSCs demonstrating no significant differences in excitatory drive between groups at P13-15 (two-way ANOVA). Data are shown as mean ± SEM (5 mice for each).

Oligodendrocyte ablation during postnatal development diminishes neuronal synchrony at adult stage.

A, Representative traces of spontaneous calcium transients (ΔF/F0) extracted from 25 regions of interest (ROIs) in the cerebellum of control and DTA mice at P62. Scale bar, 20 s. B, Correlation coefficient matrices of spontaneous calcium activity across multiple ROIs in control (top) and DTA (bottom) mice at P62. Color scale indicates the strength of pairwise correlations. C, Mean correlation coefficients between calcium traces from pairs of ROIs plotted against the distance of separation between pairs of ROIs in control (n = 6 mice) and AAV-hMAG-DTA-injected (n = 6 mice) mice at P62-80. Error bars indicate SEM. **** p < 0.0001 (two-way ANOVA). D, Scatter plot graph of CCs for ROI pairs, categorized by separation distances of 0– 80 mm and 120–200 mm (analyzed from four mice per group). Lines and plots indicate median and data from individual separation distances, respectively. Significant group differences were observed (**** p < 0.0001, Nested t-test, 4 mice for each). E-G, Frequency distributions of Frequency (E), Amplitude (F), and Area (G) of calcium transients in CTL (AAV-hMAG-mCherry) and DTA-treated (AAV-hMAG-DTA) mice at P62-80. Circles and squares indicate data from individual ROIs, and bars represent SEM. There was no significant difference between groups (two-way ANOVA).

Oligodendrocyte reduction during development leads to behavioral abnormalities.

A-C, Assessment of exploratory behavior and locomotor activity in control at P60 (CTL, orange, 12 mice for each injection site) and DTA-injected at P60 (DTA, cyan, 12 mice for each injection site) mice in a novel environment, with trajectories of mouse exploratory behavior (A), distance traveled (B), and time spent in center (C) for mice with the injection of AAVs into distinct compartments (left, center, right) of the cerebellum (LCH, CBV, RCH) at P60-80. Bars and dots indicate median and data from individual mice, respectively. D-F, Representative images and quantitative analysis of sociality in CTL and DTA mice with the injection of AAVs into distinct compartments (LCH, CBV, RCH). Diagram in (D) shows the method of sociality test; (E) displays the heatmap of mouse exploratory behavior, and (F) shows the percentage of time spent around strange mouse with data points representing individual mice at P60-80. Bars and dots indicate median and data from individual mice, respectively (CTL, 12 mice for each site, DTA, 12 mice for each site). G, Line graph showing the median of latency to fall in the rotarod test over a three-day period for CTL and DTA mice at P60-80, with separate graphs for mice with the injection of AAVs into distinct compartments (LCH, CBV, RCH). Error bars indicate 95 % CI. Statistical significance is indicated by asterisks (*p < 0.05, **p < 0.01, *** p < 0.001, **** p < 0.0001, Mann-Whitney U tests).

Oligodendrocyte reduction after 3 weeks old in mice does not affect PC activity synchrony and mouse behaviors except for motor learning.

A, B, Fluorescence microscopy images comparing ASPA expression in the cerebellum of control (CTL) and 3-week DTA-treated (3W_DTA) mice at 4 weeks (A) and 8 weeks (B) old. ASPA-positive oligodendrocytes appear in pink. Scale bar, 50µm. C, Quantification of the number of ASPA-positive cells per mm² in the cerebellum of CTL and 3W_DTA mice at 4 and 8 weeks old, with data points for each field of view overlaid on the bar graphs (mean, 4 mice for each group at 4 weeks, 5 mice for each group at 8 weeks). *** p < 0.001 (Nested t-test). D, Representative calcium imaging traces from PCs in the cerebellar cortex of control (CTL) and 3-week DTA injected (3W_DTA) mice at 4weeks old, showing 25 regions of interest (ROIs). E, Correlation coefficient matrices of calcium transients between ROIs from CTL and 3W_DTA mice at 4 weeks old, indicating the level of synchrony, with warmer colors representing higher correlation coefficients. F, Line graph depicting mean correlation coefficients with increasing inter-ROI distance for CTL and 3W_DTA mice at 4 weeks old. Error bars indicate SEM. No significant group differences were observed (two-way ANOVA, 5 mice for each group). G, Scatter plot comparing the correlation coefficients of calcium transients in PCs between adjacent ROIs in CTL and 3W_DTA mice. No significant group differences were observed (Nested t-test, 5 mice for each). H-J, Frequency distributions of Frequency (H), Amplitude (I), and Area (J) of calcium transients in CTL (AAV-hMAG-mCherry) and 3W_DTA (AAV-hMAG-DTA) mice at 4 weeks. Circles and squares indicate data from individual ROIs, and bars represent SEM. There was no significant difference between groups (two-way ANOVA). K-M, Assessment of exploratory behavior and locomotor activity in control (CTL, orange, 11 mice) and 3 weeks DTA-injected (3W_DTA, cyan, 10 mice) mice in a novel environment, with trajectories of mouse exploratory behavior (K), distance traveled (L), and time spent in center (M) at P60-80. Bars and dots indicate median and data from individual mice, respectively. No significant group differences were observed (Mann-Whitney U tests). N, O, Representative images and quantitative analysis of sociality in CTL and 3W_DTA mice. A stranger mouse is present in the right cage and a doll is in the left cage. N displays the heatmap of mouse exploratory behavior, and (O) shows the percentage of time spent around strange mouse at P60-80 (11 mice for control, 10 mice for 3W_DTA). Bars and dots indicate median and data from individual mice, respectively. No significant group differences were observed (Mann-Whitney U tests). P, Line graph showing the median of latency to fall in the rotarod test over a three-day period for CTL and 3W_DTA mice at P60-80. Error bars indicate 95 % CI. Statistical significance is indicated by asterisks (**p < 0.01, Mann-Whitney U tests).

Kir2.1 expression in CFs reduces the synchrony of PC population spontaneous activity and leads to abnormal behaviors similar to the phenotype of mice with developmental oligodendrocyte reduction.

A, Experimental scheme. AAV-Htr5B-tTA & AAV-TRE-GFP or AAV-Htr5b-tTA & AAV-Tre-kir2.1_T2A_GFP were injected into the cerebellum at P7, and AAV-L7-jGCaMP7f was injected into the inferior olive at P10 to label CFs. CF activity was imaged in the cerebellar cortex at P13-15. B, Fluorescent images of sagittal cerebellar sections showing Car8 immunoreactivity (Purkinje cells, magenta) and GFP expression (Climbing fibers (CFs), green). Scale bar, 200 µm. C, Higher magnification of the boxed area in A showing detailed GFP-labeled CF morphology. Scale bar, 100 µm. D, Scatter plot graph depicting the percentage of PCs with GFP-positive CFs among PC populations in CTL and Kir2.1 mice (17 sections from 10 mice for CTL and). E, Representative time-course of spontaneous calcium transients from regions of interest (ROIs) in the mouse cerebellum at P13 showing spontaneous PC activity in control mice (CTL, top) and mice expressing Kir2.1 in a subset of CFs (kir2.1, bottom). Scale bar, 20 s. Y-axis is ΔF/F0 = (F - F0)/(F0). F, Correlation matrices for spontaneous PC activity between ROIs in control (top) and kir2.1-expressing (bottom) mice at P13. The color scale indicates correlation strength. G, Graph displaying the correlation coefficient of Ca2+ transients among PCs as a function of distance between adjacent PCs in CTL and Kir2.1 groups at P13-15 (4 mice for CTL, 8 mice for Kir2.1). Data points indicate the mean ± SEM. H, Scatter plot comparing correlation coefficients for ROI pairs within 0-80μm and 120-200μm separation distance ranges in CTL (orange) and kir2.1-expressing (cyan) mice at P13-15, with lines representing average values. Statistical significance is indicated by asterisks (*p < 0.05, Nested t-test, 4 mice for CTL, 8 mice for Kir2.1). I-K, Frequency distributions of Frequency (I), Amplitude (J), and Area (K) of calcium transients in CTL and Kir2.1 mice at P13-15. Circles and squares indicate data from individual ROIs, and bars represent SEM. There was no significant difference between groups (two-way ANOVA). L-N, Analysis of exploratory behavior and locomotor activity in CTL (orange) and Kir2.1- expressing (green) mice at P60-80 with distance traveled (L, M) and percentage of time spent in the center (N) in an open-field test (14 mice for CTL, 18 mice for Kir2.1). Bars and dots indicate median and data from individual mice, respectively. O, P, Representative images and quantitative analysis of sociality in CTL and Kir2.1-expressing mice at P62-80. (O) displays the heatmap of mouse exploratory behavior, and (P) shows the percentage of time spent around strange mouse. Bars and dots indicate median and data from individual mice, respectively. Q, Rotarod test. The line graph shows the median latency to fall across three days for CTL and Kir2.1-expressing mice at P64-80. Error bars indicate 95 % CI. Statistical significance is indicated by asterisks (**p < 0.01, *** p < 0.001, Mann-Whitney U tests).

Optogenetic re-synchronization of PC activity rescues social and motor deficits but not anxiety-like behavior in adult mice with developmental oligodendrocyte ablation.

A, Representative fluorescence images of cerebellar sagittal sections showing rsChrmine expression (magenta) in PCs and nuclear counterstain (DAPI, blue). Middle and right are the magnified images of dot region in left. Scale bars, 500 µm (left), 200 µm (middle, right). B, Bar graphs showing the distribution of PCs expressing rsChrmine in individual cerebellar lobules of control (CTL, orange, 3 mice) and DTA (cyan, 3 mice) mice. The number of rsChrmine-positive PCs in each lobule was evaluated using a semi-quantitative 4-point scale (0 = none, 1 = few, 2 = moderate, 3 = abundant). Data are presented as mean values across lobules, with individual dots indicating scores from each animal. C, Representative trajectories of exploratory behavior in the open-field test from DTA and control (CTL) mice with or without LED stimulation at P60. D, Quantification of total distance traveled in the open-field at P60-65. Each dot represents an individual mouse (7 mice for each). E, Quantification of percentage time spent in the center of the open-field. LED stimulation did not alter anxiety-like behavior in DTA mice, whereas in CTL mice it reduced center time, indicating increased anxiety-like responses. F, Representative heat maps of social exploration during the sociality test in DTA and CTL mice at P64 with or without LED stimulation. G, Quantification of sociability, measured as the percentage of time spent around a novel mouse at P64-68 (8 mice for each). Optogenetic stimulation rescued sociability in DTA mice to control levels. H, Rotarod test performance across three days at P66-70 (9 mice for CTL-LED, 8 mice for DTA-noLED, 8 mice for DTA-LED). Optogenetic stimulation restored motor coordination in DTA mice to control levels. Data represent median ± 95% CI. Statistical significance is indicated by asterisks (*p < 0.05, **p < 0.01, ***p < 0.001, ++p < 0.01, +++p < 0.001; Mann–Whitney U test). + shows the significant difference between DTA_noLED and DTA_LED.

Targeted myelin disruption by oligodendrocyte-specific AAVs.

A, Representative images of MBP immunostaining in sagittal sections of control mice at P5 and P7, shown at low magnification (left) and higher magnification of the white matter (right). Scale bars, 500 µm (low magnification), 50 µm (high magnification). B, Bar graphs showing the mean fluorescence intensity of MBP at P5 and P7, derived from three mice in each group. C, Representative immunofluorescence images showing the specificity of MAG-mCherry expression at P14. mCherry (magenta) co-localizes with ASPA-positive oligodendrocytes (cyan) and Olig2-positive oligodendrocyte lineage cells (green). Notably, mCherry was absent in ASPA-negative, Olig2-positive immature oligodendrocytes. Scale bar, 50 µm. D, Quantification of the proportion of mCherry+ cells that were ASPA- and Olug2+ cells in total mCherry+ cells (5 mice). Dots represent individual mice. E, Quantification of the thickness of MBP bundle in CTL and DTA mice at P10, P14, P21, and adulthood (P60–P94). The thickness was significantly reduced at P10 and tended to be reduced at P14 (p = 0.13) but recovered to control levels by P21 and adulthood (n = 4 mice per group). Bars indicate mean, dots represent individual fields of view. ****p < 0.0001 (nested t-test). F, Low-magnification images of parvalbumin-stained sagittal sections showing overall cerebellar morphology and foliation in CTL and DTA mice at P14 and P21. No gross abnormalities were observed. Scale bar: 1 mm. G, Quantification of total cerebellar area measured from sagittal sections at P14 (n = 8 mice for control, n = 6 mice for DTA) and P21 (n = 10 mice for control, n = 9 mice for DTA). No significant differences were found between CTL and DTA mice (Mann–Whitney U test). Bars indicate median, dots represent the data for individual animals.

Oligodendrocyte ablation does not alter other glial or neuronal populations in the developing cerebellum.

A, B, Representative images at P14 and quantification of Iba1 immunostaining at P13-15 showing microglial density and mean fluorescence intensity in control (CTL) and DTA mice (8 mice for each). Scale bar, 50 µm. Bars indicate median, dots represent the data for individual animals. No significant differences were found between CTL and DTA mice (Mann–Whitney U test). C, D, Representative images at P21 and quantification of Iba1 immunostaining at P21-28. Bars indicate median, dots represent the data for individual animals. No significant differences were observed between CTL and DTA groups (13 mice for CTL, 15 mice for DTA, Mann– Whitney U test). E, F, Representative images at P14 and quantification of S100β immunostaining at P13-15 showing Bergmann glial density and mean fluorescence intensity (9 mice for CTL, 8 mice for DTA). Scale bar, 50 µm. Bars indicate median, dots represent the data for individual animals. No significant differences were found between CTL and DTA mice (Mann–Whitney U test). G, H, Representative images at P21 and quantification of S100β immunostaining at P21-28 (13 mice for CTL, 15 mice for DTA). Scale bar, 50 µm. Bars indicate median, dots represent the data for individual animals. No significant differences were found between CTL and DTA mice (Mann–Whitney U test). I–L, Parvalbumin (PV) immunostaining at P14. Quantification includes PC number, PC dendritic arbor size, and molecular layer PV+ interneuron density (13 mice for CTL, 15 mice for DTA). Scale bar, 50 µm. Bars indicate median, dots represent the data for individual animals. No significant differences were found between CTL and DTA mice (Mann– Whitney U test). M–P, PV immunostaining at P21. Quantification includes PC number, PC dendritic arbor size, or PV+ interneuron density (13 mice for each). Scale bar, 50 µm. Bars indicate median, dots represent the data for individual animals. No significant differences were found between CTL and DTA mice (Mann–Whitney U test).

Timeline of AAV injection and behavioral assessment for mice with the injection into distinct cerebellar regions.

A, Schematic timeline illustrating the experiments from birth to behavioral assessment, with AAV injections performed at P6–P7 and subsequent behavioral assessments conducted at around P60. B, Representative whole-cerebellum images of mCherry expression highlighting three different regions: left cerebellar hemisphere (LCH), cerebellar vermis (CBV), and right cerebellar hemisphere (RCH). Scale bars, 2 mm.

Kir2.1 expression in CFs reduces the CF activity.

A, Schematic timeline of the experimental protocol. At P7, AAV injections are administered to either overexpress kir2.1 and GFP for Kir2.1 or only GFP for control (CTL). Ca2+ imaging is conducted at P13-15. B, Representative image of jGCaMP7s-labeled CFs in the Kir2.1 experimental group, captured using two-photon laser scanning microscopy. Scale bar, 50 µm. C, Representative traces of calcium transients in CTL and kir2.1-overexpressing CFs. Each tick mark represents a single transient. Scale bar, 20 s. D, Bar graph showing the median frequency of calcium transients in CTL and Kir2.1-overexpressing CFs, with data points of individual cells. ***p < 0.001 (Mann-Whitney U tests).