Repetitive unidirectional spinal tactile stimulation engages microglial Bmal1 pathways to promote synaptic remodeling in the mPFC of adolescent VPA-exposed mice

Abstract

Background: Synaptic abnormalities are hallmark pathological features of autism spectrum disorders (ASD), contributing to the behavioral impairments frequently observed in these neurodevelopmental conditions. Microglia, as the brain's primary immune cells, are essential for synaptic refinement during adolescent development. Disrupted microglia-dependent synapse remodeling has been implicated in pathophysiology of ASDs, however, the underlying mechanisms remain incompletely elucidated. In this context, repetitive unidirectional spinal tactile stimulation (RSTS) has emerged as a promising non-invasive therapeutic strategy. This study aims to explore whether and how RSTS enhances microglia-dependent synapse remodeling in the medial prefrontal cortex (mPFC) during adolescent development in ASD mice, with a specific focus on the role of Brain and Muscle ARNT-Like 1 (Arntl1), a core circadian protein crucial for regulating this process.

Methods: ASD mice underwent RSTS treatment during adolescent brain for 21 days, administered twice daily for 10 min per session. Behavioral changes were evaluated using the three-chamber social interaction and open field tests. Synapse number and morphology were assessed through Golgi staining. Microglia-dependent synapse remodeling ability was analyzed using immunofluorescence and Western blot. Furthermore, the molecular mechanism was investigated using single-nucleus RNA sequencing (snRNA-seq) and chromatin immunoprecipitation sequencing (ChIP-seq). Finally, the role of Bmal1 was validated, confirming its involvement in the enhancement of RSTS during adolescent brain in ASD.

Results: RSTS was found to alleviate autistic-like behaviors in adolescent ASD mice. Results from snRNA-seq and ChIP-seq indicated that the therapeutic effects of RSTS may be mediated through microglial Bmal1 and its role in the transcriptional regulation of microglia-dependent synapse remodeling. Furthermore, in vivo experiments confirmed that RSTS enhances microglia-dependent synapse remodeling in mPFC of adolescent ASD mice via Bmal1. These findings suggested that Bmal1 serves as a critical target of RSTS in facilitating microglia-dependent synapse remodeling during the adolescent brain developmental period in ASD mice.

Conclusion: Our findings suggest that the therapeutic effects of RSTS are potentially mediated through the modulation of microglial Bmal1-dependent synapse remodeling and the regulation of synaptic proteins and the complement system. These results provide novel empirical evidence for RSTS in restoring synaptic balance and offer valuable insights into its potential as an intervention for ASD.

Keywords: Autism spectrum disorder༛Microglia-dependent synapse remodeling༛Bmal1༛Adolescent brain development; Repetitive unidirectional spinal tactile stimulation.

Fig. 1

Diagram of repetitive unidirectional spinal tactile stimulation

Fig. 2

Reciprocating spinal-touched stimulation improve VPA-induced social behavior deficits and microglial morphology. A Schematic description of the experimental timeline for the prenatal VPA-induced ASD mouse model, and treatment by RSTS with behavioral tests and biological assessments. B Representative trajectories and quantification of (C) total travelled distance and (D) time in the center area in open field test (n = 8 in each group). E Representative heatmap and quantification of (F) sniffing time in stage 1, (G) sniffing time in stage 2 and (H) time in each chamber in 3-chambered social test (n = 8 in each group). I. J Western blot images and quantitative analysis of IBA1 in the mPFC of ASD mouse after RSTS treatments (n = 5 in each group). K Representative images of Iba1 + cells (raw, yellow) that were used for binary and skeletonization and analyzed by using the “skeletonize” plugin in ImageJ to measure microglial morphology. Quantification for (L) branches, (M) end points, (N) triple points, and (O) maximum branch length. Scale bar = 50 (raw)/25 (Binary and Skeleton) µm. n = 15 cells from 3 mice per group. The data are shown as the mean ± SEM. and were statistically analyzed by two-way ANOVA and two-tailed unpaired t-tests. * p < 0.05; ** p < 0.01; *** p < 0.001; ^#p < 0.05 vs. the time in stranger 1 chamber within the same group; ^##p < 0.01 vs. the time in stranger 1 chamber within the same group; ^###p < 0.001 vs. the time in stranger 1 chamber within the same group. ns: non-significance. M: mouse; O: object; E:empty

Fig. 3

RSTS corrects VPA-induced deficits in dendritic spine morphology and synaptic related proteins during the early adolescent period. A Representative dendritic spines from the four groups and (B) diagram of mature and immature spines, and quantification of (C) number of spines, (D) mature spines rate and (E) immature spines rate. Scale bar = 10 (raw)/5 (zoom) µm. n = 30 slides from 3 mice per group. F-H Western blot images and quantitative analysis of PSD95 and Syn1 in the mPFC of ASD mouse after RSTS treatments (n = 5 in each group). I Representative images of Syn (green) and PSD95 (magenta) from the four groups and quantitative analysis of (J) Syn1 and (K) PSD95 expression in mPFC of adolescent ASD mouse after RSTS treatments. Scale bar = 200 μm. n = 9 slides from 3 mice per group. The data are shown as the mean ± SEM. and were statistically analyzed by two-way ANOVA and two-tailed unpaired t-tests. * p < 0.05; ** p < 0.01; *** p < 0.001. ns: non-significance

Fig. 4

RSTS upregulates the expression of microglial CD68⁺ and complement proteins in mPFC of adolescent ASD mice. A Representative microscopic images of co-immunostaining of IBA1 (green) and CD68⁺ (red, a phagocytosis marker in microglia) in mPFC of adolescent ASD mice, (B) Quantitation of microglial CD68⁺ fluorescence intensity and (C) IBA1 average intensity. Scale bar = 20 (raw)/10 (zoom) µm. n = 15 cells from 3 mice per group. (D) Representative microscopic images of co-immunostaining of C1q (green) and C3b (magenta) in mPFC of adolescent ASD mice, (E) quantitation of microglial C1q fluorescence intensity and (F) C3b average intensity. Scale bar = 20 μm. G Representative microscopic images of CR3 (red) in mPFC of adolescent ASD mice, (H) quantitation of CR3 fluorescence intensity. Scale bar = 20 μm. I-L Representative immunoblots and quantitation of the C3, CR3 and C1q protein (n = 5 in each group). β-actin was used as a loading control. The data are shown as the mean ± SEM. and were statistically analyzed by two-way ANOVA and two-tailed unpaired t-tests. * p < 0.05; ** p < 0.01; *** p < 0.001. ns: non-significance

Fig. 5

Microglial depletion via PLX5622 reverses the behavioral rescue effect of RSTS in VPA-induced ASD mice. A Representative immunofluorescence images of IBA1 staining in the mPFC showing robust microglial depletion following PLX5622 treatment. (B) Quantification of IBA1⁺ microglial density confirms depletion efficiency in each group (Scale bar = 100 μm, n = 4 per group). C Representative locomotor trajectories during the OFT across Saline, VPA, VPA-RSTS, VPA-PLX5622, and VPA-RSTS-PLX5622. Quantification of (D) total distance traveled and (E) time spent in the center area during OFT (n = 8 in each group). F Representative heatmaps and (G) quantification of sniffing time in stage 1 (S1, sociability) of the three-chamber test (n = 8 in each group). H Heatmaps and (I) sniffing time in stage 2 (S2, social novelty) across the five experimental groups (n = 8 in each group). The data are shown as the mean ± SEM. and were statistically analyzed by one-way ANOVA followed by LSD post hoc tests. * p < 0.05; ** p < 0.01; *** p < 0.001. ns: non-significance. M: mouse; O: object; E:empty

Fig. 6

Single-nucleus sequencing analysis of the RSTS in adolescent ASD mice. A Schematic diagram of the single-nucleus sequencing workflow; (B) Diagram of dimensionality reduction clustering among samples; (C) Different colors represent 9 clusters (cell types) respectively, including endothelial cells, microglia, astrocytes, oligodendrocytes, fibroblasts, mural cells, OPC, excitatory neuron, and inhibitory neuron; (D) Cell type composition histogram of the ratio; (E) Representative cell type marker genes (y-axis) with the percent of cells that express a gene (size of dot) in each cluster (distributed along the x-axis) and the average expression level (color intensity) are shown for endothelial cells (Cldn5, Vwf), microglia (P2ry12, Cx3cr1), astrocytes (Aldoc, Atp1b2 and Aqp4), oligodendrocytes (Cnp, Plp1, Mog, and Mag), fibroblasts(Dcn, Apod, and Col1a1), mural cells(Rgs5, Abcc9, and Pdgfrb), OPC(Olig1, Pdgfra, C1ql1, and Olig2), excitatory neuron (Slc17a7, Nefm), and inhibitory neuron (Gad1, Gad2) for each cluster; (F) Diagram of dimensionality reduction clustering for each group; (G) Heatmap of Saline-VPA and VPA-RSTS different express genes; (H)GO enrichment analysis of Saline-VPA and VPA-RSTS different express genes

Fig. 7

Single-nucleus sequencing analysis of the RSTS in adolescent ASD mice. A Cell communication analysis; (B) Arntl1 expression in each cell types; (C-F) expression map of synapse-related genes in various cell types. (G) GO enrichment analysis of VPA and VPA-RSTS different express genes in inhibitory cells. H-I Representative microscopic images and quantification of co-immunostaining of Neun (neuronal marker, red) and Bmal1 (green) in mPFC of adolescent ASD mice. Scale bar = 50 (raw)/25 (zoom) µm. n = 6 mice per group. J-K Representative microscopic images and quantification of co-immunostaining of Bmal1(red) and IBA1 (microglial marker, green) in mPFC of adolescent ASD mice. Scale bar = 10 μm. n = 15 cells from 3mice per group. L-N Representative immunoblots and quantitation of the Clock and Bmal1 protein. n = 6 in each group. β-actin was used as a loading control. The data are shown as the mean ± SEM. and were statistically analyzed by one-way ANOVA followed by LSD post hoc tests. * p < 0.05; ** p < 0.01; *** p < 0.001. ns: non-significance

Fig. 8

RSTS alters the genome-wide binding patterns of Bmal1 in mPFC of the adolescent ASD mice. A canonical E-box motif (CACGTG) serves as the binding site for the bHLH-PAS transcription factor Bmal1. B The immunoprecipitation (IP) experiment successfully precipitated the Bmal1 protein from the lysate system using an anti-Bmal1 antibody. C Agarose gel electrophoresis validated the integrity of immunoprecipitated DNA (most are 300–400 bp) across Saline, VPA and RSTS group. D The number of aligned reads and the number of identified peaks in Saline, VPA and RSTS group. E The genomic distribution (%) of the Bmal1 binding sites in Saline, VPA and RSTS group. F Plot heatmap of transcription start sites (TSSs) derived from the ChIP-seq analysis of three independent ASD mPFC samples immunoprecipitated with Bmal1 antibody after 21 days RSTS treatment. The top portion of each column shows the average peak profile for each sample, centered around the TSS. The bottom heatmap represents the traces of the top 1000 peaks and displays the read coverage density (greener means more reads at that location) around the TSS. G A partial list of the representative motifs enriched in the ChIP-seq peaks. H-L Representative genome browser view of Bmal1 ChIP-seq on distal genomic regions annotated to Nr1d1, Nr1d2, Cry1, Cry2, and Per1 using Integrative Genomics Viewer (IGV). A zoomed-in view is presented for each highlighted zone (zone1 and zone2) in Per1 (red dashed box) on the genome browser track. P value was calculated with exact hypergeometric probability

Fig. 9

RSTS alters the genome-wide binding patterns of Bmal1 in mPFC of the adolescent ASD mice. AVenn diagram for the target genes enriched by Bmal1 protein. B Number of differentially enriched Bmal1 peaks, include 1 st intron, 1 st exon, intergenic, other exon, other intron and promoter-tss regions between Saline, VPA and RSTS groups. C A partial list of representative motifs enriched in the ChIP-seq peaks across the Saline vs. VPA and VPA vs. RSTS comparisons. D-G Representative genome browser view of Bmal1 ChIP-seq on distal genomic regions annotated to Itgam, Grb10, Rai1 and TSC1 using Integrative Genomics Viewer (IGV). H-I GO functional enrichment and KEGG pathway analysis of Bmal1-targeted genes between Saline and VPA groups. Gene ratio refers to the percentage of total candidate genes in the given pathway. J-K GO functional enrichment and KEGG pathway analysis of Bmal1-targeted genes between VPA and RSTS groups. Gene ratio refers to the percentage of total candidate genes in the given pathway

Fig. 10

Bmal1 deficiency impairs complement-mediated synaptic remodeling in the mPFC of adolescent mice. A–D Western blot analysis and quantification of complement components C3, CR3, and C1q in the mPFC of Bmal1^+/+ and Bmal1^–/– mice. β-actin was used as a loading control (n = 5 in each group). E–F Immunofluorescence images and quantification of CR3 (magenta) co-localized with the microglial marker IBA1 (green) in the mPFC. Scale bar = 20 μm. n = 15 slides from 3 mice per group. G–I Representative immunofluorescence images and quantification showing C1q (green) and C3b (magenta) puncta in the prefrontal cortex. Scale bar = 100 μm. n = 9 slides from 3 mice per group. Merged images demonstrate reduced complement deposition in Bmal1-deficient mice. The data are presented as mean ± SEM. Statistical comparisons were performed using unpaired Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 11

RSTS improves social behaviors and synapse development in VPA-induced mice in Bmal1-dependent manner. A Representative movement traces of mice in each group. Quantification of (B) total distance traveled and time spent in the center area (n = 8 in each group). C Heatmaps showing exploration of the chamber containing an unfamiliar mouse versus an empty chamber. D Quantification of sniffing time (n = 8 in each group). E Heatmaps showing interaction with a familiar mouse versus a novel mouse. F Quantification of sniffing (n = 8 in each group). G Representative dendritic spines from the five groups and quantification of (H) number of spines, mature spines rate. Scale bar = 10 (raw)/5 (zoom) µm. n = 30 slides from 3 mice per group. I Representative images and quantification (J) of Syn (green) and PSD95 (magenta) in the mPFC. Scale bar = 100 μm. n = 9 slides from 3 mice per group. K Representative images and quantification (L) of C1q (green), C3b (magenta) in the mPFC. Scale bar = 50 μm. n = 9 slides from 3 mice per group. All data are presented as individual values and mean ± SEM. Statistical analysis was performed using one-way ANOVA or unpaired Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ns: non-significance. M1: stranger 1; M2: stranger 2; O: object; E:empty

Fig. 12

RSTS activates Bmal1-mediated Wnt/β-catenin signaling and promotes synaptic remodeling proteins in mPFC of adolescent VPA-exposed mice. A–D Western blot and quantification of synaptic remodeling-related markers C3, CR3, and C1q in the mPFC from Saline, VPA, VPA-RSTS, VPA-Bmal1-KO, and VPA-RSTS-Bmal1-KO groups (n = 5 in each group). E–J Protein expression levels of β-catenin, TCF4, c-Myc, and Cyclin D1 were examined to assess Wnt/β-catenin pathway activity under the same conditions. RSTS significantly upregulated key pathway components, which were abolished in Bmal1-deficient mice (n = 5 in each group). K–P To validate the regulatory role of Bmal1, cells were stably transfected with Bmal1 overexpression or control plasmids into BV2 cells. Western blot analysis confirmed Bmal1-induced upregulation of β-catenin, TCF4, Prox1, c-Myc, and Cyclin D1. Quantifications are shown in (L–P) (n = 3 in each group). Q–R Co-immunoprecipitation assays demonstrated a direct interaction between Bmal1 and β-catenin. Cells co-expressing Flag-Bmal1 and Myc-tagged β-catenin were subjected to reciprocal IP using anti-Flag or anti-Myc beads. All data are presented as individual values and mean ± SEM. Statistical analysis was performed using one-way ANOVA or unpaired Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001