Enriched conditioning expands the regenerative ability of sensory neurons after spinal cord injury via neuronal intrinsic redox signaling

Abstract

Overcoming the restricted axonal regenerative ability that limits functional repair following a central nervous system injury remains a challenge. Here we report a regenerative paradigm that we call enriched conditioning, which combines environmental enrichment (EE) followed by a conditioning sciatic nerve axotomy that precedes a spinal cord injury (SCI). Enriched conditioning significantly increases the regenerative ability of dorsal root ganglia (DRG) sensory neurons compared to EE or a conditioning injury alone, propelling axon growth well beyond the spinal injury site. Mechanistically, we established that enriched conditioning relies on the unique neuronal intrinsic signaling axis PKC-STAT3-NADPH oxidase 2 (NOX2), enhancing redox signaling as shown by redox proteomics in DRG. Finally, NOX2 conditional deletion or overexpression respectively blocked or phenocopied enriched conditioning-dependent axon regeneration after SCI leading to improved functional recovery. These studies provide a paradigm that drives the regenerative ability of sensory neurons offering a potential redox-dependent regenerative model for mechanistic and therapeutic discoveries.

Fig. 1. Enriched conditioning (EE + SNA) induces a significant increase in DRG regenerative growth in vitro and axon regeneration in vivo after SCI.

a Representative images of cultured DRG neurons from SH sham, EE Sham, SH SNA, and EE + SNA stained for Beta-III-tubulin. Scale bar, 100 µm. b Quantification of average neurite outgrowth per neuron (mean ± SEM, one-way ANOVA, Tukey’s post hoc, n.s. = nonsignificant, n = 4 biologically independent animals/group, average of 20 cells/replicate). c Timeline for the in vivo experiment. d Representative images of CTB-traced (magenta) dorsal column sensory axons after injury, glial fibrillary acidic protein (GFAP) (green) and DAPI (blue). Scale bar, 200 µm. e Quantification of CTB-positive regenerating axons (mean ± SEM, Two-way repeated-measures ANOVA, Tukey’s post hoc, n = 5 biologically independent animals/group. Fluorescence intensity was measured in one series of tissue for each spinal cord). f Average distance between lesion center and furthest regenerating axon/caudal tract retraction (mean ± SEM, n = 5 biologically independent animals/group. Length of the furthest regenerating five axons and main tract retraction was measured in one series of tissue for each spinal cord). g Quantification of GFAP intensity around the lesion site (mean ± SEM, one-way ANOVA, n.s. = nonsignificant, n = 5 biologically independent animals/group). Fluorescence intensity was measured in one series of tissue for each spinal cord.

Fig. 2. EE + SNA induces upregulation of signaling pathways and of NOX2 complex in DRG.

a Area proportional Venn diagram showing the number of differentially expressed (DE) genes and the extent of overlap among the three conditions SH SNA, EE Sham, and EE + SNA (each normalized to respective SH sham, P-value < 0.05, n = 3 independent biological experiments/condition). b Heatmap showing the Gene ontology analysis (molecular function, DAVID) of the DE genes modulated by SH SNA, EE Sham, and EE + SNA vs. SH Sham (modified Fisher’s exact P-value < 0.05). c Network visualization of protein–protein interaction of the NOX oxidases category with proteins belonging to signaling pathways specifically enriched upon EE + SNA. Each protein is organized in a circular layout accordingly to their shared molecular function. Nox2 subunits are organized in the center of the network. Each line (edge) represents a protein–protein interaction. d Heatmap showing mRNA expression analysis (fold change vs. SH Sham, P-value < 0.05) of the genes belonging to NOX2 complex. e Immunoblotting analysis of NOX2 complex components from DRG protein extracts after SH Sham, EE Sham, SH SNA, or EE + SNA. f Quantification of immunoblotting, glyceraldehyde phosphate dehydrogenase (GAPDH) was used as a loading control to which protein expression was normalized. (mean ± SEM, one-way ANOVA, Tukey’s post hoc, n = 3 biologically independent animals/group examined over three independent experiments).

Fig. 3. NOX2 is required for EE + SNA-dependent increase in neurite outgrowth and axon regeneration after SCI.

a Representative images of cultured DRG from NOX2^fl/fl mice previously transduced in vivo with AAV-GFP or AAV-Cre-GFP (green) and stained with Beta-III-tubulin (red) after SH Sham or EE + SNA. Scale bar, 100 µm. b Quantification of average neurite outgrowth per neuron (mean ± SEM, one-way ANOVA, Tukey’s post hoc, n.s. = nonsignificant, n = 4 biologically independent animals/group, average of 20 cells/replicate). c Timeline for the in vivo experiment. d Representative images of CTB-traced (red) dorsal column sensory axons after injury and DAPI (blue) to determine the lesion site (dashed line). Scale bar, 400 µm. e Quantification of CTB-positive regenerating axons (mean ± SEM, two-way repeated-measures ANOVA, Tukey’s post hoc, n.s. = nonsignificant, n = 5 biologically independent animals/group). Fluorescence intensity was measured in one series of tissue for each spinal cord.

Fig. 4. NOX2 is required for EE + SNA-dependent increase in ROS production and redox signaling in DRG.

a Representative images of DRGs from WT or NOX2^fl/fl mice after EE SNA or SH Sham, hydrocyanine–Cy3 (red) staining of ROS 1 day after SCI. Scale bar, 100 µm. b Quantification of hydrocyanine–Cy3 in DRG neurons (mean ± SEM, one-way ANOVA, Tukey’s post hoc, n = 3 biologically independent animals/group, examined over three independent experiments). Fluorescence intensity was measured in one series of tissue for each DRG. c Network visualization of the GO clustering analysis of the differentially oxidized proteins in EE + SNA (green) and SH Sham (red)-specific datasets run with ClueGO in Cytoscape (Bonferroni P-value < 0.05). Each node represents a GO-enriched term and the size of each node represents its enrichment significance. Edges represent interrelations between terms, defined by the Kscore. Red and green color intensity reflects proteins from a given dataset (red = Sham, green = EE + SNA, gray = common). Main functional groups are reported.

Fig. 5. pSTAT3 is induced and occupies the promoters of NOX2 subunits following EE + SNA.

a In silico motif analysis run with Pscan of the NOX2 gene member’s promoters, showing the putative transcription factors regulating their transcription ranked by p-value. STAT3 binding motif; letters represent the relative enrichment for each gene at that position. b Immunoblotting of pSTAT3 from DRG extracts after SH Sham, EE Sham, SH SNA, or EE + SNA. c Quantification of immunoblotting, pSTAT3 was normalized to levels of STAT3, GAPDH was used as a loading control (mean ± SEM, one-way ANOVA, Tukey’s post hoc, n = 3 biologically independent animals/group examined over three independent experiments). d Quantitative RT-PCR analysis of NOX2 complex genes after chromatin immunoprecipitation (ChIP) for pSTAT3, expressed as fold change of DNA enrichment compared to IgG (mean ± SEM, two-tailed unpaired Student’s T-test, n = 4 independent biological replicates/group). e Quantitative RT-PCR analysis of NOX2 complex genes after ChIP for H3K27ac, expressed as fold change of DNA enrichment compared to IgG (mean ± SEM, two-tailed unpaired Student’s T-test, n.s. = nonsignificant, n = 3 independent biological replicates/group, examined over three independent experiments).

Fig. 6. STAT3 deletion in DRG blocks EE + SNA-dependent axonal regeneration and NOX2 complex expression.

a Timeline for the in vivo experiment. b Representative images of WT (injected with control AAV-GFP) and STAT3 KO (injected with AAV-Cre-GFP) CTB-traced (red) dorsal column sensory axons after injury and DAPI (blue), to determine the lesion site (dashed line). Scale bar, 200 µm. c Quantification of fluorescence intensity of CTB-positive regenerating axons (mean ± SEM, two-way repeated-measures ANOVA, Tukey’s post hoc, ***P-value < 0.0001, n = 4 biologically independent animals/group). Fluorescence intensity was measured in one series of tissue for each spinal cord. d Representative images of cultured DRG neurons from SH Sham + vehicle (DMSO), SH Sham + PKC activator, EE + SNA + vehicle (DMSO), and EE + SNA + PKC inhibitor, stained with Beta-III-tubulin (red) and pSTAT3 (green). Scale bar, 100 µm. e Quantification of mean neurite outgrowth per neuron (mean ± SEM, one-way ANOVA, Tukey’s post hoc, n.s. = nonsignificant, n = 3 biologically independent animals/group, average of 20 cells/replicate). f Quantification of pSTAT3 fluorescence intensity (mean ± SEM, one-way ANOVA, Tukey’s post hoc, n = 3 biologically independent animals/group, average of 20 cells/replicate).

Fig. 7. Overexpression of constitutively active p47phox (p47-3X) in DRG neurons induces axon regeneration and synaptic plasticity after SCI.

a Timeline for the in vivo experiment. b Representative images of GFP-positive sensory axons (green) after spinal cord injury and DAPI (blue) to determine the lesion site (dashed line). Scale bar, 200 µm. c Quantification of fluorescence intensity of GFP-positive regenerating axons (mean ± SEM, two-way repeated-measures ANOVA, Tukey’s post hoc, n = 9 biologically independent animals/group examined over two independent experiments). Fluorescence intensity was measured in one series of tissue for each spinal cord. d Average distance between lesion center and furthest regenerating axon/caudal tract retraction (mean ± SEM, n = 9 biologically independent animals/group examined over two independent experiments. Length of the furthest regenerating five axons and main tract retraction was measured in one series of tissue for each spinal cord). e Multi-fluorescent orthogonal 3D confocal images show colocalization between vGAT-positive boutons (green) and ChAT-positive motor neurons (red) below the injury (L1-4). Scale bar 25 μm, the orthogonal planes XY, XZ, and YZ are labeled. f Quantification of vGAT-positive boutons apposed to motoneurons (mean ± SEM, two-tailed unpaired Student’s t-test, n = 9 biologically independent animals/group, average of 20 motoneurons/replicate examined over two independent experiments). g Multi-fluorescent orthogonal 3D confocal images show juxtaposition between vGlut1-positive boutons (blue), GFP-positive axons (green), and ChAT-positive motoneurons (red) below the injury (L1-4), arrowheads indicate triple colocalization. Scale bar 25 μm, the orthogonal planes XY, XZ, and YZ are labeled. h Quantification of vGlut1-positive boutons apposed to motoneurons (whole bar), vGlut1 and GFP colocalization apposed to motoneurons (cross-hatched area) (mean ± SEM, two-tailed unpaired Student’s t-test, n = 9 biologically independent animals/group, average of 20 motoneurons/replicate examined over 2 independent experiments). i Quantification of the time required to first contact an adhesive pad placed on the hind paws (mean ± SEM, two-way repeated-measures ANOVA, Tukey’s post hoc, n = 9 biologically independent animals/group examined over two independent experiments). j Quantification of the time required to remove an adhesive pad placed on the hind paws (mean ± SEM, two-way repeated-measures ANOVA, Tukey’s post hoc, n = 9 biologically independent animals/group examined over two independent experiments).