The Impact of Treadmill Training on Tissue Integrity, Axon Growth, and Astrocyte Modulation

Abstract

Spinal cord injury (SCI) presents a complex challenge in neurorehabilitation, demanding innovative therapeutic strategies to facilitate functional recovery. This study investigates the effects of treadmill training on SCI recovery, emphasizing motor function enhancement, neural tissue preservation, and axonal growth. Our research, conducted on a rat model, demonstrates that controlled treadmill exercises significantly improve motor functions post-SCI, as evidenced by improved scores on the Basso, Beattie, and Bresnahan (BBB) locomotor rating scale and enhanced electromyography readings. Notably, the training facilitates the preservation of spinal cord tissue, effectively reducing secondary damage and promoting the maintenance of neural fibers in the injured area. A key finding is the significant stimulation of axonal growth around the injury epicenter in trained rats, marked by increased growth-associated protein 43 (GAP43) expression. Despite these advancements, the study notes a limited impact of treadmill training on motoneuron adaptation and highlights minimal changes in the astrocyte and neuron-glial antigen 2 (NG2) response. This suggests that, while treadmill training is instrumental in functional improvements post-SCI, its influence on certain neural cell types and glial populations is constrained.

Keywords: astrocytes; axon growth; neuron; spinal cord injury; treadmill training.

Figure 1

Treadmill training restores motor function, maintains tissue integrity, and supports axonal growth at 28 dpi (3 weeks of motor training). (a) Locomotor recovery measured by BBB rating scores. Maximum score—21. The electrophysiology results of M-wave amplitudes (b), H-wave amplitudes (b’) and somatosensory evoked potentials (SEPs) (c). (d) Three-dimensional reconstructions of the lesions. Scale 1 mm. (e) Representative images of quantitative assessment of neural tissue integrity in longitudinal sections (5 mm rostral to 2.5 mm caudal) of Untrained and Trained groups. Digital RGB color images (raw) of the spinal cord after hematoxylin–eosin staining. Binary inverted images after conversion to 8-bit with the application of Huang’s automatic threshold method and overlay masks. Identified cavities are colored, larger than 200 µm². Scale 1 mm. Area quantification of spared tissue (f) and pathological cavities (f’). (g) Expression of growth-associated protein 43 (GAP43) and chondroitin sulfate proteoglycan 4 (CSPG4) in the spinal cord, longitudinal section, 6–8 mm caudally. Confocal microscopy. Scale 20 µm. (h) Quantitative graph of the GAP43 and CSPG4 areas. For BBB test (a): Mann–Whitney test with Bonferroni correction; for EMG (b,b’): Kruskal–Wallis test with Dunn’s multiple comparison; for tissue integrity analyze (f,f’) and assessment of axonal growth (h): Mann–Whitney test. For all (a,b,b’,f,f’,h) values expressed as medians (minimum, maximum, and first and third quartiles (25th and 75th percentiles)): Mann–Whitney test. For all (a,b,b’,f,f’,h) values expressed as medians (minimum, maximum, and first and third quartiles (25th and 75th percentiles)), x—mean, +—outliers, $n=20$ rat/Untrained and Trained, $n=10$ rat/Intact group for (a,b,b’) and $n=10$ rat/group for f, f’,h; * $p<0.05$; ** $p<0.01$; *** $p<0.001$; **** $p<0.0001$.

Figure 2

Limited impact of treadmill training on motoneurons. (a) Confocal microscopy of osteopontin (OPN) (red) and choline acetyltransferase (ChAt)⁺ (yellow) neurons in the ventral horns (VHs) of the spinal cord, 6–8 mm caudal to the epicenter of injury at 28 days post-injury (dpi). Nuclei are stained with DAPI (blue). Scale 100 µm. OPN⁺- (b), ChAT⁺- (c) and OPN⁺/ChAT⁺-cells (d) quantification at 28 dpi. (e) Confocal microscopy of bTubIII⁺ (red) and PARV⁺ (yellow) neurons in the ventral horns of the spinal cord, 6-8 mm caudal to the epicenter of injury. Nuclei are stained with DAPI (blue). Scale 100 µm. (f) PARV⁺/bTubIII⁺-cells quantification. qRT-PCR was performed to determine the expression of chat (g) and syp (h) at 28 dpi. The data are expressed as $2^{-\Delta \Delta \mathrm{Ct}}$. For all (b–d,f–h): Mann–Whitney test. Values are expressed as medians (minimum, maximum, and first and third quartiles (25th and 75th percentiles)), x—mean, $n=5$ rat/group, no significant differences between groups.

Figure 3

Treadmill training promoted the development of anti-inflammatory reactive astrocytes. (a) Combined staining for aldehyde dehydrogenase 1 family member L1 (ALDH1L1) (green) and glial fibrillary acidic protein (GFAP) (red). ALDH1L1 marks both the cell bodies and extensive processes of astrocytes in the CST in spinal cord, 6–8 mm caudal to the epicenter of injury. GFAP labels the intermediate filament cytoskeleton of astrocytes. Confocal microscopy. Scale 100 µm. Quantitative graphs of different populations of astrocytes: GFAP⁺-cells (includes GFAP⁺/ALDH1L1⁻ and GFAP⁺/ALDH1L1⁺-astrocytes) (b), only GFAP⁺/ALDH1L1⁺-cells (c), and stable population of ALDH1L1⁺-(or GFAP⁻/ALDH1L1⁺)-astrocytes (d). qRT-PCR was performed to determine the expression of gfap (e) at 28 days post-injury (dpi). The data are expressed as $2^{-\Delta \Delta \mathrm{Ct}}$. (f) Confocal microscopy S100A10 cells in the ventral horns (VHs) of the spinal cord, 6–8 mm caudal to the epicenter of injury. Nuclei are stained with DAPI (blue). Scale 100 µm (a,f). For all (b–e): Mann–Whitney test. Values are expressed as medians (minimum, maximum, and first and third quartiles (25th and 75th percentiles)), x—mean, $n=5$ rat/group, * $p<0.05$; ** $p<0.01$.

Figure 4

Stability of NG2 glia and limited astrocytic response to treadmill training post-SCI. (a) Confocal microscopy of GFAP⁺ (red) and CSPG4⁺ (green) neurons in the ventral horns of the spinal cord, 6–8 mm caudal to the epicenter of injury at 28 dpi. Nuclei are stained with DAPI (blue). Scale 100 µm. (b) Quantitative graph of the CSPG4 mean fluorescent intensity. (c) Fluorescence colocalization analysis of CSPG4⁺ and GFAP⁺ cells using the Pearson’s coefficient. For all (b,c): Mann–Whitney test. Values are expressed as medians (minimum, maximum, and first and third quartiles (25th and 75th percentiles)), x—mean, • and ▴—outliers, $n=5$ rat/group, no significant differences between groups.