Functional electrical stimulation helps replenish progenitor cells in the injured spinal cord of adult rats

Abstract

Functional electrical stimulation (FES) can restore control and offset atrophy to muscles after neurological injury. However, FES has not been considered as a method for enhancing CNS regeneration. This paper demonstrates that FES dramatically enhanced progenitor cell birth in the spinal cord of rats with a chronic spinal cord injury (SCI). A complete SCI at thoracic level 8/9 was performed on 12 rats. Three weeks later, a FES device to stimulate hindlimb movement was implanted into these rats. Twelve identically-injured rats received inactive FES implants. An additional control group of uninjured rats were also examined. Ten days after FES implantation, dividing cells were marked with bromodeoxyuridine (BrdU). The "cell birth" subgroup (half the animals in each group) was sacrificed immediately after completion of BrdU administration, and the "cell survival" subgroup was sacrificed 7 days later. In the injured "cell birth" subgroup, FES induced an 82-86% increase in cell birth in the lumbar spinal cord. In the injured "cell survival" subgroup, the increased lumbar newborn cell counts persisted. FES doubled the proportion of the newly-born cells which expressed nestin and other markers suggestive of tripotential progenitors. In uninjured rats, FES had no effect on cell birth/survival. This report suggests that controlled electrical activation of the CNS may enhance spontaneous regeneration after neurological injuries.

Keywords: cell birth; exercise; neural activity; regeneration; rehabilitation; spinal cord injury; stem cell.

Figure 1. Experimental design

Schematic diagrams of (A) the sites of injury and FES device implantation, and (B) the 43-day experimental design. In uninjured control groups (Supplement Figure 2), no SCI was induced, but other event timings were identical.

Figure 2. FES promoted new cell birth in the damaged spinal cord

BrdU-labeled cells were counted in six animals per group at six levels in coronal spinal cord sections using stereological methods (optical fractionator). Two-way ANOVA demonstrated effects of treatment group and spinal level (p<0.05). (A) FES induced a selective increase in cell birth that was confined to the lower lumbar spinal cord, segments predicted to experience increased activity from the patterned peroneal nerve stimulation. (B) This effect persisted in the cell survival group 7 days later (*p < 0.05; **p < 0.001, FES vs. control).

Figure 3. FES did not affect the proportion of BrdU+ cells that were also NG2+

NG2+/BrdU+ cell distribution (arrows; A, T1; B, L5) and morphology (C-E) in the cell birth groups was consistent with glial progenitor cells. Panels C-E show NG2 staining (C; green), BrdU staining (D, red), and their co-localization (E). There were more NG2+/BrdU+ cells above the injury level in both groups (two-way ANOVA; 47 ± 3% at T1 vs. 33 ± 3% at L5 in control animals, *p < 0.05; 51 ± 3% at T1 vs. 39 ± 3% at L5 in FES-treated animals, **p < 0.001). Scale bar (A,B) = 50 μm, (C-E) = 10 μm.

Figure 4. FES did not affect the proportion of BrdU+ cells that were also GFAP+

GFAP+/BrdU+ cell distribution (arrows; A, T1; B, L5) and morphology (C-E) in the cell birth groups was consistent with astrocytes. Panels C-E show GFAP staining (C; green), BrdU staining (D, red), and their co-localization (E). There was no difference in the number of GFAP+/BrdU+ cells above and below the injury (44 ± 4% at T1 vs. 40 ± 3% at L5 in control animals; 46 ± 3% at T1 vs. 39 ± 3% at L5 in FES-treated animals). Scale bar (A,B) = 50 μm, (C-E) = 10 μm.

Figure 5. FES selectively doubled the proportion of BrdU+ cells that were also nestin+

This increase (**p < 0.001) was only observed below the injury level, in spinal segments predicted to experience increased activity from the patterned stimulation. Nestin+/BrdU+ cell distribution (arrows; A, T1; B, L5) and morphology (C-E). Panels C-E show nestin staining (C; green), BrdU staining (D, red), and their co-localization (E). Some of the nestin+ cells showed morphologic characteristics of reactive astrocytes (A, asterisk) but these were rare at levels distant from the injury. As with NG2 (Figure 3), but in contrast to GFAP (Figure 4), there were more nestin+/BrdU+ cells above the injury level in both groups (two-way ANOVA, 24 ± 2% at T1 vs. 10 ± 3% at L5 in control animals, *p < 0.001; 27 ± 2% at T1 vs. 19 ± 2% at L5 in FES-treated animals, **p < 0.001). Scale bar (A,B) = 50 μm, (C-E) = 10 μm.

Figure 6. BrdU+/nestin+ cells exhibited the distribution, morphology, and staining specificity consistent with adult-derived bi- or tri-potential progenitor cells

(Horner et al., 2000; Shihabuddin et al., 2000). Confocal microscope images of ventrolateral white matter of an FES-treated rat in the cell birth group. Nestin immunoreactivity shown in blue, BrdU labeling shown in red. A, D, and G show overlays of nestin and BrdU; B, E, and H show overlays of BrdU and other markers; C, F, and I show triple overlays. (A-C) Nestin immunoreactivity (blue) was found predominantly in white matter (L5 shown). Less than 2% of nestin+/BrdU+ cells co-labeled with GFAP (green, arrowhead). (D-F) A few BrdU+/nestin+ cells also expressed NG2 (green, arrowhead). These triple-labeled cells were distinguished from other BrdU+/nestin+ cells by their pial location and their morphologies (L5 shown). (G-I) Macrophages, labeled with ED1 (green), were located within and surrounding the injury site (T11 shown). We did not identify a substantial number of NG2+/ED-1+, nestin+/ED1+ cells. (J-L) BrdU+ oligodendrocytes. Representative examples of the distribution pattern and cell morphology of BrdU+/APC-CC1+ cells (arrowheads) located at levels above (J, C2) and below (K, L5) the injury. These features are most consistent with oligodendrocytes, as is the fact that the number of BrdU+/APC-CC1+ cells increased substantially in the cell survival groups (see text). (L) Single confocal section of a BrdU+/APC-CC1+ cell. Scale bar (A-K) = 50 μm, (L) = 10 μm.