Rapid and robust restoration of breathing long after spinal cord injury

Abstract

There exists an abundance of barriers that hinder functional recovery following spinal cord injury, especially at chronic stages. Here, we examine the rescue of breathing up to 1.5 years following cervical hemisection in the rat. In spite of complete hemidiaphragm paralysis, a single injection of chondroitinase ABC in the phrenic motor pool restored robust and persistent diaphragm function while improving neuromuscular junction anatomy. This treatment strategy was more effective when applied chronically than when assessed acutely after injury. The addition of intermittent hypoxia conditioning further strengthened the ventilatory response. However, in a sub-population of animals, this combination treatment caused excess serotonergic (5HT) axon sprouting leading to aberrant tonic activity in the diaphragm that could be mitigated via 5HT2 receptor blockade. Through unmasking of the continuing neuroplasticity that develops after injury, our treatment strategy ensured rapid and robust patterned respiratory recovery after a near lifetime of paralysis.

Fig. 1

PNN breakdown restores ipsilateral hemidiaphragm function 12 weeks after cervical SCI. a Experimental protocol. b, c Average amplitude of b ipsilateral and c contralateral diaEMG. Filled circles = responders; unfilled circles = animals in d–g; triangles = non-responders; stars = tonic activity shown. d–g Representative EMG recordings at baseline and following the 4 weeks of treatment for all groups. Data panels presented from the same animal. For b–g n = 6 per group, baseline n = 24. h–aa Immunohistochemistry and intensity readings at the ipsilateral C4 PMP utilizing Ctb (red) and either (green) h–l 5HT, m–q TrkB, r–v 2B6, or w–aa WFA (n = 4 per group). For all EMG or graph panels, treatment groups = baseline (grey); saline+air (yellow); saline+IH (red); ChABC+air (light blue); ChABC+IH (blue). *p < 0.05, **p < 0.01 and ***p < 0.001. If no post-hoc result is shown, comparison was not-significant. Scale bar = 50 μm. For all panels: values represent mean ± SEM

Fig. 2

Matrix remodelling recovers respiratory activity at multiple time points. a Experimental protocol. b–g DiaEMG recordings at baseline and following treatment. Animals received ChABC and air (Nx) or IH conditioning for b, c 1 week; d, e 3 weeks; or f, g 5 weeks. Data panels presented from the same animal (n = 6/7). h, i Average amplitude of h ipsilateral and i contralateral diaEMG. Filled circles = responders; unfilled circles = animals in b–g; triangles = non-responders; stars = tonic activity shown. For b–i, n = 6/7 per group, baseline n = 38. j–p Immunohistochemistry and intensity readings at the ipsilateral C4 PMP utilizing Ctb (red) and 5HT (green; n = 4/5 per group). q–s Activity is present in the ipsilateral hemidiaphragm following application of exogenous 5HT (n = 9). For all EMG or graph panels, treatment groups = baseline (grey); ChABC+air (light blue); ChABC+IH (blue). ***p < 0.001. If no post-hoc result is shown, comparison was not-significant. Scale bar = 50 μm. For all panels: values represent mean ± SEM

Fig. 3

Respiratory recovery is maintained up to 6 months after PNN digestion. a Experimental protocol. b–e DiaEMG recordings at baseline and after treatment. Animals received ChABC and air (Nx) or IH conditioning and were housed for b, c 3 months or d, e 6 months without treatment. Data panels presented from the same animal. f, g Average amplitude of f 3 month and g 6 month animals diaEMG. Filled circles = responders; unfilled circles = animals in b–e; triangles = non-responders; stars = tonic activity shown. For b–g n = 6/7 per group, baseline n = 13/14. h–n Immunohistochemistry and intensity readings at the ipsilateral C4 PMP utilizing NeuN (red) and either (green) h, i 5HT; j, k 2B6; or l, m WFA 6 months after treatment (n = 6/8 per group). For all EMG or graph panels, treatment groups = baseline (grey); ChABC+air (light blue); ChABC+IH (blue). **p < 0.01 and ***p < 0.001. If no post-hoc result is shown, comparison was not-significant. Scale bar = 50 μm. For all panels: values represent mean ± SEM

Fig. 4

Excessive 5HT and receptor imbalance causes ipsilateral tonic activity. a–f Tonic diaEMG activity following a 3 or b 5 weeks conditioning. c–f Representative CTAs of ipsilateral (coloured) and contralateral (grey) activity for mean inspiration (orange) and expiration (green) where ±SEM is shown for both inspiration (yellow) and expiration (light green). IH for c, d 3 or e, f 5 weeks (n = 6, see Fig. 5). g–l Intensity readings for g 5HT, h 5HT2c/a and l representative images at the ipsilateral C4 PMP utilizing i+k NeuN (red) and 5HT (green) or j+l 5HT2c/a (green). Graphs show mean ± SEM. Data grouped on ipsilateral hemidiaphragm activity: non-responders (circle; n = 1/2), normal activity (square; n = 1/2), tonic (triangle; n = 2). m, n Tonic activity following m 5HT (n = 7) and n 5HT2c/a antagonist (n = 6) application. CTAs of ipsilateral (coloured) and contralateral (grey) activity for mean inspiration (orange) and expiration (green) where ±SEM is shown for both inspiration (yellow) and expiration (light green) of pharmacologically induced and removed tonic activity. Panels recorded in the same animal. For all EMG or graph panels: animals were treated with ChABC+IH (blue)

Fig. 5

Representative diaEMG traces of ChABC+IH (blue) animals that displayed tonic firing of the ipsilateral hemidiaphragm (n = 6). a–d Tonic activity exhibited in animals with 3 weeks of conditioning. CTAs generated with triggers at the onset of inspiration and expiration. Panels (b, c) demonstrate that the onset of tonic ipsilateral hemidiaphragm firing can reduce the firing frequency of the contralateral hemidiaphragm. e, f Tonic activity exhibited in animals with 5 weeks of conditioning. CTAs reveal inhibition of tonic activity before the onset of inspiration and expiration. CTA on panel (e) shows the effect on cardiovascular activity on the trace. Panels recorded in the same animal. For all panels: animals were treated with ChABC+IH (blue EMG traces); CTAs shown of ipsilateral (coloured) and contralateral (grey) activity for mean inspiration (orange) and expiration (green) where ±SEM is shown for both inspiration (yellow) and expiration (light green)

Fig. 6

PNN digestion recovers respiratory activity a year and a half after spinal injury. a Experimental protocol. b–d DiaEMG at baseline (grey; n = 5) and following either b NaCl (yellow; n = 2) or c ChABC (light blue; n = 3). d Average amplitude of ipsilateral diaEMG. e–h DiaEMG at baseline (grey; n = 5) and following ChABC and e air (light blue; n = 2) or f IH (blue; n = 3). g, h Average amplitude of g ipsilateral and h contralateral diaEMG. i–l Immunohistochemistry and intensity readings at the ipsilateral C4 PMP utilizing NeuN and 5HT for i injury only, ChABC and j air or k IH. Scale bar = 50 μm (n = 2/3 per group). For all EMG or graph panels, treatment groups = baseline (grey); saline+air (yellow); ChABC+air (light blue); ChABC+IH (blue). For d, g, h: filled circles = responders; unfilled circles = animals in b, c, e, f; triangles = non-responders; stars = tonic activity shown. If no post-hoc result is shown, comparison was not-significant. For all panels: values represent mean ± SEM