Locomotor deficits induced by lumbar muscle inflammation involve spinal microglia and are independent of KCC2 expression in a mouse model of complete spinal transection

Abstract

Spinal cord injury (SCI) is associated with damage to musculoskeletal tissues of the spine. Recent findings show that pain and inflammatory processes caused by musculoskeletal injury mediate plastic changes in the spinal cord. These changes could impede the adaptive plastic changes responsible for functional recovery. The underlying mechanism remains unclear, but may involve the microglia-BDNF-KCC2 pathway, which is implicated in sensitization of dorsal horn neurons in neuropathic pain and in the regulation of spinal excitability by step-training. In the present study, we examined the effects of step-training and lumbar muscle inflammation induced by complete Freund's adjuvant (CFA) on treadmill locomotion in a mouse model of complete spinal transection. The impact on locomotor recovery of each of these interventions alone or in combination were examined in addition to changes in microglia and KCC2 expression in the dorsal and ventral horns of the sublesional spinal cord. Results show that angular motion at the hip, knee and ankle joint during locomotion were decreased by CFA injection and improved by step-training. Moreover, CFA injection enhanced the expression of the microglial marker Iba1 in both ventral and dorsal horns, with or without step-training. However, this change was not associated with a modulation of KCC2 expression, suggesting that locomotor deficits induced by inflammation are independent of KCC2 expression in the sublesional spinal cord. These results indicate that musculoskeletal injury hinders locomotor recovery after SCI and that microglia is involved in this effect.

Keywords: Chloride homeostasis; Disinhibition; Neuroplasticity; Nociception; Pain; Spinal cord injury model; Training.

Figure 1:

Locomotor kinematics overview. (A) Stick figure displays individual examples of angular excursions of hind limb joints during a 750-ms recording of treadmill locomotion for each group on day 28 post-transection and in an intact mouse. (B) Relative joint angle plots of ankle excursion in function of hip and knee excursions during locomotion show decreased hind limb coordination and an abnormal locomotor kinematic pattern on day 28 compared to the intact state. These plots illustrate the decreased angular excursions in untrained CFA mice compared to untrained and trained mice and compared to trained CFA mice.

Figure 2.

Hind limb joint kinematics variations over time and across groups Hip (A), knee (B) and ankle (C) angular excursion measured between maximal flexion and maximal extension angles and MTP extension angle (D) were evaluated at each step and averaged over locomotor bout for each animal. Group median (black squares), interquartile (coloured bars) and extreme values (coloured lines) are shown on each time points (intact, day 2, day 7, day 14, day 21 and day 28). Open circles show outliers. Dashed lines over median values on each time point were added to improve visualization of joint excursion recovery for each group. After initial drops in hind limb joints excursion following spinal transection, partial recovery of hip, knee and ankle excursion and MTP extension during locomotion were observed in all groups. Regardless of training, mice that received CFA showed decreased hip, knee and ankle excursions (red brackets) compared to mice that did not received CFA. Moreover, regardless of CFA, mice that were trained showed increased knee and ankle excursion (blue brackets) compared untrained mice. * p ≤ 0.05.

Figure 3.

Variations in step length after spinal transection and during recovery period. Step length averaged over locomotor bout was evaluated for each mouse before transection and on days 2, 7, 14, 21 and 28 after transection. Group median (black squares), interquartile (coloured bars) and extreme values (coloured lines) are shown for each time points. Outliers are shown as open circles. Dashed lines joining median values was added to improve visualization of recovery for each group. Step length decreased after transection, but partly recovered during the recovery period in all groups.

Figure 4.

CFA induces an increase in microglial inflammation in the lumbar spinal cord 28 days after a complete spinal cord transection in mice. Iba1+ cells were labeled for microglia visualization in the L1-L2 spinal cord (A). Area fraction of stained tissue in dorsal (middle) and ventral horns (right) was evaluated (B). The area fraction median (black squares), 25–75% centiles (boxes) and extreme values (black lines) are shown. Regardless of step-training, Iba1 area fraction was enhanced in CFA-injected mice (both untrained CFA and trained CFA) compared to controls (both untrained and trained mice) in dorsal (p = 0.03) and ventral horns (p = 0.03).

Figure 5.

CFA and step-training fail to modulate KCC2 expression in the L2-L4 dorsal horns. (A) Individual examples of KCC2 expression in the dorsal horn. (B) The signal intensity median (black squares), 25–75% centiles (boxes) and extreme values (bars) are shown. It did not differ between groups (p’s > 0.05).

Figure 6.

CFA and step-training fail to modulate KCC2 expression in lumbar motoneurons. (A) Individual examples of KCC2 expression in the ventral horn. ChAT+ cells (red) were labeled for motoneurons visualization. (B) The KCC2 signal intensity in the membrane was averaged from 3 different ROIs and normalized to the cytosol signal intensity. (C) The signal intensity median (black squares), 25–75% centiles (boxes) and extreme values (bars) are shown. It did not differ between groups (p’s> 0.05).