Mechanisms of Rhythmicity in the Spinal cord

National Institute Of Neurological Disorders And Stroke

Investigators

Linked publications & trials

Abstract

We have been using optogenetics and calcium imaging to understand the contributions of motoneuron activity to fictive locomotion in the neonatal mouse. For this purpose, we have constructed mouse lines in which the excitatory channelrhodopsin or the inhibitory archaerhodopsin is expressed in motoneurons or in the spinal neurons that receive motoneuron inputs. We found that motoneuron hyperpolarization slowed the locomotor rhythm whereas exciting motoneurons accelerated the rhythm. These results indicate that motoneurons modulate the locomotor central pattern generator. Consistent with this interpretation, in motoneurons expressing the excitatory channelrhodopsin-2, we found that blue light could trigger or reset the locomotor rhythm. The light-induced change in frequency was not abrogated by cholinergic antagonists or gap junction blockers, indicating that neither cholinergic interneurons nor gap junctions mediated the effects. We did find, however, that the frequency changes were completely blocked by the AMPA-receptor antagonist NBQX. These findings are novel and surprising and reveal a completely new role for motoneurons in the operation of the mammalian locomotor central pattern generator and will have major implications not only for basic science but also for neurodegenerative motoneuron disease and spinal cord injury. We are now investigating how stimulation of motor axons in the ventral root can activate the locomotor central pattern generator. To accomplish this, we are imaging spinal cords in which the calcium-sensitive indicator GCaMP6 is expressed in either glutamatergic, GABAergic or cholinergic neurons. Using GCaMP6-VGluT2 animals, in which GCamp6 is expressed in glutamatergic neurons, we find that ventral root stimulation activates dorsal regions of the cord. Often this activation is patchy when viewed from the lateral or dorsal surface. Following this initial pattern of activation, a region just below the lateral curvature of the dorsal horn becomes active, initially in a restricted location, which then propagates in a spatially delineated band rostro-caudally along the cord. If locomotor like activity was triggered by the ventral stimulus this region was always activated. Lesion experiments revealed that the most dorsal region of the cord was required to activate locomotor-like activity suggesting that the ventral root stimulus activates dorsal circuits first and these circuits activate a propagating rostro-caudal network. We are currently investigating the compositions of these networks and how they are activated by stimulation of motor axons. The experiments also reveal that the classical distinction between the dorsal sensory and the ventral motor parts of the spinal cord are an oversimplification and that the functions of the two regions are much more integrated than previously acknowledged. Renshaw cells are one of two intraspinal interneuronal targets of motoneurons. They belong to the V1 population that expresses the transcription factor engrailed-1. To establish if these interneurons were involved in the regulation of locomotor frequency by motoneuronal activity, we introduced Archaerhodopsin or Channelrhodopsin-2 into the V1 population. We found that hyperpolarizing the V1 population slowed the locomotor rhythm. This result mimics the effect of reduced motoneuron discharge thereby implicating Renshaw cells and their synaptic partners in the regulation of locomotor frequency. When V1 interneurons expressing channelrhodopsin2 were activated a more complex picture emerge. In these cords, the function of the V1 class varied according to the mode of activation of the locomotor rhythm and whether the evoked rhythm was flexor- or extensor-biased. This latter finding indicates that the function of a particular neuronal class varies according to the behavioral and functional contexts in which it is studied.