Calcium-dependent Dendritic Signaling Properties in Lamprey Spinal Neurons During Locomotion

Calcium-dependent Dendritic Signaling Properties in Lamprey Spinal Neurons During Locomotion Michael Alpert, MPhil Graduate program in Neuroscience University of Illinois at Chicago Chicago, Illinois (2013) Dissertation Chairperson: Janet Richmond, Ph.D. Locomotion is generated following supraspinal activation of local, reciprocally connected neurons within the spinal cord. Rhythmic output of the spinal cord, responsible for producing swimming in fish, but also walking, running, and flying in other animals, is the result of the synaptically connected segmental microcircuits and the intrinsic membrane properties of their constituent neurons. Pharmacological activation of the isolated spinal cord using the specific glutamate receptor subtype, NMDA, reproduces the rhythmic output seen during swimming in behaving animals. NMDA receptor (NMDARs) activation leads to membrane potential oscillations in individual neurons, which are repolarized following activation of a Ca2+-dependent K+ channel (KCa2). However, the source of Ca2+ responsible for KCa2 channel activation is unknown. Similarly, because previous experiments have typically been conducted in isolated cords and used glutamate agonists to initiate the spinal network, it is unknown whether NMDARs are activated during brainstem-evoked locomotion, when descending command neurons and local spinal interneurons can release glutamate synaptically, in a physiological manner. Furthermore, the pattern of dendritic Ca2+ activity from synaptically released glutamate, and whether this Ca2+ is sufficient to activate KCa2 to promote rhythmicity and synchronization amongst locally connected neurons is unknown. Therefore several questions emerged and were answered using electrophysiology combined with Ca2+ imaging to determine the dendritic Ca2+ signaling properties of spinal neurons within the context of locomotion: 1) Are the components necessary for oscillation generation located at the glutamatergic synapse? a. This was approached by demonstrating that synaptically evoked NMDAR-dependent Ca2+ entry is localized to dendrites and sufficient to activate KCa2 in isolated spinal cords. b. Furthermore, contributions of voltage-gated Ca2+ channels in KCa2 activation were excluded on account of voltage thresholds reached during membrane potential oscillations and the use of selective VGCC antagonists. 2) Are KCa2 channels necessary for coordinated, brainstem-evoked locomotion? a. Using a whole brain-spinal cord preparation with separate recording chambers for the brain and spinal cord, the selective K¬Ca2 channel antagonist, apamin, could be selectively added to the spinal cord, leading to a highly irregular swimming pattern. 3) What is the pattern of dendritic Ca2+ activity from brainstem-evoked locomotion and how does it compare to other glutamate receptor agonists? a. Using brainstem-evoked locomotion, spinal neurons filled with Ca2+ sensitive dyes displayed dendritic Ca2+ fluctuations during locomotor episodes. Following transection of the spinal cord, other glutamate receptor agonists could be applied to the isolated spinal cord to generate pharmacologically induced fictive locomotion while imaging from the same dye-filled neurons. This demonstrated that brainstem-evoked signaling was distinct from agonist-induced signaling on account of the spatiotemporal characteristics of Ca2+ signals, their amplitudes, and spinal output recorded from the ventral root. 4) Are NMDARs or other glutamate receptors responsible for Ca2+ signals during brainstem-evoked locomotion? Are these receptors necessary for brainstem-evoked locomotion? a. During brainstem-evoked locomotion, selective antagonists to both NMDARs and metabotropic glutamate receptors (mGluR5) were applied locally to Ca2+ dye-filled neurons. While network activity persisted, inhibition of either NMDARs or mGluR5s completely abolished locomotor-associated Ca2+ signals. b. Similarly, using a dual pool recording bath, the blockade of either NMDARs or mGluR5s completely abolished the locomotor rhythm. 5) What are the synaptic mechanisms underlying mGluR5-dependent dendritic Ca2+ signaling seen during locomotion? a. Several different glutamatergic and glycinergic synapses were stimulated in isolated spinal cords using the selective mGluR5 antagonist, MPEP. While both a small potentiation of glycinergic IPSCs and inhibition of glutamatergic EPSCs was observed from local circuit interneurons, neither correspondingly impacted the evoked Ca2+ signal recorded from voltage-clamped spinal neuron dendrites. Thus, an explanatory mechanism for the inhibition of Ca2+ signaling and of the network during brainstem-evoked swimming is still being examined. Thus, I have linked NMDAR-dependent Ca2+ entry in dendrites to KCa2 activation, placing the components necessary for oscillation generation at the glutamatergic synapse. I have also demonstrated the importance of different glutamate receptors subtypes in generating dendritic Ca2+ activity, and their subsequent importance for locomotor output.