Homeostatic compensation and neuromodulation maintain synchronized motor neuron activity in the crustacean cardiac ganglion
Metadata[+] Show full item record
Animals rely on the nervous system to produce appropriate behavior throughout their lives. In sending commands to the musculature for rhythmic motor behaviors such as breathing or walking, neural networks must be stable enough to send a reliable level of drive with the proper temporal coordination. Networks must also be flexible enough to meet changing environmental demands. A network's output ultimately arises from the intrinsic excitability of its constituent neurons and the synaptic connections between them. Interestingly, neurons and networks are able to produce highly conserved output from highly variable underlying intrinsic and synaptic properties. To explore the consequences of this variability, we have used the crustacean cardiac ganglion (CG) which consists of 9 neurons: 4 pacemaker cells that give excitatory input to 5 Large Cell motor neurons (LCs) which are responsible for driving the simultaneous contraction of the musculature that makes up the walls of the animal's single-chambered heart (Alexandrowicz, 1934; Hartline, 1967; Anderson and Cooke, 1971). The intact network can be dissected from the animal in physiological saline and it continues to produce robust, reliable, and rhythmic output (Welsh and Maynard, 1951; Cooke, 2002). LCs have virtually identical synchronized activity, but their intrinsic ionic conductances can be highly variable (Ransdell et al., 2013a). In Chapter 1, we exploit this variability by pharmacologically blocking a subset of their conductances to make LCs hyperexcitable and desynchronize their activity. We find that homeostatic compensation restores synchronized activity and excitability within one hour. This happens via two synergistic mechanisms: the membrane properties of each cell are re-tuned to converge on similar voltage activity, and increased conductance of the gap junctions between the cells helps to buffer away differences in their voltage activity. A separate but related study asked whether naturalistic perturbations of network activity would also result in desynchronization. Neuromodulation provides flexibility in the output of neural networks by altering a subset of their conductances. We hypothesized that this could also cause desynchronization. We found that modulation with serotonin and dopamine both increased the excitability of the CG. Interestingly, serotonin desynchronized the CG, but dopamine did not. We found that dopaminergic modulation directly increases gap junctional conductance. By co-applying these modulators, we found dopamine was able to prevent serotonin from desynchronizing the network without occluding its effects. It was also able to prevent the desynchronization caused by ion channel blockers. Finally, to fully understand the output of LCs, we must recognize that their activity arises not only from their intrinsic properties, but also from their synaptic drive from pacemaker cells. To address how variable this can be from one animal to the next, we analyze the activity of 131 animals taken over the course of approximately 5 years. We use this to address the fundamental question of how variable networks underlying a particular behavior can be across animals. We recognize two distinct classes of pacemaker inputs to LCs, and characterize bursting patterns for both types of pacemaker spike and LC output. We conclude that LCs from different animals receive different temporal patterns of pacemaker drive, which may have important functional implications. We also compare animals from winter and summer months, and find that temperature-independent seasonal effects may explain some of the variance in our data.