Generation of Flexible Rhythmic Behaviors Resulting from Interactions Between Intrinsic Sensory Feedback and Extrinsic Neuromodulators

Abstract

While the neuronal networks that generate rhythmic motor patterns are hard‐wired, they must nonetheless generate flexible outputs to enable animals to respond appropriately to both long‐ and short‐term changes in their internal and external environments. Much of the required flexibility is thought to result from the effects of modulatory neurotransmitters, largely neuropeptides, acting on the neurons and synapses of the central pattern generators (CPGs) that underlie rhythmic motor patterns. Additionally, sensory inputs can modulate the same pattern generating networks. Using the cardiac neuromuscular system of the American lobster, Homarus americanus, we are examining the roles played by sensory inputs and peptide neuromodulators in altering pattern generator output, and asking whether and how these two systems interact with one another. The neurogenic heartbeat of crustaceans is controlled by a 9‐neuron CPG, the cardiac ganglion, which includes five motor neurons and four premotor, or pacemaker, neurons; these neurons are both electrically and chemically coupled, so that they fire nearly synchronous driver potentials and bursts of action potentials. Dendritic processes of both motor and premotor neurons appear to serve as stretch receptors that provide direct feedback to the CPG. To characterize the responses of the cardiac neurons to stretch, we isolated the cardiac ganglion together with the muscle fibers surrounding the four premotor neurons of the CPG. We then stretched these muscle fibers, thereby stretching the mechanosensitive dendrites of these neurons, while recording intracellularly from one of the motor neurons. During sustained stretch, driver potential frequency increased and duration decreased; these changes were a function of both the extent of the stretch and the baseline driver potential duration. During the rising portion of the stretch, we recorded phase delays that were a function of the rate of the imposed stretch, i.e., the strain rate. During the return from stretch, burst duration increased in neurons that had relatively long baseline driver potential durations; this increase was inversely related to baseline driver potential duration, but did not appear to be correlated with either amount or duration of the imposed stretch. Interestingly, we were unable to mimic this suite of effects of stretch by injecting current into the motor neurons. To examine the interactions between stretch and extrinsic neuropeptide modulators, we applied similar stretches to the cardiac muscles in control saline and in saline containing 10−8 or 10−9 M SGRNFLRFamide (SGRN). Preliminary data suggest that, while 10−9M SGRN had little or no effect on the response to sustained stretch,10−8M SGRN decreased this response. It is thus possible this peptide alters the responsiveness of the stretch feedback in this neuromuscular system.Support or Funding InformationNSF IOS‐1121973 and IOS‐1354567; NIH INBRE grant P20GM0103423 from NIGMS