Quantification of gastric mill network effects on a movement related parameter of pyloric network output in the lobster.

It has long been known that gastric mill network activity (cycle period 5-10 s) alters pyloric network output (cycle period approximately 1 s), but these effects have not been quantified. Many pyloric muscles extract gastric mill timed variations in pyloric motor neuron firing, and consequently produce gastric mill timed movements even though no gastric mill neurons innervate them. Determining pyloric behavior therefore requires detailed description of gastric mill effects on pyloric neural output. Pyloric muscle activity correlates well with motor neuron overall spike frequency (OSF, burst spike number divided by cycle period). We quantified OSF variation of all pyloric neurons as a function of time into the gastric mill cycle [as measured from the beginning of Gastric Mill (GM) neuron bursts] in the lobster, Panulirus interruptus. No repeating pattern within individual gastric mill cycles of Lateral Pyloric (LP) and Ventricular Dilator (VD) neuron OSF was visually apparent. Averaged data showed that VD and LP neuron OSF decreased (approximately 0.5 and 1.5 Hz, respectively) at the beginning of each gastric mill cycle. Visually apparent patterns of OSF waxing and waning within each gastric mill cycle were present for the Inferior Cardiac (IC), Pyloric Dilator (PD), and Pyloric (PY) neurons. However, when averaged as a function of phase or delay in the gastric mill cycle, the average changes were smaller than those in individual gastric mill cycles because when the OSF variations occurred varied considerably in different gastric mill cycles. We therefore used a "pattern-based" analysis in which an identifying characteristic of each neuron's repeating OSF variation pattern was defined as pattern pyloric cycle zero. The pyloric cycles in each repetition of the OSF variation pattern were numbered relative to the zero cycle, and averaged to create an average OSF variation profile. The zero cycle delays relative to GM neuron burst beginning were then averaged to determine when in the gastric mill cycle the profile occurred. This technique preserved the full extent of pyloric neuron OSF changes. Maximum PY neuron OSF occurred within the GM neuron burst, whereas maximum IC and PD neuron OSF occurred during the GM neuron interburst interval. Despite these changes, pyloric cycling did not phase lock with gastric mill activity, nor were an integer number of pyloric cycles present in each gastric mill cycle. In addition to providing data necessary to predict pyloric movement, this work shows how pattern-based analysis can successfully quantify interactions between nonphase-locked networks.