Neuronal morphologies built for reliable physiology in a rhythmic motor circuit

The neurons of the crustacean stomatogastric ganglion (STG) exhibit highly-conserved firing patterns, voltage waveforms, and circuit functions despite quantifiable animal-to-animal variability in their neuronal morphologies. In recent work, we showed that one neuron type, the Gastric Mill (GM) neuron, is electrotonically compact and operates much like a single compartment, despite having thousands of branch points and a total cable length on the order of 10 mm. Here, we explore how STG neurite morphology shapes voltage signal propagation and summation in four STG neuron types. We use focal glutamate photo-uncaging in tandem with somatic intracellular recordings to examine passive electrotonic structure and voltage signal summation in the GM neuron and three additional STG neuron types: Lateral Pyloric (LP), Ventricular Dilator (VD), and Pyloric Dilator (PD) neurons. In each neuron, we measured the amplitudes and apparent reversal potentials (Erevs) of inhibitory responses evoked with focal glutamate photo-uncaging at more than 20 sites varying in their distance (100–800 μm) from the somatic recording site in the presence of TTX. Apparent Erevs were relatively invariant (mean CVs = 0.04, 0.06, 0.05, and 0.08 for 5–6 GM, LP, PD, VD neurons, respectively), suggesting that all four neuron types are similarly electrotonically uniform and compact. We then characterized the directional sensitivity and arithmetic of voltage summation (with fast sequential activation of 4–6 sites) in individual STG neurites. All four neuron types showed no directional bias in voltage signal summation and linear voltage summation. We motivate these experiments with a proof-of-concept computational model that suggests the immense tapering of STG neurite diameters: from 10–20 μm to sub-micron diameters at the terminal tips, may explain the uniform electrotonic structures experimentally observed and contribute to the robust nature of this central pattern-generating circuit.