Projections of the wing stretch receptors to central flight neurons in the locust.

To investigate the cellular mechanisms by which proprioceptors modify the activity of a central neuronal oscillator, we identified flight motoneurons and interneurons in the thoracic ganglia of the locust Locusta migratoria, which receive constant short-latency PSPs from the wing stretch receptors (SRs). We also examined the changes in membrane potential oscillations of individual neurons when the SRs were stimulated on every cycle so as to mimic their discharge in intact animals. The SRs evoked constant short-latency (4 msec or less) EPSPs of 2 mV or less in depressor motoneurons. In general, the hindwing SRs project to both forewing and hindwing depressor motoneurons, whereas the forewing SRs project only to forewing depressor motoneurons. The SRs did not evoke IPSPs in elevator motoneurons. Twenty-one thoracic interneurons were identified that receive short-latency (4 msec or less) EPSPs from the forewing and/or hindwing SRs. Amplitudes varied from 1 to 6 mV, depending on the particular neuron. Only one neuron was found that responded with an IPSP to each SR afferent spike, but the longer latency (5-8 msec) and the different polarity suggest at least a disynaptic pathway. Fourteen interneurons received input from more than one SR. Of these, 12 neurons received EPSPs from the forewing and hindwing SRs ipsilateral to each neuron's soma, while 3 received bilateral forewing SR input. Convergence of more than one SR with the remaining 7 interneurons cannot be ruled out. Most EPSPs were evoked in interneurons depolarized in the depressor phase of the central oscillator cycle, which corresponds to the natural phase of SR activity in intact animals. However, elevator, nonrhythmic, and tonically active cells also received short-latency EPSPs from the SRs. Stimulating the SRs with trains of stimuli on every cycle produced pronounced changes in the membrane potential oscillations of several neurons, including phase-shifts, increased burst frequency, prolonged bursts, and abolition of activity. These observations suggest that the concepts of proprioceptors as error detectors and of the central oscillator as the primary generator of the movement are too simplistic to be applied to the locust flight system.