Functional specializations in Drosophila primary mechanosensory neurons: experimental methods, physiology, and models

Abstract

Primary mechanosensors encode the detailed features of mechanical stimuli, with different neurons extracting different stimulus features. How this occurs is not fully understood. This is due partially to the difficulty of recording activity from identifiable mechanosensors in situ while applying controlled forces.

To investigate these issues, we have turned to Johnston’s Organ (JO), a mechanosensory organ in the Drosophila antenna. JO contains an array of ≈ 500 primary mechanosensory neurons, called Johnston’s Organ Neurons (JONs). It encodes rotations of the distal antennal driven by wind, gravity, touch, and sound. Because all JONs connect to the distal antenna segment, we can stimulate them all by manipulating the rotational position of the antenna. Due to technical limitations, however, individual JONs have been electrophysiologically inaccessible, and previous studies have relied on population-level calcium imaging or field potential recordings. We have overcome these limitations by using laser-dissection and specialized recording techniques to target single JONs for patch-clamp electrophysiology.

We find that JONs show rapid, temporally precise, and diverse responses to mechanical stimulation. Individual neurons can show selectivity for both antennal position and velocity, and many are direction-selective. When presented with vibrational stimuli, JONs show time-dependent frequency preferences. Overall, we find individual JONs have complex, conjoint tuning to multiple mechanosensory features.

We noticed some of these response properties can be explained using a hair-cell inspired model of mechanoelectrical transduction. In this model, a gating spring is attached to directional motors, and elements which limit their travel. We show this model is sufficient to reproduce many JON tuning properties, and variations in its components can reproduce much of the diversity observed. To explain additional elements of JON diversity, we incorporate geometrical properties of mechanosensors in JO. Finally, we show that JON adaptation to vibrational stimuli can be modeled using stimulus-dependent changes in mechanosensor compliance.

In summary, our model shows how antennal motion could drive mechanotransduction differently across cells, explaining a diversity of responses. This type of "mechanical diversification" would allow different JONs to extract different features of a mechanical stimulus, allowing those features to be transmitted selectively to the appropriate downstream circuits.