A Comparison of the Linear Tuning Properties of Two Classes of Axons in the Bullfrog Lagena

Abstract

Various vertebrate inner-ear end organs appear to have switched their sensory function between equilibrium sensing and acoustic sensing over the courses of various lines of evolution. It is possible that all that is required to make this transition is to provide an end organ with access to the appropriate stimulus mode and frequency range. If, as we believe, however, the adaptive advantage of an acoustic sensory system lies in its ability to sort the total acoustic input into components that correspond to individual acoustic sources, and the adaptive advantage of an equilibrium sensory system lies in its ability to compute the total orientation and motion of the head without regard to the individual sources contributing to that orientation and motion, then it is easy to argue that the differences between acoustic and equilibrium sensors should be more profound than simply access to the appropriate stimuli. Effective signal-sorting requires high resolution in both time and frequency; to achieve this resolution, a peripheral tuning structure must be one of high dynamic order (i.e., constructed from multiple independent energy storage elements). If the peripheral tuning structure simply converts head acceleration to head displacement, velocity, or jerk (i.e., provides one or two steps of integration or differentiation with respect to time, where one energy storage element per step is required), then high dynamic order is inappropriate. Because the bullfrog lagena possesses both acoustic and equilibrium sensitive regions, it is especially suited for comparing these two sensor types and addressing the question of dynamic order of tuning. In this paper we report observations of the linear tuning properties of bullfrog lagenar primary afferent nerve fibers obtained by stimulating the lagena with random, dorsoventral micromotion over the frequency range from 10 Hz to 1.0 kHz. Tuning curves obtained by reverse correlation analysis and discrete Fourier transformation were used to estimate the dynamic order of each fiber’s associated peripheral tuning structure. We found two classes of lagenar afferent axons – those with lowpass amplitude tuning characteristics (44 units) and those with bandpass amplitude tuning characteristics (73 units). Lowpass units were found to originate at the equilibrium region of the macula, and they exhibited low dynamic order – summed low- and high-frequency slopes (absolute values) ranged from 10 dB/decade to 64 dB/decade, implying dynamic orders of less than one to three (the modal value was equal to one). Bandpass units were found to originate at the acoustic region of the macula, and they exhibited higher dynamic order than lowpass units – summed low- and high-frequency slopes (absolute values) ranged from 53 dB/decade to 185 dB/decade, implying dynamic orders of three to nine (the modal value was equal to five). It appears that while lagenar equilibrium and acoustic sensors both possess access to signals in the acoustic frequency range, lagenar acoustic sensors are tuned by means of peripheral structures with markedly greater dynamic order and consequently markedly greater physical complexity. These results suggest that steep-sloped (high-dynamic-order) tuning properties reflect special adaptations in acoustic sensors not found in equilibrium sensors, and that any evolutionary transition between the two sensor types must have involved profound structural changes.