Multiple mechanisms shape selectivity for FM sweep rate and direction in the pallid bat inferior colliculus and auditory cortex

Abstract

The inferior colliculus and auditory cortex of the pallid bat contain a large percentage of neurons that are highly selective for the direction and rate of the downward frequency modulated (FM) sweep of the bat's echolocation pulse. Approximately 25% of neurons tuned to the echolocation pulse respond exclusively to downward FM sweeps. This review focuses on the finding that this selectivity is generated by multiple mechanisms that may act alone or in concert. In the inferior colliculus, selectivity for sweep rate is shaped by at least three mechanisms: shortpass or bandpass tuning for signal duration, delayed high-frequency inhibition that prevents responses to slow sweep rates, and asymmetrical facilitation that occurs only when two tones are presented at appropriate delays. When acting alone, the three mechanisms can produce essentially identical rate selectivity. Direction selectivity can be produced by two mechanisms: an early low-frequency inhibition that prevents responses to upward sweeps, and the same asymmetrical two-tone inhibition that shapes rate tuning. All mechanisms except duration tuning are also present in the auditory cortex. Discussion centers on whether these mechanisms are redundant or complementary.

Fig. 1

The information used to predict the best FM sweep rate of a duration-tuned neuron in the inferior colliculus, by the formula of rate = kHz/ms = excitatory bandwidth/best duration. a The band-width of the excitatory tuning curve at the test intensity of 50 dB is 4 kHz. b The best duration of the neuron is 0.5 ms. The predicted best rate is therefore 4/0.5 = 8 kHz/ms. c The actual best rate of this neuron, tested with three downward FM sweeps of different bandwidth, is 12 kHz. d The actual and predicted best FM sweep rates for downward FM sweeps of 11 neurons with duration tuning. The solid diagonal line indicates a perfect prediction; the dashed line is a simple linear regression. The r² value is 0.797

Fig. 2

a A “two-tone inhibition over time” function of an inferior colliculus neuron showing the excitatory tuning curve (dark gray) and flanking inhibitory sidebands (light gray) as a function of frequency tuning and arrival time of inputs relative to excitation. The low-frequency inhibitory inputs arrive at the same time as excitation, while the high-frequency inhibitory input arrives 5 ms later. The arrows indicate two downward sweeps, one including the high-frequency sideband, the other eliminating it. b A two-tone inhibition analysis of the arrival times of low- and high-frequency inhibition. An excitatory tone is delayed and advanced with respect to a tone from each band. Note that the low-frequency tone suppresses the response much earlier (1 ms delay of the excitatory tone) than the high frequency. c The effect of eliminating the high-frequency inhibitory sideband from a downward sweep on rate selectivity. Rate selectivity is eliminated, since the neuron now responds at all sweep durations, and hence rates

Fig. 3

a The arrival time of low-frequency inhibition (LFI) predicts direction selectivity in auditory cortex neurons. Vertical dashed line indicates when excitatory and inhibitory inputs arrive at the same time. Negative values mean inhibition arrives earlier. Most neurons with early inhibition have higher direction selectivity. b The actual and predicted cutoff rates of nine inferior colliculus neurons, predicting the fastest rate at which their response would be reduced by 90% of maximum response. Rate tuning is assumed to be shaped by delayed high-frequency inhibition. Solid line indicates a perfect prediction. The dashed line shows a linear regression with an r² value of 0.996

Fig. 4

Facilitation alone may shape sweep direction and rate selectivity in this inferior colliculus neuron. a Tests used to eliminate sideband inhibition, which also shapes response selectivity. This neuron was selective for downward sweeps. The FM tuning curve (FMTC) ranged from 38 to 40 kHz. Possible inhibitory sidebands were eliminated by starting upward (38 → 60 kHz) and downward (40 → 15 kHz) sweeps within this tuning curve. b The neuron retained its direction and rate selectivity for downward sweeps when both broad- and narrowband sweeps were presented, suggesting that inhibitory sidebands did not contribute to response selectivity. c The neuron had a strongly asymmetrical TTF that may shape both direction and rate selectivity. HFR is high-frequency region; LFR is low-frequency region