The Performance of Olfactory Receptor Neurons: The Rate of Concentration Change Indicates Functional Specializations in the Cockroach Peripheral Olfactory System

Abstract

Slow and continuous changes in odor concentration were used as a possible easy method for measuring the effect of the instantaneous concentration and the rate of concentration change on the activity of the olfactory receptor neurons (ORNs) of basiconic sensilla on the cockroach antennae. During oscillating concentration changes, impulse frequency increased with rising instantaneous concentration and this increase was stronger the faster concentration rose through the higher concentration values. The effect of the concentration rate on the ORNs responses to the instantaneous concentration was invariant to the duration of the oscillation period: shallow concentration waves provided by long periods elicited the same response to the instantaneous concentration as steep concentration waves at brief periods. Thus, the double dependence remained unchanged when the range of concentration rates varied. This distinguishes the ORNs of basiconic sensilla from those of trichoid sensilla (Tichy and Hellwig, 2018) which adjust their gain of response according to the duration of the oscillating period. The precision of the ORNs to discriminate increments of slowly rising odor concentration was studied by applying gradual ramp-like concentration changes at different rates. While the ORNs of the trichoid sensilla perform better the slower the concentration rate, those of the basiconic sensilla show no preference for a specific rate of concentration increase. This suggests that the two types of sensilla have different functions. The ORNs of the trichoid sensilla may predominately analyze temporal features of the odor signal and the ORNs of the basiconic sensilla may be involved in extracting information on the identity of the odor source instead of mediating the spatial-temporal concentration pattern in an odor plume.

Keywords: cockroach; differential sensitivity; electrophysiology; food odor coding; resolving power.

FIGURE 1

Scanning electron micrograph of the distal edge of a round-shaped segment from the middle antennal region of the male cockroach. Two main types of olfactory sensilla were distinguished according to external features: sensilla basiconica and trichoidea. Based on their wall structures, Schaller (1978) subdivided the basiconic sensilla into single-walled type A (swA; length, 8–12 μm; base diameter, 2–3 μm) and single-walled type B (swB; length, 18–28 μm; base diameter, 3–4 μm). The trichoid sensilla include the single-walled type C (swC; length, 30–40 μm; base diameter, 3 μm) which contains the ON and OFF ORNs. Data taken from Schaller (1978).

FIGURE 2

Examples of recordings from a swA and swB sensillum type during slowly oscillating changes in the concentration of lemon oil odor. Oscillation periods are 6, 60, and 120 s. (A) Time course of odor concentration. (B) Recordings from the swA sensillum showed activities of two ORNs. (C) Only one ORN in the swA sensillum, producing larger amplitude spikes (pink), responded to oscillations in odor concentration. The rate of discharge of the ORN with the smaller impulse amplitudes remained unchanged (not shown in detail). (D) Recordings from the swB sensillum likewise revealed activities of two ORNs. Unlike the swA sensillum, both ORNs were activated by oscillations in odor concentration. (E) The ORN responding with the more negative amplitudes (blue) was referred to as swB1 ORN and that with the more positive amplitudes (orange) as swB2 ORN. (C,E) Off-line sorted action potentials obtained by spike detecting and template matching techniques using Spike2 software. Note that the increasing density of the action potentials with increasing duration of the oscillation period is due to the decreased time scales in the diagrams.

FIGURE 3

Responses of the swA ORN to oscillating changes in odor concentration. (A) Time course of odor concentration for three different oscillation periods. (B) Time course of impulse frequency. Thin lines are calculated frequency curves smoothing out sharp changes in direction without shifting the maxima and minima values. (C) Time course of mean impulse frequency and standard deviations for 10 swA ORNs. (D) Time course of the rate of concentration change for the three oscillation periods. Different time scales were used to demonstrate the whole oscillation periods. Dotted vertical lines indicate the phase shift between time courses of odor concentration, impulse frequency and rate of concentration change.

FIGURE 4

Responses of the swB1 and swB2 ORNs to oscillating changes in odor concentration. (A) Time course of odor concentration for three different oscillation periods. (B) Time course of impulse frequency of the swB1 ORN (blue) and the swB2 ORN (orange). Thin lines are calculated frequency curves smoothing out sharp changes in direction without shifting the maxima and minima values. (C) Time course of mean impulse frequency and standard deviations for 6 swB1 ORNs. (D) Time course of mean impulse frequency and standard deviations for 7 swB2 ORNs. (E) Time course of the rate of concentration change for the three different oscillation periods. Different time scales were used to demonstrate whole oscillation periods. Dotted vertical lines indicate the phase shift between time courses of odor concentration, impulse frequency and rate of concentration change.

FIGURE 5

Gain of response for instantaneous concentration and the rate of concentration change. (A–C) Impulse frequencies of a single swA, swB1, and swB2 ORN during three different oscillation periods of odor concentration plotted as a function of instantaneous odor concentration and its rate of change. Multiple regressions which utilize three-dimensional planes [F = y0 + a dC/dt) + bC]; where F is the impulse frequency and y0 is the height of the regression plane] were calculated to determine the gain of the responses for instantaneous odor concentration (b-slope) and the rate of concentration change (a-slope). R², coefficient of determination; n, number of points per plot.

FIGURE 6

(A–D) Simultaneously recorded responses of two ORN located in the swA sensillum to ramp-like concentration changes of odor of lemon oil. Ramp rates are 50, 25, and 5%/s. (A) Time course of odor concentration measured by flow meter. (B) Recordings from the swA sensillum showed activities of two ORNs. While the discharge of the ORN with the smaller impulse amplitude remained unchanged (not shown in detail), the ORN with the larger amplitudes (pink) is activated by the slow concentration changes. (C) Off-line sorted action potentials of the ORNs obtained by spike detecting and template matching techniques using Spike2 software. (D) Time course of odor concentration and impulse frequency (bin width, 0.1 s in 50%/s, 0.2 s in 25%/s, 0.5 s in 5%/s).

FIGURE 7

(A–D) Simultaneously recorded responses of two ORNs located in the swB sensillum to ramp-like concentration changes at rates indicated. Ramp rates are 50%/s, 25%/s, and 5%/s. (A) Time course of odor concentration. (B) Extracellular recorded activity; the action potentials from the two ORNs are different in amplitude and shape. Both ORNs are activated by slow changes in concentration. The ORN with the more negative amplitude response (blue) was termed swB1 ORN and that with the more positive amplitude response (orange) swB2 ORN. (C) Off-line sorted action potentials of the ORNs obtained by spike detecting and template matching techniques using Spike2 software. (D) Time course of odor concentration and impulse frequencies of both ORNs (bin width, 0.1 s in 50%/s, 0.2 s in 25%/s, 0.5 s in 5%/s).

FIGURE 8

Stimulus-response functions of ORNs located in the swA, swB1 and swB2 sensilla ORNs to ramp-like concentration changes at rates indicated. Single ORNs of (A) swA sensillum, (B) swB1 and (C) swB2 sensillum, pooled responses of (D) 6 swA ORNs, (E) 6 swB1 and (F) 6 swB2 ORNs. Parabolic regressions were used to approximate the stimulus-response relationships. Bin widths for impulse counts were 0.1 s for 50%/s ramps, 0.2 s for 25%/s ramps and 0.5 s for 5%/s ramps.

FIGURE 9

Impulse frequency of swB1 ORN plotted as a function of impulse frequency of swB2 ORN at the same time interval (bin width) for different rates of ramp-like concentration increase. Activity of each swB ORN pair was recorded simultaneously with the same extracellular electrode. (A) Impulse frequencies of the ORN pair shown in Figures 8B,C. (B) Responses of another pair of swB ORNs. Multiple regressions that utilize three-dimensional planes [F swB1 (imp/s) = y0 + a FswB2 (imp/s) + b (%/s)]; where F is impulse frequency and y0 the height of the regression plane] were calculated to determine the relationship of the response ratio of pairs of swB1 and swB2 ORNs (a-slope) for different rates of concentration increase (b-slope). R², coefficient of determination; n, number of points per plot. Bin width for calculating impulse frequency specifies instantaneous concentration and the first derivative, the rate of change.