Odors elicit three different oscillations in the turtle olfactory bulb

Abstract

We measured the spatiotemporal aspects of the odor-induced population response in the turtle olfactory bulb using a voltage-sensitive dye, RH414, and a 464-element photodiode array. In contrast with previous studies of population activity using local field potential recordings, we distinguished four signals in the response. The one called DC covered almost the entire area of the olfactory bulb; in addition, three oscillations, named rostral, middle, and caudal according to their locations, occurred over broad regions of the bulb. In a typical odor-induced response, the DC signal appeared almost immediately after the start of the stimulus, followed by the middle oscillation, the rostral oscillation, and last, the caudal oscillation. The initial frequencies of the three oscillations were 14.1, 13.0, and 6.6 Hz, respectively. When the rostral and caudal oscillations occurred together, their frequencies differed by a factor of 1.99 +/- 0.01. The following evidence suggests that the four signals are functionally independent: (1) in different animals some signals could be easily detected whereas others were undetectable; (2) the four signals had different latencies and frequencies; (3) the signals occurred in different locations and propagated in different directions; (4) the signals responded differently to changes in odor concentration; (5) the signals had different shapes; and (6) the rostral and caudal signals added in a simple, linear manner in regions where the location of the two signals overlapped. However, the finding that the frequency of the rostral oscillation is precisely two times that of the caudal oscillation suggests a significant relationship between the two. The location of the caudal oscillation in the bulb changed from cycle to cycle, implying that different groups of neurons are active in different cycles. This result is consistent with the earlier findings in the olfactory system of the locust (). Our results suggest an additional complexity of parallel processing of olfactory input by multiple functional population domains.

Fig. 1.

A, A schematic diagram of the olfactometer. Compressed air containing 1% CO₂ was used as the carrier gas. It was cleaned, desiccated, and then mixed with room air saturated with odorant vapor in the odor applicator. The flow rates of the air and the odorant vapor were controlled by a flow meter and a syringe pump, respectively. The odor applicator had two barrels; the outer one was normally under suction to remove the odor. Turning-off of the suction to the outer barrel releases odorant from the end of the applicator. The output of the odor from the applicator can be monitored by measuring the CO₂ of the carrier gas using a CO₂ detector. B, Time course of the odor output from the olfactometer measured by monitoring the CO₂in the carrier gas. The top trace shows the time course of the command pulse delivered to the suction solenoid of the outer barrel of the odor applicator. The bottom trace is the output of the CO₂ detector probe. The inlet of the probe was placed near the mouth of the odor applicator. There is a delay of ∼100 msec between the command pulse and the arrival of the pulse at the CO₂ detector. The odor pulse is approximately square-shaped. C, Schematic diagram of the optical imaging apparatus. The olfactory bulb was illuminated using a 100 W tungsten–halogen lamp. The incident light passed through a heat filter and a 520 ± 45 nm bandpass interference filter and was reflected onto the preparation by a 590 nm long-pass dichroic mirror. The image of the preparation was formed by a 25 mm, 0.95 f camera lens onto a 464-element photodiode array after passing through a 610 nm long-pass secondary filter. The output of each element of the array was amplified by a set of 464 amplifiers. The amplifier outputs were multiplexed, digitized, and stored in a computer. D, The chemical structure of the styryl dye RH414 that was used in these experiments. This dye was obtained as dibromide salt from Molecular Probes.

Fig. 2.

Two comparisons between simultaneous local field-potential recordings and voltage-sensitive dye recordings.A, An example of a relatively simple signal. The odorant evoked a slow DC response (that appears as a small peak after filtering) followed by oscillations in both recordings. The local field-potential and optical recordings are similar: the DC components of both recordings have similar onset latency, and the oscillations have the same frequency and latency and are phase-locked. The odorant was 15% isoamyl acetate. The arrows labeled1 through 4 indicate how we determined the latency and initial frequency of the signals. The time difference between arrows 1 and 2 was used to determine the latency of the DC signal. The time between arrows 1 and 3 was used to determine the latency of the oscillation, and the time between arrows 3 and4 was used to determine the initial frequency.B, An example of more complicated recordings from another preparation. The odorant evoked a fast oscillation followed by a substantially slower oscillation (four arrows) in the local field-potential recording. The bottom two traces are voltage-sensitive dye measurements from rostral (upper) and caudal (lower) regions of the olfactory bulb. The optical recording from the rostral region is simple and has a frequency that is similar to the earlier part of the local field-potential recording. The optical recording from the caudal region is more complex and has a lower-frequency component (four arrows) that is phase-locked with the slow component in the local field-potential recording. The horizontal line labeledodor indicates the time of the command pulse to the odor solenoid. The odorant was 15% isoamyl acetate. The recordings in A and B are filtered by a high-pass digital RC (10 Hz) and low-pass Gaussian (30 Hz) filters.

Fig. 3.

Simultaneous optical recordings from seven different areas of an olfactory bulb. An image of the olfactory bulb is shown on the left. Signals from seven selected pixels are shown on the right. The positions of these pixels are labeled with squares and numbers on the image of the bulb. All seven signals have a filtered version of the DC signal at the time indicated by the bar-labeled DC. The oscillation in the rostral region has a high frequency and relatively long latency and duration (detectors 1 and2). The oscillation from the middle region has a high frequency and short latency and duration (detector4). The oscillation from the caudal region has a lower frequency and the longest latency (detector7). The signal from detectors between these regions (3, 5, and6) appears to contain a mixture of two components. The horizontal line labeled 10% cineole indicates the time of the command pulse to the odor solenoid. The data are filtered by high-pass digital RC (5 Hz) and low-pass Gaussian (30 Hz) filters.

Fig. 4.

The locations and propagation of the four components from the trial illustrated in Figure 3. Multi-frame pseudocolor displays of the signals are overlaid on the image of the olfactory bulb. The DC component in this animal covers almost the entire bulb. The other three panels show the location and spatial spread of one cycle (indicated by the red andgreen lines) of the three oscillations. The rostral oscillation (D) started in a rostral position and propagated in a caudal direction. The caudal oscillation (C) started medially and propagated in a lateral–caudal direction. The center of the middle oscillation (B) remained relatively fixed. The red color and the black contour lines label the areas where the signals are larger than 80 and 20% of the largest signal (see Materials and Methods). The black horizontal bars indicate the time of the odor command pulse. The data are filtered by a high-pass digital RC (5 Hz) and low-pass Gaussian (30 Hz) filters. The ipsilateral olfactory bulb is outlined with blue line in B.

Fig. 5.

Graphs showing the instantaneous frequency of rostral, middle, and caudal oscillations as a function of cycle number in an individual response to odorant in one animal. The frequency of all three oscillations decreased as a function of cycle number. The data come from 17 trials and illustrate the trial-to-trial variability in the frequency of the three oscillations. In four trials, the rostral oscillations continued for more than 12 cycles; the data after cycle 12 are not shown (Animal tuo067).

Fig. 6.

The locations of the initiation sites for three different cycles of rostral and caudal oscillations in one response. The initiation sites of the caudal oscillation changed, whereas the location of rostral oscillations was more stable from cycle to cycle. The time course of the signals from the two regions is shown at thetop. The numbered lines indicate the three cycles (1, 2, 3) presented in multi-frame pseudocolor pictures below. The five pseudocolor images shown for each of the three cycles represent a time period of only 10–20% of the total rostral cycle duration. The ovalsand squares mark the location of the initiation site of cycle 1 of the caudal and rostral oscillation, respectively. In addition, the pseudocolor frames illustrate changes in relative phase of the onset of caudal and rostral oscillations. Thehorizontal line labeled 10% isoamyl acetate indicates the time of the command pulse to the odor solenoid. The data are filtered by a high-pass digital RC (5 Hz) and low-pass Gaussian (30 Hz) filters. A sketch of the ipsilateral olfactory bulb is drawn in 1.

Fig. 7.

Numerical fits of the oscillations. In each of the four panels, the top trace shows the original data. In the middle trace, a part of the original data is shown in expanded time scale (black line) and overlaid with the fit (gray line). The bottom trace shows the fitting error together with the intensity fluctuations before and after the fitted period. In this trace the original signal is displayed with the part enclosed within thebrackets substituted by the fitting error. The middle oscillation was well fit by a symmetrical waveform(p = 2). However, the rostral (B) and caudal oscillations (C) have sharper waveforms and are best fitted with p = 6 and 16, respectively. Dis a more complex waveform from a region where the rostral and caudal signals overlapped. This fit is a linear combination of the rostral and caudal fits shown in B and C. The parameters used in these fits are the following: A,Middle: α₀ = 0.66,p = 2, f₀ = 0.31,r = 0.83; B, Rostral: α₀ = 1.3, p = 6,f₀ = 0.13, r = 0.87; C, Caudal: α₀ = 0.35, p = 16, f₀ = 0.28, r = 0.87. The fit in D is a linear combination of B and C:D = 0.42C + 0.71B. In all four panels, the data were digitally filtered by a high-pass RC filter (5 Hz) and a low-pass Gaussian filter (50 Hz). The horizontal linesunder the traces indicate the time of the command pulse to the odor solenoid.

Fig. 8.

Effects of odor concentration on the four components. The responses to cineole at 10% (A) and 1.7% (B) saturation are shown. The rostral, middle, and caudal oscillations are shown on the left. Pseudocolor representations of the activity at the indicated time points (DC, R, M,C) are shown on the right. The oscillations are smaller and briefer in duration in response to 1.7% cineole. On the other hand, the location and the spatial extent after normalization of all four signals were similar at the two concentrations. In this range of odorant concentration, the middle oscillations were relatively insensitive to concentration change. The data are digitally bandpass-filtered by a high-pass RC filter (5 Hz) and a low-pass Gaussian filter (30 Hz). The horizontal line labeled odor indicates the time of the command pulse to the odor solenoid. A sketch of the ipsilateral olfactory bulb is drawn in A (DC).

Fig. 9.

A plot of the signal amplitude versus concentration of cineole in a single preparation that was tested with a wider range of odorant concentrations. The DC signal was the most concentration dependent, whereas in this animal the size of the middle signal did not appear to change over the range of concentrations used.

Fig. 10.

Comparison of a simultaneously recorded rostral oscillation (top trace) and electro-olfactogram oscillation (bottom trace). Although similar in frequency and duration, the two oscillations are not identical. In this and all other instances, the frequency of the EOG oscillation slowed in comparison with the rostral oscillation. The two vertical lines at time 1 show that the peak of the rostral oscillation (A) occurs at about the midpoint of the EOG oscillation, whereas later, at time2, the peak of the rostral oscillation (B) has moved forward, closer to the beginning of the electro-olfactogram period. In this trial, the rostral oscillation also had a longer duration. The horizontal line labeled 1.7% cineole indicates the time of the command pulse to the odor solenoid. The high-pass filter was a 5 Hz RC; the low-pass filter was a 30 Hz Gaussian.

Fig. 11.

Two different modes of initiation of the caudal oscillation. A, In some cases, the caudal oscillation begins with a higher frequency and then goes through a transition to the typical lower-frequency oscillation. Comparison with the simultaneously recorded rostral signals (below) shows that caudal and rostral oscillations are phase-shifted 180° with respect to each other at the time of the high-frequency caudal oscillations. Thus, the high-frequency oscillation that initiates the caudal oscillation and the rostral oscillation are different. B, An example of the emergence of the caudal oscillation without preceding high-frequency oscillation. The horizontal lines labeled10% isoamyl acetate indicate the time of the command pulse of the odor solenoid. The high-pass filter was a 5 Hz RC; the low-pass filter was a 30 Hz Gaussian.

Fig. 12.

Comparison of the ipsilateral (left) and contralateral (right) olfactory bulb response to odor presented to the ipsilateral nostril. Signals from the rostral and caudal regions of both hemispheres of the olfactory bulb from a single trial. Each signal is the spatial average from four diodes. The data were digitally bandpass-filtered by a high-pass RC filter (3 Hz) and a low-pass Gaussian filter (30 Hz). The turtle was anesthetized with urethane for this trial.