| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Neurobiology and Cognition Research, Faculty of Life Sciences, University of Vienna, Vienna, Austria
Submitted 29 April 2008; accepted in final form 4 August 2008
 |
ABSTRACT |
|---|
|
|
|
INTRODUCTION |
|---|
|
). Much attention has been focused on the trigger for changes in gain and the mechanism underlying gain control in the visual and auditory systems. In the temperature sensory system, however, the effect of adaptation on the input-output gain is less well understood, in part because of the difficulty in determining precisely the instantaneous values of temperature or even the rate of temperature changes at the receptive ending of the sensory cells.
Electrophysiologically, a series of experiments using rapid, step-like changes in temperature indicate two fast-acting adaptation mechanisms. One is sensitive to the magnitude of the step change with the consequence that the input-output functions tend to be steeper where the steps are smaller and to flatten as they become larger. The other is sensitive to the temperature from which the step change has been initiated. This means that the slope of the input-output functions is not same at all temperatures but steepest for one "best" temperature, flanked on both sides by functions of decreasing slopes. Examples are described for warm fibers of the facial pits of snakes (Hensel 1974, 1976) and the rat scrotum (Hellon et al. 1975), for warm cells of the tarsal organ of the wandering spider (Ehn and Tichy 1996), and also for cold fibers of the skin of rhesus monkeys (Darian-Smith et al. 1973), for cold cells on the antennae of the cockroach (Loftus 1968; Nishikawa et al. 1992), the stick insect (Tichy and Loftus 1987), and the locust (Ameismeier and Loftus 1988). The instantaneous values of temperature and its rate of change could not be determined at the receptive ending during rapid step-like temperature changes. Instead the temperature of the conditioning airstream and the difference in temperature between this airstream and a second airstream utilized to provide the steps were taken as parameters even though the rate of temperature change seemed to be the significant stimulus parameter.
In contrast to rapid step changes, where the temperature wave front is very steep, temperature values can be assigned from the airstream to the receptive ending during slow and continuous temperature changes, when the rate is low enough that the temperature of the receptive ending can be considered as locked to that of the air stream. During low rates of temperature change, the temperature of the air stream permits the determination of both the temperature of the receptive ending at each instant in time and the rate at which the temperature of the receptive ending is changing (Ehn and Tichy 1996; Gingl and Tichy 2001). This type of study, carried out on warm and cold cells of various insect and a spider species, revealed a dependence on both the rate of temperature change and the instantaneous temperature (Ameismeier and Loftus 1988; Corbière-Tichané and Loftus 1983; Ehn and Tichy 1996; Gingl et al. 2005; Loftus 1969; Loftus and Corbière-Tichané 1981). Therefore two parameters modulate the discharge rate of the thermoreceptors when temperature changes slowly.
In this study, we examined adaptation of the input-output gain to slow and continuous temperature changes in the cold cell on the antennae of the American cockroach (Fischer and Tichy 2002; Gingl and Tichy 2001; Loftus 1966, 1968, 1969; Nishikawa et al. 1992; Yokohari 1981; Zeiner and Tichy 2000). We used sinusoidal temperature changes because instantaneous temperature and the rate of change vary differently with the oscillation period. Hence it was possible to test the effect of different rates of change at the same instantaneous values and the effect of the same rate of change at different instantaneous temperatures. Specifically, we addressed the following two questions: does the cold cell change its sensitivity to instantaneous temperature or to the rate of temperature change when the period of oscillation varies over a wide range? Does adaptation perform a gain control function to improve the cold cell's ability to encode changes in these parameters? We were additionally interested in possible mechanisms underlying the observed adaptation phenomena. Specifically, we investigated if the observed behavior could be accounted for completely by the physical state of the sensory structures or whether adaptation involves more complex changes in coding characteristics. We compared the time course of the discharge rate with the time course of the temperature monitored inside small, cylindrical model objects. We found that the phase difference is positive between the oscillating discharge rate and the oscillating air temperature, but negative between the oscillating model temperature and the oscillating air temperature. This indicates that the discharge rate is the outcome of active processing of temperature information rather than being defined in terms of the temperature input itself.
|
|
METHODS |
|---|
|
The whip-like antennae of the cockroach P. americana consist of 120–180 ring-shaped segments that grow thinner and longer with increasing distance from the head. The thermoreceptive sensilla are located on the distal half of the antennae, only on the ventral side near a segment's most distal bristles. They are found often on alternating segments, 
20 per antenna and rarely more than one per segment. The sensillum is a small peg-shaped cuticular protuberance that projects at an angle of 50° from the antennal surface and points distally to the antennal tip (Fig. 1). The peg is 6 µm long and 3 µm in diameter at its base, giving a volume of 42 µm3. Using a cuticle of density 1.1*10–9 mg/µm3 (Shimozawa and Kanou 1984), the mass of the peg is 10–7 mg.
|
The adult male cockroaches used in this study were obtained from a crowded colony maintained on a 12:12 dark/light cycle at temperatures between 22 and 25°C. Only animals with antennae exceeding 50 mm in length were used. Thus the flagellum extended 
20 mm beyond the segments from which the recordings were made. Following anesthesia with CO2, an animal was fixed dorsal-side-down on a closely fitting Perspex holder with the head and the antennae protruding. The head was immobilized in this position with a notched Perspex yoke slipped between the head and the thorax. Wings and legs were immobilized with strips of Parafilm wrapped around the holder. For unobstructed stimulation with airstreams at various temperatures, the antenna was fastened with dental cement (Harvard Cement) on the edge of a narrow Perspex ridge projecting laterally from the holder. Action potentials were recorded extracellularly with electrolytically sharpened tungsten electrodes. One electrode was inserted lengthwise into the tip of the antenna and the other at the base of the sensillum. The signals from the electrodes were amplified, band-pass-filtered (0.03–3 kHz) and displayed conventionally, passed through a CED 1401-micro (Cambridge Electronic Design, 12 bit, 300 kHz) interface connected to a PC for on-line recording. The data were stored on a hard disc and analyzed off-line using commercial software (Spike 2).
Stimulation
Continuous changes in temperature were applied by a single air stream merging at 2.5 m/s from a 7-mm nozzle. The air stream flow rate was controlled by passing it through a mass flow meter. The temperature was varied by thermostically subjecting a heat exchanger to slow temperature oscillations. Single oscillation periods took from 42 to 770 s. The partial pressure of water vapor was set at reproducible, precalibrated values (for details, see Tichy 2007). Rapid step-like temperature changes were produced by switching from one stream at steady temperature to another at lower steady temperature for 2 s and then back to the initial temperature. Air stream temperature was measured within ±0.03°C by a small uncoated bead thermistor (250 x 400 µm; Fenwall Electronics, BC 32 L1) 1 mm downstream from the sensillum.
Rationales behind selecting the stimulus range
The range of body temperatures at which an insect functions effectively is quite narrow, often little more than 4°C (May 1979). Electrophysiological studies on various insects, ticks, and spiders have shown that all thermoreceptors exhibit a static discharge at constant temperatures and a dynamic response to changes in temperature (Loftus 1978; Tichy and Gingl 2001). In the cold cell of the cockroach, P. americana, the static activity occurs at least in the range from 20 to 30°C with a maximum between 24 and 26°C. The dynamic activity is determined by the amplitude of cooling, with the greatest dynamic sensitivity corresponding to the maximum static sensitivity. Thus the cold cell's response functions are characterized by a "best" temperature (Fischer and Tichy 2002; Loftus 1966, 1968; Nishikawa et al. 1992). The present experiments were already performed within this "best temperature" range. The rates of temperature changes that were used to examine the discharge characteristics under oscillating stimulus conditions were similar to previous studies of the cold cell of the cockroach (Gingl and Tichy 2001; Loftus 1969), the cold cell of the locust and the warm cell of the tick (Gingl and Tichy 2001), the cold and warm cells of mosquitoes (Gingl et al. 2005) or the warm cell of the wandering spider (Ehn and Tichy 1996). All these thermoreceptors displayed interrupted response patterns when exposed to temperature oscillations between 1 and 0.1 Hz (Ehn and Tichy 1996; Gingl and Tichy 2001; Gingl et al., 2005; Loftus 1969); this complicated the quantitative description of the discharge rates. Therefore oscillation periods <0.02 Hz were not tested in the present study. Five different sine waves were selected, with periods between 42 s and 12 min (oscillation frequencies between 0.02 and 0.001 Hz). The separation between the sine waves was large enough for a tendency to manifest itself in the course of the oscillations' periods, yet small enough to render very low the probability of a significant but unobserved bump or dip in the general course of the function. The lower limit was 0.001 Hz to finish the whole set of five sine waves within a reasonable period of 30 min. With two repetitions, we arrived at a recording period of 1.5 h.
The American cockroach is distributed throughout the temperate, tropical, and subtropical regions of the world. Although very adaptive, this insect prefers temperatures between 25 and 30°C (Baumholtz et al. 1997). The temperature in our rearing room was kept near 25°C. Provided that a water supply is constantly maintained, we had no indication that maintaining the temperature above ambient is of any advantage in rearing cockroaches. The present experiments were therefore conducted at laboratory temperatures between 22 and 25°C.
Identification
The cold cell occurs in the same sensillum with two hygroreceptive sensory cells. Most recordings (72%, n = 47) revealed the activity of all three sensory cells, but in some recordings (28%, n = 18), only two were detected. The cold cell was identified by its increase in impulse frequency to a drop in temperature, which was produced by shifting between two streams of dry air, the first at higher temperature and the second at lower temperature (Tichy 2007). A shift back to the initial air stream yielded an abrupt cessation of the activity. When the temperature change was repeated at moderate humidity, however, both the cold cell and the hygroreceptors responded by changing their discharge rates. Such reactions indicate receptors for relative humidity; because relative humidity is defined as the ratio of the partial pressure of water vapor to the saturation vapor pressure, it must change when temperature changes and the vapor pressure remains constant because the saturation vapor pressure changes with temperature.
Impulse frequency
Impulse frequency (imp/s) was determined by impulse count during 100 ms after the onset of rapid step-like stimulation. For responses to slowly changing temperature, running averages of three consecutive 4-s periods were taken to measure frequency (Corbière-Tichané and Loftus 1983). A 4-s period was used rather than the more common 1-s period because the low rate of temperature change was reflected in a slow change in the cells' discharge rate.
Model objects
We made four model objects from Plexiglas. They were cylindrical and weighed 4, 46, 115, or 440 mg. They all had the same surface-to-volume ratio of 0.5. This ratio was important for a series of yet unpublished experiments in which we tested the thermal effect of infrared radiation. Here we describe data from a parallel study of the thermal effect of convection. The air stream flowed along the longitudinal axis of the model objects or at an inclination of 60° with respect to the longitudinal axis. The model temperature was measured with a small bead thermistor (250 x 400 µm; Fenwall Electronics, BC 32 L1) positioned inside a central channel and fixed with Epoxy glue.
Transmission factor
The efficiency of the model objects in transmitting heat was determined by the ratio of air temperature to object temperature. Value 1 indicates that the model object attained air temperature. All values obtained from the different models are independent of the direction of the oncoming air flow.
|
|
RESULTS |
|---|
|
To determine the gain for the instantaneous temperature and the rate of temperature change, we exposed each of 45 cold cells to five series of temperature oscillations with periods between 42 and 770 s. In each series, a temperature range of roughly 3.5°C between 18 and 24°C was covered and the rate of change lay between –0.2 and +0.2°C/s. Figure 2 shows the results of such an experiment. The top trace in each panel represents the time course of the temperature oscillations with periods of 42, 88, 200, 385, and 770 s, and the middle trace the corresponding oscillations in impulse frequency. In general, impulse frequency tended to be higher at the lower instantaneous temperature values and lower at the higher instantaneous values. The frequency values may be interpreted as the response to the instantaneous temperature. However, the oscillations in impulse frequencies and instantaneous temperature were not in step. The frequency curves led the temperature curves. The phase difference was +5 s for an oscillation period of 42 s, and +35 s for an oscillation period of 770 s. Thus the oscillations in impulse frequency cannot be explained exclusively by oscillations in the instantaneous temperature.
|
To estimate the double dependence of the cold cell on instantaneous temperature and its rate of change, the impulse frequencies for the different oscillation periods were plotted in Fig. 2 as a function of both parameters. The frequency curves approached closed curves reminiscent of Lissajous figures in which two oscillating magnitudes are plotted, one as a function of the other. The figures indicate that the sequence of frequency values is too orderly to simply attribute phase differences during any oscillation period to random variation in the response.
Multiple regressions (F = a + bT + cdT/dt; where F is the impulse frequency and a the height of the regression plane) were calculated to determine the simultaneous effect of instantaneous temperature (b slope) and the rate of change (c slope) on the response frequency during different oscillation periods (Fig. 2). In all 45 cells thus examined, the correlation coefficients (r > 0.99) show a strong linear relationship between impulse frequency, instantaneous temperature and the rate of temperature change. The square of the correlation value (r2) indicates that an average of 98% of the variation in impulse frequency can be explained by the double regression. When the correlation coefficient is reduced by its SD (±0.005), the percentage drops only to 97%.
The orderly relationships of impulse frequency to the instantaneous temperature and its rate of change during different oscillation periods provide an opportunity to determine the cold cell's gain for each of these two parameters. In the example shown in Fig. 2, the gain for instantaneous temperature was –3.4 imp/s per °C for an oscillation period of 42 s (Fig. 2A) and –1.4 imp/s per °C for a period of 770 s (Fig. 2C); the gain for the rate of temperature change was –25.6 imp/s per °C/s for an oscillation period of 42 s (Fig. 2A) and –104.3 imp/s per °C/s for a period of 770 s (Fig. 2C, the negative values represent the downward direction of the temperature change, yielding a rise in the impulse frequency of the cold cell). The measurements show that during any oscillation period, impulse frequency can be influenced more by changing the rate of temperature change by 1°C/s than by changing instantaneous temperature by one additional degree. An increase of 1 imp/s during an oscillation period of 42 s can be elicited either by a –0.29°C decrease in instantaneous temperature (provided the rate of change is constant) or by a rate of change of –0.039°C/s. During an oscillation period of 770 s, it takes a decrease of –0.71°C in instantaneous temperature to increase impulse frequency by 1 imp/s (or a rate of change of –0.0095°C/s).
For all 45 cold cells, the values of the three parameters of the regression planes were pooled and plotted in Fig. 3 against the oscillation period. Exponential functions indicate that both the height of the regression plane (Fig. 3A) and the gain of the response for instantaneous temperature (Fig. 3C) decrease rapidly with increasing oscillation period. In contrast, the gain of the response for the rate of change increases rapidly when the oscillation period increases (Fig. 3B). Therefore the gain for the instantaneous temperature (Fig. 3B) is high when the oscillation period is brief and adapts when the oscillation period is lengthier. During adaptation of the gain for the instantaneous temperature, however, the gain for the rate of temperature change is improved (Fig. 3C). This improvement occurred during adaptation to instantaneous temperature. Adaptation is therefore a mechanism for gain control through which the cold cell increases its sensitivity to low rates of temperature change at the expense of sensitivity to the instantaneous temperature. These results indicate that the cold cell does not simply respond to the temperature changes as they occur at the receptive ending, but balances—from instant to instant—the response magnitude according to these temperature changes. To illustrate that such balance is an intrinsic property of the cold cell, we subjected small models to oscillating changes.
|
The four cylindrical models weighing 4, 16, 115, and 440 mg were exposed to oscillating changes in air temperature. A sequence of three different oscillation periods of 60, 130, and 300 s was tested on each model object with rates between –0.2 and +0.2°C/s. The temperature range covered was roughly 10°C between 19 and 25°C. Figure 4 is an example of the temperature change measured within the 115-mg model object. The top trace in each panel represent the time course of air temperature, the middle trace the time course of the model temperature, and the bottom trace the time course of the rate at which air temperature changed.
|
|
|
|
DISCUSSION |
|---|
|
; Corbière-Tichané and Loftus 1983; Ehn and Tichy 1996; Gingl et al. 2005; Loftus 1969; Loftus and Corbière-Tichané 1981), but here we varied the oscillation period. This enabled us not only to test the cold cell's responsiveness to many different combinations of instantaneous temperatures and rates of change, but also to study the combined effect of both parameters on the gain of the input-output function. We found that the gain can be changed through changes in the duration of the oscillation period. When the period is long, adaptation of the cold cell performs a gain control function by improving the ability to code the rate of temperature change at the expense of the instantaneous temperature. As shown in Fig. 6, gain control in the cold cell involves decreasing sensitivity to instantaneous temperature while increasing sensitivity to the rate of change. (The negative values for gain reflect the downward direction of the temperature change, yielding a rise in impulse frequency, and specify the receptor type—a cold cell).
|
|
While sensillum temperature is virtually the same as air temperature, the temperature of small model objects departs considerably from air temperature. The extent of departure was expressed by two means: the transmission factor, which describes the temperature that the model reached during changes in air temperature, and by the phase difference, which indicates whether the former is in step with the latter. Clearly, the mass of the model objects affects both the transmission factor and the phase difference. The lower the model mass, the smaller the difference between the two temperatures and, in addition, the smaller the phase lags between them. Notwithstanding the difference between the dimensions of the model objects and the thermoreceptive sensillum, two properties are evident. First, the phase relationship between the oscillating temperature of the models and the oscillating air stream is negative due to hysteresis, but the phase relationship between the oscillating cold cell's responses and the oscillating air stream is positive as result of its ability to respond to the rate of temperature change. Second, the sensillum temperature is locked to the air stream and takes on the maximum and minimum values of the oscillating air temperature. Thus sensillum temperature changes at the same rate as air temperature but with a positive phase shift. In the models, the temperature range covered during an oscillation period is smaller than air temperature. Thus object temperature changes at different rates than air temperature and with a negative phase shift.
The cold cell's output cannot simply be defined in terms of the temperature input itself as some kind of passive transducer. Rather the output apparently reflects the cold cell's ability to adjust its gain to different aspects of the temperature input. The signal controlling gain must be derived directly from the temperature input or from the signal of the cold cell itself. This inherently limits the accuracy with which gain can be controlled. The biophysical mechanism that allows the interaction of the incoming temperature parameters is unknown. This is not a trivial problem because, intuitively, the natural operation for a sensory cell to perform would be some kind of average or weighted sum not a product. Ultimately this gain control is part of what makes thermoreceptors so difficult to emulate with man-made devices.
|
|
GRANTS |
|---|
|
|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: H. Tichy, Dept. of Neurobiology and Cognitive Science, Faculty of Life Sciences, University of Vienna, Althanstrasse 14, 1090 Vienna, Austria (E-mail: Harald.Tichy{at}univie.ac.at)
|
|
REFERENCES |
|---|
|
Baumholtz MA, Parish LC, Witkowski JA, Nutting WB. The medical importance of cockroaches. Int J Dermatol 36: 90–96, 1997.[CrossRef][Web of Science][Medline]
Chance FS, Abbott LF, Reyes AD. Gain modulation from background synaptic input. Neuron 35: 773–782, 2002.[CrossRef][Web of Science][Medline]
Corbière-Tichané G, Loftus R. Antennal thermal receptors of the cave beetle Speophyes lucidulus Delar. II. Cold receptor response to slowly changing temperature. J Comp Physiol [A] 153: 343–351, 1983.[CrossRef]
Darian-Smith J, Johnson KO, Dykes R. "Cold" fiber population innervating palmar and digital skin of monkey: response to cooling pulses. J Neurophysiol 36: 325–346, 1973.
Ehn R, Tichy H. Response characteristics of a spider warm cell: temperature sensitivities and structural properties. J Comp Physiol [A] 178: 537–542, 1996.
Fischer H, Tichy H. Cold-receptor cells supply both cold- and warm-responsive projection neurons in the antennal lobe of the cockroach. J Comp Physiol [A] 188: 643–648, 2002.[CrossRef][Web of Science][Medline]
Gingl E, Hinterwirth A, Tichy H. Sensory representation of temperature in mosquito warm and cold cells. J Neurophysiology 94: 176–185, 2005.
Gingl E, Tichy H. Infrared sensitivity of thermoreceptors. J Comp Physiol [A] 187: 467–475, 2001.[CrossRef][Web of Science][Medline]
Haug T. Struktur, Funktion und Projektion der antennalen Thermo- und Hygrorezeptoren von Antheraea pernyi (Lepidoptera: Saturniidae). (PhD thesis). University Regensburg, Germany, 1986.
Hellon RF, Hensel H, Schäfer K. Thermal receptors in the scrotum of the rat. J Physiol 248: 349–357, 1975.
Hensel H. Properties of warm receptors in boa constrictor. Naturwissenschaften 61: 369, 1974.[Web of Science][Medline]
Hensel H. Functional and structural basis of thermoreception. Prog Brain Res 43: 105–118, 1976.[Medline]
Loftus R. Cold receptor on the antenna of Periplaneta americana. Z. vergl. Physiol 52: 380–385, 1966.
Loftus R. Differential thermal components in the response of the antennal cold receptor of Periplaneta americana to slowly changing temperature. Z Vergl Physiol 63: 415–433, 1969.
Loftus R. Peripheral thermal receptors. In: Sensory Ecology: Reviews and Perspectives, edited by Ali MA. New York: Plenum, 1978, p. 439–466.
Loftus R. Responses of the antennal cold receptor of Periplaneta americana to rapid temperature changes and to steady temperature. Z Vergl Physiol 59: 413–455, 1968.[CrossRef]
Loftus R, Corbière-Tichané G. Antennal warm and cold receptors of the cave beetle, Speophyes lucidulus Delar, in sensilla with a lamellated dendrite. I. Response to sudden temperature change. J Comp Physiol 143: 443–452, 1981.[CrossRef]
May M. Insect thermoregulation. Annu Rev Entomol 24: 313–349, 1979.[CrossRef][Web of Science]
Nishikawa M, Yokohari F, Ishibashi T. Response characteristics of two types of cold receptors on the antennae of the cockroach, Periplaneta americana L. J Comp Physiol [A] 171: 299–307, 1992.
Shimozawa T, Kanou M. The aerodynamics and sensory physiology of range fractionation in the cercal filiform sensilla of the cricket. Gryllus bimaculatus. J Comp Physiol 155: 495–505, 1984.
Tichy H. Humidity-dependent cold cells on the antenna of the stick insect. J Neurophysiol 97: 3851–3858, 2007.
Tichy H, Gingl E. Problems in hygro- and thermoreception. In: Ecology of Sensing, edited by Barth FG, Schmid A. Berlin: Springer, 2001, p. 271–287.
Tichy H, Loftus R. Response characteristics of a cold receptor in the stick insect, Carausius morosus. J Comp Physiol 132: 149–152, 1987.
Yokohari F. The sensillum capitulum, an antennal hygro- and thermoreceptive sensillum of the cockroach, Periplaneta americana L. Cell Tissue Res 216: 525–543, 1981.
Zeiner R, Tichy H. Integration of temperature and olfactory information in cockroach antennal lobe glomeruli. J Comp Physiol [A] 186: 717–727, 2000.[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
TOP
ABSTRACT
INTRODUCTION
). B: the sensillum appears as a short peg (
T/
) are indicated. Impulse frequency is always ahead of instantaneous temperature and lags behind the rate of change. Multiple regressions from values of impulse frequency with respect to corresponding values of instantaneous temperature (T) and the rate of temperature change (
