Halothane Modulates Thermosensitive Hypothalamic Neurons in Rat Brain Slices

Methods: Brain slices were obtained and prepared from Sprague-Dawley rats. Preoptic region neurons were characterized by extracellular recording of spontaneous firing rates and thermosensitivity to localized heating and cooling, before, during, and after halothane equilibrated in the perfusate and carrier gas.

Results: One hundred sixteen neurons were characterized by their thermosensitivity as: 29% warm-sensitive (> 0.8 spikes *symbol* s sup -1 *symbol* degrees Celsius sup -1); 14% cold-sensitive (< 0.6 spikes *symbol* s sup -1 *symbol* degrees Celsius sup -1); and 57% temperature-insensitive. Halothane significantly reduced the spontaneous firing rates to 64% of control and the thermosensitivity to 55% of control for warm-sensitive neurons at 1% halothane. Halothane significantly reduced the spontaneous firing rate of cold-sensitive neurons to 24 and 40% of control, and the thermosensitivity to 61 and 36% of control at 0.5, and 1% halothane, respectively. Spontaneous firing rates and thermosensitivity returned toward control values in warm-sensitive neurons (92 and 122% of control, respectively) after discontinuation of halothane, which did not occur in cold-sensitive neurons (49 and 36% of control, respectively). Halothane did not alter the thermosensitive temperature range or the set point temperature at which neurons became most thermosensitive. Halothane also did not affect the firing rates of temperature-insensitive neurons.

Conclusions: Halothane alters the firing rate and thermosensitivity of individual temperature-sensitive neurons in in vitro slices of the preoptic region of the anterior hypothalamus in the absence of afferent modulation. This disruption may result in an imprecision of thermoregulatory responses locally within the preoptic region, to thermal challenges and represents a potential mechanism by which halothane widens the thermoregulatory threshold range. (Key words: Anesthetics, volatile: halothane. Brain: anterior hypothalamus; brain slices; preoptic region. Single-unit activity, temperature: thermoregulation.)

VOLATILE anesthetics disrupt thermoregulation and diminish autonomic and behavioral responses to thermal challenges. Alterations in thermal homeostasis produced by the volatile agents interfere with both heat loss and heat gain, and may profoundly affect cardiovascular, musculoskeletal, renal, and respiratory systems as well as coagulation pathways. Multiple sites within the central nervous system (CNS) are involved in normal thermoregulation. Peripheral and deep thermal receptors converge on the neural substrates of the spinal cord and higher CNS thermoregulatory areas, and are essential to the physiologic processing of thermal input and initiation of appropriate thermoregulatory responses. However, specific neurons in the preoptic region of the anterior hypothalamus appear to be most intimately involved at the highest CNS level in the initiation and maintenance of a normal thermoregulatory response. A hierarchical model has been proposed in which phylogenetically older centers subserve newer, more precise regions and signal processing occurs successively in the spinal cord, mid-brain, and hypothalamus. .

Disrupting normal preoptic region neuronal function alters the extent of thermoregulatory responses. Partial ablation of these neurons attenuates the magnitude of a response to a thermal challenge, whereas complete ablation of these neurons severely blunts normal thermoregulatory responses. Local preoptic region heating increases body heat loss through various mechanisms including vasomotor adjustments, sweating, and respiratory and behavioral alterations. Similarly, local preoptic region cooling results in decrease in the rate of body heat loss or increase in the rate of heat production by vasomotor alterations and the initiation of shivering. Previous studies from this laboratory have demonstrated that in chronically instrumented cats, volatile anesthetics produce an imprecision in the control of thermoregulatory responses at the level of the anterior hypothalamus at the preoptic region. Halothane, isoflurane, and enflurane abolished normal thermoregulatory responses to preoptic region thermal challenges and preoptic region heating and cooling was found to appropriately modulate postanesthetic shivering. The presence of specific preoptic region neurons that alter firing rates and patterns in response to local thermal challenges is well established. While the ability of anesthetics to alter normal thermoregulation is well documented, only a previous study from this laboratory has specifically examined the effects of volatile anesthetics on single thermosensitive neurons in the CNS. Poterack and coworkers demonstrated that halothane decreases both the spontaneous firing rate and thermosensitivity of warm-sensitive preoptic region neurons in cats anesthetized with chloralose and urethane. However, the effects of the baseline anesthetics and the confounding influence of potent afferent thermal input from peripheral or lower CNS sites was not eliminated. Studies using preoptic region tissue slice preparations have confirmed the existence of distinct populations of warm- and cold-sensitive and temperature-insensitive neurons. Subtypes of preoptic region neurons differ in terms of morphology, intracellular second messengers, synaptic dependency, and ionic conductances. There are no previous studies examining the effects of halothane on these preoptic region subpopulations.

The purpose of the current investigation was to examine the effects of graded concentrations of halothane on firing rates and thermosensitivity of temperature-sensitive and temperature-insensitive preoptic region neurons in tissue slices, while eliminating confounding thermal afferent modulation and additional baseline anesthetics. In addition, the differential effects of halothane on warm-sensitive and cold-sensitive neurons was examined.

Experimental procedures were approved by the Medical College of Wisconsin Animal Use and Care Committee, and protocols were completed in accordance with the Guiding Principles in the Care and Use of Laboratory Animals of the American Physiological Society and in accordance with National Institutes of Health guidelines.* All animals were housed within the animal facilities of the Medical College of Wisconsin, accredited by the American Association for the Accreditation of Laboratory Care.

Male Sprague-Dawley rats, weighing 250–325 g, with no significant weight differences between the groups, were used in these experiments. Many of the techniques used in this tissue slice preparation are similar to those described by Hatton et al. Rats were placed in a chamber and lightly anesthetized with halothane (less than 1% on vaporizer dial), decapitated, and their brains were quickly removed and rinsed with nutrient medium of the following composition (in mM): NaCl 124, KCl 5, CaCl²2.4, MgCl²1.3, glucose 10, KH²PO⁴1.24, NaHCO sub 3 26, gaseous equilibration with 95% Oxygen²and 5% CO², pH 7.4, 300 mOsm/kg. Nutrient medium was prepared daily and all electrophysiologic recordings were performed on the same day as tissue slices preparation. Brains were cut freehand to small blocks containing the ventrorostral forebrain. Coronally oriented tissue slices (300-micro meter thick) containing the preoptic region were sectioned with a vibratome mechanical tissue slicer. Typically, two tissue slices containing both the anterior commissure and the optic chiasm, in addition to at least one slice immediately caudal and/or rostral were evaluated for recordings. During the slicing procedure, tissue remained bathed in the oxygenated nutrient medium at or slightly below room temperature. Time from decapitation to placement in nutrient medium was less than 5 min. Subsequently, the slices were transferred to a thin nylon mesh, which rested on the methyl methacrylate polymer slice chamber (Figure 1).

The slice chamber employed in these studies was a Hass type chamber (Medical Systems, Greenvale, NY), physically modified after the method of Kelso et al. With this modification, selective and precise thermal challenges of very small portions of the neuronal tissue being concurrently used for extracellular neuronal recording may be performed. The center recording chamber containing the slices was surrounded by an outer, temperature-controlled water bath to provide a humidified oxygenated atmosphere above the slice. Slice temperature was maintained at 37 degrees Celsius by autoregulation heating of the water bath and perfusion by a sensing thermistor throughout the experiment. Slices were maintained in this chamber, continuously perfused with the oxygenated nutrient medium for 1–2 h before the initiation of the experimental protocol to allow for neuronal recovery after decapitation and slicing. The nutrient medium flowed at a rate of 1.5–3.0 ml/min, under and around the tissue slices, which allowed the volume of the slice chamber to be completely exchanged 2–3 times/min by switching the perfusion medium via solenoid controlled valves. Filter paper was placed at the edge of the nylon mesh to direct the perfusion medium out of the chamber. Precise, rapid, local preoptic region tissue temperature was controlled by passing water through a thermoelectric Peltier device (Melcor, Trenton, NJ) coupled to the thermode cannula. The thermode cannula was located within the perfusion chamber methyl methacrylate polymer base immediately below the preoptic region of the brain slice (as modified and described by Kelso et al. ) This allowed local, precise cooling and heating of the preoptic region neurons without altering the temperature of the perfusion medium or the surrounding areas of the slice. A complete heating and cooling thermal challenge was performed over approximately 10 min.

After the slices had equilibrated, single-unit neuronal activity was recorded with 1- to 2-micro meter tip, glass carbon and filament microelectrodes or 10-micro meter tip polytetrafluorethylene-insulated microelectrodes. Electrodes were mechanically advanced (Burleigh Instruments, Fishers, NY) within tissue slices at 5–20-micro meter increments until a consistent single extracellular single unit was identified. Unit activity was sequentially amplified at 10–20,000 times by alternating current coupled amplifiers. A microthermocouple was inserted into the slice tissue for precise measurement of local temperature at the single unit recording site. Amplified electrical activity was displayed on an oscilloscope and monitored by a speaker. Voltage and time/voltage window discriminators were used for single-unit isolation and elimination of background noise and the subsequent window discriminator directed to an 8-bit frequency to voltage converter to generate a spike histogram. Halothane concentrations in the artificial cerebrospinal fluid bath were intermittently sampled and measured with the use of a Perkins Elmer (Norwalk, CT) model Sigma 3B gas chromatography system.

After single neuronal identification, specific heating and cooling of the preoptic region (30–40 degrees Celsius) was performed as described earlier, and control measurements of firing rates at 36 degrees Celsius and throughout the thermal challenge were recorded. Graded concentrations of halothane (0.25%, 0.5%, and 1.0%) were introduced into the perfusion medium and into the humidified oxygenated environment immediately surrounding the slice chamber. After introduction of halothane or any change in concentration, 30 min was allowed for equilibration. Preliminary experiments in this system demonstrated that 15–22 min was required to achieve steady-state halothane concentrations within the slice chamber. After halothane equilibration, heating and cooling temperature curves were again performed and firing rates recorded. At the end of the experiments, halothane was discontinued for 30–45 min and temperature curves again repeated and the results quantified.

All data were recorded on FM tapes using a Vetter tape deck (Rebersburg, PA). In addition to the raw spike firing rate, integrated firing rate, spike histogram, and tissue slice temperature were recorded continuously on a polygraph. Data were subsequently analyzed on an IBM-compatible microcomputer equipped with a Metrabyte Dash-16 12-bit analog-to-digital convertor and Asyst analysis software (Keithly Metrabyte, Taunton, MA) developed in this laboratory. Thermosensitivity was calculated as impulses or spikes *symbol* s sup -1 *symbol* degrees Celsius sup -1. Neurons were classified as warm-sensitive; generally increasing firing rates with increasing temperature (> 0.8 spikes *symbol* s sup -1 *symbol* degrees Celsius sup -1), cold-sensitive; generally increasing firing rates with decreasing temperature (< 0.6 spikes *symbol* s sup -1 *symbol* degrees Celsius sup -1) and temperature-insensitive based on well-accepted criteria. After establishing thermosensitive neuronal responses, linear regression was determined over the most thermosensitive temperature range. Thermosensitivity was determined over a 2–5 degrees Celsius temperature range which included 37 degrees Celsius during the control, prehalothane conditions. Linear regression was used to analyze and describe the thermosensitivity for all neurons evaluated. Although several units exhibited more nonlinear curves throughout the entire temperature range tested, complex computer-assisted analysis of these various order exponential and logarithmic regressions, although more precisely descriptive of firing patterns, resulted in data that were difficult to average and made interneuronal comparisons impossible. Therefore, whereas many units were analyzed in this fashion, the data are not presented here because of their complexity. Linear regression curves were performed over a 2–5 degrees Celsius temperature range in which these neurons exhibited their greatest thermosensitivity. The specific range was determined by computer regression analysis of multiple temperature ranges before determination of the maximum thermosensitive range. Linear regression correlation (R value) of > 0.6 was required to be considered temperature-sensitive. During each experimental condition, linear response curves were the result of 2 or 3 computerized merged heating and cooling cycles from an individual neuron. As evaluations were performed over the most thermosensitive temperature range, the temperature ranges were not necessarily consistent between experimental conditions, i.e., control or halothane concentrations. Therefore, the threshold temperatures were also recorded and compared between experimental conditions. These threshold temperatures represent the lowest temperature at which a warm-sensitive neuron entered the most thermosensitive range and the highest temperature at which a cold-sensitive neuron entered the most thermosensitive range. All thermosensitivity calculations were determined by 2 or 3 evaluators, one of whom was blinded to the experimental conditions.

Spontaneous firing rate at 36 degrees Celsius, firing rate at 37 degrees Celsius, maximum firing rate within the most thermosensitive temperature range, and threshold temperatures, were calculated as a mean of 2 or 3 samples, for each neuron, for each experimental condition. The spontaneous firing rate at 36 degrees Celsius was obtained by an average of 5–7 samples of 10-sec counts on a ratemeter from the window discriminator output, while the firing rate at 37 degrees Celsius and the maximum firing rate within the most thermosensitive temperature range were obtained by graphic analysis of the linear regression curves.

The millimolar concentrations of halothane measured in the bath were converted to equivalent partial pressures in the solution and expressed as percentages of the volatile agent in the gas phase (Table 1).

Statistical analysis of data during control conditions and during anesthetic or postanesthetic conditions for each firing rate parameter, threshold temperatures, and for thermosensitivity was performed with a repeated-measures analysis of variance on the absolute values. Pairwise comparisons of interventions were performed with contrasts derived from the repeated-measures analysis of variance adjusted for multiplicity by Dunnett's method using the SAS procedure GLM software (SAS Institute, Cary, NC). Changes at each anesthetic concentration and after discontinuing halothane were compared to control and considered significant when the probability value was less than 0.05. Data are expressed as the mean plus/minus standard error of the mean. Data were also analyzed after normalization, with reference to baseline-control firing rates and thermosensitivities to allow more direct comparisons between the neuronal subpopulations (temperature sensitive and insensitive). Data distributions did not change after data normalization.

A total of 150 preoptic region neurons, taken from 190 slices, obtained from 158 rats, were located. Of these, 77%(n = 116) were reliably characterized according to thermosensitivity by demonstrating reproducible firing patterns in response to heating and cooling challenges. The remaining unclassified units were either lost before a temperature change was completed or their response was not repeatable on successive temperature challenges. Of the 116 characterized neurons, 34 (29%) were warm-sensitive, 16 (14%) were cold-sensitive, and 66 (57%) were temperature-insensitive. Of these 116 neurons, 13 warm-sensitive, 6 cold-sensitive, and 5 temperature-insensitive neurons were successfully challenged with halothane and completed the experimental protocol.

Successful experiments were completed with halothane in 13 warm-sensitive neurons. The control spontaneous firing rate at 36 degrees Celsius of the warm-sensitive neurons was 11.2 plus/minus 2.1 Hz and the firing rate at 37 degrees Celsius, (measured graphically from the linear regression curve), was 12.9 plus/minus 2.0 Hz (Table 2). Responses to thermal challenges varied from linear over the entire temperature range (31–41 degrees Celsius) to biphasic increases, with threshold temperatures (initiation of the most thermosensitive temperature range) commonly at 35.5–36.5 degrees Celsius. Administration of 0.25% halothane did not change the spontaneous firing rate at 36 degrees Celsius (Figure 2) or the firing rate at 37 degrees Celsius (T2-15). Increasing concentrations of halothane resulted in firing rates of 84% plus/minus 11 at 36 degrees Celsius; 83% plus/minus 15 at 37 degrees Celsius (0.5% halothane) and 64% plus/minus 12 (P < 0.05) at 36 degrees Celsius; and 58% plus/minus 12 (P < 0.05) at 37 degrees Celsius (1% halothane) as % of control. The firing rates returned to 92–95% of control levels after halothane was discontinued.

The temperature range in which warm-sensitive neurons were most thermosensitive was 35.8 plus/minus 0.3 degree Celsius to 38.6 plus/minus 0.4 degree Celsius at control. Neither the temperature range nor the threshold temperature (lowest temperature within the most thermosensitive temperature range) was significantly altered by halothane administration. The maximum firing rate within the thermosensitive temperature range evaluated (19.6 plus/minus 2.6 Hz at control) was significantly decreased to 66 plus/minus 10% of control at 1% halothane, and returned to 94 plus/minus 9% of control after discontinuation of halothane administration. The calculated thermosensitivity of the warm-sensitive units before halothane administration was 2.9 plus/minus 0.5 spikes *symbol* s sup -1 *symbol* degrees Celsius sup -1 (Figure 3). Thermosensitivity was significantly attenuated (55 plus/minus 16% of control) by halothane administration only at the highest concentration used (1%). After discontinuation of halothane, thermosensitivity returned to control (122 plus/minus 26% of control). Figure 4shows representative responses to thermal challenges in a single warm-sensitive neuron.

Successful experiments were completed with halothane in six cold-sensitive neurons. The control spontaneous firing rate at 36 degrees Celsius of the cold-sensitive neurons was 11.9 plus/minus 2.6 Hz and the firing rate at 37 degrees Celsius was 9.0 plus/minus 2.9 Hz (Table 3). Responses to thermal challenges varied from linear (2 of 6) over the entire temperature range (31–41 degrees Celsius) to previously well-described, bell-shaped curves with the greatest thermosensitivity above 36 degrees Celsius. Administration of 0.25% halothane did not significantly affect the spontaneous firing rate at 36 degrees Celsius or the firing rate at 37 degrees Celsius. However, further increases in halothane decreased firing rates to 24 plus/minus 9% at 36 degrees Celsius (0.5% halothane) and 40 plus/minus 18% at 36 degrees Celsius (1% halothane) of control (T3-15, F2-15). In contrast to the recovery of firing in warm-sensitive neurons on discontinuation of halothane, the spontaneous firing rate of cold-sensitive neurons only returned to 49 plus/minus 11% of control 30 min after discontinuation of halothane.

The temperature range in which cold-sensitive neurons were most thermosensitive was 36.5 plus/minus 0.5 degree Celsius to 38.8 plus/minus 0.4 degree Celsius at control. As observed with the warm-sensitive neurons, neither the most thermosensitive temperature range nor the threshold temperature (highest temperature within the thermosensitive temperature range) was significantly altered by the administration of halothane. The maximum firing rate within the thermosensitive temperature range evaluated was 20.5 plus/minus 2.9 Hz at control. Halothane administration significantly decreased the maximum firing rate (expressed as % of prehalothane control to 54 plus/minus 7%(P < 0.05), and 61 plus/minus 17%(P < 0.05) at 0.5, and 1% halothane, respectively. Thirty minutes after discontinuing halothane, the maximum firing rate did not return toward control values and remained significantly attenuated at 59 plus/minus 15% of control. The average calculated thermosensitivity during control conditions of the cold-sensitive units was -5.3 plus/minus 0.8 spikes *symbol* s sup -1 *symbol* degrees Celsius sup -1. Maximum thermosensitivity was blunted by all concentrations of halothane administered (61 plus/minus 12% and 36 plus/minus 9% of control at 0.5, and 1% halothane, respectively)(F3-15). After discontinuation of halothane, thermosensitivity did not recover and remained significantly decreased from control values (36 plus/minus 7% of control). Figure 5illustrates representative responses to thermal challenges in a single cold-sensitive neuron.

Halothane was administered to five neurons classified as temperature-insensitive, which subsequently completed the experimental protocol. The effects of halothane on various firing rate and thermosensitivity parameters in temperature-insensitive neurons are found in Table 4. The spontaneous firing rate of these neurons at 36 degrees Celsius during control conditions was similar to the firing rate observed at 37 degrees Celsius during a thermal challenge (4.1 plus/minus 0.6 Hz and 3.7 plus/minus 0.6 Hz, respectively). Maximum firing rate during prehalothane control was 7.2 plus/minus 1.1 Hz and varied both between and within neurons as to the temperature at which the maximum firing rate occurred. The maximum firing rate was not associated with specific temperatures in this subset of neurons, but rather was a random occurrence. Administration of increasing concentrations of halothane did not significantly affect the thermosensitivity of these cells. In addition, halothane administration did not significantly modulate the spontaneous or maximum firing rate (T4-15) of temperature-insensitive neurons during a thermal challenge.

Temperature-sensitive neurons exist in the preoptic region of the anterior hypothalamus and are intimately involved in thermoregulatory functions. The role of the anterior hypothalamus in thermoregulation is well documented. Previous investigations have demonstrated a system in which a subpopulation of these neurons respond both to extrahypothalamic thermal challenges and, appropriately, to the administration of a variety of pharmacologic agents. These findings have led to the widely accepted concept that an essential role for these neurons is the initiation and modulation of thermoregulatory responses. The current investigation demonstrates that the volatile anesthetic, halothane, significantly alters both the spontaneous and maximum firing rate and the thermosensitivity of single, isolated temperature-sensitive preoptic region neurons in an in vitro preparation without afferent modulation. The alterations in thermosensitivity by halothane were not accompanied by modifications in the threshold temperatures or thermosensitive temperature range. In addition, halothane differentially affects warm-sensitive and cold-sensitive neurons and only minimally alters firing of temperature-insensitive neurons in the preoptic region of the anterior hypothalamus.

It is widely believed that volatile anesthetics, including halothane, produce a poikilothermic state in which thermoregulatory inhibition results in deep core temperatures approaching the prevailing ambient temperature. It also has been suggested that under general anesthesia, thermoregulation still occurs, but at altered thermoregulatory thresholds or with a greater imprecision in thermoregulatory responses. Farber et al. have shown that volatile anesthetics abolish the thermoregulatory responses that typically occur during heating and cooling of the preoptic region in chronically prepared cats. Also, postanesthetic shivering during emergence was attenuated or augmented by preoptic region heating or cooling, respectively. Previous investigations have shown that general anesthetics disrupt normal thermoregulatory responses by inhibiting shivering and sweating, attenuating peripheral vasoconstriction, disrupting basal metabolism and altering behavioral thermoregulation. However, only one previous study has attempted to delineate the actions of halothane specifically on the CNS neurons that mediate and modulate these thermoregulatory responses. Poterack et al. examined the direct effects of halothane on thermosensitive neurons in the preoptic region of cats anesthetized with alpha-chloralose and urethane. Halothane administration produced progressive decreases in spontaneous firing rate and calculated thermosensitivity of preoptic region, warm-sensitive neurons without changing the periodicity of firing. These findings supported those of multiple clinical studies demonstrating the disrupting effects of volatile anesthetics on thermoregulation and provided the first evidence that these alterations are also observed in thermosensitive units within the preoptic region. However, in that previous study, possible confounding effects of additional baseline anesthetics and the extrahypothalamic afferent thermal and nonthermal input into the CNS or on other thermoregulatory centers did not allow definitive conclusions to be drawn about the specific effects of halothane directly on these thermosensitive neurons. In addition, only one cold-sensitive neuron was completely evaluated in that study and thus differential effects on warm and cold-sensitive neurons could not be evaluated.

To evaluate the direct role of halothane on preoptic region neurons, the current studies were performed in tissue slices devoid of distant afferent inputs, yet the neurons remained embedded in and affected by their normal microenvironment (neuronal and glial framework). This technique obviates the need for additional baseline anesthetics and allows precise localization of brain regions and subsequent placement of recording electrodes, which is not possible with in vivo studies. Numerous investigations have demonstrated that hypothalamic tissue slices contain the same proportions of warm-sensitive, cold-sensitive, and temperature-in-sensitive neurons as those recorded in intact animals.

While preoptic region neurons receive extensive thermal afferent input, they have also been shown to have an integral role in osmolar, glucose, steroid regulation, and cardiovascular pressure/volume responses. Further, it is not only the thermoinsensitive neurons that interact with other regulatory systems, but nearly half of the thermosensitive neurons respond to nonthermal stimuli. As an integrative structure important in the control of several homeostatic systems including body temperature regulation, and fluid and metabolite balance, the direct effects of volatile anesthetics on the preoptic region is highly consequential. The actions of halothane on preoptic region neurons are thus important both in terms of mechanisms of thermoregulatory control as well as a potential model for cellular actions of volatile anesthetics on other regulatory or neuronal functions that may be more difficult to examine in vitro.

Three populations of preoptic region neurons were identified in this study using well-accepted criteria: warm-sensitive, cold-sensitive, and temperature-insensitive. Further subsets of thermosensitive neurons have been previously characterized, including primary preoptic region thermodetectors with linear relationships between firing rate and temperature, those with thermosensitivity above or below a certain threshold, and those with bell-shaped thermosensitive curves. While all of these types of firing responses were observed in the current investigation, there were too few in each classification to allow evaluation of the differential effects of halothane on these subpopulations of preoptic region neurons. The results of the current study demonstrate that halothane differentially disrupts warm and cold neuronal thermosensitivity, in that cold-sensitive neurons were more susceptible to the effects of lower halothane concentrations.

It has been suggested that neurons with the highest firing rates are most thermosensitive in the hypothermic range, may be most influenced by extrahypothalamic temperature changes and are primarily responsible for heat production responses. In contrast, peripheral temperatures may have limited influence on heat-loss responses, which are regulated by hypothalamic neurons receiving relatively little synaptic input or the input is primarily inhibitory in nature. While a single central thermoregulatory integrator with multiple inputs and output may exist, a multiple thermostat system, in which warm and cold afferent information elicit responses from separate neural networks, has been postulated. Within this framework, warm-sensitive neurons would subserve heat loss mechanisms while cold-sensitive neurons may modulate heat gain mechanisms. Several studies have shown that the local application of various pharmacologic agents results in changes in the activity of thermosensitive preoptic region neurons that are appropriate to those observed in whole body thermoregulatory responses. Microinjection of a variety of pyrogens into the preoptic region attenuates heat loss and augments heat production responses. Similarly, local application of these substances inhibited warm-sensitive neuronal firing and excited cold-sensitive preoptic region neurons. Conversely, substances such as serotonin and Delta⁹-tetrahydrocannabinol, which evoke hypothermic responses in terms of whole body thermoregulation, increased the thermosensitivity of heat-sensitive neurons, while diminishing both the thermosensitivity and spontaneous firing of cold-sensitive neurons. .

Results of the current investigation support the findings of Poterack et al. that halothane attenuates the thermosensitivity and spontaneous firing rate of warm-sensitive preoptic region neurons in intact anesthetized animals. Also, the current finding that cold-sensitive neurons are more sensitive to the inhibitory effects of halothane suggests a mechanism for the clinical observation that warm defense thresholds may not decrease to the same extent as cold defense thresholds. Similarly, the halothane-mediated attenuation of both cold- and warm-sensitive neuronal firing and thermosensitivity may be associated with the widening of the interthreshold thermoregulatory range seen clinically during the administration of the volatile anesthetics. However, several clinical studies have demonstrated that although halothane disrupts threshold temperatures, thermosensitivities and maximum response intensities remain relatively well preserved. The reason for the differences between those of Sessler and colleagues and those of the current study are unclear, but may be related to the lack of distant afferent input in our study, the possibility that the altered thermosensitivity (gain) and firing rates in isolated neurons does correlate with a wider threshold range in an intact organism, or that thermomodulatory changes by halothane are also mediated at centers other than the anterior hypothalamus. Halothane has been shown to dose-dependently attenuate thermoregulatory responses to spinal cord cooling in brain-stem-transacted cats, suggesting a spinal cord or lower brain stem site of thermomodulatory action. Studies have previously demonstrated convergence of thermal information at mid-brain and medullary areas where thermoregulatory, vasomotor, respiratory, and somatic responses may be initiated and/or integrated.

Neuronal thermosensitivity may be either an intrinsic excitability or a characteristic dependent on local synaptic circuits. While the firing rate of some warm-sensitive preoptic region neurons is inhibited by synaptic blockade, the thermosensitivity of all warm-sensitive preoptic region neurons is preserved, suggesting that these neurons possess an intrinsic thermosensitivity. In contrast, cold-sensitive preoptic region neuronal thermosensitivity was abolished during synaptic blockade, indicating that cold sensitivity may be due to local synaptic input and inhibition from local warm-sensitive neurons. The differential effects of halothane on these neuronal subpopulations may be related to these specific characteristics. Analyses of thermosensitive neurons in other diencephalic regions have shown that populations of both intrinsically and synaptically mediated warm- and cold-sensitive neurons exist. .

Previous findings using extracellular recordings have been confirmed with intracellular analysis of warm- and cold-sensitive neurons. Warm-sensitive neurons exhibit a slow, depolarizing temperature-dependent prepotential, which disappeared during hyperpolarization, suggesting intrinsic thermosensitivity. Indeed, a halothane-mediated hyperpolarization may be the mechanism for the attenuation of thermosensitivity. Activity of cold-sensitive preoptic region neurons correlated with excitatory and inhibitory postsynaptic potentials, suggesting synaptic-dependent thermosensitivity. Such an interpretation, that cold-sensitive, preoptic region neurons represent synaptically mediated interneurons, is supported by the observation that halothane produced greater and more prolonged depressant effects on these cells, as compared to effects on warm-sensitive neurons, in as much as halothane has been previously demonstrated to attenuate synaptic transmission in some but not all, neuronal models. .

The prolonged effects of halothane on cold-sensitive neurons, and thus an attenuation of heat gain rather than heat loss responses, may be related to the findings of Sessler and coworkers that thermoregulatory vasoconstriction was sufficient to reduce the magnitude of core cooling, despite a continually decreasing body heat content. If postanesthetic shivering were mediated at the level of the anterior hypothalamus, one would predict a hypersensitivity of cold-sensitive neurons or a prolonged depressant effect on warm-sensitive neurons after discontinuing halothane. In contrast, our results provide further evidence that postanesthetic tremor may arise as a result of either disinhibition of, or direct action on other CNS or spinal cord thermoregulatory sites.

In contrast to the modulation of thermosensitive neurons, halothane did not alter the firing frequency of thermoinsensitive neurons. While some temperature-insensitive neurons may display warm-sensitive characteristics when depolarized, temperature-insensitivity appears to be maintained by hyperpolarization and a Sodium sup +-Potassium sup + pump. While the functional importance of temperature-insensitive neurons is poorly understood, early studies suggesting a primary role in modulating nonthermoregulatory homeostatic systems, is unlikely because these neurons do not represent a majority of osmosensitive, glucosensitive, steroid-sensitive, or baro/volume sensitive neurons. .

In summary, various intravenous anesthetics have been found to decrease the thermosensitivity and firing rate of warm-sensitive preoptic region neurons and cold-sensitive neurons in the reticular formation. The results of the current study show that halothane alters the firing rate and thermosensitivity of individual temperature-sensitive preoptic region neurons. This modulation of neuronal sensitivities may result in an imprecision of thermoregulatory responses to thermal challenges. In addition, these findings strongly suggest that effects on preoptic region neurons represent an important mechanism by which volatile anesthetics disrupt normal thermoregulatory responses. Further studies are needed to delineate whether halothane alters the thermosensitivity of these cells by direct cellular action or an effect on neurotransmission.

The authors thank David Schwabe, B.S.E., for technical assistance, and Dr. Pragati Ganjoo, for aid in data analysis.

*Guide for the Care and Use of Laboratory Animals. DHEW (DHHS) publication no. (NIH) 85–23, revised 1985.