| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Neurosurgery, 2 Department of Neuroscience, and 3 Department of Anesthesiology and Critical Care Medicine, The Johns Hopkins Medical School, Baltimore, Maryland 21287-7509
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Dougherty, P. M., Y. J. Li, F. A. Lenz, L. Rowland, and S. Mittman. Correlation of effects of general anesthetics on somatosensory neurons in the primate thalamus and cortical EEG power. J. Neurophysiol. 77: 1375-1392, 1997. The effects of two types of general anesthetic on the neurophysiological properties of the primate somatosensory thalamus were correlated with effects on frontal cortex electroencephalographic (EEG) power and spectral properties. Graded doses of the intravenous agent methohexital sodium (METH) were studied in 12 cells in three monkeys on a halothane baseline anesthetic. Low doses of METH (0.2-1.0 mg/kg) produced a reduction of EEG power but had no effects on spontaneous or evoked thalamic activity. EEG power showed maximal attenuation after 2.0 mg/kg METH, whereas decreases in thalamic activity were first noted over a similar moderate dose range (2.0-5.0 mg/kg). The physiological parameter most sensitive to METH was the spontaneous activity, which showed initial changes in rate and pattern at moderate doses followed by marked inhibition at higher doses. Finally, the high dose of METH (10.0 mg/kg) produced marked reduction in all neurophysiological parameters with recovery over the following 30-45 min. The effects of the volatile anesthetic halothane were studied on 15 cells in four monkeys anesthetized with pentobarbital sodium. The low dose of halothane (0.25%) produced a facilitation of responses to cutaneous stimuli as well as a decrease in the rate and burst patterns in the spontaneous activity. The power in the EEG was not affected at this concentration. The responses of the cells to the mechanical stimuli at moderate doses (0.5-1.0%) of halothane returned to the baseline magnitude, whereas spontaneous activity remained unaffected compared with initial effects. EEG power was reduced by 1% halothane. Finally, all neurophysiological parameters showed profound reduction at the highest halothane concentrations (2.0-3.0%) with recovery over the next 30-45 min. In conclusion, the two classes of anesthetics most commonly used for acute neurophysiological studies in the primate show well-defined thresholds at which changes in the response properties of thalamic neurons are produced. This threshold for the barbiturates and halothane can be predicted by monitoring of cortical EEG.
Neurophysiological studies conducted in anesthetized animals have contributed to our understanding of the mechanisms underlying somatosensation over the past several years. Nevertheless, anesthesia presents the potential to confound the interpretation of a single experiment as well as the more general application of the findings on the understanding of sensory mechanisms in awake animals. For example, although primary afferent fibers are relatively resistant to the effects of pentobarbital sodium, chloroform, ether, urethan, and alcohol at doses that produce surgical anesthesia (Larrabee and Posternak 1952 Surgical procedures
All procedures were consistent with the guidelines of the International Association for the Study of Pain and the National Institutes of Health guide for the care and use of laboratory animals. Adult male monkeys (Macaca fascicularis) between 5.0 and 10.0 kg were tranquilized with ketamine (10.0 mg/kg im) during transportation to the laboratory and insertion of an intravenous canula. Surgical anesthesia was then either induced by an intravenous dose of pentobarbital (10.0 mg/kg) and later maintained by intravenous infusion of pentobarbital (5.0 mg·kg Recording procedures
Extracellular activity of thalamic neurons was monitored with a low-impedance (500 k Cutaneous stimuli
After collection of 2 min of spontaneous activity, the responses of all cells to a series of mechanical stimuli applied to the skin were determined. Before application of these stimuli, crude RF boundaries were mapped with innocuous and noxious mechanical stimuli with the use of a camel hair brush, finger taps, and brief applications of arterial clips to the skin. Depending on RF size, anywhere from two to five test points spanning the RF were chosen and marked with ink for delivery of the cutaneous stimuli. One site near the center of the RF (most responsive site) was chosen for all cells. If the RF size permitted, two other sites, one proximal and one distal, within the RF were also selected. Finally, one or two sites on or very near the initial boundaries of the RF (minimally responsive sites) were chosen. All mechanical stimuli were then applied to all sites throughout the experiments on each cell.
Anesthetic administration
After the baseline recordings the animals were exposed to graded administrations of either the intravenous anesthetic methohexital sodium (METH) or the volatile anesthetic halothane. The animals maintained on a baseline of halothane received bolus infusions of methohexital at 0.2, 1.0, 2.0, 5.0, and 10.0 mg/kg at intervals of ~30 min. The pentobarbital-maintained animals received halothane by inhalation at 0.25, 0.50, 1.0, 2.0, and 3.0% vol/vol concentrations. The responses to all cutaneous stimuli were recorded over a period of 2-5 min following each METH dose, whereas the responses to the mechanical stimuli after each halothane dose were recorded over a similar time span but after also allowing 15 min of equilibration for the new gas concentrations. After the recording of responses at the highest anesthetic dose a recovery period was allowed until mean EEG frequency and spectral edge returned to the baseline level, and responses to the cutaneous stimuli were again collected.
Data analysis
MECHANICAL RESPONSES.
The stored digital record of unit activity was retrieved and analyzed off-line. Accumulated frequency histograms were generated for the responses to all sensory stimuli. Background activity was calculated as a mean value (and analyzed in detail as described further below) and subtracted from the responses obtained after each stimulus. The responses evoked at each cutaneous site were broken into 100-ms segments for the entire 10-s duration of each stimulus. Significance of responses to each sensory stimulus at each site was determined by comparisons between the segment values during 10 s of spontaneous activity compared with the equal number of segments obtained at each response (Student-Neuman-Keuls test). Cells were defined as NS when statistically significant responses to the small clip but not to brush or the large clip were found. LT cells were defined as those with a significant increase in firing to brush but not to the small clip. WDR cells were defined as those with a significant increase in firing to all three mechanical stimuli and a significant increase in firing rate with increasing stimulus intensity. MR cells were defined as those with a significant increase in firing to all three mechanical stimuli compared with background but without a significant increase with graded stimulus intensity.
EEG POWER AND SPECTRAL ANALYSIS.
A 1-min block of EEG activity at the baseline recording and at each treatment period was collected before application of any sensory stimuli. A periodogram of the power in each signal (sample frequency 200 Hz, high-pass filter open, low-pass filter 400 Hz, no decimation) was generated by fast Fourier transform (Welch method, 8 segments, 4,096 points). Total power, median frequency, spectral edge frequency, and the power in each band were calculated. The ranges used to define each band were as follows: very low: 0-1 Hz; ANALYSIS OF SPONTANEOUS ACTIVITY PATTERNS.
Stationarity of each signal was first determined by the median runs test. After computation of firing rate and interspike interval (ISI) histograms, spontaneous activity was analyzed first to determine whether cells were undergoing bursting activity and to determine the character of that bursting activity. The presence of two peaks in the ISI histogram was taken as preliminary evidence that the cell was firing in bursts of action potentials. Higher-order ISI histograms were then computed. An nth-order histogram measures the interval between n + 1 consecutive action potentials so that high-frequency bursts containing more than n action potentials were identified by looking at nth-order interspike histograms (Lenz et al. 1994 Effects of anesthetic baseline on physiological organization of thalamus
The nonhuman primate thalamus has previously been shown in awake animals to be physiologically organized such that cells responding exclusively to stimulation of deep tissues, joints, and limb movements are anterior to regions of cells responding to cutaneous stimuli (Kaas et al. 1984
Effects of anesthetics on responses to mechanical stimuli
EFFECTS OF METH.
Figures 2-4 show representative responses of two cells and a summary of the responses for all cells tested after infusion of bolus doses of METH while the animals were maintained on a background of halothane. In Fig. 2 the responses of two representative neurons, one an NS cell the other an LT cell, to graded doses of METH are illustrated. The responses of the NS neuron are shown in Fig. 2, left. The RF of the NS cell was a small area confined to the tip of digit 4 as shown on the drawing of the limb in Fig. 2, bottom left. As illustrated in the top horizontal line of histograms, the cell showed no or very little response to the brush stimulus at either sites 1 or 2, but the compressive stimuli elicited clear responses from site 1 only. Site 2 was therefore considered as outside the RF of this cell. The horizontal lines of histograms going down show the responses to the same stimuli 2-5 min after three graded doses of methohexital. Comparing between lines it can be seen that the doses tested had little effect on the responses of this cell. The only noteworthy effect is the modest reduction in spontaneous discharge rate following 2 mg/kg methohexital (compare the 1st 10 s of line 4 with the 1st 10 s of other lines).
EFFECTS OF HALOTHANE.
Figure 5 shows a representative cell illustrating the effects of graded concentrations of halothane on the responses to mechanical stimuli, whereas the effects of halothane on all cells are shown in Figs. 6 and 7. The representative cell shown in Fig. 5 had spontaneous activity when isolated that averaged 2.8 spikes/s and responses to all mechanical stimuli when applied to sites on the posterior aspect of the left hindlimb as illustrated in Fig. 5, bottom. The cell responded best to innocuous brush, but also showed modest responses to the compressive stimuli when they were applied to the center of the RF (sites 2 and 3). Administration of 1% halothane for 15 min produced a small decrease in the spontaneous discharge rate of the cell to 2.4 spikes/s, but produced a noticeable increase in the discharges evoked by the cutaneous stimuli. Of note, most cells showed these increases in responses to the cutaneous stimuli at lower doses of halothane (see below). When the halothane concentration was raised to 2% for 15 min, the spontaneous activity became nearly absent, now averaging only 0.98 spikes/s. Similarly, the responses to the mechanical stimuli, although present at all previously sensitive sites, showed clear reductions from the baseline magnitudes. The decrease in activity became profound after administration of 3% halothane for 15 min, so that spontaneous activity (0.27 spikes/s) and the responses to the mechanical stimuli were all but eliminated. The responses to the mechanical stimuli showed complete recovery 30 min after termination of the halothane administration; however, the spontaneous discharge rate did not resume over the recovery period for this cell.
Effects of general anesthetics on cortical EEG power
EFFECTS OF METH.
EEG periodograms showing representative effects of graded doses of METH are presented in Fig. 8A. The gross EEGs in all baseline recordings were composed of large-amplitude slow-frequency waves with superimposed high-frequency components. The baseline periodogram reflects this by the presence of significant power over the frequency range of 0-50 Hz. The lowest dose of METH had little apparent effect on either the gross EEG or power distribution in the periodogram. The gross EEG following the 1.0-mg/kg dose showed an absence of high-frequency waves and periodic halts in slow wave activity consistent with the occurrence of burst suppression. The periodograms reflected these effects with noticeable decreases in power in the 5- to 20-Hz and 30- to 50-Hz bands. METH at 2.0 mg/kg produced an essentially flat EEG that is reflected by the sharp decrease in the entire power distribution in the periodograms. The level of decrease in the EEG power produced by this amount of METH was not exceeded by higher doses of METH (see below). All effects of METH showed recovery to baseline levels 30-45 min after the final METH dosage.
EFFECTS OF HALOTHANE.
Periodograms showing representative effects of halothane are presented in Fig. 8B. The animals in this series, as in the previous one, showed gross EEG patterns composed of both low-amplitude high-frequency waves and high-amplitude low-frequency waves. The periodograms showed power concentrated over the frequency range of 0-50 Hz. However, comparison of the scales indicates that the baseline EEG power for animals on the pentobarbital baseline (later receiving graded halothane) was considerably lower than that for animals held on the halothane baseline (later receiving graded methohexital doses). This difference in baseline states can also be seen by examination of the summary figures (Fig. 8, D-F). The lowest concentrations (0.25 and 0.50%) of halothane did not produce marked changes in either gross EEG or in the periodograms for power (not illustrated). At 1% halothane the gross EEG showed the first noticeable reduction in high-frequency waves, and examination of the periodograms over the middle-frequency ranges (15-40 Hz) also shows a noticeable decrease in EEG power. The gross EEG at 2% halothane showed an absence of high-frequency waves and periods of loss in the slow wave activity consistent with burst suppression. These changes are noted in the periodograms by the decreases in the distributions of power above 20 Hz. Finally, the gross EEG was essentially flat at 3% halothane, which is also reflected by the marked decrease in EEG power noticeable across the periodograms of the whole frequency band. Recovery of EEG power to near the baseline level was shown 45 min after termination of halothane administration.
Effects of the general anesthetics on the structure of spontaneous discharge activity in the somatosensory thalamus
EFFECTS OF METH.
As noted in the first section of RESULTS, administration of graded doses of METH ultimately resulted in an attenuation of mean spontaneous discharge rate of the cells in the primate sensory thalamus. However, a thorough analysis of the effects of anesthetics on the spontaneous activity of cells in the primate thalamus requires more than just an examination of the overall discharge rate; rather, it requires an analysis of the discharge pattern. For example, neurons in thalamus have two modes of discharge, the relay and burst modes, and the burst mode may be composed of a particular type of burst associated with low-threshold calcium spikes (Deschênes et al. 1984
EFFECTS OF HALOTHANE.
The ISI analyses for the cells in the halothane series also showed that all baseline recordings and all but one of the recordings obtained after halothane administration showed non-Poisson distributions. One cell showed conversion to a Poisson distribution after administration of 0.50% halothane, sustained this pattern with increases in halothane to 2.0%, and then resumed a non-Poisson pattern in the recovery period. Further examination shows that the percentage of cells with first-order ISIs occurring in the range of 5-10 ms was 15-20% of the population, which was much lower than that observed for the cells in the METH series. Similarly, the percentage of second-order ISIs occurring in the range of 10-15 ms was also somewhat less than in the METH series. The halothane series was further different, in that as the concentrations of halothane were increased the overall tendency shown by the cells studied was to increase the percentage of cells with short first- and second-order ISIs. Indeed, the percentage of cells with first-order ISIs <5 and 10 ms was significantly greater than baseline when the animals were maintained at 2 and 3% halothane. The increase in the percentage of short first-order ISIs was not fully reversed in the recovery period.
One main finding of this study was that anesthesia with pentobarbital versus halothane did not affect the general organization of response profiles of neurons in the primate thalamus. Additionally, the proportions of each type of thalamic neuron (LT, WDR, MR, NS) were equivalent under both types of baseline anesthetics. In each group, the most anteriorly located cells responsive to sensory stimuli were activated by stimulation of deep tissues or by movement of the limbs. Posterior to this, a layer of cells responsive to deep stimuli was observed in the dorsal part of the thalamus, with cells responsive to innocuous cutaneous stimuli more ventral. Finally, in the most posterior areas of the thalamus we found a mixture of cells responsive to both innocuous cutaneous stimuli alone (LT cells) and cells responsive to noxious cutaneous stimuli (WDR, MR, and NS cells). The most medial regions of the thalamus were characterized by cells responsive to stimulation of the face and perioral body surfaces, whereas cells in the most lateral regions responded to stimulation of the hindlimb and tail. This type of organization in the macaque thalamus is consistent with earlier reports from anesthetized preparations (Mountcastle and Henneman 1949
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
), the responses of these fibers are increased by comparable concentrations of halothane (Campbell et al. 1995). As reviewed below, the literature demonstrates that cells at each level of the somatosensory system may show even more pronounced effects of anesthesia.
, 1970), and monkeys (deJong and Wagman 1968). Thiopental and propofol produced many of the same effects as halothane on cat dorsal horn and medullary neurons over a dose range of 2.5-10.0 mg/kg body wt (Kawahara et al. 1982; Kishikawa et al. 1995; Kitahata et al. 1975; Uchida et al. 1995). However, complicating the picture, nitrous oxide had little effect on spontaneous activity while producing a decrease in cutaneous evoked responses of primate spinal neurons (deJong and Wagman 1968); and still others found little effect of supplemental doses of pentobarbital on receptive field (RF) properties of monkey dorsal horn neurons until after administration of excessive doses (10-20 mg/kg) (Hori et al. 1984). Complicating the issue further, in some cases responses to noxious thermal and mechanical stimuli of spinal neurons in cats could be unmasked by administration of anesthetic doses of pentobarbital (Collins and Ren 1987; Collins et al. 1990).
reported that RF characteristics of primate thalamic neurons did not change with increasing anesthetic depth, whereas Perl and Whitlock (1971) reported a decrease of responses of these same cells at deeper levels of anesthesia. The responses of cat thalamic neurons and nociceptive rat thalamic neurons were suppressed by increasing doses of chloralose, pentobarbital (Harris 1978), or urethan (Guilbaud et al. 1981). However, it was also reported that RF sizes and responses of low-threshold rat thalamic neurons are increased in parallel with anesthetic depth (Guilbaud et al. 1981). Finally, a recent report indicated that the frequency of nociceptive specific cells sampled in the thalamus of rats was increased for animals under pentobarbital versus awake animals (Montagne-Clavel and Oliveras 1995), whereas no such differences in the types of raccoon thalamic cells were found under varying anesthetic states (Simone et al. 1993).
; Armstrong-James and George 1988; Collins and Roppolo 1980; Diamond et al. 1992) and monkeys (Duncan et al. 1982) show decreases in spontaneous activity and RF size with increasing anesthetic depth. Anesthetic-induced effects in cat primary somatosensory cortex were reported to be less with chloralose than with pentobarbital (Harding et al. 1979). Others (Chapin and Woodward 1981) reported that strong arousal produced a slight increase in responses of primary somatosensory neurons in awake rats, and that awake animals had a higher level of spontaneous activity than anesthetized animals (Nicolelis and Chapin 1994). These results contrast with those of ethanol, which had little effect on the properties of primary somatosensory cortical neurons up to a dose of 2.5 g/kg body wt (Collins and Roppolo 1980).
; Jessop and Jones 1992; Nayak et al. 1994). So we sought to determine in this study whether EEG power or spectral properties of the frontal cortex ipsilateral to the thalamic nucleus under study could be used as a tool for predicting anesthetic effects on thalamic neurons in the laboratory.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
1·h1), or induced by a bolus dosage of methohexital (5.0 mg/kg) and later maintained by inhalation of halothane (1.5-2.0%). A surgical level of anesthesia was determined by establishment of areflexia to noxious stimulation. Once anesthesia was deep, the monkeys were paralyzed with pancuronium bromide (0.1 mg/kg) and artificially ventilated. Throughout the remainder of the surgical preparation this level was periodically checked by examining pupillary size and continuously monitored by recording expired CO2, blood pressure, and heart rate. End-tidal CO2 was kept between 3.5 and 4.5%, and core temperature was maintained between 37 and 38°C by a servo-controlled heating blanket. The animals were mounted in a stereotaxic apparatus and a partial craniectomy was made. At the conclusion of the surgical procedures, the rate of anesthetic administration was reduced to establish the recording baseline. The infusion of pentobarbital was set at 3-5.0 mg·kg1·h1 and the halothane concentration was set to 0.75-1.0% (1 minimum alveolar concentration). Adequacy of this baseline was established by occasionally allowing the neuromuscular blockade to dissipate and ensuring areflexia to all stimuli. At all other times adequacy of the baseline level was determined by ensuring the stability of expired CO2, blood pressure, and heart rate.
to 1 M) glass-coated platinum-iridium electrode. Unit activity was amplified and observed on storage oscilloscopes and fed to a computer running a neurophysiological data acquisition and analysis package (Brainwave Systems, Discovery package, Version 3.0). Unit activity was discriminated from background and stored for later analysis. Spike size and configuration were continuously monitored to confirm that activity of the same cell was recorded throughout the experiment. The EEG was recorded by a stainless steel monopolar electrode placed 5 mm into the frontal cortex (apex of the genu of the arcuate gyrus). This region of cortex was chosen primarily because of the shift in dominant EEG power to the frontal regions from the posterior regions of cortex in monkeys as they enter "surgical" levels of anesthesia (Tinker et al. 1977) and thus ensured the recording of a "maximal" cortical signal. Also, the anatomy of the arcuate gyrus allowed placement of the EEG electrode at a precise, consistent location in each experiment. The EEG signal was also digitized and stored on computer for later analysis. A taped backup of all data was collected (Vetter model 4000 PCM recording adapter and Panasonic model VRG260 VCR).
: 1-3.5 Hz; 
: 3.5-7.5 Hz; 
: 7.5-12.5 Hz; sleep spindle: 12-14 Hz; 
1: 14-17.5 Hz, 2: 17.5-25 Hz; very high: 25-100 Hz. The changes in responses for individual EEG recordings were based on inspection of the raw traces alone without statistical analysis. The effects for the population were calculated by the use of the values for each recording as detailed above combined into treatment groups. Differences were then calculated with the use of the Wilcoxon test.
; Perkel et al. 1967). A quantitative measure of high-frequency bursting activity was derived by calculating the percentage of ISIs in nth-order histograms that were <5 ms. If bursting activity was identified in this manner, subsequent tests were carried out to determine the characteristics of that activity.
5 ms) and progressive lengthening of interspike times throughout the burst (Deschênes et al. 1984; Jahnsen and Llinás 1984; Roy et al. 1984). These bursts were identified by a computerized algorithm on the basis of the durations of a sequence of ISIs (Lenz et al. 1989, 1994; McCarley et al. 1983). Indexes of burst firing such as rate of bursting, numbers of action potentials in a burst, duration of the first ISI in a burst, duration of the burst, etc. were determined for individual cells (Domich et al. 1986). Finally, the techniques of primary event analysis were applied to determine whether the increase in bursting activity of cells was accompanied by an increase in nonburst firing of these cells (McCarley et al. 1983).
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
; Poggio and Mountcastle 1963). Our results and those of previous investigators have shown that this basic organization is not altered when studied in the anesthetized preparation (Friedman and Jones 1981; Kenshalo et al. 1980;Mountcastle and Henneman 1952; Perl and Whitlock 1971). Cells with nociceptive inputs in primates have been reported as scattered throughout the cutaneous receptive zone (Kenshalo et al. 1980) but more often as concentrated toward the posterior-inferior zone of the cutaneous receptive area (Lenz et al. 1988). Figure 1 summarizes the basic physiological organization of thalamic cells consistent with each of these previous studies as we observed under both pentobarbital and halothane anesthesia. An additional note to this basic organization uncovered in this study was that the distance anterior to the intraural zero point at which cells responding only to innocuous cutaneous stimuli were located (the ventroposterolateral nucleus core) correlated with body weight of the animals. The core region of ventroposterolateral nucleus in the smallest animals studied (5-6 kg) was located 8.0 mm anterior to the intra-aural zero, whereas in larger animals this site became more anteriorly displaced. In addition, we discovered that the baseline anesthetics did not have any influence on the frequency with which cells of different classification were encountered. As detailed in Table 1, the proportions of LT, WDR, MR, and NS cells were not different between the two groups as determined with a 
2 analysis.
View larger version (50K):
[in a new window]
FIG. 1.
Summary of the somatotopic organization of the primate thalamus as mapped with the animals under light barbiturate (pentobarbital sodium) and halothane anesthesia. Heavy vertical lines on the outlines of the thalamic nuclear boundaries (A-C) indicate the positions of 3 representative tracks each for barbiturate-maintained ( 
) and halothane-maintained (· · ·) animals at different anteroposterior distances through the thalamus. D: anterior-posterior coordinates of the location of the most anterior trajectory in which units responding to innocuous touch were identified. Shaded regions on the drawings in E (pentobarbital animals) and F (halothane animals) indicate the types of receptive fields (RFs) that were encountered; numbers by the drawings indicate the depth from the surface of the brain at which the corresponding neurons for these RFs were recorded. LT, low-threshold (brush only) responses; MR, multireceptive neurons; WDR, wide-dynamic-range neurons, NS, nociceptive specific neurons.
View this table:
TABLE 1.
Frequencies of thalamic cell types sampled with pentobarbital versus halothane anesthesia
View larger version (29K):
[in a new window]
FIG. 2.
Responses of a representative NS and LT thalamic neuron to a series of mechanical stimuli as graded doses of methohexital sodium (METH) are given on top of a halothane anesthetic background. The outlines of the limbs at the bottom of each set of histograms show the RFs for each cell. In addition, several superimposed traces of both spikes are shown at bottom of the figure. Top line: baseline record. Bottom line: record obtained after recovery. Middle lines of histograms: responses obtained after the indicated doses of METH. The stimuli were delivered to the sites indicated by the arrows or dots on the outlines of the limbs, and the times at which the stimuli were delivered are shown by the labeled horizontal lines in the histograms (bin size 1 s). Stimuli evoking the responses shown for the NS cell are indicated by the labels over the stimulus time bar. Responses shown for the LT neuron are to the brush stimulus only.
View larger version (25K):
[in a new window]
FIG. 3.
Summaries of the effects of graded doses of METH on the spontaneous activity and the responses of thalamic units to mechanical stimulation of the skin. A: bar graphs summarizing the mean (spontaneous activity) or total spikes/s (mean ± SE) evoked by each stimulus at each experimental condition such that the left group of bars shows the data for the spontaneous (Spont.) activity, whereas the groups of bars from left to right show the data for activity evoked by brush (BR), large clip (LC), medium clip (MC), and small clip (SC), respectively. Data at each treatment condition are shown by the bars of different shading patterns (see key at top right of the figure). B: spontaneous and evoked activity of each experimental condition expressed as mean ± SE of percentage of the baseline (100%) value. The data in B for each stimulus modality are summarized by the different symbol and line drawings (see key at top left of the figure); different doses of METH are indicated by the points on the X-axis. The data in C are spike rate data as in A, but here the data are organized so that effects of the various doses on the mean stimulus responses functions of the neurons can be observed. The responses to each stimulus appear in columns labeled across the bottom of the figure; each line and symbol plot correspond to data at different anesthetic concentrations. Points of statistical difference are only shown for the bar graphs. In each case the comparisons are between the responses at a given dose and the baseline value. Single asterisk: P < 0.05. Double asterisk: P < 0.01.
View larger version (42K):
[in a new window]
FIG. 4.
Symbol and line plots summarizing the effects of each drug treatment on the mean total spikes/s evoked by brush (A), the large clip (B), the medium clip (C), and the small clip (D) at each of the test sites in the RFs of the cells. The data were compiled by aligning each of the most responsive sites to the center position (C, on X-axis). P: sites proximal to the center. D: sites distal to the center. Lower numbers indicate sites nearest the center. The curves for each dose are keyed by the symbols in A, top left. Error bars have been omitted for clarity.
View larger version (27K):
[in a new window]
FIG. 5.
Rate histograms illustrating the responses of a representative MR neuron to mechanical stimulation of the skin before and then after graded concentrations of halothane. The outline of the limb at the bottom of the set of histograms shows the RF for this cell (· · ·). In addition, several superimposed traces of the spike are shown at the bottom right of the figure. The histograms are arranged in 4 sets corresponding to the responses after the level of halothane indicated. In each histogram set, the top line shows the responses to brush, the 2nd line to the large clip, the 3rd line to the medium clip, and the 4th line to the small clip. Filled circles on the outline of the limb: sites to which stimuli were delivered. Labeled horizontal lines: times at which the stimuli were delivered. Bin size 1 s.
View larger version (25K):
[in a new window]
FIG. 6.
Summaries of the effects of graded doses of halothane on the spontaneous activity and the responses of thalamic units to mechanical stimulation of the skin. A: bar graphs summarizing the mean (spontaneous activity) or total spikes/s (mean ± SE) evoked by each stimulus at each experimental condition such that the left group of bars shows the data for the spontaneous (Spont.) activity whereas the groups of bars from left to right show the data for activity evoked by the brush, large clip, medium clip, and small clip, respectively. The data at each treatment condition are shown by the bars of different shading patterns (see key at top left). B: spontaneous and evoked activity of each experimental condition expressed as mean ± SE of percentage of the baseline (100%) value. The data in B for each stimulus modality are summarized by the different symbol and line drawings (see key at top left); different doses of halothane are indicated by the points on the X-axis. The data in C are spike rate data as in A, but here the data are organized so that effects of the various doses on the mean stimulus responses functions of the neurons can be observed. The responses to each stimulus appear in columns labeled across the bottom of the figure; each line and symbol plot correspond to data at different anesthetic concentrations. Points of statistical difference are only shown for the bar graphs. In each case the comparisons are between the responses at a given dose and the baseline value. Single asterisk: P < 0.05. Double asterisk: P < 0.01.
View larger version (38K):
[in a new window]
FIG. 7.
Symbol and line plots summarize the effects of each drug treatment on the mean total spikes/s evoked by brush (A), the large clip (B), the medium clip (C), and the small clip (D) at each of the test sites in the RFs of the cells. The data were compiled by aligning each of the most responsive sites to position 4. Lower numbers: more proximal sites. Higher numbers: more distal sites. Responses for each drug concentration are keyed in A, top left. Error bars have been omitted for clarity.
View larger version (40K):
[in a new window]
FIG. 8.
Summary of the electroencephalographic (EEG) data. Representative periodograms of the EEG power (Y-axis, log scale) vs. frequency (X-axis, linear scale) before and after graded METH are shown overlaid in A; those for halothane are shown in B. The treatment conditions illustrated are keyed by the line types. The line-symbol plots summarize the power in each band of the EEG for the METH (C) and halothane (D) experiments for each dose as a percentage of baseline power. The data for each band are indicated by the different symbol types keyed in D, bottom right. The data at each dose or concentration are indicated by the labels on the X-axis. E and F: line plots show the summarized effects of METH (circles and solid lines) and halothane (squares and dotted lines) on the mean ± SE total EEG power and on the mean ± SE spectral edge over each treatment condition. Points corresponding to the various METH doses are indicated by the labels on the X-axis across the bottom of the figure; points corresponding to the various halothane concentrations are labeled on the X-axis across the top of the figure. Single asterisk: P < 0.05.
-, -, -, and high-range frequency bands. Finally, at the highest dose (10.0 mg/kg) further marked decreases in power for each band were produced, all of which showed recovery to near the baseline values within 30-45 min.
-bands were comparable with that observed in the METH administration experiments. The differences between the gross EEG patterns and mean total power are explained by the relatively lower levels of power in the -, -, and high-frequency bands for the halothane-treated animals compared with the METH-treated animals. Halothane at a concentration of 0.25% produced a noticeable decrease in power in the 0- to 1-Hz and -bands. The next noticeable changes in power of the bands was observed in the -band at 1% halothane, and then in the - and -bands at 2%. A marked decrease in power for all bands was produced by 3% halothane. Interestingly, power in the low-frequency bands was reduced to the same values by the maximal halothane concentrations as by the maximum METH doses, whereas halothane had a smaller maximal effect on the high-frequency bands than did METH.
; Jahnsen and Llinás 1984; Roy et al. 1984; Steriade and Llinas 1988). Each of these states may be distinguished on the basis of discharge pattern.
15 ms. Thus a very significant proportion of the cells observed in the baseline records had ongoing high-frequency burst activity. At the lowest doses of METH a trend toward an increase in the percentage of first-order ISIs in the 5- to 10-ms window was observed. However, over the middle dose ranges a trend toward a decreasing percentage of first-order ISIs in the 5- to 10-ms range was observed, with a significant decrease in this percentage found following the 10.0-mg/kg dose. There was no obvious change in the frequency of second-order ISIs until the depression in frequency following the highest METH dose (10.0 mg/kg). The frequency of both first-order and second-order ISIs returned to the baseline values in the recovery period. This pattern of ISI changes suggests an increase in bursting with increasing anesthetic concentration.
View larger version (24K):
[in a new window]
FIG. 9.
Line diagrams showing the mean ± SE for the percentages of 1st-order interspike intervals (ISIs) (1st ISI) and 2nd-order ISIs (2nd ISI) occurring in different time distributions of the spontaneous activity recorded at each of the experimental conditions for the METH (top, open symbols) and halothane (bottom, filled symbols) studies. Circles: percentage of 1st ISIs <5 ms. Squares: 1st ISIs <10 ms. Triangles: percentage of 2nd-order ISIs <10 ms. Diamonds: 2nd ISIs <15 ms.
). Specifically, these indexes defining a burst were the following: the interval preceding the first action potential in the burst had a duration of 20 ms; the interval after the first action potential had a duration of 6 ms, and the burst ended on an action potential preceding an ISI of 15 ms. The group mean for the total number of bursts per minute detected with this algorithm over the drug series is summarized in Fig. 10A. The baseline recordings were found on average to contain 31.3 bursts per minute. This number was significantly increased after the addition of METH over the doses of 0.2-2.0 mg/kg. However, a profound reduction in bursts as defined above was produced by 5-10.0 mg/kg METH. This inhibition of bursting was only partially restored over the recovery period.
View larger version (32K):
[in a new window]
FIG. 10.
Line diagrams showing the summaries for the spike train analyses from segments of the records of spontaneous activity at each treatment condition for the METH (filled symbols) and halothane (open symbols) experiments. For each panel the points corresponding to the data for each METH dose are shown by the labels on the lower X-axis; points for the various halothane concentrations are shown by the points labeled on the upper X-axis. LTCS, low-threshold calcium spike; Base, baseline; Recov, recovery.
; Jahnsen and Llinás 1984; Roy et al. 1984). The indexes used to define this type of burst were also as used previously in studies of human thalamus (Lenz et al. 1994). The burst patterns were analyzed for an initial period of inhibition followed by a high-frequency pattern that was followed by decelerating frequency of the spikes (Domich et al. 1986; McCarley et al. 1983). This type of bursting was present in 23% of the baseline records and increased to 67% after 0.2 mg/kg METH. The incidence of the low-threshold calcium spike patterns was more sensitive to inhibition by additional barbiturates than overall bursting, because this property was observed in only 39% of the cases after 1.0 mg/kg METH and in only 31% of the cases after 2.0 mg/kg METH. Low-threshold calcium spike patterns became nearly absent after 5- and 10.0-mg/kg boluses of METH, but returned to near baseline levels in the recovery period.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
, 1952; Perl and Whitlock 1971), as well as with results later described in the unanesthetized monkey (Loe et al. 1977) and human (Lenz et al. 1988).
) with only few exceptions (Chung et al. 1986). An explanation for the difficulty in finding this type of cell in the primate thalamus is particularly elusive when one considers that spinothalamic cells identified by antidromic activation from this same region are almost universally nociceptive. For example, a series of >300 spinothalamic neurons recorded over several years' time yielded only 4 cells that were classified as nonnociceptive (P. M. Dougherty and W. D. Willis, unpublished observations). It is possible that NS neurons in the thalamus represent a physically small-sized cell population that is poorly sampled with our choice of microelectrodes. Alternatively, the NS cell population is not found in the ventral posterior region proper, but instead may be concentrated in a recently described more medially located thalamic nuclear group (Craig et al. 1994).
; Jahnsen and Llinás 1984; Roy et al. 1984). It was only after high methohexital doses that all activity was inhibited. Halothane over the barbiturate baseline, on the other hand, produced a unimodal suppression of overall background rate and bursting. Thus barbiturates appear to produce increased bursts, whereas halothane reduces these events.
-, sleep spindle, and very high-frequency bands until the highest doses. In contrast, supplemental halothane tended to suppress the middle-frequency band powers and spared the very low- and very high-frequency bands until the highest doses. This clear sparing of high-frequency power is most likely the reason supplemental halothane failed to affect the spectral edge, and further suggests that the baseline level of barbiturate suppressed the high-frequency bands whereas the halothane baseline left these intact. Thus the two classes of anesthetics investigated in this study had remarkable differences in almost all neurophysiological parameters we measured until very high doses of each were given resulting in global suppression of all indexes.
) and thalamic units in the monkey after increasing doses of pentobarbital (Poggio and Mountcastle 1963). Similarly, the decrease in spontaneous activity followed by decrease of the discharges evoked by cutaneous stimulation is consistent with the effects of thiopental on spinal cord neurons in cats (Kitahata et al. 1975). However, in one detail our results differ from those previously reported by Hori et al. (1984) on primate spinothalamic neurons: we never found any enhancement of responses to mechanical stimuli after low doses of METH as they did with pentobarbital. In fact, the observation of Hori et al. is most consistent with the observation that barbiturates lower pain thresholds in humans until an anesthetic dose is achieved (Clutton-Brock 1961; Dundee 1960). An explanation for our apparent discrepancy with these findings may be that neurons at different levels of the neuraxis have varying sensitivities to the effects of the barbiturates. Alternatively, a more plausible explanation may be based on the differences in the expected pharmacokinetics of the two drugs. The more lipophilic methohexital would be expected to have a faster penetration to the nervous system and thus a shorter, more abrupt onset of effects in the CNS. Pentobarbital, on the other hand, would have a slower penetration and onset of full effects. Thus, of the two agents, a potential excitatory effect at a low dose would be more readily observed with pentobarbital than with methohexital. However, we cannot conclude that the barbiturates reduced RF size of the cells in our study, because the spontaneous activity of most cells dropped to almost nothing at lower doses than those at which the barbiturates had their effects on the mechanical responses. Thus, even at very high doses of the barbiturates, a few spikes could be elicited from all areas of skin that were originally sensitive by most mechanical stimuli, with the possible exception of the lightest compressive stimulus.
) and with the onset of normal sleep from wakefulness (Glenn and Steriade 1982; Steriade et al. 1971). Burst frequency with METH dropped precipitously after the 1.0-mg/kg bolus, which was half to one fifth the amount needed to suppress the mechanical-evoked discharges of the cells. Thus burst frequency in the monkey maintained on barbiturate could be used as an indicator for a level of anesthesia at which somatosensory neurons could be studied while avoiding a confounding influence of anesthesia. This same physiological parameter was previously reported as an indicator of stable anesthesia for somatosensory studies in the rat maintained on urethan (Armstrong-James and Callahan 1991). On the other hand, the burst pattern of thalamic cells would not be as reliable a predictor of possible confounding effects of halothane in the thalamus.
; Mori et al. 1972) and in monkeys (deJong and Wagman 1968). Similar increases in the responses of both A-mechano-heat and C-mechano-heat primary afferent fibers of monkeys have been observed with halothane but not pentobarbital (Campbell et al. 1995). Inhibition of the background activity followed by decreases in the responses to cutaneous stimuli was reported in the cat ventroposterolateral nucleus at 0.5-2.0% (Simons and Carvell 1989), and at 0.5-3% in the cat dorsal horn (deJong et al. 1969; Namiki et al. 1980) and monkey dorsal horn (deJong and Wagman 1968). Finally, the RFs in monkey ventroposterolateral nucleus were found in this study to be reduced in size, similar to the RF size in monkey and cat dorsal horn (deJong and Wagman 1968; deJong et al. 1969).
). However, the median frequency parameter does not appear as useful when assessing anesthesia level when opioids are used as the principal anesthetic agent, as during cardiac surgery (Nayak et al. 1994). For example, intravenous midazolam combined with ketamine produces an increase in median EEG frequency at doses producing sedation (Hering et al. 1994). Power spectral analysis of the EEG has been suggested as the better measure of anesthetic depth in humans (Nayak et al. 1994); and one parameter extracted from such an analysis, the EEG spectral edge (the point defining the 95% limit of the total EEG power), has proven the most reliable indicator of anesthetic depth in horses (Johnson et al. 1994). In the monkey, surgical anesthetic depth has been assessed by measuring the relative power between the anterior and posterior cortical areas, with a shift to anterior dominance correlating to adequate levels of anesthesia (Tinker et al. 1977). On the basis of the sum of these findings we analyzed the EEG by application of power spectral analysis to the EEG measured from the anterior cortical region.
). Thus EEG parameters appear to provide very sensitive indicators that could be used to reliably titrate the anesthetic depth to a point at which production of alterations in the physiological properties of thalamic somatosensory neurons is avoided.
1·h1 infusion of pentobarbital) was somewhat lower than that established in the METH-treated group (0.75-1.0% halothane). The baseline responses to the mechanical stimuli agree well with this hypothesis. On the other hand, the near equivalence between the baseline total power dampens this straightforward conclusion, but rather suggests that halothane affects the EEG power spectrum in a very different manner than the barbiturates. Examination of the power in the individual bands supports this contention. Of final note, analysis of the power in the individual bands does not appear a reliable predictor of anesthetic depth for either barbiturate or halothane anesthesia. Each treatment group showed small but inconsistent changes in power of individual bands until at the very highest doses tested. However, as detailed above, anesthetic administration at this level produced marked changes in the physiological properties of somatosensory thalamic neurons.
| |
ACKNOWLEDGEMENTS |
|---|
This work was supported in part by National Institute of Neurological Disorders and Stroke Grant PO1 NS-32386-Project 2.
| |
FOOTNOTES |
|---|
Address for reprint requests: P. M. Dougherty, Johns Hopkins Medical School, Meyer 5-109, 600 N. Wolfe Street, Baltimore, MD 21287-7509.
Received 6 May 1996; accepted in final form 29 October 1996.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Weiss, S. Ohara, K. O. Johnson, and F. A. Lenz The Human Thalamic Somatic Sensory Nucleus [Ventral Caudal (Vc)] Shows Neuronal Mechanoreceptor-Like Responses to Optimal Stimuli for Peripheral Mechanoreceptors J Neurophysiol, February 1, 2009; 101(2): 1033 - 1042. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Engelmann, J. Bacelo, E. van den Burg, and K. Grant Sensory and Motor Effects of Etomidate Anesthesia J Neurophysiol, February 1, 2006; 95(2): 1231 - 1243. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Vahle-Hinz and O. Detsch What can in vivo electrophysiology in animal models tell us about mechanisms of anaesthesia? Br. J. Anaesth., July 1, 2002; 89(1): 123 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. D. Davis, C. L. Kwan, A. P. Crawley, and D. J. Mikulis Functional MRI Study of Thalamic and Cortical Activations Evoked by Cutaneous Heat, Cold, and Tactile Stimuli J Neurophysiol, September 1, 1998; 80(3): 1533 - 1546. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. T. Wallace, M. A. Meredith, and B. E. Stein Multisensory Integration in the Superior Colliculus of the Alert Cat J Neurophysiol, August 1, 1998; 80(2): 1006 - 1010. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
