Modulation of the Gating of Auditory Evoked

BIOL PSYCHIATRY 1988 ;24; 179-190

179

Modulation of the Gating of Auditory Evoked Potentials by Norepinephrine: Pharmacological Evidence Obtained Using a Selective Neurotoxin Lawrence E . Adler, Kevin Pang, Greg Gerhardt, and Greg M. Rose

Central mechanisms of sensory gating were assessed in Sprague-Dawley rats using an evoked potential technique similar to one that we have previously employed to show diminished sensory gating in psychotic patients. Gating mechanisms were examined using a conditioning-testing paradigm in which pairs of74-dB clicks were delivered; the interval between the conditioning and test stimuli was 0.5 sec. A middle latency auditory evoked response (N50) recorded from the skull of unanesthetized, freely moving rats demonstrated significant suppression to the test click. Systemic administration of amphetamine (1 mglkg, ip) significantly reduced the amount of suppression of the response to the test stimulus; haloperidol (1 mglkg), injected after the amphetamine, returned the conditioning-testing suppression ratio toward normal values. Amphetamine also decreased the latency and amplitude of the conditioning response, an effect that was also reversed by haloperidol. Both decreased suppression of the test response and reduced amplitude and latency of the conditioning response have been observed in schizophrenia. To aid in determining the underlying mechanism of these effects, the animals were treated with two doses, given at a 1-week interval, ofN-(2-chloroethyl-N-ethyl-2-bromobenzylamine) (DSP4; 50 mglkg, ip), an agent that selectively depletes central norepinephrine. The extent and selectivity of the depletions were confirmed by chemical analysis. Following DSP4, the effects of amphetamine on the amplitude and latency of the conditioning response were largely unchanged. However, pretreatment with DSP4 significantly attenuated the reduction in conditioning-testing suppression observed following the administration of amphetamine. The data suggest a specific role for norepinephrine in the modulation of sensory processing.

Introduction Abnormalities in sensorineural processing of sensory information have been suggested as possible contributing factors to the etiology of psychotic symptoms. Venables (1964) suggested that schizophrenics cannot filter stimuli adequately and so are "flooded" by sensory information. Epstein and Coleman (1970) have suggested that schizophrenics

From the Departments of Psychiatry ( L . E . A . , G . G . ) and Phannacology ( K . P . , G . G . , G . M . R . ) , University of Colorado Health Sciences Center, and the Medical Research Service ( L . E . A . , K . P . , G . M . R . ) , V A M C , Denver, C O . Supported by USPHS Grants MH38321 and AA03527 and a grant from the Veterans Administration Medical Research Service. Address reprint requests to Dr. Lawrence E . Adler, Department of Psychiatry, C268, University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Denver, C O 80262. Received June 17, 1987.

© 1988 Society of Biological Psychiatry

0006-3223/88/503.50

BIOL P S Y C H I A T R Y 1988 ;24; 179-190

i

L . E . Adler et al.

suffer from a "low threshold of disorganization under increasing stimulus input." Evoked potential paradigms have been used by different investigators to demonstrate altered information processing in schizophrenics, but few investigators have been able to correlate similar electrophysiological measures in laboratory animals. However, recently, Swerdlow et al. (1985) have demonstrated similar prepulse inhibitory phenomena (decreased startle to a loud stimulus preceded by an earlier auditory stimulus) in both humans and laboratory rats. Our research group has been particularly interested in the involvement of inhibitory mechanisms in the gating of auditory sensory information. We utilize a conditioningtesting paradigm that is well known in animal electrophysiology in the study of inhibitory pathways (Eccles 1969). In this paradigm, closely paired auditory click stimuli are presented to a subject. I f inhibitory pathways are functioning normally, the amplitude of the response to the second stimulus (test response) is decreased because of inhibitory pathways that are activated in response to a first (conditioning) stimulus. The amplitude of the test response can be divided by the amplitude of the response to the conditioning stimulus to give a "conditioning-testing ratio," which is expressed as a percent. Lower conditioningtesting ratios indicate stronger inhibition. In human subjects, we focused initially on the P50 waveform of the auditory evoked potential, as this middle latency waveform is less dependent on changes in attention than are later waveforms (Hillyard et al. 1973; McCallum et al., 1983). We found in our initial studies that normal controls had P50 conditioning-testing ratios of 14%, as compared to 90% for schizophrenics, when a 0.5-sec interval was used between paired click stimuli (with 10 sec between presentations of paired stimuli to allow for complete recovery) (Adler et al. 1982). P50 amplitude and latency were also decreased in unmedicated schizophrenic patients compared to normal controls. Neuroleptics increased P50 latency and amplitude to normal, but did not normalize conditioning-testing ratios in schizophrenics (Freedman et al. 1983). Acutely psychotic manics had P50 conditioning-testing ratios very similar to those of schizophrenic patients, but unlike the schizophrenics, the P50 conditioning-testing ratios of manics were indistinguishable from those of normal controls when these patients were euthymic. Furthermore, manics had normal P50 latencies and amplitudes (Franks et al. 1983). In a previous study (Adler et al. 1986), we showed that the N50 waveform of the auditory evoked potential of the Harlan Sprague-Dawley rat had many characteristics in common with the P50 waveform of the human. We recorded auditory evoked potentials in a conditioning-testing paradigm from chronically implanted rats in a sound-isolated recording chamber. Like the human P50, the rat N50 showed a low conditioning-testing ratio when two auditory stimuli were paired 0.5 sec apart. Also like the human P50, recovery of the test response amplitude was observed as the conditioning-testing interval increased. Rats that were new to the recording chamber, and therefore hyperaroused, had N50 conditioning-testing ratios averaging 56%; after they had become used to the environment and less aroused, it averaged 25.3%. This effect of hyperarousal could be blocked by haloperidol, suggesting that it might be mediated by catecholamines. A role for catecholamines in the modulation of sensory processing was also suggested by experiments with amphetamine. Intraperitoneally injected amphetamine, which enhances the action of both central nervous system (CNS) dopamine and norepinephrine, caused an augmentation of conditioning-testing ratios to 176% of control, as well as a significant decrease in amplitude and latency of the conditioning response. These effects were

Noradrenergic Modulation of Sensory Gating

BIOL PSYCHIATRY 1988;24:179-190

181

reversed by haloperidol, which can block the action of both dopamine and norepinephrine at the postsynaptic receptor. Our assessment was that catecholamines were significant mediators of auditory sensory processing. The purpose of the present experiment was to specifically assess the relative role of norepinephrine in auditory sensory processing. DSP4 [ A^-(2-chloroethyl-A^-ethyl-2-bromobenzylamine) ] , a selective neurotoxin that causes long-lasting depletion of norepinephrine in the central nervous system, but which is without major effects on serotonin or dopamine (Jonsson et al. 1981), was used to determine the noradrenergic influences on the N50 and its conditioning-testing suppression.

Methods

/

'

'

_

'

Animals and Surgery Male Sprague-Dawley albino rats were obtained from Harlan Laboratories (Indianapolis, IN) and weighed 290-350 g at the time of surgery. Animals were pretreated with atropine methyl nitrate (0.4 mg/kg) to minimize secretions; 30 min later, they were anesthetized with sodium pentobarbital (50 mg/kg) and were supplemented with chloral hydrate (150 mg/kg) as necessary. Once functional anesthesia was established, as determined by lack of response to tail pinch and loss of corneal blink reflex, the rat's head was secured in a stereotaxic frame and the skull exposed for implantation of electrodes. The recording electrode consisted of a 00-90 x 1/8 inch stainless steel screw to which a length of 0.(X)5-inch Teflon-covered stainless steel wire had been soldered; the other end of the - ?i; wire was crimped into a Cannon C T A series pin; this electrode was placed 4 mm posterior to bregma on the midline ("vertex"). The reference electrode was a pair of 0.010-inch Teflon-covered stainless steel wires crimped into a single Cannon pin that had the insulation removed from the last 4 mm of the tips. The uninsulated portions of these wires : w e r e placed into the frontal cortex, 3.0 mm anterior to bregma and 1.5 mm to either side of the midline. Four to five additional screws were placed around the periphery of the exposed skull to serve as anchors. The pins from the recording and reference electrodes were gathered into a connector made from two four-socket Cannon centi-loc strips that had been glued into a 2 x 4 array. This headstage was then cemented to the skull using dental acrylic. Animals were allowed a minimum of 1 week to recover following the implantation surgery. No animal was used for recording before it had recovered its :»;r preoperative body weight. Rats were group housed and were given free access to food and water during all phases of the experiments. «

Apparatus

s-^i ; r-.-

All recordings took place with the animals in a 21 x 22 cm plexiglass chamber, which was located in a sound-attenuated enclosure. The signal from the recording electrode was amplified 10 times using a F E T operational amplifier that plugged into the headstage; the reference electrode was connected to ground. The amplified signal was led, via a cable through a slip-ring commutator of the type described by Micco (1977), to a second stage amplifier that further boosted the signal gain to 10(X) times its original amplitude. A second-order high-pass filter installed between the two amplifier stages limited the bandwidth of the recorded signal from 0.5 to 10,000 Hz. ' Auditory stimuli were presented from a speaker positioned 45 cm off the floor of the


BIOL P S Y C H I A T R Y 1988;24:179-190

recording chamber. Evoked potentials were elicited by 500-|jLsec clicks of 74-dB (RMS) intensity; the clicks were delivered in pairs, with an interclick interval of 0.5 sec. Click pairs were presented every 15-35 sec. Stimulus presentations were controlled by a Nova 3/12 computer, which also digitized the evoked waveforms over 512-msec epochs (at 1msec intervals), beginning 150 msec preceding click onset. Digitized waveforms were stored for later averaging and analysis.

Pharmacology

,

.

Drugs used in this study included dextroamphetamine sulfate (1.0 mg/kg) and haloperidol (1.0 mg/kg). Both agents were dissolved in 0.9% NaCl. Injections were made into the intraperitoneal cavity. Sessions following drug administration were begun 20-30 min after the injection. DSP4 [ A^-(2-chloroethyl-A'-ethyl-2-bromobenzylamine) ] was used to selectively deplete central norepinephrine (Ross 1976; Jonsson et al. 1981). DSP4 was dissolved in 0.9% NaCl to make a 10 mg/ml solution; the drug was administered in 2 doses, each 50 mg/kg ip, spaced 1 week apart. As DSP4 is very unstable in solution, every attempt was made to inject the drug-containing solution as rapidly as possible (Ross 1976). Injections were routinely delivered to the animals (anesthetized with ether) within 30 sec after the solution was made.

Procedure The animal selected for a given day's experiment was removed from its group cage and handled for 5-10 min prior to being placed in the recording chamber. Once in the chamber, the rat was allowed an additional 15-30 min of acclimatization before the experiment proceeded. A single session involved presentation of up to 200 pairs of click trials. The rat's behavior at the moment of onset of the first click of each pair was noted for every trial. Behavioral state was classified as either still alert, moving, or asleep. The animal was first recorded after 30 min in the recording chamber, followed by a second unmedicated recording 30-60 min later. The rat was then injected with dextroamphetamine and recorded. After that, the rat was injected with haloperidol (1 mg/kg) and rerecorded. A minimum of 3 days after drug treatment, the rat was injected intraperitoneally with DSP4. One week later, the rat received a second injection of DSP4. The animals were again recorded with the same drug treatment (unmedicated, amphetamine, haloperidol) at 1 and 2 weeks after the second DSP4 injection. For these experiments, only trials recorded during still, alert behavior were collected for averaging in order to minimize the possibility of variations in perception of the conditioning and testing clicks.

Data

Analysis

The auditory evoked potentials from single trials were sorted according to the behavioral criteria described above, and averages of still, alert trials were made for each session. Amplitude and latency measurements of major waveform components were made by cursor measurements from computerized files of these averages. Statistical comparisons were made using the one-tailed paired Student's r-test, paired for individual animals before and after DSP4 treatments, as the direction of expected effects was predicted by our previous experiment (Adler et al. 1986). To assess overall treatment effects. Analysis of Variance (ANOVA) was used.

Noradrenergic Modulation of Sensory Gating

BIOL PSYCHIATRY 1988;24;179-190

183

Table 1. N50 Conditioning-Testing Ratio (% ± SEM) 1 Week

2 Weeks

Pre-DSP4

post-DSP4

post-DSP4

Unmedicated 1

42 ±

16.4

37 ± 5.2

45 ± 1 5 . 6

Unmedicated 2

22 ± 8.0"

37 ± 7.9

30 ± 6.7

Trial

Amphetamine Haloperidol

107 ± 22.1*

46 ±

27 ± 4.1'-

12.0

47 ± 15.1

59.1 ± 19.2"^ 45 ± 20.2

T h e N50 C T ratio in the second unmedicated trial decreased significantly (( = 1.79, df = 6, p < 0.03 by one-tailed ttest) compared to the first unmedicated trial. ''Amphetamine significantly increased the N50 C T ratio compared to the second unmedicated trial (t = 3.44, df = 6, p < 0.004, one-tailed /-test). •^Haloperidol significantly decreased the N50 C T ratio compared to the amphetamine trial (r = 4.04, df = 6, p = 0.002 by one-tailed /-test). •Two weeks after DSP4, amphetamine caused a significant increase in N50 C T ratio by one-tailed /-test (/ = 1.76, df = 4, p < 0.04). However, this value (59.1 ± 19.2%) was still significantly less than the impairment in sensory gating caused by amphetamine prior to DSP4 (107 ± 22.1%, one-way /-test, / = 1.55, df = 10, p < 0.038).

Determination

of CNS Monoamine

Levels

At the conclusion of the recording experiments, the animals were overdosed with ether and decapitated. The brain was quickly removed, and samples from the caudate nucleus, hippocampus, and cerebellum were taken. The tissue was frozen on dry ice, weighed, and stored at — 70°C until monoamine levels could be measured. In addition to the DSP4treated rats, samples were taken from 6 control animals that had received intraperitoneal injections of 0.9% NaCl. Whole tissue levels of dopamine, norepinephrine, and serotonin were measured by high-performance liquid chromatography (HPLC) with electrochemical detection ( E C ) , using the method of McKay and coworkers (1984) and Spuhler et al. (1987). Dihydroxybenzylamine was used as an internal standard to calculate recovery. A l l monoamine values were calculated as total nanograms per gram wet weight of tissue (ng/g). Significance between treated (DSP4) and untreated groups was also analyzed by one-tailed ttest.

Results

*' .\i r ;

The N50 waveform was visible as the second of two negative waveforms in the auditory evoked potential. The latency for the N50 waveform on the second of the two unmedicated trials was 49 ± 5.2 msec (mean ± standard error of the mean, average for 7 rats). The N50 amplitude of the second unmedicated trial was 627 ± 173 |xV. When two clicks were delivered 0.5 sec apart, the response following the second click was invariably smaller than that evoked by the first (conditioning) stimulus. The conditioning-testing ratio for the second of two unmedicated trials prior to DSP4 was 22% ± 8%. When rats were first placed in the recording chamber, they exhibited behavioral signs of moderately high arousal (moving about the recording chamber, periods of grooming and teeth gnashing, etc.). Although only nonmoving, alert trials were recorded, there were still observable differences in N50 amplitudes and conditioning-testing ratios between the first unmedicated trial, when the rat was still new to the recording chamber, and the second unmedicated trial (about 1 hr later), when the rat was less aroused and more acclimated to the recording chamber (Tables 1-3). The N50 amplitude increased between the first and second unmedicated trials prior to DSP4 {p < 0.03, by one-tailed f-test). In comparing the conditioning-testing ratio ( C T ratio) between the first and second



Table 2. N50 Amplitude in Microvolts ( ± SEM) "

1 Week post-DSP4

Pre-DSP4

Trial

Unmedicated 1 Unmedicated 2 Amphetamine Haloperidol

379 627 229 525

± ± ± ±

395 486 356 474

57.2 173" 59* 97'^

± 43 ± 66.1 ± 71.4'' ± 68.8'

2 Weeks post-DSP4 432 509 280 718

±119 ± 105 ± 89* ± 251'"

"A significant increase in N50 amplitude from thefirstuiimedicated trial to the second uiimedicated trial (/ = 1.83, df = 6, p < 0.03, one-tailed r-test). *A significant decrease in N50 amplitude from the second unmedicated trial to the amphetamine effect noted (« = 2.68, df = 6, p < 0.01, one-tailed Mest). •^A significant increase in N50 amplitude between the amphetamine and haloperidol trials (t = 2.50, df = 6, p < 0.01, one-tailed r-test). ''A significant decrease in N50 amplitude between the second unmedicated and the amphetamine trials {t = 1.49, df = 6, p < 0.05, one-tailed /-test). 'A significant increase in N50 amplitude between the amphetamine and haloperidol trials (/ = 1.77, df = 6, p < 0.03, one-tailed Mest). fA significant increase in N50 amplitude between the first and second unmedicated trials (r = 1.97, df = 4, p < 0.03, one-tailed /-test). 'A significant decrease in N50 amplitude between the second unmedicated and amphetamine trials (/ = 3.03, df = 4, p < 0.01, one-tailed /-test). ''A significant increase in N50 amplitude between the amphetamine and haloperidol trials (/ = 1.80, df = 4, p < 0.04, one-tailed/-test). . . ... , . , ,,. .

...

...-^ . ,

n: -

-

unmedicated trials, we found that there was a tendency for the N50 conditioning-testing ratio to decrease (mean ± S E M : from 42% ± 16.4% to 22% ± 8% for 7 rats; p < 0.03, one-tailed /-test). N50 latency also increased significantly between the first and second unmedicated trials (Table 3). Intraperitoneal amphetamine (1.0 mg/kg) significantly increased the conditioningtesting ratio from the second unmedicated trial (22% ± 8% to 105% ± 2 2 . 1 % for 7 rats; p < 0.(X)4, one-tailed /-test). Comparison of conditioning and test responses before and after amphetamine are shown in Figure 1. Amphetamine also significantly decreased N50 amplitude from the second unmedicated trial (p < 0.01, one-tailed Mest). These changes were paralleled by an extreme hyperaroused state induced by amphetamine. State-dependent changes included increased locomotion, stereotypies (gnawing, sniffing a comer of the recording chamber), and increased grooming behavior. Haloperidol not only caused a reversal of the amphetamine-induced behavior effects (rats became placid, nonmoving), but reversed the amphetamine-induced increased C T ratio to baseline levels (Table I , p < 0.002). Haloperidol also significantly increased N50 amplitude over the amphetamine-induced decrease (/ = 2.5, p < 0.02, df = 6; Figure I and Table 2) and similarly increased N50 latency (Table 3; t = 2.35, p < 0.03, df = 6).

Effects

of DSP4

One week after the second injection of DSP4, significant N50 conditioning-testing differences between unmedicated and amphetamine trials were no longer observed (Table 1). Analysis of Variance showed a significant effect of drug (second unmedicated trial, amphetamine, or haloperidol) on N50 conditioning-testing ratio (F = 5.131, df = 3, p < 0.003), no effect of DSP4 by itself ( F = 0.361, df = 2, /? < 0.698), and a significant drug-DSP4 interaction (F = 2.29, df = 6, /? < 0.046). As DSP4 also causes a peripheral noradrenergic depletion during the first week, we repeated the experiment 2 weeks

CONTROL

AMPHETAMINE

HALOPERIDOL

Figure 1. Evoked potentials recorded before (left) and after (middle) amphetamine (1 mg/kg). The response to haloperidol (1 mg/kg ip) given 0.5 hr after amphetamine is also shown (right). Both conditioning ( C ) and testing ( T ) waves are shown. The stimulus artifact occurs at S; small artifacts can be seen in most tracings. The amplitude was measured peak to peak from the crest of P35 to the trough of N50. The amplitude of N50 is shown as a line from the point used as baseline by the computerized analysis system. Before amphetamine, there was significant suppression of N50 in the test response. After amphetamine, N50 in the conditioning response was smaller and earlier; the suppression in the test response was also diminished. Haloperidol returns the suppression of the test response to normal limits and increases the amplitude to the conditioning stimulus.



H

CONTROL AMPHETAMINE