Introduction

Top Abstract Introduction Methods Results Discussion References

It has been suggested that inadequate central processing mechanisms underlie schizophrenics' self-reported inability to filter incoming sensory information (Judd et al., 1992; Waldo et al., 1994). Auditory filtering mechanisms can be assessed, in both humans and animals, using a condition-test paradigm in which two identical stimuli are presented close together (Adler et al., 1982; Freedman et al., 1987; Braff and Geyer, 1990). In most humans, the second of two vertex-recorded P50 auditory evoked potentials is markedly attenuated compared with the first when two clicks are presented at a 0.5-second interval (Braff and Geyer, 1990; Freedman et al., 1987). In contrast, P50 responses recorded from schizophrenic patients show minimal or no attenuation (filtering) in this paradigm (Freedman et al., 1987). In rats, a midlatency auditory evoked potential, N40, which has filtering properties analogous to the human P50 wave, can be recorded from brain surface. Similar to observations in humans, most rats show filtering of the N40 wave in response to paired clicks (Stevens et al., 1993, 1995). However, administration of psychogenic agents (e.g., amphetamine or phencyclidine) produces a nonfiltering, schizophrenic-like response pattern (Adler et al., 1986). Thus, the N40 wave in rats can serve as a model to study the neurobiological substrates of sensory filtering.

The central mechanisms responsible for sensory filtering have not yet been determined. However, it is known that auditory filtering is not a function of peripheral registration of the auditory stimulus because potentials recorded in the cochlear nucleus are not filtered (Bickford et al., 1993). Several neurotransmitter systems are known to be involved in the modulation of auditory filtering. For example, activation of central nicotinic cholinergic receptors has been shown to improve filtering in both amphetamine-treated rats and schizophrenic humans (Adler et al., 1992; Stevens et al., 1995). Conversely, activation of catecholamine systems disrupts auditory filtering (Adler et al., 1988; Stevens et al., 1993, 1996).

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Ketanserin reduces filtering of rat midlatency auditory evoked potential. Averages are of condition and test response waveforms across animals at 15 to 45 min after injection of ketanserin. Top, base-line average n = 8. Bottom, average waveforms 15 to 45 min after ketanserin (2.5 mg/kg i.p.), n = 5. The effect of ketanserin was to slightly decrease the condition response, while increasing the test response, thereby increasing the T/C ratio.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   A, Ketanserin dose-dependently disrupted auditory filtering. After 0.5 and 2.5 mg/kg ketanserin, T/C ratios were different than base-line values at P < .01 and P < .001, respectively. Data were taken 15 to 45 min after drug administration. N = 5 for each dose. B, The duration of ketanserin-induced reduction of auditory filtering also was dose dependent, with the effect of the higher dose lasting longer. * P < .05; ** P < .01; n = 5 for each group.

View larger version (48K): [in this window] [in a new window]   Fig. 3.   The effects of ketanserin on condition and test response amplitudes are presented as a percent of baseline recorded during the 15 to 45-min time interval. A, Condition response amplitudes decreased with increasing doses of ketanserin; n = 9 and n = 5 for base line and all doses of ketanserin, respectively. B, Test response amplitudes increased with increasing doses of ketanserin. N = 9 and n = 5 for base line and all doses of ketanserin, respectively. * P < .05, ** P < .01 vs. base line.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   DOI enhanced, and attenuated the disruptive effect of amphetamine on auditory filtering. Waveforms are grand averages from 6 animals: (A) in the unmedicated state, (B) 45 min after 1.83 mg/kg amphetamine, (C) 30 min after 2.5 mg/kg DOI and (D) 45 min after 1.83 mg/kg amphetamine plus 30 min after 2.5 mg/kg DOI.

View larger version (45K): [in this window] [in a new window]   Fig. 5.   DOI prevented the amphetamine-induced disruption of auditory filtering. Shaded bars illustrate the time course of amphetamine (1.83 mg/kg) effect on T/C values. Amphetamine significantly reduced auditory gating at the 15-, 30-, 45- and 60-min intervals. However, the combination of amphetamine plus DOI (2.5 mg/kg, ) was not different from base line at any time point. Data are the mean ± S.E.M. responses for 6 animals, recorded over a 3-hr period. For amphetamine alone, T/C values were compared with base line; amphetamine + DOI T/C values were compared with amphetamine. * P < .05 and ** P < .01.

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Comparison of the effects of DOI, amphetamine and the combination of amphetamine and various doses of DOI on condition and test response amplitudes and the resulting T/C ratio. A, Condition N40 responses were reduced by amphetamine and by the 2.5 and 5.0 mg/kg doses of DOI given in the presence of amphetamine. The lower doses of DOI prevented the amphetamine-induced reduction in condition response amplitude. B, Except for the 0.1 mg/kg dose, DOI given after amphetamine dose-dependently reduced test response N40 amplitudes. C, The overall effect of DOI (2.5 mg/kg) was to significantly reduce the T/C ratio (i.e., enhance the filtering of N40), whereas the effect of amphetamine (1.83 mg/kg) was in the opposite direction. When the two drugs were coadministered, DOI reversed the increase in T/C ratio caused by amphetamine. * P < .05, ** P < .01 vs. base line.

Evidence is emerging that disruption of central serotonergic systems plays an important role in schizophrenia. Studies of post mortem human brain tissue comparing schizophrenics and control subjects have shown both increases and decreases in 5-HT₂ receptor density in differing brain regions (Bleich et al., 1988; Arora and Meltzer, 1993; Joyce et al., 1993). Atypical neuroleptics have purported antagonistic activity at 5-HT₂ receptors (Leysen et al., 1992, 1993; Schotte et al., 1993), and it has been suggested that this activity is responsible for the increased clinical efficacy of drugs such as clozapine and risperidone (Huttunen, 1995; Meltzer, 1995; Svensson et al., 1995). Studies in rats of prepulse startle inhibition, another test of sensory filtering, have shown that 5-HT₂ receptors play a modulatory role in this paradigm (Rigdon and Weatherspoon, 1992; Sipes and Geyer, 1995). To further understand the role of serotonin in sensory filtering, the present study evaluated the effects of the selective 5-HT₂ agonist DOI and the selective 5-HT₂ antagonist ketanserin on auditory evoked potential filtering in both unmedicated and amphetamine-treated rats.

	Methods

Top Abstract Introduction Methods Results Discussion References

Animals and surgery. Male Sprague-Dawley rats (Harlan Laboratories, Indianapolis, IN) were kept in a room maintained at 21°C and had unlimited access to Purina Rodent Chow and water. Lighting was cycled at 12-hr intervals (lights on at 6:00 a.m.). The rats were group housed (3 per cage) until surgery and individually housed thereafter. All animals were 60 to 80 days old and weighed 300 to 350 g at the time of surgery. Details of electrode implantation and recording procedures have been previously reported (Stevens et al., 1995). Briefly, stereotaxic implantation of the recording and reference electrodes was accomplished with secobarbital anesthesia (50 mg/kg i.p.). The recording electrode was a 00-90 stainless steel screw with an attached Teflon-coated stainless steel wire (75-µm diameter) that was implanted through the skull at a point approximating human vertex (4.0 mm posterior to bregma, 0.5 mm lateral to midline). A bilateral reference electrode (125-µm diameter Teflon-coated stainless steel wire, uninsulated over the last 3 mm) was placed in contact with the cortex ~3.0 mm anterior to bregma and was grounded. The electrode contacts were gathered into a plastic head plug (Carlton University, Ottawa, Canada), which was then cemented to the rat's skull with acrylic dental cement. Rats were allowed to recover for at least 1 week before initiation of recording sessions.

Chronic recordings. Animals were handled for 10 to 15 min, and then placed in a Plexiglas chamber (21 × 22 cm floor), which was located in a sound-attenuating enclosure. The head plug was connected to a cable, terminating in an FET operational amplifier, which was attached to the top of the recording chamber by a swivel assembly to allow full freedom of movement. The animal was allowed to acclimate to the chamber for 10 to 15 min before recording was begun. The signal from the recording electrodes were fed through a unity gain headstage amplifier and then to a second-stage amplifier, which increased the signal to 5000 times its original amplitude. Auditory stimuli, delivered through a speaker mounted in the ceiling of the recording chamber, consisted of two 600-µsec duration, 87-dB (SPL) clicks, delivered 500 msec apart; click pairs were presented every 15 sec. Evoked potentials were computer digitized at 1 kHz over 450-msec epochs.

Drug administration. Ketanserin tartrate, a selective 5-HT₂ antagonist (Fuller and Snoddy, 1984; Pranzatelli, 1991), and DOI (±-2-5-dimethoxy-4-iodoamphetamine hydrobromide), a selective 5-HT₂ agonist (Buckholtz et al., 1988; Pranzatelli, 1991), were obtained from Research Biochemicals (Natick, MA). Amphetamine sulfate was obtained from Sigma Chemical (St. Louis, MO). All drugs were prepared in physiological saline (pH 7.4) and administered intraperitoneally. Recordings began immediately after injection and continued for consecutive 15-min intervals for 75 min after injection. Additional records were obtained at 2 and 3 hr after injection. In some experiments, rats received both amphetamine and DOI. This protocol involved amphetamine administration 15 min before DOI, and recordings were initiated after the DOI injection according to the above schedule.

Data analysis. The N40 auditory evoked potential was identified as the largest negative going wave with a peak occurring between 30 and 50 msec. Thirty waveforms were averaged to provide the following parameters: CAMP and TAMP, T/C (the measure of filtering) and CLAT and TLAT. Waveform amplitudes were measured from the peak of the preceding positivity. Possible changes in each of the five variables due to ketanserin, amphetamine, DOI and amphetamine plus DOI were analyzed by repeated measures analysis of variance (MANOVA), with the Tukey-Kramer or Newman-Kuell a posteriori analysis as appropriate (SPSS/PC+ 5.0, 1992). The threshold level for statistical significance was set at  = .05 for all comparisons.

	Results

Top Abstract Introduction Methods Results Discussion References

As previously reported, the brain surface-recorded evoked potential elicited by a suprathreshold auditory stimulus consisted of several components (Knight et al., 1985). The primary midlatency complex was composed of a pair of negative-going peaks with latencies to peak occurring between 20 and 30 msec (N25) and 30 and 50 msec (N40) after stimulus onset (fig. 1). When two clicks were presented 500 msec apart, the midlatency responses recorded to the second click were routinely attenuated, or filtered. However, it was observed that filtering of N25 was both more variable and less pronounced than filtering of N40. Therefore, N40 was chosen for evaluation of the effects of serotonin 5-HT₂ compounds on auditory filtering.

The administration of ketanserin, a selective 5-HT-2 antagonist, dose-dependently reduced filtering (figs. 1 and 2; table 1), as was indicated by a significant increase in the TC ratio (F_4,24 = 4.922, P = .005). Furthermore, the duration of ketanserin's effect on the TC ratio was also dose dependent (F_8,64 = 2.43, P = .023; fig. 2B). The ketanserin-induced disruption of sensory filtering was the result of changes in both condition and test amplitudes (fig. 3). Ketanserin administration resulted in a dose-dependent decrease in condition amplitude (F_4,23 = 5.298, P = .004), which was significant at individual doses of 0.5 and 5.0 mg/kg (fig. 3A). In contrast, ketanserin produced increases in test amplitude (F_4,23 = 3.760, P = .017), although individual significance was observed only for the 2.5 mg/kg dose.

The effects of 5-HT₂ receptor activation upon sensory filtering were examined under three conditions. DOI, a selective 5-HT₂ agonist, was given alone, as well as to rats that had received either amphetamine or ketanserin. Amphetamine and DOI produced opposite effects on filtering. As we previously reported, 1.83 mg/kg amphetamine increased the TC ratio by reducing condition response amplitude (Stevens et al., 1995). However, DOI (2.5 mg/kg) reduced the amplitude of the test N40, resulting in a decreased TC ratio. The effect of combined amphetamine and DOI was similar to what was observed when DOI was administered alone and was not significantly different from the base line. These data are shown in figures 4 and 5. In the presence of amphetamine, DOI maintained the TC ratio at base-line levels by causing significant decreases in the test amplitude responses (F_(7,63) = 2.062, P = .019; see fig. 4 and table 2). The normalization of amphetamine-induced loss of filtering was dose-dependent; only the lowest dose (0.1 mg/kg) failed to significantly reverse amphetamine's effect (fig. 6).

DOI's action in the presence or absence of amphetamine can be broken down into effects on both CAMP and TAMP. Administration of DOI alone produced nonsignificant reductions in both CAMP and TAMP (fig. 6, b and c). However, because the decrease in TAMP was proportionately greater, filtering was improved. DOI plus amphetamine resulted in a decrease in CAMP (F_5,27 = 4.89, P = .0026); the lower doses (0.1 and 0.625 mg/kg) of DOI reversed this effect, although the higher doses of DOI (2.5 and 5.0 mg/kg) did not (table 2). Similarly, TAMP was significantly decreased when DOI was given in combination with amphetamine at the 0.625, 2.5 and 5.0 mg/kg DOI compared with both base line and amphetamine-alone test amplitude responses (F_5,27 = 4.90, P = .0026). The effect of DOI on TAMP was sufficient to improve filtering performance in the presence of amphetamine, regardless of whether a change in CAMP took place.

Although DOI did not completely normalize ketanserin's disruptive effect on filtering of the N40, the effect of ketanserin was lessened by DOI (2.5 mg/kg ketanserin alone was 186 ± 11% of base line, and ketanserin plus 2.5 mg/kg DOI was 133 ± 11% of base line) (table 1). The DOI-induced improvement of filtering in ketanserin-treated rats was the result of both a reduction in the condition amplitude and a still greater reduction in the test amplitude (table 1). Latency to peak for the condition or test responses was not significantly altered after ketanserin administration (table 1).

The DOI-induced improvement of amphetamine- or ketanserin-induced reductions in auditory filtering was not a result of the second injection alone because filtering in rats given either drug followed by saline injections did not differ from those given amphetamine or ketanserin alone (e.g., TC ratios for amphetamine and amphetamine plus saline at 45 min were 0.72 ± 0.07 and 0.80 ± 0.02, respectively, n = 2). The latency to peak for both conditioning and test N40s did not differ significantly from base line under any treatment condition (table 2).

	Discussion

Top Abstract Introduction Methods Results Discussion References

The primary finding of this study was that manipulations of the serotonin 5-HT₂ system profoundly affect auditory filtering in the unanesthetized rat. Administration of the 5-HT₂ antagonist ketanserin impaired filtering in a dose-dependent manner. In contrast, the 5-HT₂ agonist DOI slightly but significantly improved filtering in otherwise unmedicated rats and dose-dependently reversed the disruptive effect of amphetamine on auditory filtering.

The current work provides additional confirmation of the disruptive effect of amphetamine on auditory filtering (Stevens et al., 1991, 1993; Adler et al., 1988). The duration of amphetamine's effect was approximately 3 hr at the dose used. When the selective 5-HT₂ agonist DOI was administered in the presence of amphetamine, reversal of the amphetamine-induced reduction of auditory filtering occurred. The ability of DOI to reduce the TC ratio was much greater in amphetamine-treated rats than in unmedicated animals, indicating that the DOI-related improvement of auditory filtering in the presence of amphetamine was more than simply additive.

The impairment of filtering induced by ketanserin was achieved by a different mechanism than for amphetamine. Consistent with previous observations, amphetamine significantly reduced condition response amplitude (Stevens et al., 1991). Ketanserin, however, only rarely produced significant changes in either the condition or test responses. Instead, the combination of a reduction in the condition response and an increase in the test response resulted in a significant increase in the TC ratio. The reduction in filtering caused by ketanserin was reversed by DOI, but this was achieved through a significant reduction in the test response (similar to what was observed when DOI was coadministered with amphetamine). Furthermore, DOI did not increase condition response amplitude, as might have been expected. This may indicate that the action of either ketanserin or DOI was not strictly limited to 5-HT₂ receptors. It is known, for example, that ketanserin also has weak antagonistic effects at brain noradrenergic alpha-1 receptors (Battaglia et al., 1983; Hoyer et al., 1987; Tsuchihashi and Nagamoto, 1989). However, our previous work has shown that phentolamine, a general noradrenergic  receptor antagonist, had no effect on auditory filtering by itself, and actually improved auditory filtering that had been disrupted by amphetamine (Stevens et al., 1991). These effects appear to be incompatible with the observed action of ketanserin on auditory filtering. Further work will be necessary to resolve these issues.

Because numerous reports have demonstrated impaired sensory filtering in schizophrenics (Adler et al., 1982; Freedman et al., 1987; Snyder, 1973; Solomon et al., 1981; Sorensen et al., 1993), the results of the present study suggest that activation of 5-HT₂ receptors could improve sensory filtering in these individuals. In support of this theory, decreased numbers of 5-HT₂ receptors in the frontal cortex of schizophrenic patients have been reported (Arora and Meltzer, 1991; Hashimoto et al., 1993). Thus, it is possible that increased activation of remaining 5-HT₂ receptors might restore normal sensory filtering in schizophrenics.

There appear to be many mechanisms for correcting the amphetamine-induced reduction of auditory filtering, which is brought about by the decrease in condition response amplitude. For example, haloperidol (Adler et al., 1986) or the dopamine D₁ antagonist SCH23390 (Stevens et al., 1991) counteracts the effect of amphetamine directly by restoring the amplitude of the conditioning response. The activation of 5-HT₂ receptors improved filtering by a less direct mechanism in that DOI reduced the amplitude of the test response to a point at which the TC ratio was similar to base line despite the small condition response amplitude. Adrenergic receptor antagonists (Stevens et al., 1991) or nicotine (Stevens et al., 1995) have similar actions in the presence of amphetamine. Clozapine initially (first month of treatment) improves filtering by increasing the condition amplitude, yet the corrected clinical profile (14 months) correlates with a significant reduction in the test amplitude (Nagamoto et al., 1997). Because clozapine does not return the condition and test evoked potentials to their base-line states yet corrects filtering and clinical profile, suggests the DOI-induced reduction of TC ratio represents a "normalized" situation in terms of auditory information processing.

Because activation of either 5-HT₂ or nicotinic-cholinergic receptors attenuates the test response amplitude, it suggests that these systems may be acting in concert with each other to improve gating status. Consistent with this notion is the synergistic action of nicotinic-cholinergic mechanisms in the modulation of serotonergic mediated behavior and levels (Codignola et al., 1994; Takada et al., 1995). For example, Riekkinen et al. (1993) reported that the cholinergic impairment on water maze performance was further aggravated after lesions of the serotonergic system with PCPA. In addition, Ribeiro et al. (1993) reported an increase in frontocortical serotonin levels after systemic nicotine administration. These two examples clearly suggest a synergistic action between the cholinergic and serotonergic system.

Although the effects of serotonin 5-HT₂ agents on auditory filtering in the present experiments were dose dependent and internally consistent, they were quite different from what has been observed for another sensory filtering paradigm, prepulse-startle inhibition (PPI). Prepulse inhibition of the acoustic startle response (PPI) is similar to the auditory filtering described in this study in that both paradigms evaluate the effect of a prior auditory stimulus presentation on the response produced by a second stimulus presented shortly thereafter. However, in our auditory filtering paradigm, a central representation of stimulus registration is measured (i.e., an auditory evoked potential), whereas in the PPI paradigm, the output of central nervous system processing (muscular startle) is the measured response. The administration of DOI impairs PPI, whereas ketanserin restores DOI-disrupted PPI (Sipes and Geyer, 1994).

It has been suggested that PPI and the current auditory filtering paradigm share similar mechanisms. It is true that both PPI and auditory filtering are disrupted by amphetamine (Swerdlow et al., 1990; Adler et al., 1986). However, some reports demonstrate normalization of amphetamine's disruptive effects on PPI with nicotinic agents (Curzon et al., 1994) (Stevens et al., 1995), whereas other studies have observed the opposite result (Acri et al., 1991). The effects of 5-HT₂ receptor system activation appears to represent another dichotomy between these two forms of sensory filtering. It seems apparent that although PPI and auditory filtering may share some features, they differ in certain aspects of their pharmacological modulation.

Interestingly, the new atypical neuroleptics have purported antagonistic activity at 5-HT₂ receptors (Leysen et al., 1992, 1993; Schotte et al., 1993). This result would not have been predicted based on the outcome of the present study. One possible explanation may lie in the fact that these novel antipsychotic agents are not pure antagonists at the 5-HT₂ receptors. Rather, these compounds appear to be active at several neurotransmitter receptors, including other serotonergic receptor subtypes (Corbett et al., 1993; Duinkerke et al., 1993; Gerlach, 1991; Matsubara et al., 1993; Svensson et al., 1995). An alternate explanation for the efficacious nature of atypical neuroleptics is that these agents exert their effects differentially at 5-HT₂ receptors subtypes (e.g., 5-HT_2A vs. 5-HT_2C). Studies are currently under way to evaluate the respective contributions of these receptor subtypes to the modulation of auditory filtering.