Introduction

Top Abstract Introduction Materials and methods Results Discussion References

N-Methyl-D-aspartate (NMDA) receptor stimulation is considered a critical event in several pathological conditions, including ischemia-induced neurotoxicity (Choi, 1990) and epilepsy (Meldrum, 1988). The development of clinically viable NMDA receptor antagonists (NMDA-RAs) has therefore been of great interest (Rogawski, 1993). A major factor compromising the therapeutic use of NMDA-RAs is their psychotomimetic profile; noncompetitive NMDA-RAs such as phencyclidine (PCP) produce in healthy humans a psychosis that closely resembles schizophrenic symptomatology (Javitt and Zukin, 1991). Preclinical behavioral studies, however, have indicated a possible functional dissociation between the noncompetitive NMDA-RAs, which bind to a site inside the cation channel, and the competitive NMDA-RAs, which reversibly bind to the glutamate recognition site outside the channel (Watkins, 1994). For example, noncompetitive and competitive NMDA-RAs do not substitute for each other in drug discrimination paradigms (Gold and Balster, 1993; Wiley and Balster, 1994).

One paradigm with particular relevance for the sensory disturbances associated with drug-induced psychotic states and in which striking differences have been observed between noncompetitive and competitive NMDA-RAs is prepulse inhibition (PPI). PPI refers to the normal inhibition of a startle response when a weak stimulus immediately precedes a startling stimulus and is thought to provide an operational measure of sensorimotor gating, one of the processes by which an organism filters information from its surroundings (Geyer et al., 1990). Deficits in PPI are observed in schizophrenic patients (Braff et al., 1992) and in rats or healthy humans treated with noncompetitive NMDA-RAs such as PCP, dizocilpine, or ketamine (Mansbach and Geyer, 1989; Geyer et al., 1990; Karper et al., 1994). In contrast, multiple studies demonstrate that competitive NMDA-RAs either do not alter PPI (Mansbach, 1991; Wedzony et al., 1994; McCloskey et al., 1995) or decrease PPI only as an artifact of severe startle reactivity depression (Furuya and Ogura, 1997), leading to the suggestion that these compounds may be neuroprotective agents devoid of psychotomimetic or abuse (vide supra) potential. Some recent reports, however, indicate that the competitive antagonists d-2-amino-5- or d-2-amino-7-phosphonopentanoic acid (AP-5 and AP-7, respectively) decrease PPI (Reijmers et al., 1995; Kretschmer and Koch, 1997; Wan and Swerdlow, 1997) and generalize to the PCP discriminative stimulus (Tricklebank et al., 1987) when administered intracerebrally.

In the present study, we sought to examine systematically the possibility that the putative inability of competitive NMDA-RAs to disrupt PPI reflects poor penetration into the brain (Tricklebank et al., 1987, 1989) rather than a veritable difference between the competitive and (psychotomimetic) noncompetitive NMDA-RAs. Two novel competitive NMDA-RAs (SDZ 220-581 and SDZ EAB-515) have been synthesized recently, with a hydrophobic moiety permitting superior absorption into the central nervous system (CNS) and with an excellent neuroprotective profile in vivo (Urwyler et al., 1996a,b). These compounds were tested for their ability to disrupt PPI and elicit hyperactivity, another cardinal feature of the behavioral profile of the noncompetitive NMDA-RAs in rats. In addition, the classic competitive antagonist D-CPPene (SDZ EAA-494; Lowe et al., 1994) was tested for its ability to alter PPI after either systemic or intracerebral administration.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion References

Animals

A total of 186 male Sprague-Dawley rats (Harlan Laboratories, San Diego, CA), weighing 300 to 400 g, were used in the present study. Animals were housed in pairs in clear plastic cages located inside a temperature- and humidity-controlled animal colony and were maintained on a reversed day/night cycle (lights on from 7:00 PM to 7:00 AM). Food (Harlan Teklad, Madison, WI) and water were available continuously, except during behavioral testing, which occurred between 9:00 AM and 5:00 PM. On arrival in the colony, all animals were handled gently by the experimenter for 1 to 2 min every 3 to 4 days to minimize stress during behavioral testing. Animal facilities were AAALAC approved; protocols were in accordance with the "Guiding Principles in the Care and Use of Animals" (provided by the American Physiological Society) and the guidelines of the National Institutes of Health.

Drugs

The following drugs were used: SDZ 220-581 (0, 2.5, or 5.0 mg/kg), SDZ EAB-515 (0, 3.0, 10.0, or 30.0 mg/kg), SDZ EAA-494 (D-CPPene; 0, 0.5, 1.0, or 5.0 mg/kg), clozapine (0 or 5.0 mg/kg), and haloperidol (0 or 0.1 mg/kg). All drugs except haloperidol were synthesized by Novartis Pharma AG (Basel, Switzerland). Haloperidol was obtained from Sigma Chemical Co. (St. Louis, MO). All of the NMDA-RAs were dissolved in saline. Clozapine was dissolved initially in 0.1 N hydrochloric acid and isotonic saline and then titrated to a final pH of 6.0 with 0.1 N NaOH. Haloperidol was diluted with saline from a stock solution of 5 mg/ml. All doses were calculated as the salt. Injection volume was 1 ml/kg for all drugs.

Surgical Procedure

Rats were anesthetized with sodium pentobarbital (50 mg/ml; Abbott Labs, North Chicago, IL) and treated with 0.1 ml of methyl atropine bromide (0.5 mg/ml; Sigma Chemical) to minimize respiratory distress from the anesthetic agent. Bilateral stainless steel cannulas (23 gauge) were fixed to the skull using light-curable dental cement (Henry Schein, Port Washington, NY) and skull screws (Small Parts, Miami Lakes, FL). Coordinates, based on the atlas of Paxinos and Watson (1986), were as follows: AP, 2.6 mm from bregma; LM, ±4.9 mm from bregma; DV, 5.4 mm from skull surface. Wire stylets were placed inside the cannulas to prevent occlusion. Postoperative health checks and gentle handling were performed daily after surgery.

Intracerebral Infusion Procedure

On the test day, stylets were removed, cannulas were cleaned with a dental broach, and stainless steel injectors (30 gauge) were lowered through the cannulas so that the injector tip was positioned in the amygdala, 3 mm below the cannulas. Polyethylene tubing connected the injectors to 10-µl Hamilton syringes, which were mounted on a microdrive pump (Harvard Apparatus, South Natick, MA). The infusion bolus (1 µl) was delivered at a constant rate over 93 s, with a subsequent 60-s diffusion period to allow for absorption into the tissue before the removal of injectors. Immediately after the infusions, stylets were replaced, and the animals were placed into startle chambers.

Startle Apparatus

All testing occurred within San Diego Instruments (San Diego, CA) startle boxes, which consisted of clear nonrestrictive Plexiglas cylinders resting on a platform inside of a ventilated and illuminated chamber. A high-frequency loudspeaker inside the chamber produced both a continuous background noise of 65 db as well as the various acoustic stimuli. As described previously (Mansbach et al., 1988), the whole body startle response of the animal caused vibrations of the Plexiglas cylinder, which were then converted into analog signals by a piezoelectric unit attached to the platform. These signals were digitized and stored by a microcomputer and interface unit. Weekly calibrations were performed on the chambers to ensure the accuracy of the sound levels and measurements. Sound levels were measured as described previously (Mansbach et al., 1988) using the db (A) scale.

Startle Testing

One week after arrival, all rats underwent a brief startle session to create matched treatment groups. In this session and the subsequent test session, the background noise (65 db) was presented alone for 5 min and then continued throughout the remainder of the session. A total of 20 trials were presented in a pseudorandom order: 17 presentations of a 40-ms 120-db broadband burst and 3 trials in which a 77-db burst preceded the 120-db burst by 100 ms. Treatment groups were established using the mean startle response to the 120-db Pulse-Alone trial so that all groups had comparable baseline startle reactivity. At 1 to 2 days after the baseline session, drug testing began. The test session used in all of the experiments except the intra-amygdala infusion study contained six different trial types: a Pulse-Alone trial in which a 40-ms, 120-db broadband burst was presented; four Prepulse + Pulse trials in which 20-ms noises that were either 2, 4, 8, or 16 db above the background noise were presented 100 ms before the onset of the 120-db pulse; and a No Stimulus trial, which included only the background noise. All trial types were presented several times in a pseudorandom order for a total of 70 trials. In addition, several Pulse-Alone trials, which were not included in the calculation of PPI values, were presented at the beginning of the test session to achieve a relatively stable level of startle reactivity for the remainder of the session (based on the observation that the most rapid habituation of the startle reflex occurs within the first few presentations of the startling stimulus; Geyer et al., 1990). An average of 15 s (ranging from 9 to 21 s) separated consecutive trials. For the study of the amygdala, the same testing parameters were used, except that prepulse intensities were 3, 6, or 12 db above background noise. This session was used because it has been found previously to detect decreases in PPI produced by intracranial microinfusion of the noncompetitive NMDA-RA dizocilpine (Bakshi and Geyer, 1998).

Locomotor Activity Apparatus

Unless otherwise specified, the experimental chamber, computer, data reduction and analyses, pattern description, and pattern analyses were identical with those described previously (Flicker and Geyer, 1982; Geyer et al., 1986). Briefly, each behavior pattern monitor (BPM) consisted of a 30.5- × 61.0- × 38.0-cm black Plexiglas box with a stainless steel floor and a wall touch plate located 15 cm above the floor. Infrared photobeams were arranged in a grid pattern along the bottom of the walls, enabling the localization of the position of each animal with a resolution of 3.8 cm. Rearings were detected when the animals contacted the wall touch plate with their forepaws, making a connection between the floor and the touch plate. Each chamber was also equipped with seven wall holes (three per side wall, one on the back wall) and three floor holes, each 2.5 cm in diameter and each containing an infrared photobeam. Every 55 ms, a microprocessor system checked the status of all beams (broken or unbroken) and recorded and stored this information.

Locomotor Activity Testing

Six measures of locomotor activity were obtained from the BPM: crossings, rearings, holepokes, time in center, spatial CV, and spatial d (see below). The pattern of photobeam interruptions was used to calculate the x, y position of the animal and to thereby assign the rat to one of the eight square sectors and one of nine unequally sized regions into which the BPM was divided by the photobeams (compare with Flicker and Geyer, 1982). Crossings were defined as the total number of entries into any sector and used to assess the amount of horizontal locomotion. Rearings, as explained above, consisted of the animal standing on its hindlegs and touching the wall with its forepaws. Holepokes were detected as the interruption of an infrared photobeam located within any hole. The time in center referred to the total amount of time that the animal spent in the center regions of the BPM. Rearings, holepokes, and the time in center are conventionally used as measures of investigatory behavior.

Detailed characterization of the pattern of locomotor activity was obtained through calculation of spatial CV and spatial d. The spatial distribution and sequential patterns of locomotor activity were also examined graphically. Based on the x, y position data, a variable-speed plot program displayed the successive positions of the animal by the movement of a cursor inside a two-dimensional reconstruction of the chamber on a video terminal. Together with visual observations, this method provided a comprehensive description of the pattern of movements for each animal. The degree of redundancy in these spatial patterns and the frequency of transitions between any of the nine regions of the BPM were calculated to determine the coefficient of variation (spatial CV), which measured the distribution of these transition frequencies, as detailed elsewhere (Geyer, 1982; Geyer et al., 1986). Thus, as the animal repeated certain transitions preferentially, the spatial CV increased, whereas a more random pattern of transitions produced a lower spatial CV. The spatial CV therefore reflected the extent to which an animal exhibited a preferred or rigid pattern of locomotor activity.

Analysis of the average geometrical structure of movement patterns was performed by calculating the spatial scaling component, d. This measure was based conceptually on fractal geometry, describing the relative smoothness or roughness of the locomotor path (Paulus and Geyer, 1991). Briefly, the length of the entire locomotor path taken by an individual rat through the BPM was calculated from the raw data using several different spatial resolutions. The rate with which the calculated path length changed as a function of spatial resolution was obtained by least-squares fitting procedures. The fitted coefficient of the exponent of that function corresponds to d. Values for d increase when the locomotor path is rough (many directional changes per path length) and decrease when the path is smooth (fewer directional changes).

Data Analysis

The startle response to the 120-db burst was recorded for each Pulse-Alone and Prepulse + Pulse trial. Two measures were calculated from these data for each animal. First, the amount of PPI was calculated as a percentage score for each Prepulse + Pulse trial type: % PPI = 100  {[(startle response for Prepulse + Pulse trial)/(startle response for Pulse-Alone trial)] × 100}. Second, startle magnitude was calculated as the average response to all of the Pulse-Alone trials. PPI data were analyzed with either two-factor (treatment and trial type) or three-factor analysis of variance (ANOVA) with pretreatment and treatment as between-subject factors and trial type (prepulse intensity) as a repeated measure. Startle magnitude data were analyzed with one-factor (treatment) or two-factor (pretreatment and treatment) ANOVA. For analysis of locomotor activity, separate one-way ANOVAs were carried out for each of the activity indices (crossings, holepokes, rearings, spatial CV, spatial d, and time in center) with treatment as a between-subject factor. Posthoc analyses were carried out using Tukey's test. The  level was set at .05.

Histology

On completion of startle testing, cannulated animals were deeply anesthetized with sodium pentobarbital and perfused transcardially with isotonic saline followed by 10% formalin (Fisher Scientific, Pittsburgh, PA). Brains were removed, stored in formalin, and then sectioned into 60-µm slices using a freezing-stage sliding microtome (Leica Instruments, Deerfield, IL). Slices were mounted onto gelatin-coated slides, stained with Cresyl violet, and examined under a microscope to verify the placement of injector tips. During this analysis, the experimenter remained blind to the pharmacological treatment as well as the behavioral data. The data from animals that were determined to possess injector placements outside of the amygdala were excluded from statistical analyses.

Experimental Design

PPI Testing. Three startle studies were conducted using separate groups of rats. In the first, the effects of SDZ 220-581 and SDZ EAB-515 were assessed. Rats were given s.c. injections of either saline vehicle or 2.5 or 5.0 mg/kg SDZ 220-581 30 min before entering startle chambers. A separate group of animals was given the highest dose 75 min before testing to compare the relative efficacies of short versus long pretest injection intervals. One week later, all animals were retested with SDZ EAB-515. Each treatment group for this second experiment contained two animals from each of the previous (SDZ 220-581) treatment groups, so that all the previous treatments were equally represented in each of the new groups. In this experiment, rats received either saline or 3.0, 10.0, or 30 mg/kg SDZ EAB-515 s.c. 30 min before testing.

Locomotor Activity Testing

Rats from the first startle study (see study 1 of PPI testing) were placed into BPMs immediately after the startle session for assessment of locomotor activity. Thus, two locomotor activity experiments were conducted: first, examining the effects of SDZ 220-581, and, second, examining the effects of SDZ EAB-515. The duration of the prelocomotor activity startle session was 30 min. For both experiments, animals were placed gently by the experimenter into the front left quadrant of the BPM, and behavior was measured for a total of 30 min.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

Startle Testing

Study 1: Effects of SDZ 220-581 and SDZ EAB-515. Figure 1A illustrates the effects of SDZ 220-581 on PPI. As can be seen from the graph, SDZ 220-581 produced a robust and dose-dependent reduction in PPI. ANOVA revealed a main effect of treatment [F(3,28) = 29.97, P < .001]. Subsequent analyses indicated that at the 4- and 8-db prepulse intensities, both doses reduced PPI (P < .01, P < .05). At the highest prepulse intensity, however, only the 5.0 mg/kg dose decreased PPI (P < .01). It should be noted that this graded drug effect on PPI with increasing prepulse intensity is characteristic of several compounds, including the noncompetitive NMDA-RA PCP (Mansbach and Geyer, 1989; Bakshi et al., 1994). The high dose of SDZ 220-581 was very effective in disrupting sensorimotor gating, nearly abolishing PPI at the two middle prepulse intensities. Moreover, no apparent difference in efficacy was observed between the short (30 min) and long (75 min) pretest injection intervals, suggesting that a relatively short amount of time is sufficient for SDZ 220-581 to achieve bioactivity. In contrast to the marked effects on PPI, startle magnitude was left unaffected by SDZ 220-581 treatment [F(3,28) = 0.54, NS], indicating that the ability of this compound to disrupt PPI cannot be attributed simply to changes in baseline startle reactivity (Table 1).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Effects of (A) SDZ 220-581 and (B) SDZ EAB-515 on prepulse inhibition. Values represent mean ± S.E.M. for each group. Doses are in mg/kg. SAL, saline. *P < .05, **P < .01, compared with saline group.

Study 2: Effects of D-CPPene. The effects of the classic competitive NMDA-RA D-CPPene (SDZ EAA-494) on PPI are depicted in Fig. 2, A-C. In contrast to the previous two compounds, D-CPPene treatment had no effect on PPI with either a short (10 min, Fig. 2A), or long (60 min, Fig. 2B) pretest injection interval. No main effect of drug treatment on PPI was observed in either experiment [F(3,20) = 0.10, NS, short pretest injection interval; F(3,28) = 1.93, NS, long pretest injection interval]. A tendency to decrease startle magnitude was seen, particularly in the second experiment of this study (Table 1); however, this trend did not reach statistical significance for either the short [F(3,20) = 1.94, NS] or long [F(3,28) = 2.33, NS] pretest injection interval.

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Effects of D-CPPene on prepulse inhibition. A, s.c. administration, 10 min before testing. B, s.c. administration, 60 min before testing. C, Intra-amygdala infusion, immediately before testing. Values represent mean ± S.E.M. for each group. Doses for systemic experiments are in mg/kg; dose for amygdala experiment in µg/µl. SAL, saline. *P < .05, compared with saline group.

View larger version (58K): [in this window] [in a new window]   Fig. 3.   Reconstructions of coronal cross sections depicting the location of injector tips within the amygdala. Each dot represents injector placement for a different animal. Numbers denote distance (in mm) from bregma in the anterior-posterior plane. Sketches based on the atlas by Paxinos and Watson (1986).

Study 3: Effects of Antipsychotics on SDZ 220-581. The ability of various antipsychotic drugs to prevent the SDZ 220-581-induced deficit in PPI is shown in Fig. 4. A main effect of SDZ 220-581 treatment was revealed by ANOVA [F(1,76) = 71.69, P < .001], replicating the findings of study 1. In addition, a significant main effect of pretreatment was also seen [F(2,76) = 4.37, P < .02]. Subsequent separate three-way (with pretreatment, treatment, and prepulse intensity as factors) ANOVAs indicated that this effect was due to a main effect of haloperidol [F(1,52) = 8.95, P < .005] but not clozapine [F(1,48) = 2.83, NS] pretreatment. Although the pretreatment × treatment interaction did not reach statistical significance [F(2,76) = 1.64, NS], post hoc analyses were deemed justifiable because of the large main effects as well as a strong a priori hypothesis that antipsychotic administration would block NMDA-RA-induced deficits in PPI (Bakshi et al., 1994; Swerdlow et al., 1996). Thus, subsequent analyses indicated that animals treated with SDZ 220-581 had markedly lower levels of PPI at the 2-db (P < .05), 4-db (P < .01), and 8-db (P < .01) prepulse intensities than vehicle-treated controls. Because no drug effects were seen for the 16-db prepulse intensity, these data are not shown. Examination of the graph demonstrates the tendency of haloperidol to increase PPI at the 8-db prepulse intensities. This trend, however, was not statistically significant in the post hoc analyses. Most importantly, animals that received either clozapine (P < .05) or haloperidol (P < .05) before receiving SDZ 220-581 had significantly higher levels of PPI at the 8-db prepulse intensity than animals that received only SDZ 220-581 (Fig. 4). This effect was also seen at the 4-db prepulse intensity (for haloperidol, P < .05). These findings indicate that either clozapine, an atypical antipsychotic, or haloperidol, a traditional antipsychotic, can prevent the disruption of PPI produced by a competitive NMDA-RA. ANOVA of startle magnitude data revealed neither a significant main effect of pretreatment [F(2,76) = 2.12, NS] nor a pretreatment × treatment interaction [F(2,76) = 0.79, NS] but did indicate a main effect of treatment [F(1,76) = 20.15, P < .001]. Although post hoc analyses failed to confirm reliable differences, examination of the mean startle values indicates that SDZ 220-581 tended to increase startle magnitude and that this trend was not affected by antipsychotic pretreatment (Table 1).

View larger version (31K): [in this window] [in a new window]   Fig. 4.   Effects of clozapine and haloperidol on SDZ 220-581-induced deficit in prepulse inhibition. Values represent mean ± S.E.M. for each group. Doses are in mg/kg. VEH, vehicle; SAL, saline; CLOZ, clozapine; HP, haloperidol; 220-581, SDZ 220-581 (2.5 mg/kg). *P < .05, **P < .01, compared with VEH/SAL group. +P < .05, compared with VEH/220-581 group.

Locomotor Activity Testing. As illustrated in Table 2, the amount of activity induced by SDZ 220-581 and SDZ EAB-515 was significantly above control, as measured by the number of crossings [SDZ 220-581: F(3,27) = 12.49, P < .01; SDZ EAB-515: F(3,27) = 10.38, P < .001]. Measures of investigation (rearings and holepokes) were significantly decreased by SDZ 220-581[holepokes: F(3,27) = 17.69, P < .01; rearings: F(3,27) = 59.71, P < .01]. Rearings were also attenuated by SDZ EAB-515 [F(3,27) = 5.19, P < .01]. In contrast, SDZ EAB-515 produced a biphasic effect on holepoking, with a significant increase at the 3.0 mg/kg dose and a decrease at the highest dose tested (30.0 mg/kg) [F(3,27) = 5.93, P < .01]. Furthermore, SDZ 220-581-treated animals spent significantly less time in the center of the chamber than did controls [F(3,27) = 4.78, P < .01]. The same was true for animals treated with SDZ EAB-515 [F(3,27) = 4.24, P < .05] (Table 2). Both SDZ 220-581 and SDZ EAB-515 produced characteristic patterns of locomotion in which the animals rarely moved away from the walls. Figure 5 shows the typical pattern of locomotion exhibited by animals treated with vehicle, SDZ 220-581, or SDZ EAB-515. Control animals typically exhibited a preference for one area of the chamber, the "home" area, from which they made excursions to various parts of the BPM and back, following progressively more fixed routes over time (Geyer et al., 1986). Both SDZ 220-581 and SDZ EAB-515 disrupted this structure and produced abnormally repetitive patterns of locomotion. For example, 5.0 mg/kg SDZ 220-581 and 30.0 mg/kg SDZ EAB-515 produced very prominent patterns of stereotyped or perseverative ambulation, as displayed in Fig. 5. Computer reconstructions of the ambulatory pattern revealed that in most animals, the pattern consisted of circling the perimeter. ANOVA was applied to the spatial measures of the coefficient of variation (spatial CV) to quantify these differences in ambulation patterns; the effects of the compounds on the spatial CV are illustrated in Table 2. Both SDZ 220-581 and SDZ EAB-515 produced significant increases in the spatial CV, which reflect the repetitive patterns of locomotion induced by these drugs [SDZ 220-581: F(3,27) = 3.93, P < .05; SDZ EAB-515: F(3,26) = 8.99, P < .001].

View larger version (58K): [in this window] [in a new window]   Fig. 5.   Computer reconstructions of the patterns of locomotor activity induced by (A) saline, (B) 5.0 mg/kg SDZ 220-581, and (C) 30 mg/kg SDZ EAB-515.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

The primary findings of this work are that newly synthesized competitive NMDA-RAs SDZ 220-581 and SDZ EAB-515, which were designed to rapidly cross the blood-brain barrier, cause robust and reliable decreases in PPI without affecting baseline startle reactivity. In contrast, the classic competitive NMDA-RA D-CPPene failed in two separate studies to disrupt PPI when administered systemically, despite a tendency to decrease startle magnitude. Direct intracerebral (intra-amygdala) application of the same compound, however, produced a significant deficit in PPI that was independent of changes in startle magnitude. The disruption in PPI produced by SDZ 220-581 was blocked by both the typical and atypical antipsychotic agents, haloperidol and clozapine, respectively. Finally, SDZ 220-581 and SDZ EAB-515 also increased motor activity. Taken together, these findings indicate that if they gain rapid and sufficient access to the brain, competitive NMDA-RAs are capable of causing profound behavioral deficits similar to those seen after administration of the psychotomimetic noncompetitive NMDA-RA PCP.

Behavioral Effects of Systemically Administered NMDA-RAs. To date, several studies have indicated consistently that competitive NMDA-RAs do not disrupt PPI, even at doses that significantly decrease startle magnitude (Mansbach, 1991; Wedzony et al., 1994, McCloskey et al., 1995). Our results with s.c. administered D-CPPene are consistent with these findings; D-CPPene had no effect on PPI in the present study, regardless of the pretest injection interval or the tendency to decrease startle magnitude. This lack of efficacy contrasts markedly with the pattern of effects on PPI that is produced by noncompetitive channel-blocking NMDA-RAs such as phencyclidine, dizocilpine, or ketamine (Mansbach and Geyer, 1989; Geyer et al., 1990; Mansbach and Geyer, 1991; Bakshi et al., 1994): these compounds induce a dose-dependent and robust decrease in PPI that can be independent of startle magnitude effects (Bakshi et al., 1994).

Centrally Acting Competitive NMDA-RAs Disrupt Sensorimotor Gating: Implications for Psychotomimetic Potential. To the best of our knowledge, this is the first report of disrupted PPI after the systemic administration of competitive NMDA-RAs in which deficient PPI was not simply an artifact of severely decreased basal startle reactivity (Furuya and Ogura, 1997). Both SDZ 220-581 and SDZ EAB-515 produced a profound deficit in sensorimotor gating, nearly abolishing PPI at almost every prepulse intensity without affecting startle magnitude. Interestingly, the classic and highly selective competitive NMDA-RA D-CPPene (SDZ EAA-494) had no effect on PPI when administered systemically but did disrupt PPI when infused directly into the amygdala. Future studies examining the effects of CPPene infusion into other brain regions will aid in determining the anatomical specificity of the intracerebral CPPene effect. Our finding is consistent with three previous reports that have shown that another, less potent competitive NMDA-RA, AP-5 (Watkins, 1994; Urwyler et al., 1996a), also can disrupt PPI after intracerebral administration (Reijmers et al., 1995; Kretschmer and Koch, 1997; Wan and Swerdlow, 1997). Taken together, these findings indicate that if competitive NMDA-RAs gain access to the CNS, they produce marked deficits in sensorimotor gating. Interestingly, AP-7 has been found to generalize to the PCP discriminative stimulus when delivered into brain ventricles (Tricklebank et al., 1987). The present results thus provide compelling evidence for the hypothesis that the previously reported inability of competitive NMDA-RAs to disrupt PPI was simply an artifact of poor penetration into the brain. Given that the PPI paradigm is thought to provide a model for the disturbances in sensorimotor gating that are seen in schizophrenia (Geyer et al., 1990) and that the deficits in PPI induced by SDZ 220-581 were sensitive to the antipsychotics clozapine and haloperidol, the present results indicate that competitive NMDA-RAs might be hampered by the same considerations that compromise the clinical use of the noncompetitive NMDA channel blockers such as PCP, which disrupts PPI and produces psychosis in humans (Mansbach and Geyer, 1989; Javitt and Zukin, 1991).

Effects of Centrally Acting Competitive NMDA-RAs on Patterns of Locomotor Activity. The effects of SDZ 220-581 and SDZ EAB-515 on locomotor activity were generally similar to those seen previously with PCP and related noncompetitive NMDA-RAs (Tricklebank et al., 1989; Lehmann-Masten and Geyer, 1991). The administration of either competitive NMDA-RA caused increases in crossings and decreases in rearings; both of these effects have been observed with PCP-like compounds. In addition, a pattern of activity consistent with decreased exploratory behavior was noted with SDZ 220-581 and SDZ EAB-515: both drugs decreased holepokes as well as the time spent in the center of the chamber. The induction of certain stereotyped or perseverative behaviors might, through behavioral competition, account for this decrease in investigatory behavior. Specifically, both SDZ 220-581- and SDZ EAB-515-treated animals frequented the walls of the BPM in a perseverative manner, as reflected by the increase in the spatial CV measure, which is similarly increased by PCP and other channel-blocking NMDA-RAs (Lehmann-Masten and Geyer, 1991). Although this pattern of increased in locomotor activity, decreased investigatory responses, and increased perseverative patterns of ambulation is very similar to the profile of effects seen with PCP-like noncompetitive NMDA-RAs, it is important to note that certain subtle effects that differentiate SDZ 220-581 and SDZ EAB-515 from PCP were also observed. Specifically, the spatial d measure, which essentially gives an estimate of the smoothness of the animal's path (low d indicates smoothness), is known to be altered in a biphasic fashion by PCP-like noncompetitive antagonists, with low doses decreasing d and high doses increasing d (Lehmann-Masten and Geyer, 1991). This measure was dose-dependently decreased by either SDZ 220-581 or SDZ EAB-515 in the present dose ranges. It will be of interest to determine in future studies whether even higher doses of these compounds will increase d and thereby mimic the pattern of effects produced by PCP on this measure.

Development of NMDA Receptor Ligands Lacking Effects on Sensorimotor Gating. The objective in designing the new competitive NMDA-RAs SDZ EAB-515 and SDZ 220-581 was to create compounds that were less polar and had higher affinity for the glutamate recognition site than the older competitive NMDA-RAs such as AP-5 (Urwyler et al., 1996a). Ironically, it appears from the present study that the same structural modifications that are thought to impart a superior blood-brain barrier penetration and neuroprotective profile to these compounds may also impart a greater potential for eliciting psychotic symptoms. Although these compounds have not yet been tested clinically, it has been reported in one study that in epileptic patients, D-CPPene can produce cognitive impairments (Sveinbjornsdottir et al., 1993). Thus, either compounds that block the cation channel (PCP) or antagonize the glutamate recognition site (SDZ 220-581, SDZ EAB-515) of the NMDA receptor complex disrupt sensorimotor gating. It has been suggested that low affinity blockade of the NMDA channel (with compounds such as memantine) will not produce psychotomimetic effects (Grant et al., 1996); it may be that low affinity compounds are less likely to be psychotomimetic because they do not achieve rapid and sufficient brain concentrations to fully block the receptor. Interestingly, antagonists for the glycine modulatory site on the outside of the NMDA receptor complex have been found to actually improve deficient sensorimotor gating in rats (Bristow et al., 1995), a finding that is consistent with earlier suggestions that this site may be a good target for the development of novel neuroprotective and antipsychotic drugs (Iversen and Kemp, 1994). Nevertheless, a recent report suggests that when administered intracerebrally, this type of NMDA-RA can also disrupt PPI under certain conditions (Kretschmer and Koch, 1997). It remains to be seen whether a meaningful separation between the likely therapeutic benefits of blocking the NMDA receptor and psychotomimetic liability can be achieved in compounds having selectivity for the potentially numerous subtypes of the receptor that have been recently identified (Moriyoshi et al., 1991; Monyer et al., 1992).