Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

An accumulating body of evidence has indicated that recruitment of activity in the prefrontal cortex (PFC) is necessary for higher cognitive functions, particularly those that require short-term working memory (Baddeley, 1992; Frith and Dolan, 1996). In addition, deficits in working memory have been identified in several neurological and psychiatric disorders and are often associated with decreased functioning of the PFC (Baddeley et al., 1991; Weinberger and Berman, 1996). Several lines of evidence suggest that deficits in glutamatergic excitatory transmission in PFC may be involved in working memory impairment. For example, working memory performance in rodents is diminished by reducing glutamate release or its postsynaptic action on -amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors of PFC neurons (Romanides et al., 1999). Other experiments have shown that administration of N-methyl-D-aspartate (NMDA) glutamate receptor antagonists or agonists can, respectively, reduce or enhance retention times in working memory tasks in monkeys (Dudkin et al., 1997). Analogous brain imaging experiments in humans have shown that administration of NMDA receptor blockers (e.g., phencyclidine, ketamine) disrupt working memory in conjunction with a selective decrease in PFC activity (Krystal et al., 1994). In light of these findings, enhancement of glutamatergic transmission in PFC has been proposed as a strategy for the treatment of certain types of cognitive dysfunction (Yamada, 1998).

Historically, efforts to enhance glutamatergic transmission have focused on compounds from two chemical classes: pyrrolidinones (e.g., aniracetam, piracetam) and benzothiadiazines (e.g., cyclothiazide), which positively modulate AMPA receptor-mediated currents by suppressing the desensitization process of these receptor/channels (Ito et al., 1990; Copani et al., 1992; Yamada and Tang, 1993). The functional consequence of these compounds is to augment glutamatergic excitatory postsynaptic potentials (EPSPs) by increasing ion flux through AMPA receptors (Ito et al., 1990; Yamada and Tang, 1993). Similar potentiation of AMPA receptor activity and glutamatergic EPSPs has been reported for a newer class of benzoylpiperidine compounds typified by 1-(quinoxalin-6-ylcarbonyl)-piperidine (CX516), which can penetrate the blood-brain barrier (Stäubli et al., 1994a,b; Arai and Lynch, 1998a,b). Moreover, CX516 has been shown to facilitate the induction of NMDA-dependent, long-term potentiation (LTP) in the hippocampus in vivo (Stäubli et al., 1994a). Although not established, the facilitation of LTP has been postulated to arise from an increase in the activity of voltage-sensitive NMDA receptors as a consequence of the greater depolarization produced by AMPA receptor potentiation (Stäubli et al., 1994a,b). Such changes in synaptic plasticity may provide a cellular substrate through which positive modulation of AMPA receptors could promote memory encoding. Consistent with this hypothesis, behavioral studies have demonstrated that CX516 and other potentiators can improve performance in rodents on a variety of memory tasks including those that require working memory (Pontecorvo and Evans, 1985; Stäubli et al., 1994b; Hampson et al., 1998). More importantly, several clinical studies have shown the utility of AMPA receptor potentiators as nootropic agents (Dimond et al., 1979; Oepen et al., 1985; Ingvar et al., 1997). Thus, it is possible that some of the actions of these AMPA receptor potentiators may be mediated in the PFC.

Recently, a novel class of biarylpropylsulfonamide compounds (e.g., LY392098, LY404187) have been developed that positively modulate AMPA receptors (Bleakman et al., 2000; Baumbarger et al., 2001) (Fig. 1). These compounds have been shown to augment the amplitude of the desensitized current recorded from human recombinant homomeric AMPA receptors with nanomolar potency compared with millimolar potency for CX516 (Bleakman et al., 2000). In addition, the potentiation by biarylpropylsulfonamides of both recombinant and native receptors is markedly different from that by CX516 and other potentiators in that the magnitude of the enhancement is time-dependent, increasing with continued agonist exposure (Bleakman et al., 2000; Baumbarger et al., 2001). These data suggest that biarylpropylsulfonamides may affect the desensitization process of AMPA receptors via a distinct mechanism of action. In the present experiments, the actions of the biarylpropylsulfonamide LY404187 were compared with CX516 on the activity of native AMPA receptors of PFC neurons with respect to potency, biophysical mechanisms of action, and ability to potentiate AMPA and NMDA receptor-mediated synaptic responses in vitro and in vivo.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Chemical structures of LY404187 (A) and CX516 (B).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Acutely Isolated PFC Pyramidal Neurons: Preparation and Recording Procedures

Neurons from young (14- to 22-day-old) rats were acutely isolated from the PFC using standard procedures. Male Sprague-Dawley rats were deeply anesthetized with methoxyflurane and decapitated. Their brains were removed rapidly from the skull and immersed in a cold (~2°C) NaHCO₃-buffered saline solution (concentrations in mM): NaCl 126.0, KCl 3.0, MgCl₂ 1.5, Na₂PO₄ 1.25, CaCl₂ 2.0, NaHCO₃ 26.0, glucose 10.0; pH = 7.4, osmolarity = 300 ± 5 mOsm/l. The brains were blocked, and 400-µm-thick coronal sections were cut through the rostrocaudal extent of the PFC using a Vibroslice (Campden Instruments, London, England). Slices then were incubated at room temperature (20-22°C) for 0.5 to 6 h in a holding chamber containing the continuously oxygenated (95% O₂:5% CO₂) NaHCO₃-buffered saline solution. Following the incubation period, slices were transferred to a glass Petri dish containing a low Ca²⁺ HEPES-buffered saline solution (concentrations in mM): NaHOCH₂CH₂SO₃ (sodium isethionate) 140.0, KCl 2.0, MgCl₂ 4.0, CaCl₂ 0.1, glucose 23.0, HEPES 15.0; pH = 7.4, osmolarity = 300 ± 5 mOsm/l and placed under a dissecting microscope. The PFC from each hemisphere was dissected from the surrounding cortex. The tissue was placed into a holding chamber containing protease type XIV (1 mg/ml; Sigma, St. Louis, MO) dissolved in a HEPES-buffered Hanks' balanced salt solution (no. 6136; Sigma) maintained at 37°C and oxygenated (100% O₂), pH = 7.4, osmolarity = 300 ± 5 mOsm/l. Following 30 to 40 min of incubation in the enzyme solution, the cortex was rinsed three times with the low Ca²⁺ HEPES-buffered saline solution and triturated using two fire-polished Pasteur pipettes having tips of decreasing diameter. For experiments examining NMDA responses, tissue was incubated in trypsin (10 mg/ml) and the NMDA receptor antagonist 3-[(RS)-2-carboxypiperazin-4-yl]-propyl-1-phosphonic acid (3 µM) to preserve the response. Prior to whole-cell recording, the cell suspension was placed into a 50-mm transparent plastic Petri dish that was mounted onto the stage of an inverted microscope. Pyramidal neurons were selected on the basis of their triangular somatic shape, soma size (~20-30 µm in diameter), and presence of some apical and basal dendrites.

The whole-cell variant of the patch-clamp technique was used for recording current from acutely isolated PFC pyramidal neurons. Electrodes were pulled from borosilicate capillary tubing (Corning 7052, WPI Inc., Sarasota, FL) using a multistage puller (Sutter Instruments Inc., Novato, CA). The electrodes were fire-polished using a microforge (Narishige Inc., Tokyo, Japan) prior to use. The internal electrode filling solution contained (concentrations in mM): N-methyl-D-glucamine 160.0, MgCl₂ 4.0, HEPES 40.0, BAPTA 3.0, phosphocreatine 12.0, Na₂ATP 2.0, GTP 0.2; pH was adjusted to 7.2 with KOH and osmolarity adjusted to 270 to 280 mOsm/l. The extracellular solution contained (concentrations in mM): sodium isethionate 140.0, KCl 1.0, BaCl₂ 5.0, MgCl₂ 1.0, HEPES 10.0, tetrodotoxin 0.001; pH adjusted to 7.4 with 1.0 M NaOH; osmolarity adjusted to 300 ± 5 mOsm/l with glucose.

Upon placing the recording electrode in the bath, offset potentials were corrected and electrode resistances ranged between 2 and 7 M. Voltage-clamp recordings were made using an Axon Instruments 200B amplifier (Axon Instruments Inc., Foster City, CA). The membrane potential of cells was held at 80 mV unless stated otherwise. Currents were digitized and monitored with pCLAMP software version 8.0 (Axon Instruments Inc.) running on a PC Pentium computer. A small amount of constant positive pressure (2-3 cm of H₂O) was applied to the electrodes as they were advanced through the bath. After achieving the whole-cell configuration, series resistance was compensated (70-85%) and monitored periodically.

Application of drugs was accomplished using one of two methods. A 16-barrel pipette array made from small diameter (~600 µm) glass capillary tubing was used to study the desensitized AMPA responses in PFC neurons. Solutions were contained in 10-ml syringes and positioned approximately 12 inches above the recording chamber. Gravity-induced flow of each solution from the syringe to the corresponding barrel was controlled by electronic valves. The pipette array was positioned 100 to 200 µm from the cell prior to seal formation. The solutions from the drug array were changed (~100 ms) by altering the array position with a d.c. actuator (Newport Inc., Irvine, CA). A fast application system was used to study the desensitization kinetics of AMPA receptor-mediated responses in PFC neurons. The application barrels were constructed from double-barrel borosilicate glass theta tubing. The theta tubing was pulled so that the final diameter of each barrel was approximately 125 µm and the septum between barrels was approximately 80 µm. A piezoelectric actuator (Burleigh Instruments Inc., Fishers, NY) was used to shift the theta tubing. The charging time of the actuator was set to 0.3 ms, and the onset-to-offset time of junction potential measurements (0.1 M and 1.0 M NaCl) was approximately 700 µs. A six-to-one manifold was attached to the input of each barrel of the theta tubing to permit multiple solutions to be applied to an individual neuron. Solutions were delivered to the theta tubing using the gravity-induced flow method described above. All experiments were conducted at room temperature.

PFC Slices: Preparation and Recording Procedures

Coronal PFC slices were prepared from young (14- to 22-day-old) male Sprague-Dawley rats. Animals were sacrificed, and their brains were removed and sectioned (at 300 µm) as described above. Slices were placed into the continuously oxygenated NaHCO₃-buffered saline solution warmed to 32°C for 30 min and then maintained at room temperature. After at least 1 h of incubation, individual slices were transferred to a recording chamber mounted on an upright microscope (BX50WI, Olympus Optical Co., Tokyo, Japan) and continuously superfused (2-3 ml/min) with the oxygenated saline solution maintained at 30 ± 0.2°C. Differential interference videomicroscopy was used to visualize pyramidal neurons in the PFC. Whole-cell current-clamp recordings were conducted using procedures similar to those described above. Patch pipettes were fabricated from thin-walled borosilicate glass and had resistances of 1 to 4 M when lowered into the extracellular solution. The pipette solution contained (in mM): K⁺-gluconate 130.0, KCl 10.0, MgCl₂ 2.0, EGTA 1.0, HEPES 10.0, Na₂ATP 2.0, Na₂GTP 0.3; pH adjusted to 7.3 with 1 M NaOH, osmolarity of 290 to 300 mOsm. For some experiments, voltage-dependent Na⁺ currents were blocked by inclusion of QX314 (1 mM) in the internal solution. The extracellular solution contained (in mM): NaCl 125.0, KCl 3.0, CaCl₂ 2.4, MgCl₂ 1.3, NaHCO₃ 26.0, glucose 10.0; pH adjusted to 7.4 with 1 M NaOH, osmolarity of 300 ± 5 mOsm. For some experiments, -aminobutyric acid_{A and/or B} receptors were blocked with bicuculline methiodide (1-2 µM) and SCH59011 (2 µM), respectively. Voltage signals were amplified by an Axoclamp 200B amplifier, low-pass filtered at 5 kHz, and stored on the computer hard-disk for off-line analysis (Clampfit 8.0, Axon Instruments Inc.). Series resistance (10-30 M) compensation was monitored, and recordings displaying >30% change in resistance were not included in subsequent analyses. Voltage errors due to the liquid junction potential were subtracted during analysis. Postsynaptic potentials were evoked by constant current single stimulation pulses (100 µs, 50-500 µA) delivered with a 20-s interstimulus interval using bipolar stimulating electrodes positioned in the cell body layer.

Extracellular Single-Unit Recording from PFC

Male Sprague-Dawley rats (250-320 g) were anesthetized with urethane (1.5 g/kg, i.p.) and mounted on a stereotaxic frame. Core temperature was monitored by a rectal probe and maintained at 37°C by a heating pad. Burr holes were drilled into the skull over the PFC and the ventral subiculum of the hippocampus. The stereotaxic coordinates for the PFC and the ventral subiculum were: PFC: A-P = 2.7 to 3.0 mm anterior to the bregma, L-M = 1 to 1.2 mm, D-V = 2.0 to 3.5 mm; ventral subiculum: A-P = 6.0 mm posterior from the bregma; L-M = 5 mm from the midline, and D-V = 7 mm from the cortical surface. A concentric bipolar stimulation electrode (NE-100, Rhodes Medical Instruments, Inc., Woodland Hills, CA) was positioned stereotaxically in the ventral subiculum. Stimulation currents (20-µs duration, 300-900 µA) were delivered via an optically isolated stimulation unit (Isoflex, A.M.P.I., Jerusalem, Israel). Single stimulation pulses were delivered at 1 Hz.

Conventional extracellular single-unit recordings were made using filament-filled borosilicate glass capillary tubing. The electrodes were pulled from a glass micropipette (o.d. = 1.5 mm, i.d. = 1.17 mm; Hilgenberg GmbH, Malfeld, Germany) using a multistage microelectrode puller (Sutter Instruments Inc.) in two stages. The electrodes were filled with 0.5% sodium acetate in 2% Pontamine Sky blue and 0.5 µM bicuculline methiode to partially block local GABA_A receptor-mediated inhibitory responses. The electrode was advanced by a single-axis Narishige MHW-40 hydraulic micromanipulator (Narishige Inc.) mounted onto the stereotaxic frame.

Extracellular single-unit activity was amplified by an Xcell-3 Plus amplifier (Frederick Haer & Co., Brunswick, ME). Single-unit activity was isolated using a window discriminator (model 74-60-3, Frederick Haer & Co.). The output signals from the window discriminator were digitized and multiplexed by an analog to digital converter (1401 mini, Cambridge Electronics Design Ltd., Cambridge, UK) and were sampled at 10 kHz by a PC-based computer using Spike 3 software (Cambridge Electronics Design Ltd.).

At the end of each recording session, direct current (100 µA for 1 min) was delivered to make an iron deposit at the stimulating electrode tip. Direct current (10 µA for 15 min) also was delivered to iontophorese Pontamine Sky blue through the recording electrode to mark the recording site. The animal then was perfused with saline, followed by buffered formalin. Brain sections (70 µm) containing the PFC and the ventral subiculum were cut using a freezing microtome, and the sections were washed, dehydrated with alcohol, and stained with cresyl violet to permit examination of the recording and stimulation sites.

Drug Preparation and Administration

In Vitro Experiments. Concentrated stock solutions for all compounds were prepared by dissolving in 100% dimethyl sulfoxide. The solution then was aliquoted and stored at 20°C until needed. On the day of recording, solutions were diluted to the desired concentrations in the NaHCO₃-buffered saline solution. When low nanomolar concentrations were tested, serial dilutions were performed to assure accuracy.

In Vivo Experiments. Stock solutions of LY404187 (0.1 mg/ml) and CX516 (10.0 mg/ml) were made by dissolving the compound in ethanol (200 µl) and then adding 0.9% NaCl and 15% 2-hydroxypropyl--cyclodextran (HC; Sigma) to bring the final stock solution volume to 5 ml. The solutions then were aliquoted and stored at 20°C until used.

Data Analysis

In Vitro Data. Concentration-response profiles for all potentiators were constructed by measuring the peak current amplitude during a 10-s coapplication of compound and glutamate (100 µM), calculating the percent increase relative to the glutamate alone response and plotting the data as a function of potentiator concentration. The plotted points then were fit with a logistic equation of the form: percent potentiation = E_max/(1 + ([LY404187]/EC₅₀))ⁿ, where the maximal percent potentiation is relative to the current evoked by glutamate alone, EC₅₀ is the concentration equal to 50% of the maximally effective concentration, and n is the Hill coefficient. The best fit was chosen using the Marquardt-Levenberg algorithm. Average EC₅₀ and E_max values were determined and reported as mean ± standard deviation.

In Vivo Data. For in vivo recording experiments, peristimulus time histograms (PSTH, bin width = 1 ms) were constructed for spontaneous and evoked responses of PFC neurons to single-pulse stimulation (at 1 Hz, 300 sweeps) of the ventral subiculum. The mean bin height and the standard deviation 50 ms before the onset of the stimuli was defined as the prestimulus control response for each recording. The beginning of a significant poststimulus response period was defined as the first three consecutive bins whose heights were 1 standard deviation above (for excitatory responses) or below (for inhibitory responses) the mean prestimulus control bin height. Likewise, the end of the significant response period was defined as the first three consecutive bins whose heights were within 1 standard deviation of the control mean.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Potency and Time Course of Potentiation by LY404187 and CX516. Initial studies evaluated the potency and efficacy of LY404187 (0.03-10 µM) on the glutamate-evoked responses of PFC neurons by recording the inward current evoked in response to application of glutamate (100 µM, 10-s duration; holding potential = 80 mV) alone and in the presence of the compound. Preliminary experiments demonstrated that this concentration of glutamate was equal to 30% of the maximal response (EC₃₀). Because of the relatively slow solution switching speed (~100 ms) of the actuator used in these experiments, the responses recorded reflect desensitized currents. At all concentrations tested, application of LY404187 alone had no effect on the holding current (Fig. 2A). However, when applied in the presence of glutamate, LY404187 enhanced the evoked current in a concentration-dependent manner (Fig. 2A). As previously reported for biarylsulfonamides on both recombinant and native receptors (Bleakman et al., 2000; Baumbarger et al., 2001), the potentiated response by LY404187 displayed a marked time dependence such that a steady-state level was never achieved during the 10-s glutamate stimulus. Because of this property, the data were expressed as a percent change in peak amplitude from that of the glutamate response alone and plotted as a function of compound concentration. As such, the values for potency and efficacy are estimates. The concentration-response profile shows that LY404187 potentiated glutamate responses with an EC₅₀ value of 1.3 ± 0.3 µM and a maximal efficacy (E_max) of a 45.3 ± 8.0-fold (n = 6) increase relative to glutamate alone (Fig. 2C). Consistent with a selective effect on AMPA receptors (also see below), the potentiator also enhanced AMPA-evoked currents (5 µM = EC₃₀) with the same time dependence as that of glutamate-evoked currents (Fig. 2A, inset). In addition, the concentration-response profile revealed that LY404187 potentiated AMPA-evoked currents with similar potency (EC₅₀ = 1.2 ± 0.4 µM) and efficacy (E_max = 45.2 ± 6.8-fold increase, n = 5) to that of glutamate-induced responses.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Concentration-response profile for LY404187 and CX516 potentiation of AMPA receptor-mediated currents. A and B, the concentration-response profiles for LY404187 (0.03-10.0 µM) and CX516 (0.1-30.0 mM) potentiation were constructed by measuring the responses to 100 µM glutamate (GLU) alone and in the presence of potentiator. Both compounds enhanced GLU responses in concentration-dependent manner. Insets, both compounds potentiated responses to AMPA (5 µM) in a manner similar to that of GLU. C, plot of the average degree of potentiation as a percentage of the GLU (100 µM) response for each concentration of compound tested. Analysis revealed that LY404187 was more potent and efficacious than CX516 with the compounds having EC50 values of 1.2 µM and 2.8 mM and maximal efficacies of 42- and 5-fold, respectively. The effects of cyclothiazide are presented for comparison. D, the concentration threshold for potentiation of AMPA responses by LY404187 was assessed by comparing the peak responses to application of AMPA (5 µM) alone and in the presence of the potentiator (3.0 and 10.0 nM). Plot of the average peak amplitude response in each condition showed that significant increases in the AMPA-evoked responses were observed with coapplication of 10.0 nM LY404187.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Activity and time dependence of enhancement by AMPA potentiators. A, the activity and time dependence of the AMPA receptor potentiation produced by LY404187 (1 µM), cyclothiazide (300 µM), and CX516 (3 mM) were evaluated by applying 5 µM AMPA (10-s duration) alone and in the presence of the potentiator. None of the compounds produced an effect alone, but all compounds potentiated the AMPA-evoked currents, suggesting an activity dependence to their action. Unlike cyclothiazide and CX516, the potentiation produced by LY404187 displayed a marked time dependence during the 10-s application. B, this property of the potentiation was explored by applying pulses of AMPA with durations ranging from 3 to 40 s in the presence of 3 µM LY404187. Even in response to a 40-s AMPA pulse, the potentiation by LY404187 never reached a steady-state level. C, inclusion of LY404187 (3 µM) in the recording pipette solution followed by a 5-min dialysis of the neuron did not potentiate extracellular AMPA-evoked responses. However, subsequent application of LY404187 extracellularly did potentiate AMPA responses, indicating an extracellular site of action. D, responses to application of AMPA (5 µM) before, during, and after application of LY404187 (10 µM). Following elimination of LY404187, the responses to AMPA returned to control levels. E, plot of the recovery from potentiation by LY404187. Points reflect the average responses to AMPA before, during, and after application of the compound expressed as a percentage of the response amplitude before potentiator and plotted as a function of time. Points represent mean ± S.D.

LY404187 Selectively Potentiates AMPA Receptor-Mediated Currents. The selectivity of LY404187 for potentiating AMPA receptor-mediated currents was tested by attempting to block the potentiation with the selective AMPA receptor antagonist LY300168 (50 µM). Application of 100 µM glutamate elicited a small inward current (101.7 ± 25.6 pA, n = 7) that was reduced 90.4% by LY300168 (9.3 ± 6.4 pA). Following wash of the antagonist, subsequent application of LY404187 (3 µM) potentiated the response to AMPA (1689.2 ± 562.8 pA) in the same neurons. This potentiated response was reduced 98.0% by LY300168 (34.3 ± 11.3 pA) (Fig. 4, A and C).

View larger version (30K): [in this window] [in a new window]   Fig. 4.   LY404187 selectively potentiates AMPA receptor-mediated currents. A, the response to application of 100 µM glutamate (GLU) was blocked by 50 µM LY300168. Following elimination of LY300168, subsequent application of 3 µM LY404187 potentiated the GLU response. This potentiated response was blocked by 50 µM LY300168. B, when the membrane potential of neurons was held near 0 mV, application of 200 µM NMDA (in the presence of 30 µM glycine) evoked an inward current that was not potentiated by LY404187. In these same neurons, when the membrane potential was held at 80 mV, the response to 5 µM AMPA was potentiated by 3 µM LY404187. C and D, plots of the average peak amplitudes for each condition in A and B. E and F, plots of the peak whole-cell potassium and sodium currents recorded during control conditions and in the presence of LY404187 (0.1-10.0 µM). No change in these conductances was observed.

LY404187 Increases Agonist Potency. Previous studies have shown that positive modulators of AMPA receptors (e.g., cyclothiazide) can exert their effects in part by producing a leftward shift in the agonist concentration-response relationship (Yamada and Tang, 1993; Partin et al., 1994). The possibility that LY404187 produced similar effects on the responses of pyramidal neurons was investigated. The normalized concentration-response profile for glutamate (0.3-100 µM) alone had an average EC₅₀ value of 272.7 ± 178.5 µM (n = 6). The glutamate concentration-response relationship was shifted leftward by LY404187 (1 µM) approximately 6-fold having an EC₅₀ value of 44.1 ± 16.1 µM (n = 9) (t = 3.6, p < 0.01) (Fig. 5). The magnitude of AMPA receptor responses also was increased dramatically by LY404187; the response to 3 mM glutamate increased by 12-fold from 249.5 ± 68.2 pA (n = 6) during control conditions to 3057.2 ± 1054.8 pA (n = 9) (t = 6.1, p < 0.01) in the presence of the potentiator. Tests for a similar shift in agonist potency by CX516 were conducted. Results showed that the potency of glutamate for PFC AMPA receptors was increased approximately 3-fold by CX516 (3 mM) (EC₅₀ = 80.3 ± 20.2 µM; n = 8, t = 8.5, p < 0.0001). These data also can be interpreted within the context of results from previous receptor binding experiments with biarylpropylsulfonamides (Lindén et al., 2001). In these studies, specific binding of [³H]LY359153 (a structural analog of LY404187) to cortical membranes was low but increased 5- to 10-fold in the presence of escalating concentrations of AMPA with an EC₅₀ value of 5.5 µM. A similar observation was reported for [³H]AMPA binding in the absence and presence of aniracetam (Nicoletti et al., 1992). The present results demonstrate a functional correlate to these binding experiments, such that the potentiating effect of LY404187 on AMPA currents increased as a function of agonist concentration.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   LY404187 increases the potency of glutamate at PFC AMPA receptors. A, concentration-response of glutamate (0.01-10.0 mM) alone. B, concentration-response of glutamate in the presence of LY404187 (3 µM). Note that the activity and time dependence of LY404187 potentiation is evident regardless of glutamate concentration. C, plot of the normalized peak responses as a function of glutamate concentration in the absence and presence of LY404187 (3 µM). The potentiator produced an approximately 6-fold leftward shift in the concentration-response relationship of glutamate at PFC pyramidal cell AMPA receptors.

LY404187 and CX516 Differentially Alter Desensitization of AMPA Receptors. The pronounced augmentation of glutamate responses by LY404187 suggested that the compound should alter the desensitization process of AMPA receptors in PFC pyramidal neurons, whereas the more modest potentiation by CX516 suggested that it would have much less of an effect on the degree of desensitization. To test these hypotheses, AMPA receptor desensitization was studied in isolated PFC neurons by delivering 500-ms glutamate (1 mM) pulses alone and in combination with the potentiators using a piezoelectric fast perfusion system. The responses to glutamate were due entirely to activation of AMPA receptors as evidenced by their complete blockade in the presence of 50 µM LY300168 (Fig. 6A, inset). The 500-ms glutamate pulse evoked an inward current that reached a peak within milliseconds and then decayed to a steady-state value (measured at the end of the glutamate pulse) that was 8.4 ± 5.5% of the peak response (Fig. 6A). A single exponential function was used to fit the current decay, and the average time constant of desensitization was 15.4 ± 3.7 ms (n = 10) during control conditions. Subsequent application of concentrations of LY404187 (0.1, 0.3, and 1.0 µM) at or lower than the EC₅₀ value did not alter the time constant of desensitization (0.1 µM = 15.1 ± 4.8 ms; 0.3 µM = 17.2 ± 6.1 ms; 1.0 µM = 18.2 ± 8.2 ms) (Fig. 6B). However, application of LY404187 enhanced the magnitude of the steady-state current relative to the peak current amplitude in a concentration-dependent manner (0.1 µM = 23.3 ± 9.9%; 0.3 µM = 32.3 ± 13.7%; 1.0 µM = 56.9 ± 19.3%, F_3,35 = 24.1, p < 0.0001) (Fig. 6C). The peak amplitude also was increased by LY404187 relative to control levels (0.1 µM = 17.1 ± 19.2%; 0.3 µM = 39.5 ± 25.0%; 1.0 µM = 72.3 ± 23.2%) (F_3,2 = 9.5, p < 0.0001) (Fig. 6D). The increase in peak amplitude most likely indicates the presence of receptor desensitization in the rising phase of the current during control conditions (Raman and Trussell, 1995; Partin et al., 1996).

View larger version (39K): [in this window] [in a new window]   Fig. 6.   LY404187 and CX516 differentially affect desensitization of AMPA receptors in PFC pyramidal neurons. A and E, whole-cell responses to 500-ms application of glutamate (1 mM) in the absence and presence of LY404187 (0.1-1.0 µM) and CX516 (0.3-3.0 mM) are presented. Inset, the glutamate-evoked current was completely blocked by LY300168 (50 µM). B and F, the time constant of desensitization of glutamate-evoked responses in the absence and presence of potentiator was determined by fitting the current decay with a single exponential function. Because of the time-dependent nature of LY404187 potentiation, for the 1 µM LY404187 responses, the boundaries of the fit were from the peak of the response to 200 ms after the stimulus. CX516, but not LY404187, increased the time constant of desensitization. C and G, the degree of desensitization was evaluated by expressing the amplitude of the steady-state current as a percentage of the peak current amplitude in the absence and presence of the potentiators. LY404187, but not CX516, significantly decreased the magnitude of desensitization. D and H, a change in the initial peak response to glutamate application was evaluated by expressing the peak amplitude in the presence of each concentration of potentiator as a percentage of the control peak amplitude. LY404187 and CX516 enhanced the peak response, indicating that desensitization occurs during the rising phase of the response.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   LY404187 decreases the time course of recovery from desensitization. A and C, pairs of glutamate pulses (10 ms, open boxes) having ISIs of 30 to 500 ms were delivered to PFC pyramidal neurons in the absence and presence of LY404187 (1.0 µM) and CX516 (3.0 mM). B and D, an average recovery ratio was measured for each ISI by dividing the peak of the second response by that of the first response in the absence and presence of potentiator for each cell tested. The average data points for each condition then were fit with a single exponential function to determine the time constant of recovery from desensitization. Results showed that 1.0 µM LY404187 reduced the time course of the recovery process (control = 109.5 ± 57.6 ms; 1.0 µM LY404187 = 49.8 ± 13.8 ms).

LY404187 Enhances AMPA Receptor-Mediated Synaptic Responses in Vitro. The robust enhancement of glutamate-evoked currents by LY404187 in the fast perfusion experiments suggested that the compound should similarly potentiate AMPA-mediated synaptic responses in PFC pyramidal neurons in a slice preparation. To test this hypothesis, synaptic responses of PFC pyramidal neurons were evoked by stimulation of fibers in the proximal apical dendritic region of these neurons. Because of the dense recurrent excitatory collateral network among PFC neurons, the synaptic response is comprised of both mono- and polysynaptic components, yielding a complex EPSP (Fig. 8A). In initial experiments, application of 0.1 to 3.0 µM LY404187 enhanced the subthreshold synaptic response in PFC pyramidal neurons to the point of action potential discharge. The effects of LY404187 were reversible within 15 to 30 min of eliminating the potentiator from the extracellular solution (Fig. 8A). Subsequent experiments focused on the selective potentiation of AMPA receptor-mediated EPSPs by recording from slices in which GABA_{A and B} receptors and NMDA receptors were blocked with bicuculline (1 µM), SCH50911 (2 µM), and APV (50 µM), respectively. In addition, action potentials were blocked by including QX314 (1 mM) in the internal recording solution. Stimulation of glutamatergic inputs elicited a complex EPSP having an average peak amplitude of 9.6 ± 3.6 mV in control solution. This EPSP was enhanced in a concentration-dependent manner to 11.5 ± 10.5 mV (19.8%), 15.2 ± 6.3 mV (58.3%), 18.1 ± 7.4 mV (88.5%), and 23.2 ± 6.0 mV (141.7%) in the presence of 0.1, 0.3, 1.0, and 3.0 µM LY404187, respectively (F_4,33 = 18.2, p < 0.0001) (Fig. 8, B and D). The selectivity of the effect of LY404187 for AMPA receptors was confirmed by blocking the potentiation with LY300168 (50 µM; n = 4) (Fig. 8C).

View larger version (34K): [in this window] [in a new window]   Fig. 8.   LY404187 enhances glutamatergic synaptic and spontaneous AMPA receptor-dependent EPSPs in vitro. A, single stimulation of glutamatergic afferents to PFC pyramidal neurons evokes a complex EPSP composed of both mono- and polysynaptic components. Addition of 1 µM and 3 µM LY404187 enhanced the excitatory response to induce spike discharge. B, application of LY404187 (0.1-3.0 µM) increased EPSP amplitude and duration in a concentration-dependent manner. Recordings were made following blockade of GABAA and B and NMDA receptors and voltage-dependent sodium channels. C, LY404187 potentiation of the complex EPSP was abolished by the AMPA receptor antagonist LY300168 (50 µM). D, plot of the average AMPA receptor-mediated EPSP amplitude from recordings as in B as a percentage of control amplitude. E, spontaneous EPSPs recorded following blockade of GABAA and B and NMDA receptors in an extracellular solution containing elevated potassium concentration (6 mM) in the absence and presence of 0.3 µM LY404187. F, frequency histogram plotting the total number of occurrences of EPSPs recorded as a function of EPSP amplitude for four cells during control and drug conditions. The amplitude threshold for detection was 0.4 mV, and the bin width was 0.2 mV. LY404187 significantly shifted the distribution toward larger amplitudes.

LY404187 Enhances NMDA Receptor-Mediated Synaptic Responses in Vitro. The pronounced enhancement of AMPA receptor-mediated synaptic responses by LY404187 suggested that a secondary effect of the greater depolarization would be to recruit NMDA receptor-dependent input by reducing the Mg²⁺ block of these channels. To test this hypothesis, the NMDA component of the EPSP, defined as the portion of the response sensitive to 40 µM APV, was measured before and during application of LY404187. For these experiments, the monosynaptic component of the EPSP was isolated by raising the extracellular concentrations of Ca²⁺ and Mg²⁺ to 4 mM; GABA_A receptors were blocked with bicuculline (1 µM), and the recording electrode solution contained QX314 (1 mM) to eliminate action potentials. Recordings were performed while the somatic membrane was held at approximately 65 mV with constant current. To better evaluate the contribution of the NMDA conductance, the integrated area under the curve (AUC) delineated by the depolarizing response was measured for each cell in each condition and normalized relative to the control value. During control conditions, single stimulation evoked an EPSP having an average AUC of 1085.1 ± 295.2 mV/ms (n = 5). Application of APV (40 µM) reduced the AUC of the response to 78.3 ± 5.0% (AUC = 853.0 ± 244.7 mV/ms) of the control value (Fig. 9, A and C). APV then was eliminated from the extracellular solution, and the evoked EPSP returned to control levels after 10 to 15 min (Fig. 9A, gray trace). Following the washout period, LY404187 (1 µM) was administered and increased the average AUC of the EPSP to 153.0 ± 7.2% (AUC = 1664.8 ± 494.1) of the control area (t = 4.2, p < 0.01) (Fig. 9B). Subsequent application of APV reduced the EPSP to control levels (95.7 ± 7.9% of control) (AUC = 1053.1 ± 346.1). Isolation of the APV-sensitive response during control and LY404187 conditions using subtraction procedures revealed that the APV-sensitive response was enhanced in the presence of the potentiator (Fig. 9B, inset). In addition, the EPSP evoked in the presence of LY404187 and APV (95.7 ± 7.9% of control) was significantly larger than the response recorded during application of APV alone (gray trace) (78.3 ± 5.0%) (t = 4.2, p < 0.01), demonstrating the enhancement of the AMPA receptor-dependent component of the synaptic response (Fig. 9, B and C).

View larger version (21K): [in this window] [in a new window]   Fig. 9.   LY404187 enhances NMDA receptor-dependent synaptic responses in PFC pyramidal neurons. A, single stimulation of glutamatergic afferents to PFC pyramidal neurons in the presence of elevated calcium (4 mM) and magnesium (4 mM) concentrations evoked a monosynaptic EPSP. Application of APV (40 µM) reduced the AUC of the response by approximately 20%. The effects of APV were eliminated within 15 min of washout (gray trace). B, addition of LY404187 (1 µM) enhanced the EPSP by approximately 50%. In the presence of LY404187, APV reduced the EPSP by approximately 50%, indicating an augmentation of the NMDA component of the response. This residual response reflected the potentiated AMPA response and was significantly larger than the control AMPA response (APV trace). Inset, APV-sensitive responses generated during control and LY404187 conditions by subtracting the responses in the presence from those in the absence of APV. C, plot of the average area under the curve for EPSPs recorded during the conditions indicated.

LY404187 Potentiates Excitatory Synaptic Responses of PFC Pyramidal Neurons Evoked by Stimulation of Hippocampal Afferents in Vivo. The results from the in vitro slice experiments demonstrated that LY404187 markedly enhanced excitatory synaptic inputs to PFC pyramidal neurons at nanomolar concentrations, suggesting that this compound should have similar actions on specific inputs to these neurons in vivo provided that LY404187 crosses the blood-brain barrier. The central nervous system permeability and potentiation of synaptic responses by LY404187 were tested in vivo using extracellular recording of spontaneous and evoked action potential discharge of PFC neurons before and during i.v. administration of the compound. Previous neuroanatomical and electrophysiological studies have demonstrated that the ventral subiculum of the hippocampal formations provides monosynaptic, glutamatergic afferents to PFC pyramidal neurons (Jay et al., 1992). In addition, pharmacological experiments have shown that these excitatory synaptic responses are primarily mediated by AMPA receptor activation (Jay et al., 1992). Therefore, electrical stimulation of the ventral subicular inputs to PFC was used to assess the effects of LY404187 and compare with that of CX516, a compound that has previously been shown to be centrally active (Stäubli et al., 1994a).

View larger version (11K): [in this window] [in a new window]   Fig. 10.   Dose-dependent potentiation by LY404187 of the excitatory input to PFC neurons from ventral subiculum. A, control PSTH showing the total number of spike discharges of a single PFC neuron in response to stimulation of ventral subiculum (300 trials delivered at 1 Hz). Bin size is 1 ms. Arrow denotes stimulus presentation (time = 0). B and C, LY404187 (10 and 100 µg/kg, i.v.) increased the probability of spike discharge to synaptic stimulation in a dose-dependent manner. Note that spontaneous activity (defined as the number of spikes during the 50-ms prestimulus period) of this neuron also was enhanced by LY404187.

View larger version (36K): [in this window] [in a new window]   Fig. 11.   Group data showing the effects of LY404187 and CX516 on the spontaneous and evoked responses of PFC neurons in vivo. A and C, plots of the average peak probability of evoked spike discharge to stimulation of ventral subiculum as a percentage of the control probability of discharge for LY404187 (0.1-100 µg/kg) and CX516 (0.1-10 mg/kg). The threshold for potentiation of evoked spike discharge by LY404187 (1 µg/kg) was 1000-fold lower than that for CX516 (1 mg/kg). The maximal increase produced by both compounds was approximately 200% of control. B and D, plots of the average peak probability of spontaneous firing as a percentage of the control probability of firing for LY404187 (0.1-100 µg/kg) and CX516 (0.1-10 mg/kg). Neither compound significantly enhanced (or depressed) spontaneous discharge.

View larger version (17K): [in this window] [in a new window]   Fig. 12.   AMPA and NMDA receptor antagonists suppress the potentiation of evoked spike discharge produced by LY404187. A, plot of the average peak probability of evoked spike discharge to stimulation of ventral subiculum as a percentage of the control probability of discharge for LY404187 (10.0 µg/kg) administered alone or followed by either LY235959 (5 mg/kg, i.v.) or LY300168 (1 mg/kg, i.v.). B, plot of the average peak probability of spontaneous firing as a percentage of the control probability of firing for LY404187 (10.0 µg/kg) administered alone or followed by either LY235959 (5 mg/kg, i.v.) or LY300168 (1 mg/kg, i.v.).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Potency Differences between LY404187 and CX516. The present results demonstrate that the biarylpropylsulfonamide LY404187 is a highly potent, selective, and centrally active potentiator of AMPA receptors in PFC pyramidal neurons. In comparison to the prototypic benzoylpiperidine CX516, LY404187 was approximately 3000 times more potent and 10-fold more efficacious at potentiating AMPA receptor responses of isolated PFC neurons. However, previous studies have shown that AMPA receptor modulators can preferentially affect specific receptor subunits and/or splice variants, giving rise to considerably different potency and efficacy values depending on the receptor tested (Partin et al., 1996; Sekiguchi et al., 1997). Molecular cloning techniques have identified four distinct AMPA receptor subunits (GluR1-4 or GluRA-D) with each subunit having two alternative splice variants termed "flip and flop" (Keinänen et al., 1990; Sommer et al., 1990; Hollmann and Heinemann, 1994). Evidence suggests that cortical pyramidal cells predominantly express heteromeric receptors composed of GluR2 and GluR3 subunits and both flip and flop splice variants (Geiger et al., 1995; Kondo et al., 1997). Therefore, some of the disparity between the potencies of LY404187 and CX516 may be attributed to differences in their relative activity at the particular receptors expressed in PFC neurons. Conversely, the low potency of CX516 also has been reported for receptors in hippocampal CA₁ pyramidal neurons (Arai and Lynch, 1998b), which express different subunits and splice variants compared with PFC pyramidal neurons (Sommer et al., 1990). In addition, recent studies showed that LY404187 was similarly more potent and efficacious than CX516 on recombinant homomeric GluR4 flip receptors (Bleakman et al., 2000). Collectively, these data suggest that LY404187 is intrinsically more potent than CX516.

Possible Mechanisms of LY404187 Potentiation. A common attribute of positive AMPA receptor modulators is that they exert their effects by altering the desensitization process. However, evidence from kinetic modeling experiments of GluR1 receptors indicates that AMPA potentiators can suppress the desensitization process through different mechanisms. For example, cyclothiazide potentiates AMPA receptors by directly slowing the onset rate of desensitization and by increasing agonist affinity, which indirectly slows the rate of channel closure (deactivation) (Yamada and Tang, 1993; Partin et al., 1994, 1996). In contrast, aniracetam appears to directly slow the rate of deactivation (without a change in agonist affinity), which indirectly slows the onset of desensitization (Partin et al., 1996).

LY404187 Potentiation of AMPA and NMDA Receptor-Mediated Synaptic Responses. The potentiation of somatic AMPA receptor responses by LY404187 in isolated cell recordings suggested that the compound should similarly enhance synaptic AMPA receptors activated by glutamatergic inputs to PFC neurons. Indeed, results showed that LY404187 potentiated synaptic AMPA receptor-mediated EPSPs in PFC slices in a concentration-dependent manner such that concentrations as low as 300 nM enhanced EPSPs by approximately 60%. In addition, the amplitude, but not the frequency, of spontaneous EPSPs was enhanced by the potentiator, indicating a postsynaptic site of action. Previous studies have demonstrated that the duration and amplitude of AMPA receptor-mediated EPSPs are influenced to a greater extent by modulators that preferentially affect deactivation (e.g., CX516) as opposed to desensitization (e.g., cyclothiazide) (Arai and Lynch, 1998b). Given the robust potentiation by LY404187, these results suggest that LY404187 also modulates the deactivation process.

Therapeutic Implications: AMPA Receptor Potentiators Facilitate Working Memory. Considerable evidence has implicated the PFC in a variety of higher cognitive functions, many of which require the mnemonic process of working memory or the ability to "hold information in mind" for later retrieval (Baddeley, 1992). Typically, working memory performance is assessed using delayed-response tasks in which a delay period is introduced between presentation of cue and a required appropriate response. Neurophysiological experiments in primates have shown that subpopulations of PFC neurons exhibit enhanced discharge during the delay period of working memory tasks, suggesting a possible cellular correlate for this form of memory (Funahashi et al., 1989). In addition, lesions of PFC selectively disrupt performance on working memory tasks (Kolb et al., 1982). Clinical studies also have linked activity of the PFC with working memory performance. Brain imaging experiments