MITSUHARU YOSHIYAMA, JAMES R. ROPPOLO and WILLIAM C. de GROAT

Transcription

1 /97/ $03.00/0 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 280, No. 2 Copyright 1997 by The American Society for Pharmacology and Experimental Therapeutics Printed in U.S.A. JPET 280: , 1997 Effects of LY215490, a Competitive -Amino-3-Hydroxy-5- Methylisoxazole-4-Propionic Acid (AMPA) Receptor Antagonist, on the Micturition Reflex in the Rat 1 MITSUHARU YOSHIYAMA, JAMES R. ROPPOLO and WILLIAM C. de GROAT Department of Pharmacology, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania Accepted for publication October 7, 1996 ABSTRACT The effects of glutamate receptor antagonists on urinary bladder and external urethral sphincter- (EUS) electromyogram (EMG) activity were evaluated in unanesthetized decerebrate rats. In normal rats, LY215490, an -amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor antagonist, in small i.v. doses (1 3 mg/kg) decreased bladder contraction amplitude (BC-Amp) by 29% and EUS-EMG by 41%; whereas a large dose (10 mg/kg) completely abolished bladder and EUS-EMG activity. LY injected intrathecally in small doses ( g) decreased BC-Amp by 20% and EUS- EMG by 62%; whereas large doses (1 10 g) completely abolished bladder and EUS-EMG activity. LY (0.1 g i.t.) increased bladder capacity by 28% and decreased voiding efficiency by 44%. Combined i.t. administration of small doses of LY (0.1 g) and MK-801 (1 g), an N-methyl-D-aspartate (NMDA) receptor antagonist, which individually had little Various drugs and preparations have been used to evaluate the role of glutamate in the central nervous control of micturition. The i.v. or i.t. administration of competitive (LY274614) and noncompetitive (MK-801) NMDA receptor antagonists depressed reflex bladder and EUS activity in urethane-anesthetized rats (Maggi et al., 1990; Yoshiyama et al., 1991, 1993a,b). However, in freely moving awake rats (Vera and Nadelhaft, 1991), unanesthetized decerebrate rats (Yoshiyama et al., 1994) or anesthetized chronic spinal rats (Yoshiyama et al., 1993a,b), NMDA receptor antagonists did not depress bladder reflexes but still depressed EUS activity. The i.v. or i.t. administration of an NMDA receptor antagonist also blocked the bladder contractions evoked by electrical stimulation of the PMC, indicating that NMDA receptors mediate excitatory transmission in the descending limb of Received for publication March 19, This work was supported by National Institutes of Health Grants DK and DK (W.D.). effect on BC-Amp, markedly suppressed bladder activity. In chronic spinal rats, LY (10 mg/kg i.v.) abolished EUS- EMG activity and decreased BC-Amp by 41%. Intrathecal injections of LY were also less effective in chronic spinal rats; a 10- g dose producing only a partial block (53%) of BC-Amp, but complete block of EUS-EMG. In chronic spinal rats, MK-801 (1 mg/kg i.v.) abolished EUS-EMG activity and decreased BC-Amp by 36%. Pretreatment with MK-801 (1 mg/kg i.v.) did not enhance the effect of LY on bladder activity in chronic spinal rats. These data suggest that AMPA glutamate receptors have a major role in the excitatory pathways controlling bladder and EUS activity in spinal cord intact rats. However, in chronic spinal rats, AMPA and NMDA receptors are essential for EUS reflexes, but are responsible for only a part of reflex bladder activity. spinobulbospinal micturition reflex pathway (Matsumoto et al., 1995a). The i.v. administration of GYKI 52466, a noncompetitive AMPA/kainate receptor antagonist also suppressed reflex bladder and EUS activity in urethane-anesthetized rats (Yoshiyama et al., 1995a) and the bladder activity evoked by electrical stimulation of the PMC (Matsumoto et al., 1995b). Combined administration of MK-801 and GYKI also produced a synergistic inhibitory effect on bladder and EUS reflexes, suggesting that an interaction between NMDA and AMPA/kainate receptors is important in the control of micturition (Yoshiyama et al., 1995b). Further analysis of the role of AMPA/kainate glutamate receptors in voiding function has been limited by the low solubility of GYKI and the necessity to use a very acidic solution (ph 2.8) to dissolve the compound (Yoshiyama et al., 1995a). In the present experiments, LY215490, a potent, selective, competitive AMPA receptor antagonist that is highly water soluble and readily enters the central nervous ABBREVIATIONS: AMPA, -amino-3-hydroxy-5-methylisoxazole-4-propionic acid; NMDA, N-methyl-D-aspartate; CMG, cystometrogram; CSF, cerebrospinal fluid; EMG, electromyogram; EUS, external urethral sphincter; ICI, intercontraction interval; V T, volume threshold; V E, voiding efficiency; PMC, pontine micturition center; CNS, central nervous system; i.t., intrathecal. 894

2 1997 Effects of LY on Micturition 895 system after systemic administration (Ornstein et al., 1993a,b), has been used to evaluate glutamatergic mechanisms involved in voiding function in unanesthetized decerebrate, spinal cord intact and chronically spinalized rats. Recent studies indicate that the systemic administration of LY elicits anticonvulsant (Ornstein et al., 1993a,b; Schoepp et al., 1995), neuroprotective (Gill, 1994; Gill and Lodge, 1994) and antinociceptive effects (Kakizaki et al., 1996). A preliminary account of this work has been presented in an abstract (Yoshiyama et al., 1995c). Materials and Methods Animal preparation. Female Sprague-Dawley rats weighing 200 to 320 g were used in this study. The animals were anesthetized with halothane (2%) in oxygen during surgery prior to decerebration. The trachea was cannulated with a polyethylene tube (PE-240) to facilitate respiration; and a cannula (PE-50) was placed in the left external jugular vein for i.v. drug administration. Decerebrations were performed according to published methods (Sapru and Krieger, 1978) that included ligating both carotid arteries followed by a precollicular decerebration using a blunt spatula. Halothane was then discontinued. Cotton and Avitene (MedChem Products Inc., Woburn, MA) were placed in the intracranial cavity and covered with agar. Experiments were started 2 to 8 hr after the decerebration. In 28 decerebrate rats, an i.t. catheter was also inserted according to the technique of Yaksh and Rudy (1976). The occipital crest of the skull was exposed and the atlanto-occipital membrane was incised at the midline using the tip of an 18-gauge needle as a cutting edge. A catheter (PE-10) was inserted through the slit and passed caudally to the L 6 -level of the spinal cord. The volume of fluid within the catheter was kept constant at 6 l in all animals. Single doses of drugs were administered in a volume of 1 l followed by 7.5 l flush with artificial CSF (Feldberg and Fleischhauer, 1960). At the end of the experiment, a laminectomy was performed to verify the location of the catheter tip. A transurethral bladder catheter connected to a pressure transducer was used to record the bladder pressure isovolumetrically with the urethral outlet ligated or to record pressure during a CMG when the bladder was filled with a constant infusion of physiological saline and allowed to empty around the catheter. To evaluate V E (% of bladder volume voided) saline was infused into the bladder (0.04 ml/min) until the peak of a voiding bladder contraction; then the infusion was stopped and the saline voided from the bladder was collected and measured to determine the voided volume. The bladder was then emptied to measure the residual volume. This procedure was repeated two to four times under control conditions and after each drug dose. Continuous CMGs were also performed using a constant, more rapid infusion (0.21 ml/min) of saline into the bladder to elicit repetitive voidings, which allowed collection of data for a large number of voiding cycles (Maggi et al., 1986). PE-90 and PE-50 cannulae were used for isovolumetric recording and constant infusion CMGs (continuous and single), respectively. In all animals during isovolumetric recording, the ureters were tied distally, cut and the proximal ends cannulated (PE-10) and drained externally. This procedure prevented the bladder from filling with urine during the experiment. TABLE 1 Urodynamic parameters during CMGs in spinal cord intact rats Twenty-nine rats were spinalized under halothane-anesthesia. After a T 8 9 laminectomy, the dura and spinal cord were cut with scissors, and a sterile sponge (Gelfoam, The Upjohn Company, Kalamazoo, MI) was placed between the cut ends. The bladders of spinal rats were expressed manually two or three times daily, and perigenital stimulation with a cotton swab was performed to encourage reflex bladder emptying (Mallory et al., 1989). The experiments on spinalized rats were performed 3 to 4 wk postspinalization. In seven spinalized rats, i.t. catheterization was performed after a T laminectomy. Drugs were administered i.t. as described above. In the majority of experiments, epoxy-coated stainless steel wire (50 m, M.T Giken Co., Ltd., Tokyo, Japan) EMG electrodes were placed percutaneously in the EUS to examine synergy between bladder and EUS. This was performed using a 30-gauge needle with a hooked EMG electrode positioned at the tip. The needle was inserted into the sphincter approximately 5 to 10 mm lateral to the urethra and then withdrawn leaving the EMG wires embedded in the muscle (Kruse et al., 1990). The EMG activity was passed through a discriminator/ratemeter, the output of which was recorded on a chart recorder. The peak firing rate during each micturition contraction was measured. Evaluation and statistical analysis. The effects of drugs in our studies on amplitude, duration and frequency of reflex bladder contractions were recorded under isovolumetric conditions or with the urethra opened allowing the bladder to empty. The interval between two voiding cycles termed the ICI (Maggi et al., 1986) was also measured during a continuous CMG. During single CMGs, each i.t. dose of LY was given 10 min before the first test. Two to four CMGs were obtained after each dose. The effects of the drug on the V T to induce micturition and the volume of fluid released (voided volume) during each voiding reflex were measured. Based on these values, V E (%) [(voided volume/v T ) 100] could be estimated. All values are expressed as mean S.E.M. Repeated measures analysis of variance, Dunnett multiple comparisons test, Student s t test and Mann-Whitney U test were used when appropriate for statistical data analysis. For all statistical tests, P.05 was considered significant. Drugs. Drugs used include: halothane (Ayerst Lab. Inc., Philadelphia, PA), LY ((3SR,4aRS,6RS,8aRS)-6-[2-(1H-tetrazol-5-yl- )ethyl]decahydroisoquinoline-3-carboxylic acid, Lilly Res. Labs., Indianapolis, IN), LY (( )-decahydro-6-(phosphonomethyl)-3- isoquinolinecarboxylic acid, Lilly Res. Labs.), MK-801 (dizocilpine, Merck, Sharp & Dohme Res. Labs., West Point, PA) and GYKI HCl (1-(4-aminophenyl)-4-methyl-7,8-methylenedioxy-5H- 2,3-benzodiazepine hydrochloride, Institute for Drug Research, Budapest, Hungary). LY215490, LY and MK-801 were dissolved in sterile saline or artificial CSF and these solutions were then adjusted to ph 7.4. GYKI HCl was dissolved in sterile water for i.v. administration (ph 2.8). Drug doses were calculated for the base of each compound. Results Effects of i.v. and i.t. administration of LY in spinal cord intact rats. Before drug administration, control urodynamic parameters were measured during continuous or single CMGs in spinal cord intact rats (table 1). Amplitude (cm H 2 O) Duration (sec) ICI (sec) V T (ml) V E (%) Continuous CMGs (n 54) (22 44) (16 88) (14 420) Single CMGs (n 6) (20 33) (21 37) ( ) (87 97) Numbers in parentheses at each parameter are ranges.

3 896 Yoshiyama et al. Vol. 280 After a 1- to 2-hr control period, LY was administered in doses between 1 and 10 mg/kg i.v. to generate a dose response curve. Preliminary experiments in which a single large dose (10 mg/kg i.v.) was injected, revealed that the drug had a slow onset of action producing detectable changes in bladder and EUS activity after 15 min and maximal effects after 45 to 65 min (figs. 1A and 2). After this large dose, bladder and EUS activity were abolished in all animals (n 5). The depression persisted for the duration of the experiment (5 hr). In four of five animals, 10 to 15 min before complete suppression of bladder activity, the duration of bladder contractions and the ICI were decreased by 25 to 59% (mean 41%) and 26 to 72% (mean 52%), respectively. A depression of respiratory depth and rate was observed visually 30 to 50 min after the administration of 10 mg/kg of LY This depression lasted for the duration of the experiment. Smaller doses of LY (1 and 3 mg/kg i.v., n 5 for each dose) did not alter respiration but did elicit small, but significant decreases in bladder contraction amplitude (range: 11 37% decrease, mean 28% at 1 mg/kg; range: 11 50% decrease, mean 31% at 3 mg/kg). The amplitude of the EUS EMG was also depressed (range: 11 52% decrease, mean 29% at 1 mg/kg; range: 28 73% decrease, mean 53% at 3 mg/kg) (fig. 2). The effect on EUS activity persisted for 90 to 120 min, whereas the effect on the bladder persisted for the duration of the experiment. The duration of bladder contractions and ICI were not changed by these doses of the drug. The i.t. injection of LY produced a rapid onset suppression of bladder and EUS activity (figs. 1B and 3). The depressant effects of LY on bladder contraction amplitude and magnitude of EUS EMG were dose dependent (fig. 3). Large doses (1 and 10 g i.t., n 5 for each dose) produced complete inhibition of bladder and EUS activity that occurred within a few minutes after injection and lasted for 90 min (1 g) to 10 hr (10 g). Smaller doses (0.01 and 0.1 g i.t., n 5 for each dose) produced slower onset and smaller decreases in bladder contraction amplitude (range: 3 24% decrease, mean 14% at 0.01 g; range: 18 40% decrease, mean 26% at 0.1 g). EUS EMG activity was also decreased (range: 26 78% decrease, mean 59% at 0.01 g; range: 39 96% decrease, mean 66% at 0.1 g) but the duration of bladder contractions and the ICI were not altered. In three rats, 10 g i.t. of LY was injected at the C 2 -level of spinal cord to determine if the LY inhibition of bladder contraction amplitude occurred as a result of the distribution of the drug to the brainstem (i.e., the PMC: see de Groat et al., 1993 for a reference). LY (10 g i.t.) administered at the C 2 -level elicited a delayed onset (75 min postinjection) slight suppression of bladder contraction amplitude (range: 8 35% decrease, mean 21%). In these same rats 2 hr after the first injection, LY (10 g i.t.) was administered at the L 6 -level. This injection completely blocked bladder activity after 5 20 min (mean 12 min). The 10 g i.t. dose of LY administered at L 6 -level elicited a delayed (40 60 min) suppression of respiration. This was not observed after injection of lower doses of the drug. To determine if the inhibitory effect of the largest i.t. dose of LY could be due to the absorption of the drug into the systemic circulation, 10 g of the drug were administered i.v. (n 3). No change was seen in any bladder or sphincter parameters after this dose. In six animals in which urinary bladder activity was recorded during single CMGs performed by a slower rate of transurethral infusion (0.04 ml/min), LY was tested over a range of doses ( g i.t.). The bladder was emptied at the end of each CMG to determine V E. The control parameters during single CMGs are shown in table 1 and figure 4. LY in a dose of 0.01 g was ineffective, but a Fig. 1. The effects of LY on the reflex bladder contractions and EMG activity of the EUS muscle during a continuous filling (0.21 ml/min) CMG in spinal cord intact rats. Note that bladder and EUS EMG activity were abolished much earlier after i.t. injection (B) than i.v. administration (A). Abscissa, Time scale (min). Ordinate, Bladder pressure in cm H 2 O and EUS activity, ratemeter output in pulses/sec.

4 1997 Effects of LY on Micturition 897 Fig. 2. Graphs showing the time course of the effects of graded doses of LY (1, 3, 10 mg/kg i.v., n 5 for each dose) on the bladder contraction amplitude and EUS EMG activity during continuous filling (0.21 ml/min) CMGs in spinal cord intact rats. Abscissa, Time scale (min) after injection of LY Ordinate, Contraction amplitude (A) or EUS EMG (B) as a% of control. Fig. 3. Graphs showing the time course of the effects of graded doses of LY (0.01, 0.1, 1 g i.t., n 5 for each dose) on the bladder contraction amplitude and EUS EMG activity during continuous filling (0.21 ml/min) CMGs in spinal cord intact rats. Abscissa, Time scale (min) after injection of LY Ordinate, Contraction amplitude (A) or EUS EMG (B) as a% of control.

5 898 Yoshiyama et al. Vol. 280 Fig. 4. The effect of LY ( g i.t.) on bladder activity during single slow infusion (0.04 ml/min) CMGs in spinal cord intact rats (n 6). LY increased volume threshold and decreased voiding efficiency and bladder contraction amplitude. Abscissa, The dose of LY ( g i.t.). Ordinates, (A) volume threshold (ml), (B) voiding efficiency (%), (C) bladder contraction amplitude (cm H 2 O). Individual doses are compared to control using Dunnett multiple comparisons test (*P.05, **P.01). C control. O.I. (on A) overflow incontinence. dose of 0.1 g significantly increased bladder capacity (V T ), decreased V E and slightly decreased the amplitude of bladder contractions. A large dose (1 g) of the drug completely blocked bladder contractions and induced overflow incontinence at bladder volumes of more than 1.5 ml. The largest dose (10 g) did not produce a further increase in bladder volume (data not shown). In two rats, LY ( g) was injected in the same manner at the C 2 -level of spinal cord to determine if the inhibitory effect on bladder capacity was due to the distribution of the drug to brainstem. In one rat, LY injected in a range of doses 0.01 to 10 g i.t. at the C 2 -level did not change any parameter, whereas, in the other rat, small doses ( g i.t.) were inactive but larger doses did elicit an increase in V T (55% at 10 g) and a decrease in V E (25% at 1 g and 62% at 10 g). Interactions between NMDA antagonists and LY in spinal cord intact rats. LY ( mg/kg i.v.), a competitive NMDA receptor antagonist, was administered prior to LY injection. Doses of 3 to 30 mg/kg of LY significantly decreased (17 52%) in a dose-dependent manner, bladder contraction amplitude and EUS EMG activity (n 6) (fig. 5). However, these effects were much less than those previously described in urethaneanesthetized animals (Yoshiyama et al., 1993b). LY in these doses ( mg/kg) did not alter the duration of bladder contractions or the ICI. Forty five min after the last dose (30 mg/kg) of LY274614, the AMPA receptor antagonist, LY was administered in a dose of 3 mg/kg i.v. that produced a small depression of bladder and EUS activity in untreated animals (see fig. 2). Before the injection of LY215490, the amplitude of bladder contractions and EUS EMG activity were, respectively, 62% (range: cm H 2 O, mean S.E. 27 3cmH 2 O) and 50% (range: pulses/sec, mean S.E pulses/sec) of the control level before LY administration (n 6, P.05 each, Student s t test). Both bladder and EUS EMG activity were completely abolished 10 to 150 min (mean 65 min) after i.v. injection of LY (3 mg/kg i.v.) indicating that the effect of the AMPA receptor antagonist was enhanced by the pretreatment with the NMDA receptor antagonist. Recovery did not occur during the 3- to 8-hr postdrug observation period. Interactions between i.t. administration of MK-801, a noncompetitive NMDA antagonist and LY were also studied during continuous CMGs, to determine whether the spinal cord is a site at which NMDA and AMPA receptor mechanisms interact synergistically to control the micturition reflex. No change in bladder and EUS activity was observed at 30 min after the injection of vehicle or 1 g of MK-801, whereas, significant suppression of both bladder contraction amplitude and EUS EMG activity occurred after 10 g of MK-801 (fig. 6). The mean bladder contraction amplitude and EUS EMG activity were decreased by LY (0.1 g i.t.) to a much greater extent, in rats pretreated with 1 g of MK-801 than in those animals treated with vehicle (fig. 6). In two of the five rats treated with 1 g of MK-801, LY (0.1 g) completely abolished bladder and EUS activity. However, in all animals pretreated with 10 g of MK-801, LY (0.1 g i.t.) completely abolished bladder and EUS activity. Effects of LY in chronic spinal rats. As shown in table 2, control bladder activity recorded under isovolumetric conditions was markedly different in spinal cord intact and chronic spinal rats. In the latter animals, the amplitude and duration of bladder contractions were less than those in spinal intact animals, but the frequency of contractions was higher. Isovolumetric recording techniques that measured nonvoiding contractions were used in this study, due to the necessity of isolating bladder activity from influ-

6 1997 Effects of LY on Micturition 899 Fig. 5. Log dose-response curves showing the effects of increasing doses of LY ( mg/kg i.v.) on the bladder contraction amplitude (A) and the EUS EMG activity (B) during continuous fast infusion (0.21 ml/min) CMGs. Abscissa, The dose of LY (mg/kg i.v.). Ordinate, (A) the bladder contraction amplitude (n 6) and (B) the EUS EMG activity (n 3) as % of control. LY (3 30 mg/kg i.v.) significantly decreased both bladder contraction amplitude and EUS EMG activity (*P.05, **P.01, Dunnett multiple comparisons test). ence of EUS activity that is abnormal in chronic spinal rats (Yoshiyama et al., 1993b) and can markedly changed voiding parameters. The time course of the effect of LY (10 mg/kg i.v.) on the bladder contraction amplitude and EUS EMG activity under isovolumetric conditions in spinal cord intact and chronic spinal rats is shown in figure 7. In chronic spinal rats, LY suppressed bladder contraction amplitude by only 41% (maximal depression range: 23 53%) while completely suppressing EUS EMG activity. The i.t. administration of LY (10 g) decreased bladder contraction amplitude by 53% (maximal depression range: 30 84%) in chronic spinal rats although it completely abolished bladder contraction amplitude in spinal cord intact rats (fig. 8). In three spinalized rats, large cumulative doses of LY (30 and 100 g i.t.) or 10 g i.t. of MK-801 did not produce further suppression of bladder contraction amplitude. Neither i.v. nor i.t. administration of LY altered the duration or the frequency of bladder contractions. Combined administration of MK-801 (1 mg/kg i.v.) and LY (10 mg/kg i.v.) was performed to examine whether a synergistic interaction between NMDA and AMPA antagonists occurred in chronic spinal rats (fig. 9). MK-801 alone decreased by 36% the bladder contraction amplitude and abolished EUS EMG activity. LY administered in MK-801-treated rats produced only a small additional inhibitory effect reducing bladder contractions to 51% of control. This contrasts with the complete inhibition of bladder activity produced by LY in spinal cord intact rats pretreated with MK-801 (fig. 9). Interactions between GYKI and MK-801 in chronic spinal rats. The possible interaction between non- NMDA and NMDA receptor antagonists in chronic spinal animals was also tested using another drug GYKI 52466, a short acting, noncompetitive AMPA/kainate receptor antagonist. Neither the vehicle (n 6) for the drug nor GYKI (0.5 8 mg/kg i.v., n 6) significantly altered bladder or EUS activity in chronic spinal rats. In four animals, 2 hr after the last dose (8 mg/kg i.v.) of GYKI 52466, MK-801 (1 mg/kg i.v.) was administered and significantly decreased the amplitude of bladder contractions in three animals 20 to 38% (mean 30%) and one animal showed no effect (EUS not recorded). Ninety min after MK-801, GYKI was administered to examine possible synergistic interactions between the two drugs, as seen in our previous report in spinal cord intact rats (Yoshiyama et al., 1995b). However, the combined administration of MK-801 and GYKI did not significantly inhibit bladder contractions in the chronic spinal rat. Discussion Our results indicate that AMPA glutamatergic transmission plays an important role in the central nervous control of lower urinary tract function in the unanesthetized, decerebrate rat. Administration of AMPA receptor antagonists suppressed the amplitude of reflex bladder and EUS activity, increased the bladder V T for inducing micturition and decreased the V E. These effects are mediated at least in part by drug actions on the lumbosacral spinal cord and depend in part on the integrity of the spinobulbospinal micturition reflex pathway. In chronic spinal animals, the depressant effects of AMPA antagonists on reflex bladder activity were markedly reduced, indicating that other transmitter mechanisms contribute to voiding function after spinal injury. The i.t. injection at the L 6 level of the spinal cord as well as

7 900 Yoshiyama et al. Vol. 280 Fig. 6. Graphs showing the time course of the effects of LY (0.1 g i.t.) on bladder contraction amplitude and EUS EMG activity during continuous filling (0.21 ml/min) CMGs, in spinal cord intact rats treated i.t. with artificial CSF (as vehicle, n 5) or MK-801 (1 and 10 g, n 5 each). Abscissa, Time scale (min). Ordinates, Bladder contraction amplitude (A) and EUS EMG (B) as a% of control. Open arrow, Injection of artificial CSF or MK-801. Closed arrow, Injection of LY MK MK-801. TABLE 2 Urodynamic parameters under isovolumetric condition in spinal intact and chronically spinalized rats Amplitude (cm H 2 O) Duration (sec) Frequency (numbers/min) Intact (n 10) (48 85) (44 120) ( ) Spinalized (n 29) (14 44) (12 55) ( ) Numbers in parentheses at each parameter are ranges measured in present experiments. All parameters are significantly different between spinal cord intact and chronically spinalized rats (P.0001, Mann-Whitney U test). the systemic administration of LY reduced or completely abolished the rhythmic bladder contractions in the spinal cord intact rat, indicating that actions on either the peripheral nervous system or on the bladder smooth muscle are not essential for the effects of the drug. The complete suppression by LY of bladder reflexes in unanesthetized decerebrate rats contrasts with the effects of NMDA receptor antagonists (MK-801, Yoshiyama et al., 1994 and LY274614, this study) that elicited only a partial reduction in the amplitude of reflex bladder activity in decerebrate rats. In the spinal cord, it is likely that LY acts at least in part on AMPA excitatory transmission in the descending limb of the spinobulbospinal micturition reflex pathway since previous studies (Matsumoto et al., 1995b) showed that the i.v. administration of GYKI 52466, an AMPA/kainate receptor antagonist depressed the bladder contractions elicited by electrical stimulation of the PMC in urethane-anesthetized rats. EUS EMG activity was also inhibited by LY given either systemically or intrathecally. In comparison to effects on the bladder, the effects on EUS were seen at lower doses and with a shorter onset of action. Previous studies from this laboratory (Yoshiyama et al., 1993b, 1994) have shown that NMDA receptor antagonists, LY and MK-801, also are more effective in inhibiting EUS EMG activity than bladder activity in unanesthetized decerebrate rats. However, these drugs only reduce and never completely inhibit EUS function. Thus, EUS activity is controlled by both NMDA and AMPA glutamatergic mechanisms. It is important to note that the maximal depressant doses of LY used in this study (10 mg/kg i.v. and 1 g i.t.) which were more than 10 times the threshold dose might have an effect on NMDA as well as AMPA receptors, because receptor binding as well as in vitro physiological experiments indicate that the drug has only a 10-fold selectivity for AMPA vs. NMDA receptors (Schoepp et al., 1995). However, in vivo experiments indicate that doses between 25 and 75 mg/kg, i.p. which are larger than the doses used in our study elicited a selective antagonism of the neurotoxic effect of AMPA in the rat striatum, without affecting the neurotoxicity to

8 1997 Effects of LY on Micturition 901 Fig. 7. Graphs showing the time course of the effects of LY (10 mg/kg i.v.) on bladder contraction amplitude and EUS EMG activity under isovolumetric conditions in chronic spinal (n 5) and spinal cord intact (n 5) rats. Abscissa, Time scale (min). Ordinates, Bladder contraction amplitude (A) and EUS EMG (B) as a% of control. Open arrow, Injection of artificial CSF. Closed arrow, Injection of LY NMDA (Ornstein et al., 1993a; Schoepp et al., 1995). Indeed, even much larger doses (up to 320 mg/kg i.p.) in the mouse did not alter NMDA toxicity (Ornstein et al., 1993, a and b). Unfortunately, it is not possible to determine from our data if NMDA receptor blockade makes any contribution to the suppression of lower urinary tract reflexes by LY We have previously reported that the i.v. administration of MK-801 and GYKI produced a synergistic inhibition of bladder contraction amplitude in spinal cord intact rats (Yoshiyama et al., 1995b). In our study, the combined i.t. administration of MK-801 and LY showed that NMDA and AMPA receptor antagonists interact synergistically at synapses in the spinal cord to control bladder activity when the spinobulbospinal micturition reflex pathway is intact. A synergistic interaction between LY and MK-801 was also noted in studies in which these drugs inhibited c-fos expression in the spinal cord induced by bladder irritation (Kakizaki et al., 1996). Electrophysiological studies in the spinal cord slice preparation of the neonatal rat suggest that the synergistic interactions between NMDA and non-nmda antagonists on bladder reflexes might occur at synapses on parasympathetic preganglionic neurons (Araki and de Groat, 1996). Patch clamp recordings in preganglionic neurons in the L 6 -S 1 spinal cord revealed that stimulation of single interneurons in the region of the sacral parasympathetic nucleus elicited short duration and long duration excitatory postsynaptic currents that were blocked, respectively, by AMPA/kainate and NMDA receptor antagonists. Stimulation of axons in the lateral funiculus in the region containing the axons of the descending limb of the micturition reflex pathway also elicited short and long duration excitatory postsynaptic currents, raising the possibility that interactions between AMPA and NMDA glutamatergic transmission within the sacral parasympathetic nucleus are also essential for the supraspinal control of bladder activity (Araki and de Groat, 1996). Synergistic interactions between NMDA and AMPA receptor antagonists have also been noted in other neural systems in the spinal cord (Honoré et al., 1988) and brain (Löscher et al., 1993), using in vitro as well as in vivo preparations (Foutz et al., 1994). However, synergism has not been detected in regard to the neuroprotective effect of these drugs (Gill and Lodge, 1992). Because the PMC seems to function as the switching circuit in the spinobulbospinal micturition reflex (de Groat et al., 1993), a depressant action of a drug distal to the PMC (i.e., on the descending limb of the micturition reflex) should reduce the magnitude of the reflex but not alter the V T for triggering voiding. However, the i.t. injection of LY increased the V T during single CMGs, indicating the existence of an AMPA glutamatergic mechanism in the excitatory afferent pathway or in the spinal interneuronal component of the ascending limb of the micturition reflex. This is consistent with recent studies in which it was shown that

9 902 Yoshiyama et al. Vol. 280 Fig. 8. Graph showing the time course of the effects of LY (10 g i.t.) on bladder contraction amplitude under isovolumetric conditions in chronic spinal (n 7) and spinal cord intact (n 2) rats. Abscissa, Time scale (min) after injection of LY Ordinates, Bladder contraction amplitude as a%of control. LY reduced the c-fos expression in the spinal cord elicited by acetic acid-induced activation of bladder afferents (Kakizaki et al., 1996). However, LY did not significantly change the ICI during continuous CMGs. Although ICI is dependent primarily on V T, it can also be influenced by other factors, such as V E. Thus, if a drug only increased V T, the ICI should increase proportionally; however, if the drug also decreased V E and increased residual urine in the bladder at the end of voiding, this would decrease the infusion volume necessary to induce the next micturition reflex and thereby reduce ICI. In addition, the faster infusion rate used during continuous CMGs, can alter the effects of drugs on the ICI (Yoshiyama et al., 1994). Thus, the conditions of the experiment as well as an effect of LY on V E could have masked a suppressant effect of the drug on V T and negated the expected increase in the ICI. In contrast to the effects of LY215490, the i.t. administration of MK-801 in unanesthetized decerebrate animals (Yoshiyama et al., 1994) reduced the V T for micturition, indicating that NMDA glutamatergic transmission is involved in a spinal inhibitory mechanism controlling the sensory limb of the micturition reflex. A role for NMDA receptors in the spinal inhibitory control of micturition has also been noted in neonatal rats (Sugaya and de Groat, 1994). NMDA in contrast to AMPA glutamatergic mechanisms are therefore involved in the neural control of urine storage as well as voiding. Because AMPA receptor antagonists also suppress EUS activity, the effects of LY to increase bladder capacity and decrease V E during single CMGs might be due in part to an indirect effect on the EUS. In the rat, rhythmic EUS activity during micturition is important for the release of urine. The EUS may function: 1) as a pump to aid in emptying the bladder (Conte et al., 1991) and/or 2) to periodically close the urethral outlet producing isovolumetric bladder contractions that facilitate bladder afferent firing and in turn enhance the micturition reflex. Thus, the inhibitory effect of LY on the EUS may indirectly decrease V E. Although LY completely suppressed both bladder and EUS activity in rats with intact spinal cords, even very large doses of LY only partially inhibited bladder activity in chronic spinal rats. Even the combination of an AMPA antagonist with an NMDA antagonist (MK-801), which in an intact animal enhanced the inhibition of bladder contractions, did not completely suppress bladder reflexes. EUS activity, however, was completely inhibited by LY in doses (10 mg/kg i.v. or 10 g i.t.) similar to those that were effective in spinal intact rats. These data suggest that even though AMPA and NMDA glutamatergic mechanisms are still important in the chronic spinal animal, they represent only part of the transmitter mechanisms controlling bladder contractions. Thus, the reorganization of synaptic connections in the spinal cord after spinal transection (de Groat et al., 1993) unmasks a nonglutamatergic spinal cord mechanism for the control of bladder activity. This new transmitter mechanism has not been identified. In contrast to bladder reflexes, EUS reflexes are dependent primarily on glutamatergic mechanisms in both intact and chronic spinal animals. Because either AMPA or NMDA receptor antagonists alone or in combination can completely block EUS activity in both preparations, it is tempting to speculate that AMPA and NMDA receptor mechanisms are located at the same synapse and have a synergistic interaction or are arranged in a serial manner at different synapses in the neural pathway controlling the EUS. The greater reflex depression that occurred with the combined administration of AMPA and NMDA antagonists would be consistent with either synaptic arrangement. The effects of two other AMPA/kainate receptor antagonists, CNQX and GYKI 52466, on bladder and EUS reflexes have also been studied in our laboratory (Suzuki et al., 1991; Yoshiyama et al., 1995a). CNQX administered i.v. in a high dose (1 mg/kg) had no effect in either urethane-anesthetized or -unanesthetized decerebrate rats. The lack of effect of CNQX is probably due to its inability to enter the CNS in significant amounts after systemic administration (Ornstein et al., 1993b). A report from another laboratory indicated that CNQX can inhibit bladder reflexes after both i.t. and i.v. administration (Matsumoto et al., 1991). GYKI is effective after systemic administration but because of its low aqueous solubility, it was difficult to test high doses of the drug with i.v. administration (Yoshiyama et al., 1995a). However, GYKI in small to moderate doses inhibits both bladder and EUS reflexes in urethane-anesthetized rats. The onset of the GYKI effect is rapid (1 3 min to peak) and the duration of action is rather short, lasting for only 10 to 15 min. In the unanesthetized decerebrate rat, GYKI had virtually no effect on bladder reflexes but inhibited EUS activity with the same time course as in urethane-anesthetized rats (Yoshiyama et al., 1995a). The lack of effect of GYKI on bladder reflexes in the decerebrate rat contrasts with the depressant effect of LY on both bladder and EUS activity, suggesting that the dose of GYKI was insufficient to produce significant inhibition of bladder activity. Because doses of GYKI that sup-

10 1997 Effects of LY on Micturition 903 Fig. 9. Graphs showing the time course of the effects of combined administration of MK-801 (1 mg/kg i.v.) and LY (10 mg/kg i.v.) on bladder contraction amplitude and EUS EMG activity under isovolumetric conditions in chronic spinal (n 5) and spinal cord intact (n 3) rats. Abscissa, Time scale (min). Ordinates, Bladder contraction amplitude (A) and EUS EMG (B) as a% of control. Open arrow, Injection of MK-801. Closed arrow, Injection of LY pressed EUS activity had no effect on the bladder, this would support the data presented above that the EUS pathways are more sensitive to AMPA antagonists. In this study, the systemic administration of high doses of LY (10 mg/kg) produced a marked change in respiration, an effect reported previously by others (Browne and McCulloch, 1994; Gill and Lodge, 1994). However, a possible indirect influence on the bladder and EUS activity due to a change in cardiorespiratory function, can be excluded by the results of the experiments in which LY was injected via the i.t. route. In these experiments, the drug did not alter respiration, but did abolish bladder and EUS activity. Regarding cardiovascular function, it has been reported that the active (-)-isomer of LY (LY293358) (Ornstein et al., 1993b; Schoepp et al., 1995) does not alter blood pressure significantly (Browne and McCulloch, 1994). Thus, it seems unlikely that changes in respiration or blood pressure can account for the inhibition of bladder and EUS activity by LY In summary, these experiments together with results from previous studies suggest that glutamic acid is an important neurotransmitter in the micturition reflex pathway in the rat. It is clear that this substance acts via NMDA and AMPA receptors in the lumbosacral spinal cord when the descending pathways from the brainstem to the cord are intact. However, in chronic paraplegic rats after the descending pathways are disrupted, glutamatergic transmission is essential for only part of the spinal reflex mechanisms controlling micturition. Thus, other neurotransmitter mechanisms must be postulated to account for the remaining reflex bladder activity. Acknowledgments The authors are grateful to Eli Lilly and Company, Merck, Sharp & Dohme Research Laboratories and Dr. I. Tarnawa, Institute for Drug Research for gifts of LY215490, MK-801 and GYKI HCl, respectively. We thank Dr. Paul L. Ornstein of the CNS Research Division of Lilly Research Laboratories (Eli Lilly and Company) for his helpful advice and Ms. Elaine R. Black for her assistance in the preparation of the manuscript. References ARAKI, I. AND DE GROAT, W. C.: Unitary excitatory synaptic currents in preganglionic neurons mediated by two distinct groups of interneurons in neonatal rat sacral parasympathetic nucleus. J. Neurophysiol. 76: , BROWNE, S. E. AND MCCULLOCH, J.: AMPA receptor antagonists and local cerebral glucose utilization in the rat. Brain Res. 641: 10 20, CONTE, B., MAGGI, C. A., PARLANI, M., LOPEZ, G., MANZINI, S. AND GIACHETTI, A.: Simultaneous recording of vesical and urethral pressure in urethaneanaesthetized rats: Effect of neuromuscular blocking agents on the activity of the external urethral sphincter. J. Pharmacol. Methods 26: , DE GROAT, W. C., BOOTH, A. M. AND YOSHIMURA, N.: Neurophysiology of micturition and its modification in animal models of human disease. In: The Autonomic Nervous System: Nervous Control of the Urogenital System, vol. 3, ed. by C. A. Maggi, pp , Harwood Academic Publishers, London, U.K., 1993.

11 904 Yoshiyama et al. Vol. 280 FELDBERG, W. AND FLEISCHHAUER, K.: Penetration of bromophenol blue from the perfused cerebral ventricles into the brain tissue. J. Physiol. (Lond.) 150: , FOUTZ, A. S., PIERREFICHE, O. AND DENAVIT-SAUBIÉ, M.: Combined blockade of NMDA and non-nmda receptors produces respiratory arrest in the adult cat. NeuroReport 5: , GILL, R.: The pharmacology of -amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA)/kainate antagonists and their role in cerebral ischaemia. Cerebrovasc. Brain Metab. Rev. 6: , GILL,R.AND LODGE, D.: The neuroprotective effects of a combination of MK-801 and NBQX in focal ischaemia. Br. J. Pharmacol. (Proc. Suppl.) 107: 59P, GILL, R. AND LODGE, D.: The neuroprotective effects of the decahydroisoquinoline, LY ; a novel AMPA antagonist in focal ischaemia. Neuropharmacology 33: , HONORÉ, T., DAVIES, S. N., DREJER, J., FLETCHER, E. J., JACOBSEN, P., LODGE, D. AND NIELSEN, F. E.: Quinoxalinediones: Potent competitive non-nmda glutamate receptor antagonists. Science 241: , KAKIZAKI, H., YOSHIYAMA, M. AND DE GROAT, W. C.: Role of NMDA and AMPA glutamatergic transmission in spinal c-fos expression after urinary tract irritation. Am. J. Physiol. 270: R990 R996, KRUSE, M. N., NOTO, H., ROPPOLO, J.R.AND DE GROAT, W. C.: Pontine control of the urinary bladder and external urethral sphincter in the rat. Brain Res. 532: , LÖSCHER, W., RUNDFELDT, C. AND HÖNACK, D.: Low doses of NMDA receptor antagonists synergistically increase the anticonvulsant effect of the AMPA receptor antagonist NBQX in the kindling model of epilepsy. Eur. J. Neurosci. 5: , MALLORY, B., STEERS, W. D. AND DE GROAT, W. C.: Electrophysiological study of micturition reflexes in rats. Am. J. Physiol. 257: R410 R421, MAGGI, C. A., GIULIANI, S., GIACHETTI, A. AND MELI, A.: The effect of MK-801 on the micturition reflex in anesthetized rats. Eur. J. Pharmacol. 181: , MAGGI, C. A., SANTICIOLI, P. AND MELI, A.: The nonstop transvesical cystometrogram in urethane-anesthetized rats. J. Pharmacol. Methods 15: , MATSUMOTO, G., HISAMITSU, T. AND DE GROAT, W. C.: Role of glutamate and NMDA receptors in the descending limb of the spinobulbospinal micturition reflex pathway of the rat. Neurosci. Lett. 183: 58 61, 1995a. MATSUMOTO, G., HISAMITSU, T. AND DE GROAT, W. C.: Non-NMDA glutamatergic excitatory transmission in the descending limb of the spinobulbospinal micturition reflex pathway of the rat. Brain Res. 693: , 1995b. MATSUMOTO, G., KAWATANI, M. AND TAKESHIGE, C.: Inhibitory effect of 6-cyano- 7-nitroquinoxaline-2,3-dione (CNQX) on the micturition reflex in rat. Neurosci. Lett. 133: , ORNSTEIN, P. L., ARNOLD, M. B., AUGENSTEIN, N. K., LEANDER, J. D., LODGE, D. AND SCHOEPP, D. D.: In vitro and in vivo characterization of a novel, systematically active competitive AMPA antagonist, LY Soc. Neurosci. Abstr. 19: 293, 1993a. ORNSTEIN, P. L., ARNOLD, M. B., AUGENSTEIN, N. K., LODGE, D., LEANDER, J. D. AND SCHOEPP, D. D.: (3SR,4aRS,6RS,8aRS)-6-[2-(1H-tetrazol-5-yl)ethyl]decahydroisoquinoline-3-carboxylic acid: a structurally novel, systemically active, competitive AMPA receptor antagonist. J. Med. Chem. 36: , 1993b. SAPRU, H. N. AND KRIEGER, A. J.: Procedure for the decerebration of the rat. Brain Res. Bull. 3: , SCHOEPP, D. D., LODGE, D., BLEAKMAN, D., LEANDER, J. D., TIZZANO, J. P., WRIGHT, R. A., PALMER, A. J., SALHOFF, C. R. AND ORNSTEIN, P. L.: In vitro and in vivo antagonism of AMPA receptor activation by (3S,4aR,6R,8aR)-6-[2-(1(2)Htetrazole-5-yl)ethyl]decahydroisoquinoline-3-carboxylic acid. Neuropharmacology 34: , SUGAYA, K. AND DE GROAT, W. C.: Effects of MK-801 and CNQX, glutamate receptor antagonists, on bladder activity in neonatal rats. Brain Res. 640: 1 10, SUZUKI, T., BIRDER, L. A., YOSHIYAMA, M., ROPPOLO, J. R. AND DE GROAT, W. C.: Electrophysiological studies of the effects of glutamate antagonists (MK801 and CNQX) on the micturition reflex in the rat. Soc. Neurosci. Abstr. 17: 1002, VERA, P. L. AND NADELHAFT, I.: MK-801, a non-competitive NMDA receptor antagonist, produces facilitation of the micturition reflex in awake, freelymoving rats. Neurosci. Lett. 134: , YAKSH, T. L. AND RUDY, T. A.: Chronic catheterization of the spinal subarachnoid space. Physiol. Behav. 17: , YOSHIYAMA, M., ROPPOLO, J. R. AND DE GROAT, W. C.: Effects of MK-801 on the micturition reflex in the rat possible sites of action. J. Pharmacol. Exp. Ther. 265: , 1993a. YOSHIYAMA, M., ROPPOLO, J. R. AND DE GROAT, W. C.: Alteration by urethane of glutamatergic control of micturition. Eur. J. Pharmacol. 264: , YOSHIYAMA, M., ROPPOLO, J. R. AND DE GROAT, W. C.: Effects of GYKI and CNQX, AMPA/kainate receptor antagonists, on the micturition reflex in the rat. Brain Res. 691: , 1995a. YOSHIYAMA, M., ROPPOLO, J. R. AND DE GROAT, W. C.: Interactions between NMDA and AMPA/kainate receptors in the control of micturition in the rat. Eur. J. Pharmacol. 287: 73 78, 1995b. YOSHIYAMA, M., ROPPOLO, J. R., RIHMLAND, J., BLASTOS, B. AND DE GROAT, W. C.: The effects of MK-801, an NMDA receptor antagonist, on the micturition reflex in the rat. Neurosci. Lett. 126: , YOSHIYAMA, M., ROPPOLO, J. R., THOR, K. B. AND DE GROAT, W. C.: Effects of LY274614, a competitive NMDA receptor antagonist, on the micturition reflex in the urethane-anaesthetized rat. Br. J. Pharmacol. 110: 77 86, 1993b. YOSHIYAMA, M., ROPPOLO, J. R., THOR, K. B. AND DE GROAT, W. C.: Effects of AMPA/kainate (LY215490) and metabotropic (MCPG) glutamatergic receptor antagonists, on the micturition reflex in rats. Soc. Neurosci. Abstr. 21: 1872, 1995c. Send reprint requests to: Dr. Mitsuharu Yoshiyama, University of Pittsburgh, School of Medicine, Department of Pharmacology, E1303A Biomedical Science Tower, Pittsburgh, PA