Intraseptal injection of GABA and benzodiazepine receptor ligands alters highaffinity choline transport in the hippocampus

BrainResearchLWeiin,Vol. 3 1,QQ.267-27 I, 1993

0361-9230193 $6.00 + .OO Copyright 0 1993 Pergamon Press Ltd.

Printed in the USA. AI1rights reserved.

Intraseptal Injection of GABA and Benzodiazepine Receptor Ligands Alters HighAffinity Choline Transport in the Hippocampus THOMAS J. WALSH,’ ROBERT W. STACKMAN,

DWAINE F. EMERICH AND LISA A. TAYLOR

Department of Psychology, Rutgers University, New Brunswick, NJ 08903 Received 20 April 1992; Accepted 6 October 1992 WALSH, T. J., R. W. STACKMAN, D. F. EMERICH AND L. A. TAYLOR. IntruseptulinjectionofGAlL and benzudiuzepine receptorligands alter high-affmitycholine transportin the hippocampus.BRAIN RES BULL 31(3/4) 267-271, 1993.--Injection of GABA and benzodiazepine (BDZ) agonists and antagonists into the medial septum produced bidirectional alterations in hip~rn~ hid-affinity choline transport (HAChT). Male Sprague-Dawley rats were injected in the medial septum with either drug vehicle, a BDZ agonist, Zionist, or inverse agonist, or with a GABA-A or GABAB agonist or Zionist and sacrificed 1h later for assessment of HAChT in hip~ampal synaptosomes. The GABA-A agonist muscimol, the GABA-B agonist baclofen, and the BDZ agonist ~hlordi~e~xide (CDP) produced dose-related decreases in HAChT 1h following injection into the septum. The muscimol-induced decrease in HAChT was prevented by prior intraseptal injection of the GABA-A antagonist, bicuculline. Intraseptal injection of GABA-A (bicuculline) or GABA-B (2-hydroxysaclofen) antagonists did not alter HAChT, whereas the BDZ antagonist flumazenil (RO15,1788) and the BDZ inverse agonist methyl-gcarboline-3-carboxylate (&CCM) increased this measure up to 30% in a dose-dependent manner. These results demonstrate that cholinergic neurons in the medial septum can be modulated in a bidirectional way through the pharmacological manipulation of GABA-A, GABA-B, and BDZ receptors. The

potential functional and therapeutic consequences of these interactions are discussed. Hippocampus

CHOLINERGIC entire

Benzodiazepines

Inverse agonists

Flumazenil

/3-Carbolines

High affinity choline transport

activation of GABA-A receptors in the medial septum should exert an inhibitor influence on choliner~~ neurons. In fact, injection of GABA-A agonists into the septum has been shown to decrease the turnover of acetylcholine in the hippocampus and reduce the incidence of theta activity (1). A variety of other neurotransmitters (catecholamines, opiates, glutamate, neuropeptides) might also exert their influence on cholinergic neurons in the septum through a GABAergic pathway [reviewed in (11,34)]. Thus, inhibitory GABAergic neurons in the septum serve as a final common pathway for the regulation of septal cholinergic neurons that project to the HPC. The GABA-A receptor is a molecular complex that includes a) a GABA receptor, b) a benzodiazepine receptor (BDZ), and c) a chloride ion channel. BDZ receptors are found in the septum, and the concentration of the endogenous BDZ receptor ligand N-desmethyldi~epam is highest in this structure (35). Agonistinduced activation of the GABA-A/BDZ receptor opens the chloride channel and either hyperpolarizes the postsynaptic membrane or neutralizes the effects of an existing depolarization ( 12). The BDZ receptor also recognizes a variety of /3-carbolines that exhibit antagonist or inverse agonist properties. Inverse agonists produce pharmacological and behavioral actions that are opposite to those induced by the BDZ agonist (6,21). In contrast, the GABA-B receptor mediates an inhibitory response via two

neurons in the medial septum project to the

sept~tem~ml extent of the HPC ( 13) and modulate the

inherent synaptic plasticity of this structure by inducing the theta rhthym (4,l.S). Disruption of the cholinergic component of the septohippocampal pathway by drugs, lesions, neurotoxins, or neurodegenerative diseases, results in a selective profile of memory impairments (33). A better understanding of the synaptic regulation of this pathway should help to define the neural substrates of memory and perhaps lead to new therapeutic approaches for the treatment of cognitive disorders with an underlying cholinergic pathology. Cholinergic neurons within the medial septum appear to be synaptically regulated by a number of transmitter systems (9,11,36). In particular, GABAergic neurons in the lateral septum exert a powerful inhibitor influence on cholinergic cells in the medial septum (7,17). Leranth and Fro&her (16), using doublelabeling techniques, have shown that GABAergic fibers originating from neurons in the lateral septum project to and synapse on cholinergic neurons in the medial septum. Chu and colleagues (10) have recently demonstrated that the medial septum does contain GABA receptors, but the ratio of GABA-A to GABAB receptors is approximately 4 to 1. Furthermore, in all brain areas examined GABA-A receptors have a threefold lower Ko and a twofold higher B,,, than GABA-B receptors. Therefore,

’ To whom requests for reprints should be addressed. 267

WALSH ET AL.

268 G protein-linked mechanisms that produce a presynaptic depression of Ca’+ conductance, or a postsynaptic increase in K+ conductance (5). Because GABA, endogenous BDZs (endozepines), and their receptors are located in the septum, it is likely that they participate in the local modulation of septal cholinergic neurons. The following experiments examined whether pharmacological manipulations of specific GABA and BDZ receptors in the medial septum would alter the activity of cholinergic neurons that innervate the hip~campus. METHOD

TABLE I PHARMCOLOGICAL AGENTS USED IN EXPERIMENTS 1. 2, AND 3

nmol GABA-A

Agonist Muscimol

Cannula Implantation

Animals were anesthetized with sodium pentobarbital (50 mg/kg) and positioned in a stereotaxic apparatus (David Kopf, Tujunga, CA). A sagittal incision was made in the scalp, and one hole was drilled through the skull for placement of a 26 gauge (0.46 mm o.d. and 0.26 mm i.d.) stainless steel guide cannulae (22). The cannulae were placed 1.5 mm dorsal to the medial septum at the following coordinates; 2.0 mm anterior to bregma, 0.0 mm lateral to the sag&al suture, and 4.0 mm beneath the dura (22). This placement allowed for the internal cannulae to extend an additional 1.5 mm beyond the tip of the guide, into the medial septal region (5.5 mm below dura). The internal cannulae were constructed from 32 gauge stainless steel hypodermic tubing (0.23 mm o.d. and 0.10 mm i.d.) (Small Parts lnc, Miami, FL). A 32 gauge obdurator, also constructed of stainless steel hypodermic tubing, was present in each guide cannula until the time of injection. Animals were allowed a minimum of 2 weeks to recover from surgery before the pharmacological manipulations were initiated. Intraseptal Injection Procedure

In the first set of studies rats were infused with either a GABA or BDZ agonist or artificial cerebrospinal fluid (CSF) into the medial septum. The doses of these compounds are presented in Table 1. Rats were unanesthetized and were gently held during the injection process. They were injected with either the GABAA agonist muscimol, the GABA-B agonist baclofen, the BDZ agonist chlordiazepoxide, or artificial cerebrospinal fluid (CSF: 147 mM NaCl; 2.9 mM KCI; 1.6 mM MgC12; 1.7 mM CaCl,; 35.9 mM HC09 and 2.2 mM dextrose). In the second set of studies, rats were injected with either a GABA or BDZ antagonist or CSF into the medial septum (see Table 1 for doses). Rats were injected with either the GABA-A antagonist (-)bicuculline methiodide, the BDZ antagonist flumazenil, the BDZ inverse agonist methyl &carboline-3-carboxylate (@CM), polyethylene glycol (PEG) vehicle (S.O%),or CSF. The BDZ antagonist and inverse agonist solutions were prepared with 5.0% PEG as the vehicle. The final experiment examined the dose-related effects of intraseptal injection of 2-hydroxysaclofen, a GABA-B antagonist or the drug vehicle (distilled water), on HAChT. As a positive control another group of rats was injected into the septum with 10 nmol of flumazenil. Muscimol, baclofen, 2-hydroxy~clofen, and @-CCM were obtained from RBI (Natick, MA). ~h~ordi~epoxide hydrochlo-

0.75 3.00

BDZ agonist Chlordiazepoxide

Subjects

Adult male Sprague-Dawley rats (Charles River Breeders, Wilmington, MA), approximately 120 days old and weighing between 250 and 300 g, were used in the following experiment. The animals were individually housed and maintained in a temperature-controlled colony room with a 12L: I2D cycle (lights on at 0700 h). Food and water were continuously available.

Grams

15.00 30.00

90 ng 340 ng 5.0 /*s 10.0 rg

GABA-B

agonist Baclofen

GABA-A antagonist Bicuculline

3.00 6.00

640 ng 1,280 ng

0.50

250 ng

I .oo

500 ng

5.00

750 ng 1,500 ng

BDZ antagonist

Flumazenil

IO.00 BDZ inverse agonist R-CCM GABA-B antagonist 2-Hydroxysaclofen

I.00 2.00

230 ng 460 ng

9.63 19.25 38.50

2.50 fig 5.00 pg

10.0 /,lg

ride and bicuculline methiodide, were obtained from Sigma (St. Louis, MO), and flumazenil was generously provided by Hoffmann-LaRoche (Nutley, NJ). All doses of drugs were infused in a volume of 0.5 ~1 and infused over a 3-min interval. Injection cannulae remained in place for an additional 2 min following the injection to allow for diffusion of the compounds. Animals were returned to home cages for a I h delay prior to assessment of HAChT. Doses of all the drugs were chosen based upon values in the literature and preliminary studies in our laboratory.

Animals were sacrificed by decapitation 1 h following intraseptal injection. Brains were rapidly removed and the hippocampi were dissected free. HAChT was assessed in hippocampal synaptosomes according to the procedure of Yamamura and Snyder (37). Hippocampal samples were homogenized on ice in 20 volumes of 0.32 M sucrose using hand-held Potter Elvehjem tissue grinders. Homogenates were then centrifuged at 1,000 X g for 10 min (4°C). The supernatant was then recentrifuged at 20,000 X g for 20 min, and the resulting P2 pellet was resuspended in the original volume of sucrose. Duplicate aliquots (50 ~1) of this suspension were then added to 450 ~1 of ice-cold buffer that contained 0.2 FM choline, 0.25 &i f3H]choline, 126 mM NaCl, 9.6 mM KC& 4.2 mM MgSO,, 2.4 mM CaC12-H20, 252 mM dextrose and 40.0 mM Tris base. After incubation for 4 min at 37”C, tubes were placed on ice and quenched immediately with the addition of 3 ml of ice-cold buffer. Tissue was collected on glass fiber filters Type A/E (Gelman) by vacuum filtration. After washing twice with 3 ml of ice-cold buffer, the filters were placed in scintillation vials and 10 ml of scintillation fluid (Hydrofluor, Nationaf Diagnostics) were added to each. Vials were assayed

INTRASEPTAL

GABA/BDZ

LIGANDS

TABLE

AFFECT

2

EFFECT OF INTRASEPTAL GABA AND BENZODIAZEPINE AGONISTS ON HIPPOCAMPAL HIGH-AFFINITY CHOLINE TRANSPORT

Percent of DW

HAChT

COlltd

CSF Muscimol 0.75 nmol 3.0 nmol Bicuculline 1.O nmol and Muscimol 3.0 nmol Baclofen 3.0 nmol 6.0 nmol Chlordiazepoxide 15.0 nmol 30.0 nmol

36.75 (4.76)

-

-

81.81 (4.55) 66.95 (7.75)*

-

92.73 (9.66)

-

81.19 (10.46) 70.91 (3.82)*

-

67.14 (4.49)* 48.80 (10.5 I)*

Intraseptal bicuculline was administered 5 min prior to the muscimol injection to those rats in the combination condition. Rats were sacrificed I h after injection. Data are expressed as the mean percent of control

(+SEM). The control (CSF) value represents the mean pmol choline transported/mg protein/4 min. * p < 0.05 vs. CSF-treated controls, Fisher’s LSD test. for radioactivity in a Packard TriCarb (Model 4530) scintillation counter with a 66% counting efficiency for ‘H. HAChT was defined as the amount of choline transported into tissue at 37°C minus that accumulated at 4°C. HAChT is both temperature and sodium dependent (29,30). Simon and Kuhar (29) demonstrated that the low-affinity choline uptake system is insensitive to changes in temperature, although the sodium-dependent high-affinity system is inhibited greater than 80% at 4°C. It is important to note that sodium-dependent highaffinity choline uptake is inhibited less than 80% in a sodiumfree buffer media in hippocampal synaptosomes at 1.0 PM of choline (30). Therefore, the sensitivity of both commonly used assays of choline uptake (temperature dependent and sodium dependent) is comparable. We have observed similar results with these two assays using a variety of experimental manipulations. The [3H]choline was obtained from NEN DuPont and had a specific activity of 80.0 Ci/mmol. The concentration of choline (0.2 wM) used in our experiments was chosen based upon previous studies conducted in our lab and upon literature evidence indicating linearity of HAChT into tissue over choline concentrations of approximately 0. l- 1.O PM (28). Tissue protein content was determined according to the procedure of Bradford using bovine serum albumin (Sigma, St. Louis, MO) as the reference standard. RESULTS

Visual inspection of the brain at time of sacrifice revealed that all cannulae were placed within the dorsal aspect of the medial septum. In a previous study we found that intraseptal injection of 0.5 ~1 of a solution of cresyl violet produced very slight diffusion to the most medial aspect of the lateral septum in a small percentage of rats, but not to other adjacent areas (3 1). Intraseptal Administration

269

HACHT

of GABA and BDZ Agonists

Hippocampal HAChT was significantly decreased by intraseptal administration of the GABA/BDZ agonists (see Table 2). A one-way ANOVA revealed a significant overall treatment ef-

feet, F(7,86) = 3.38, p = 0.0033. The results ofpost hoc Fisher’s LSD comparisons are presented in Table 2. There was no significant difference between noncannulated controls and CSFtreated controls. Implantation of the guide cannulae and injection of CSF produced no alteration in hippocampal HAChT; thus, all drug manipulations were compared to CSF-treated controls. Muscimol(3.0 nmol), baclofen (6.0 nmol), and chlordiazepoxide (15 and 30 nmol) significantly decreased hippocampal HAChT 30-5 1% 1 h after intraseptal injection. In a separate group of rats, CSF or the GABA-A antagonist bicuculline ( 1.O nmol) was injected into the septum 5 min prior to injection of CSF or 3.0 nmol of muscimol, creating the following groups: CSF-CSF, bicuculline-muscimol. Rats were sacrificed 1 h after the second injection. It was found that bicuculline pretreatment prevented the muscimol-induced decrease in HAChT. The bicuculline-muscimol group was comparable the CSF-CSF group, t( 1, 11) = 0.62; p > 0.05.

to

Intraseptal Injection of GABA and BDZ Antagonists Hippocampal HAChT was significantly increased by the intraseptal administration of BDZ antagonists (see Table 3). A one-way ANOVA revealed a significant overall treatment effect, F(7, 85) = 2.28, p = 0.037. The results of post hoc Fisher’s LSD comparisons are presented in Table 3. Infusion of 5.0% PEG produced no significant alteration in hippocampal HAChT as compared to CSF-treated controls; thus, all drug manipulations were compared to CSF controls. Intraseptal infusion of the BDZ antagonist flumazenil (10 nmol) and the BDZ inverse agonist @-CCM (2.0 nmol) increased hippocampal HAChT approximately 25-30%. In contrast, intraseptal injection of bicuculline had no effect on HAChT. Intraseptal injection of the GABA-B antagonist 2-hydroxysaclofen had no effect on HAChT (see Table 4). In contrast, intraseptal injection of 10 nmol of flumazenil significantly increased HAChT by 30% compared to the vehicle-injected group. A one-way ANOVA revealed an overall significant treatment effect, F(4, 47) = 2.86, p < 0.05. The results of post hoc comparisons using Fisher’s LSD test are presented in Table 4. TABLE 3 EFFECT OF INTRASEPTAL GABA AND BENZODIAZEPINE ANTAGONISTS ON HIPPOCAMPAL HIGH-AFFINITY CHOLINE TRANSPORT Drug

HAChT

Percent Control

CSF PEG 5.0% Bicuculline 0.5 nmol I .O nmol Flumazenil 5.0 nmol 10.0 nmol B-CCM 1.O nmol 2.0 nmol

37.30 (1.32) -

109.96 (10.34)

-

114.33 (8.04) 100.27 (5.95)

-

94.29 (5.96) 128.32 (11.83);

-

110.58 (9.67) 124.34 (6.05)*

The flumazenil and &CCM solutions were prepared in a 5.0% PEG vehicle. Rats were sacrificed and HAChT assessed 1 h after receiving the injection. Data are expressed as the mean percent of control (?SEM). The control (CSF) value represents pmol of choline transported/mg protein/4 min. * p < 0.05 vs. CSF-treated controls, Fisher’s PLSD test.

270

WALSH

TABLE

4

EFFECT OF INTRASEPTAL 2-HYDROXYSACLOFEN &ABA-B ANTAGONIST) AND FLUMAZENIL (BDZ ANTAGONIST) ON HIPPOCAMPAL HIGH-AFFINITY CHOLINE TRANSPORT

DW

HAChT

Distilled water 2-Hydroxysaclofen 9.6 nmol 18.25 nmol 38.5 nmol Flumazenil IO nmol

30.26 f 3.28 -

Percent Control

-

111.11 (18.11) 109.5 (7.67) 91.06 (5.42)

-

130.0 (10.55)*

Rats were sacrificed I h after injection. Data are expressed as the mean percent of control (?SEM). * p < 0.05 vs. distilled water-injected controls, Fisher’s LSD test. DISCUSSION

These experiments demonstrate that HAChT in the hippocampus can be either increased or decreased by discrete pharmacological manipulation of GABA/BDZ receptors in the medial septum. This measure was used to provide an in vitro index of the in vivo activity of the septohippocampal cholinergic pathway at the time of sacrifice (19,20,29,30). Intraseptal injection ofthe GABA-A or-B agonist. muscimol or baclofen, and the BDZ agonist, CDP, significantly decreased HAChT in the hippocampus. Pretreatment of rats with intraseptal bicuculline (1 .O nmol) effectively antagonized the effect of muscimol(3.0 nmol) on HAChT. This suggests that muscimol decreased HAChT through an interaction with GABA-A receptors in the medial septum. Injection of drug vehicles (artificial CSF or 5% PEG) into the septum had no effect on HAChT. Intraseptal infusion of 3.0 nmol of baclofen had no effect on hippocampal HAChT, although 6.0 nmol produced a significant decrease. Such a dose-dependent effect suggests that there may be a threshold of septal GABA-B receptor stimulation required before cholinergic neurons are affected. This is consistent with the observation that the density of GABA-B receptors in the medial septum is considerably less that that of GABA-A receptors (10). These data are in agreement with previous reports demonstrating that systemic or intraseptal injection of muscimol and systemic injection of CDP produces a dose-related decrease in several measures of presynaptic cholinergic function in the hippocampus (3,19,20). Intraseptal administration of the BDZ antagonist flumazenil (10.0 nmol) and the inverse agonist /3-CCM (2.0 nmol) increased hippocampal HAChT by approximately 25% 1 h after infusion. Recently, Miller and Chmielewski (18) reported that systemic injection of several fl-carbolines enhanced HAChT in both the cortex and hippocampus. Furthermore, different &carbolines produced regionally selective effects with DMCM (methyl-6,7dimethyl-4-ethyl+carboline-3-carboxylate), increasing HAChT in the hippocampus but not cortex, and both @-CCE (ethyl-pcarboline-3-carboxylate) and @-CCM increasing choline transport in the cortex but not the hippocampus. The regional selectivity of different /3-carbolines was suggested to depend on the differential distribution of BDZ receptor subtypes (BZ 1, BZ2) in these brain areas (18). More discrete pharmacological manipulations in different brain areas (i.e., medial septum and nucleus basalis) will be required to appreciate the regulation of different populations of cholinergic neurons by these receptor types. Injection of the GABA antagonists bicuculline (0.5 and 1.O nmol) or 2-hydroxysaclofen (9.63, 19.25, or 38.5 nmol) into the medial septum had no effect on HAChT. This is consistent with

ET AL.

previous reports demonstrating that intraseptal bicuculline attenuates the depressive effects of both muscimol and pentobarbital on ACh turnover rate, but has no effect on its own (9,33). However, other studies have found that high doses of intraseptal bicuculline (5.0 rg) can enhance HAChT (38). In our studies, intraseptal injection of 1.O nmol(500 ng) of bicuculline prevented the muscimol-induced decrease in HAChT. Therefore, this dose has pharmacological activity. However, it would be useful to construct a more complete dose-response curve in future studies. The ability to enhance HAChT with BDZ antagonists or inverse agonists but not with GABA antagonists suggests that cholinergic neurons might be differentially regulated by components of the GABA-A/BDZ receptor complex. Furthermore, BDZ ligands might exert actions that are not dependent upon the presence of GABA (23,24). BDZs have been shown to produce a potent, and perhaps clinically relevant, GABA-independent inhibition of adenosine uptake, as well as effects on membrane properties and ion channels [see (24) for review]. The ability to enhance the activity of cholinergic neurons by disinhibiting them with GABA-A or BDZ antagonists or inverse agonists may represent a useful approach to treat cognitive disorders that are characterized by a prominent cholinergic hypofunction such as Alzheimer’s disease. Chohnomimetic drugs have proven to be relatively ineffective due to a) the prevalence of side effects, b) their short half lives, c) their pharmacodynamic variability between patients. and d) their nonselective effects on different cholinergic systems and receptor subtypes (2). Furthermore, the indiscriminate activation of muscarinic receptors by cholinergic agonists might not enhance the functional activity of cholinergic systems which operate in a phasic and rhythmic way (i.e., generation of theta activity). A more useful approach might be to amplify the ongoing rhythmic activity ofcholinergic neurons by restricting the inhibitory influence of GABA-A/BDZ receptors [see (27)]. In fact, flumazenil and a variety of p-carbolines have been shown to enhance memory retention, and to attenuate the memory deficits induced by anticholinergic drugs or damage to the basal forebrain cholinergic complex (8,2528,32). A potential limitation of these compounds is that they can produce anxiogenic effects. However, there appears to be a better dose-response profile for memory enhancement with limited anxiogenesis for flumazenil compared to the p-carbolines. To fully appreciate these compounds it will be important to determine their range of activities in different models of dementia and the relationship between their behavioral effects (anxiogenesis and memory enhancement) and cholinergic activity. The discovery of specific BDZ receptors has lead to an extensive search for endogenous ligands for these sites. A variety of endogenous molecules with either agonist (N-desmethyldiazepam, diazepam), or inverse agonist (diazepambinding inhibitor (DBI) and its biologically active fragments. ODN and TTN, and n-butyl-P-carboline-3-carboxylate (@CCB) properties, are candidate ligands for these receptors. (14,34). Diazepam and n-desmethyldiazepam induce anxiolytic effects in animal models, although DBI, ODN, TTN, and &CCB are anxiogenic [see (34)]. It is also possible that there might be several endogenous BDZ ligands with either agonist or inverse agonist properties that can bidirectionally modulate the GABA-A receptor complex. Recently Wolfman and colleagues (35) reported that endogenous BDZs are altered in a regionally specific and task-dependent manner (35). These ligands are found in the highest concentration in the septum and are likely to play a role in the modulation of cholinergic neurons and the behaviors they mediate. Izquierdo and co-workers (14) suggest that endogenous BDZs might inhibit the ongoing activity of brain mechanisms

INTRASEPTAL

GABA/BDZ

LIGANDS

involved in memory. Therefore, ~ha~acolo~~ inhibition of endogenous BDZ-mediated mechanisms should enhance memory [for a review see ( 14)] through a disinhibition of cholinergic neurons that project to the hippocampus. our laboratory support this contention.

271

AFFECT HACHT

Preliminary

data from

ACKNOWLE~EME~~

The authors would like to thank Zaida Diaz for her expert technical assistance. This research was supported by NSF grant BNS-9109163

to T.J.W. REFERENCES

1. Allen, C. N.; Crawford, I. L. GABAergic agents in the medial septal

2.

3. 4. 5. 6.

7.

8.

9.

IO.

II.

12.

13.

14.

15.

16.

17. 18.

19.

20.

nucleus affect hippocampal theta rhythm and acetylcholine utilization. Brain Res. 322:261-267; 1984. Becker, R. E.; Giacobini, E. Mechanisms of cholinesterase inhibition in senile dementia of the Alzheimer type: Clinical, pha~acolo~cal, and therapeutic aspects. Drug Dev. Res. 12163-195; 1988. Blaker, W. D.; Cheney, D. L.; Gandolfi, 0.; Costa, E. Simultaneous modulation of hippocampal choline& activity and extinction by intraseptal muscimol. J. Pharmacoi. Exp. Ther. 225:361-365; 1983. Bland, B. H. The physiology and pharmacology of hippocampal formation theta rhythms. Prog. Neurobiol. 26: l-54; 1986. Bowery, N. GABA-B receptors and their significance in mammalian pharmacology. Trends Pharmacol. Sci. 10:40 l-407; 1989. Braestrup, C.; Nielsen, M.; Olsen, C. E. Urinary and brain betacarboline-3-carboxylates as potent inhibitors of brain benzodiazepine receptors. Proc. Natl. Acad. Sci. USA 77:2288-2292; 1980. Brashear, H. R.; Zaborszky, L.; Heimer, L. Distribution of GABAergic and cholinergic neurons in the rat diagonal band. Neuropharmacology 17:439-45 1; 1986. Brioni, J. D.; Arolfo, M. P.; Je~~insky, D.; Medina, J. H.; Izquierdo, I. The effect of flumazenil on a~uisition, retention, and retrieval of spatial information. Behav. Neural Biol. X:329-335; 1991. Bnmello, N.; Cheney, D. L. The septal-hip~ampal cholinergic pathway: Role in antagonism of pentobarbital anesthesia and regulation by various afferents. J. Pharmacol. Exp. Ther. 2 19:489-495; 1981. Chu, D. C. M.; Albin, R. L.; Young, A. B.; Penney, J. B. Distribution and kinetics ofGABA-B bindings sites in rat central nervous system: A quantitative autoradiographic study. Neuroscience 34:341-357; 1990. Costa, E.; Panula, P.; Thompson, H. K.; Cheney, D. L. The transynaptic regulation of the septal-hippocampal cholinergic neurons. Life Sci. 32:165-179; 1983. Duman, R. S.; Sweetnam, P. M.; Gallombardo. P. A.; Tallman, J. F. Molecular neurobiology of inhibitor amino acid receptors. Mol. Neurobiof. 1:155-1X9: 1987. Fro&her, M.; Leranth, C. Choline&c inne~ation of the rat hippocampus as revealed by choline acetyltransferase immunocytochemistry: A combined light and electron microscopic study. J. Comp. Neurol. 239:237-246; 1985. Izquierdo, I.; Da Cunha, C.; Medina, J. H. Endogenous benzodiazepine modulation of memory processes. Neurosci. Biobehav. Rev. 14:419-424; 1990. Larson, J.; Wong, D.; Lynch, G. Pattern stimulation at the theta frequency is optimal for the induction of hippocampal long-term potenfiation. Brain Res. 368:347-350; 1986. Leranth, C.; Frotscher, M. Organization of the septal region in the rat brain: Choline&-GABAergic interconnections and the termination ofhip~am~-~pta1 fibers. J. Comp. Neurol. 289:304-314; 1989. Mc~nMn, H.; Miller, J. J. GABA and inhi~tion in the septal nuciei of the rat. J. Physiol. 237:625-633; 1974. Miller, J. A.; Chmielewski, P. A. The regulation ofhi~-a~nity choline uptake in vitro in rat cortical and hippocampal synaptosomes by beta-carbolines administered in vivo. Neurosci. I_&. I 14:351355; 1990. Miller, 1. A.; Richter, J. A. Effects of anticonvulsants in vivo on high affinity choline uptake in vitro in mouse hippocampal synap tosomes. Br. J. Pharmacol. 84: 19-25; 1985. Miller, J. A.: Richter, J. A. Effects of GABAergic drugs in vivo on high-affinity choline uptake in vitro in mouse hippocampal synap tosomes. J. Neurochem. 47:1916-1918; 1986.

21. Ninan, P. T.; Insel, T. M.; Cohen, R. M.; Cook, J. M.; Skolnick, P.; Paul, S. M. Benzodiazepine receptor-mediated experimental ‘anxiety’ in primates. Science 2 18:1332-I 334; 1982. 22. Pellegrino, L. J.; Pellegrino, A. S.; Cushman, A. J. A stereotaxic atlas of the rat brain. New York: Plenum Press; 1979. 23. Phillis, J. W. Adenosine’s role in the central actions of the benzodiazepines. Prog. Neuro~ychoph~~ol. Biol. Psychiatry 8;495502; 1984. 24. Pole, P. GABA-independent mechanisms of benzodiazepine action. In: Briley, M.; File, S. E., eds. New concepts in anxiety. Boca Raton, FL: CRC Press; 1991:21 l-236. 25. Raffalli-Sebille, M.-J.; Chapouthier, G.; Venault, P.; Dodd, R. H. Methyl &carboline-3-carboxylate enhances performance in a multiple-trial learning task in mice. Pharmacol. Biochem. Behav. 35: 281-284; 1990. 26. Sarter, M.; Bodewitz, G.; Stephens, D. N. Attenuation of scopolamine-induced impairment of spontaneous alternation behavior by antagonist but not inverse agonist and agonist @carbolines. Psychopharmacology (Berlin) 94491-495; 1988. 27. Sarter, M.; Bruno, J. P.; Dudchenko, P. Activating the damaged basal forebrain cholinergic system: Tonic stim~ation versus signal ampli~cation. P~chopha~a~ol~y (Berlin) 101:l-l 7; 1990. 28. Sarter, M.; Steckler, T. Spontaneous exploration of a 6-arm radial maze by basal forebrain lesioned rats: Effects of the benzodiazepine receptor antagonist &carboline ZK 93 426. Psychopharmacology (Berlin) 98: 193-202; 1989. 29. Simon, J. R.; Kuhar, M. J. High affinity choline uptake: Ionic and energy requirements. J. Neurochem. 27:93-99; 1976. 30. Simon, J. R.; Atweh, S.; Kuhar, M. J. Sodium-dependent high affinity choline uptake: A regulatory step in the synthesis of acetylcholine. J. Neurochem. 26:909-921; 1976. 31. Stackman, R. W.; Walsh, T. J. Chlordiazepoxide-induced working memory impairments: Site specificity and reversal by flumazenii (R015,1788). Behav. Neural Biol. 57:233-243; 1992. 32. Venault, P.; Chapouthier, G.; Prado de Carvalho, L.; Simiand, J.; Morre, M.; Dodd, R. H.; Rossier, J. ~nz~~epine impairs and & car~iine enhances performance in learning and memory tasks. Nature 321:864-866; 1986. 33. Walsh, T. J.; Chrobak, J. J. Animal models of Alzheimer’s disease, Role of hippocampal cholinergic system in working memory. In: Dachowsky, L.; Flaherty, C., eds. Current topics in animal learning, brain, emotion &cognition. Hillsdale, NJ: Lawrence Erlbaum; 1991: 347-374. 34. Walsh, T. J,; Stackman, R. W. Modulation of memory by benzodiazepine-acetylcholine interactions. In: Levin E.; Butcher, I-., eds. Neurotransmittor interactions and cognitive function. Boston: Birkhauser; 1992:312-328. 35. Wolfman, C.; Da Cunha, C.; Jerusalinsky, D.; Levi de Stein, M.; Viola, H.; Izquierdo, I.; Medina, J. H. Habituation and inhibitory avoidance training alter brain regional levels of benzodiazepine-like molecules and are a&c&d by intracerebral flumazenil mi~oinj~on. Brain Res. 548:74-80; 1991. 36. Wood, P. L.; Cheney, D. L.; Costa, E. An inv~ti~tion of whether septal Emma-aminobutymte~on~ning interneurons are involved in the reduction in the turnover rate of acetylcholine elicited by substance P and @-endorphin in the hippocampus. Neuroscience 4: 1479-1484; 1979. 37. Yamamura, H. I.; Snyder, S. H. High affinity transport of choline into synaptosomes of rat brain. J. Neurochem. 2 1: 1355- 1374; 1973. 38. Zucker, J.; Calkins, D.; Zabawska, J.; Lai, H.; Ho&a, A. Effects of intraseptal drug administration on pentobarbital-induced narcosis and hippocampal choline uptake. Pharmacol. B&hem. Behav. 28: 433-436; 1987.