Neurons of the Bed Nucleus of the Stria Terminalis (BNST) Electrophysiological Properties and Their Response to Serotonin DONALD G. RAINNIE a Harvard Medical School and Department of Psychiatry, Brockton Veterans Affairs Medical Center, Brockton, Massachusetts 02301, USA INTRODUCTION A descending pathway from the bed nucleus of the stria terminalis (BNST) to nuclei within the hypothalamus and brainstem has been implicated in the generation of anxiety. 1 Drugs such as the 5-HT 1A agonist, buspirone, and specific serotonin reuptake inhibitors (SSRIs), which are used to treat anxiety behaviors, act to modulate serotonin (5-HT) transmission. 2 Hence, 5-HT may act as a neuromodulator in the BNST to regulate anxiety states. The BNST receives serotonergic afferents from the midbrain raphé nuclei, 4 and has high densities of 5-HT1-like receptors. 3 Moreover, infusion of buspirone into the BNST blocks light-enhanced startle. 1 It is, therefore, important to know how excitation is regulated within the BNST, and how 5-HT may affect neuronal excitation and hence signal transduction. The aim of this study was to determine the electrophysiological properties of individual neurons, and to examine the effects of serotonin on these intrinsic properties, using whole-cell patch clamp recording from BNST neurons, in vitro. METHODS Whole-cell patch clamp records were obtained from neurons of the dorsal subdivisions of the BNST in an in vitro slice preparation of the rat using standard techniques. Briefly, 25 35-day-old Long-Evans rats were decapitated, and the brains were rapidly removed and placed in ice-cold, oxygenated, ACSF containing in mm: NaCl, 124; KCl, 2.2; KH 2 PO 4, 3.0; MgCl 2, 1.3; CaCl 2, 2.5; NaHCO 3, 26; and glucose, 10. Slices, 500 µm thick, containing the BNST were prepared and incubated in oxygenated ACSF at room temperature for at least 1 hour prior to experimentation. Patch electrodes (6 8 MΩ) were filled with recording solution containing, in mm: KGluconate, 120; KCl, 10; phosphocreatinine, 10; MgCl 2, 3; HEPES, 10; MgATP, 2; NaGTP, 0.2; biocytin 0.40% and titred to ph 7.2 and 280 mosm. Whole-cell patch clamp records were obtained using an Axopatch-1D preamplifier and pclamp 6.0 a Address for correspondence: Dr. Donald G. Rainnie, Dept. of Psychiatry, Neuroscience Laboratory 151C, Brockton Veterans Affairs Medical Center, 940 Belmont Street, Brockton, MA 02401. Voice: 508-583-4500, ext. 2589; fax: 508-895-0059; donald_rainnie@hms.harvard.edu 695
696 ANNALS NEW YORK ACADEMY OF SCIENCES FIGURE 1. Typical whole-cell patch clamp records obtained from a neuron located in the medial-dorsal BNST. (A 1 ) Voltage excursions of a medial BNST neuron in response to transient hyperpolarizing current injection of increasing amplitude ( 10 to 50 pa). (A 2 ) In the same neuron depolarizing current injection evoked a depolarizing prepotential subthreshold to action potential generation. (B 1 ) In control ACSF hyperpolarizing voltage commands of increasing amplitude (upper trace) activate a time- and voltage-dependent inward current. (B 2 ) Application of CsCl 2 (3 mm) to the ACSF blocked the expression of the inward current and revealed a time-independent rectification. (C 1 ) In the presence of TTX (0.6 µm), a dualpulse protocol elicited a voltage-dependent rapidly activating inward current. (C 2 ) The inward current was blocked by reducing the ACSF calcium concentration from 3 to 0.3 mm.
RAINNIE: NEURONS OF THE BNST 697 FIGURE 2. Exogenous serotonin activates both an outward and an inward conductance in BNST neurons. In 39% of BNST neurons 5-HT (50 µm) evoked a monophasic membrane hyperpolarization ( 4.2 ± 0.34 mv, n = 15). (A 1 ) In 33% of BNST neurons a biphasic hyperpolarization-depolarization was observed in current clamp. The hyperpolarization decayed during the period of 5-HT application and was followed by a period of rebound excitation following washout. (A 2 ) In voltage clamp, in the presence of TTX, the 5-HT-induced membrane hyperpolarization was associated with an outward current ( 50 pa) at 60 mv, and an increased conductance. (A 3 ) Subtraction of the steady state conductances dem-
698 ANNALS NEW YORK ACADEMY OF SCIENCES software for data acquisition and analysis. Only those cells that showed a stable resting membrane potential (< 55 mv) for at least 5 minutes, and an overshooting action potential were accepted for drug application. Drugs were applied in the ACSF at known concentrations. Serotonin stock solution was made fresh on the day of experimentation and stored on ice until used. RESULTS Patch clamp records were obtained from 53 neurons of the BNST, dorsal to the anterior commissure, between 0.0 and 0.6 Bregma. Of the 22 neurons successfully recovered for visualization, 10 were located in the medial subdivision (mbnst), and 12 were located in the lateral subdivision (lbnst). Neurons of the mbnst had a significantly higher R m (680 ± 120 MΩ) than those of the lbnst (250 ± 26 MΩ; p = 0.001, F = 12.727). Similarly, τ was significantly greater for mbnst neurons (34 ± 4.8 ms) than that for lbnst (21 ± 1.1 ms, n = 11, p = 0.024, F = 5.93). In contrast, no significant difference was observed in Vm ( 63 ± 0.6 mv) for each neuronal population. Seventy percent of all BNST neurons examined expressed a depolarizing sag in the voltage response to transient hyperpolarizing current injection (FIG. 1A 1 ), that was mediated by activation of the time- and voltage-dependent nonspecific cation conductance, I h (FIG. 1B 1 2 ). Upon transient depolarization, these same neurons expressed a depolarizing envelope on which rode two or more action potentials (FIG. 1A 2 ). The depolarizing envelope was mediated by activation of the low threshold calcium current (I T ; FIG. 1C 1 2 ). At the resting Vm, electrical stimulation of the stria terminalis evoked an excitatory postsynaptic potential (EPSP), or an EPSP followed by an inhibitory postsynaptic potential (IPSP). The IPSP was abolished by the GABA A receptor antagonist, bicuculline methiodide (30 µm), whereas the EPSP was almost totally abolished by the AMPA receptor antagonist, DNQX (10 µm). Serotonin (5-HT; 50 100 µm) was applied to 38 BNST neurons; 15 responded with a membrane hyperpolarization, 13 with a hyperpolarization followed by a depolarization, and 7 with a membrane depolarization. The remaining 3 neurons were unresponsive to 5-HT application. No difference was observed in the response of mbnst or lbnst neurons to 5-HT. The response persisted in the presence of tetrodotoxin (TTX, 1.2 µm) suggesting activation of postsynaptic 5-HT receptors. In current clamp, the hyperpolarization was associated with a reduction of the membrane input resistance (FIG. 2A), and by an increased outward current in voltage clamp mode (FIG. 2B). The outward current reversed polarity at 80 mv, which is close to the potassium reversal potential for this preparation. The depolarization was associated with a small increase in membrane input resistance, and was mediated by an inward current that reversed polarity at 40 mv. onstrated that the pure I 5-HT was an inwardly rectifying conductance. (B 1 ) A typical example of the 18% of BNST neurons that expressed a 5-HT-induced membrane depolarization. (B 2 ) In voltage clamp, a comparison of the steady state conductance before and during 5-HT application revealed an inward current at all points on the ramp protocol ( 100 to 40 mv). (D) The pure I 5-HT current was inward at 60 mv ( 30 pa) and showed a reversal potential close to 40 mv.
RAINNIE: NEURONS OF THE BNST 699 In 7/8 neurons examined, application of 5-HT (50 µm) also reduced the amplitude of evoked EPSPs by 41 ± 8%, irrespective of the postsynaptic response of BNST neurons. In the remaining neuron the EPSP amplitude was increased by 20%. In voltage clamp, the 5-HT-induced reduction of evoked EPSP amplitude was mediated by an equivalent reduction in the amplitude of the EPSC. These data suggest that serotonin also may regulate BNST excitability at a presynaptic locus. Experiments are in progress to determine the 5-HT receptor subtype/s involved in the preand postsynaptic response of BNST neurons to 5-HT application. DISCUSSION These data demonstrate for the first time that the excitability of neurons of the mbnst and lbnst can be directly, and indirectly, regulated by serotonin. Hence, activation postsynaptic 5-HT receptors can result in either (1) a membrane hyperpolarization mediated by activation of an inwardly rectifying potassium conductance, and/or (2) a membrane depolarization, the mechanism of which has yet to be determined. In addition, 5-HT acts at a presynaptic locus on glutamatergic afferents to reduce the release of glutamate and, hence, indirectly reduce the excitability of the BNST. Although a continuum was observed in the response of BNST neurons to 5-HT, a correlation has yet to be established between the relative expression of either I h and/or I T in any given neuron, and its response to 5-HT. However, these results suggest that the input-output responses of these two BNST subdivisions are regulated by multiple 5-HT receptor subtypes. If activation of these 5-HT receptors has a direct impact on the behavioral expression of anxiety, the duality of response to 5-HT, demonstrated here, may act as a feedback control mechanism. Moreover, Davis and coworkers (1997) have reported that perfusion of the AMPA receptor antagonist, NBQX, into the BNST attenuates light-enhanced startle. Attenuation of the AMPA receptor-mediated EPSC by 5-HT, reported here, may represent an endogenous anxiolytic mechanism. Experiments are in progress to identify the 5-HT receptor subtypes responsible for the responses reported here. It is hoped that identification of these receptors may ultimately lead to the development of more specific treatments for anxiety. REFERENCES 1. DAVIS, M. et al. 1997. Amygdala and bed nucleus of the stria terminalis: differential roles in fear and anxiety measured with the acoustic startle reflex. Phil. Trans. R. Soc. Lond. 352: 1675 1687. 2. DE VRY, J. 1995. 5-HT 1A receptor agonists: recent developments and controversial issues. Psychopharmacology 121: 1 26. 3. PAZOS, A. & J.M. PALACIOS. 1985. Quantitative autoradiographic mapping of serotonin receptors in the rat brain. I. Serotonin-1 receptors. Brain Res. 346: 205 230. 4. PHELIX, C.E. et al. 1992. Monoamine innervation of bed nucleus of stria terminalis: an electron microscopic investigation. Brain Res. Bull. 28: 949 965.