Median preoptic nucleus neurons: an in vitro patch-clamp analysis of their intrinsic properties and noradrenergic receptors in the rat

Pergamon

PII:

Neuroscience Vol. 83, No. 3, pp. 905–916, 1998 Copyright ? 1998 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/98 $19.00+0.00 S0306-4522(97)00435-1

MEDIAN PREOPTIC NUCLEUS NEURONS: AN IN VITRO PATCH-CLAMP ANALYSIS OF THEIR INTRINSIC PROPERTIES AND NORADRENERGIC RECEPTORS IN THE RAT D. BAI and L. P. RENAUD* Neurology/Neurosciences, Loeb Research Institute, Ottawa Civic Hospital and University of Ottawa, Ottawa, Ontario, Canada, K1Y 4E9 Abstract––The median preoptic nucleus is recognized as an important forebrain site involved in hydromineral and cardiovascular homeostasis. In the present study, whole cell patch-clamp recordings in parasagittal slices of adult rat brain were used to obtain information on the properties of median preoptic neurons. Lucifer Yellow-labelled cells demonstrated small ovoid somata with two to three aspiny main dendrites and axons that branched sparingly. Median preoptic neurons displayed varying degrees of hyperpolarization-activated time-dependent and/or time-independent inward rectification, and 86% of cells demonstrated low threshold spikes. Median preoptic nucleus is known to receive a prominent noradrenergic innervation from the medulla, and 59% of 156 tested neurons were found to respond to bath applied noradrenaline (1–100 µM). In the majority (n=62) of cells, the response was an á2 adrenoreceptor-mediated, tetrodotoxin-resistant, membrane hyperpolarization that was associated with a 43&6% increase in membrane conductance. The net noradrenaline-induced current (5–45 pA) was inwardly rectifying, cesium-resistant but barium sensitive. Current reversal at "102&4 mV in 3.1 mM [K]o and "62&3 mV in 10 mM [K]o implied opening of potassium channels. By contrast, a minority (n=27) of cells responded to noradrenaline with an á1-mediated, tetrodotoxin-resistant membrane depolarization. These observations imply a functional diversity among median preoptic neurons, and the prevalence of hyperpolarizing á2 and, to a lesser extent, depolarizing á1 adrenoreceptors on median preoptic neurons suggests that noradrenergic inputs can exert a prominent influence on their cellular excitability. ? 1998 IBRO. Published by Elsevier Science Ltd. Key words: lamina terminalis, membrane properties, á1 receptors, á2 receptors.

In the mammalian brain, the lamina terminalis along the anterior wall of the third cerebral ventricle is recognized as an important forebrain region for cardiovascular and hydromineral homeostasis.13,22 The lamina terminalis actually contains three distinct groups of neurons, each contributing to different aspects of body fluid balance. At the midpoint of the lamina terminalis adjacent to the anterior commissure is the median preoptic nucleus (MnPO), recognized through a variety of lesion studies as a key location where information of a sensory, circulating and/or osmotic nature can become integrated into an appropriate cardiovascular, behavioural and hormonal response.8,11,20 Functions attributed to MnPO arise in part through an interaction with the subfornical organ (SFO), a circumventricular organ situated along the dorsal edge of the lamina terminalis that is deemed essential for initiating central and behavioural responses to circulating angiotensin.17,19 *To whom correspondence should be addressed. Abbreviations: ACSF, artificial cerebrospinal fluid; MnPO, median preoptic nucleus; OVLT, organum vasculosum lamina terminalis; SFO, subfornical organ.

MnPO also interacts with the organum vasculosum lamina terminalis (OVLT), also a circumventricular organ but positioned at the ventral edge of the lamina terminalis and containing neurons that participate in CNS responses to circulating pyrogens33 and osmoregulation.21 Behavioural deficits consequent to various lesions have served to identify the relative contributions of each of these three regions to hydromineral homeostasis. However, few studies have focused on the cellular electrophysiology and neuropharmacology of the neurons themselves, information that is important to understand the function of this region. The present study is focused on MnPO neurons, which are perhaps the most important cells for neural integration along the lamina terminalis. Indeed, MnPO neurons have the appropriate neural connectivity to regulate neuroendocrine and autonomic function. They project to SFO and OVLT,6,28 the hypothalamic paraventricular31 and supraoptic nuclei,2,28 and several brainstem nuclei involved in cardiovascular regulation.41 MnPO’s inputs come from SFO, OVLT, medial hypothalamus5,17,23,24,29 and several brainstem regions, including lateral

905

906

D. Bai and L. P. Renaud

parabrachial nucleus,18,29 dorsomedial and ventrolateral medulla.29,30 The latter provides a rich catecholaminergic and peptidergic innervation,15,29,30 and 6-hydroxydopamine-induced depletion of catecholamines in the MnPO area results in major deficits in angiotensin-induced pressor and drinking behaviour.4 Whereas these medullary inputs arise from both adrenergic and noradrenergic neurons,29,30 several observations indicate that noradrenaline has a major role: the deficits cited above can be corrected by infusions of noradrenaline;7 injections of noradrenaline into MnPO induce an increase in urine output and sodium excretion, an elevation in arterial blood pressure, and bradycardia;9 polyethylene glycol-induced reduction in extracellular volume promotes release34 and turnover39 of noradrenaline in the MnPO area. To date, there is little detailed information on the function of noradrenergic receptors on MnPO neurons, mostly derived from studies using extracellular recordings in vivo.35 To achieve an understanding of the neural mechanisms operating within MnPO, the present investigation used patch-clamp techniques to record from MnPO neurons in parasagittal slices of rat brain. This study identified morphological features, certain intrinsic membrane properties, and evaluated their noradrenergic receptors. A portion of these results has been briefly reported.3 EXPERIMENTAL PROCEDURES

Preparation Following induction of anaesthesia with methoxyflurane, male Long–Evans rats weighing 50–200 g were decapitated, the brain quickly removed and immersed in gassed (95% O2, 5% CO2) artificial cerebrospinal fluid (ACSF; see below) at 4)C. A block containing the forebrain was mounted on a vibratome and sectioned in the sagittal plane at 400 µm, retaining the single midsagittal slice that contained the anterior part of the third ventricle and lamina terminalis. This section was transferred to a submerged recording chamber, continuously superfused at 6–7 ml/min with gassed ACSF at 23–25)C and allowed to equilibrate for at least 1.5 h prior to a recording session. Standard ACSF contained (in mM) NaCl 126, KCl 3.1, NaHCO3 26, MgCl2 1.3, NaH2PO4 1.2, CaCl2 2.4, -glucose 10 at pH 7.4 and osmolality of 300–305 mOsm. In experiments using barium and cesium, NaH2PO4 was omitted to avoid precipitation. In ACSF containing high [K+]o, an equimolar amount of NaCl was deducted. Electrophysiology Data were obtained with whole cell patch-clamp techniques using an Axopatch-1D, DMA Interface and pClamp software (Axon Instruments, Foster City, CA). Recordings were obtained through micropipettes (resistance of 4–8 MÙ) filled with a solution that contained (in mM) Kgluconate 130, KCl 10, NaCl 10, HEPES 10, Na3EGTA 0.5, MgATP 2, GTP 0.2 at pH 7.2 (adjusted with 1 M NaOH) and osmolality of 285–295 mOsm. The inclusion of Lucifer Yellow (2 mM) in the pipette solution permitted visualization of the location and morphology of recorded cells. Current and voltage signals were fed to a chart recorder and stored on video tape for offline analysis. Input resistance was determined from the slope of voltage-current plots (at potentials close to resting potential) derived from a

series of 500-ms depolarizing and hyperpolarizing current pulses applied to the recording pipette. Only cells with a series resistance under 40 MÙ and membrane potentials more negative than "50 mV were selected for analysis. Series resistance was compensated through bridge balance under current clamp. Under voltage-clamp, the series resistance was not compensated since this was small relative to cell input resistance (21.0 GÙ). The error caused by series resistance was less than 2.6 mV for a 65-mV voltage command (the maximum voltage step used in the present study) and therefore much smaller than the inevitable voltage errors in space clamp with a neuron having extensive processes. All voltage-clamp data presented in this study was obtained in the presence of tetrodotoxin (1 µM). Liquid junction potential was measured25 for each solution and corrected accordingly. Results are expressed as means&S.E.M. Drugs A syringe pump (2400-004, Harvard Apparatus) was used to apply agonists into the perfusion line for bolus application. The speed of the syringe pump was equal to, or less than, 1/10 of the perfusate flow rate. To minimize oxidation, the agonist solutions in the syringe were freshly prepared in non-gassed ACSF, and mixed with gassed ACSF for 210 s before they reached the slice preparation. The final concentration of drug was calculated from the flow rates of the perfusion media and the speed of the syringe pump.37 Considering the fact that catecholamines oxidize rapidly in oxygenated ACSF and uptake by the transporters, the actual concentration to which cells were exposed is likely to be less than the calculated concentration. Antagonists were always applied by bath-perfusion. Noradrenaline hydrochloride, phenylephrine hydrochloride, UK14,304 [5-bromo-6 (2-imidoline-2-yl-amino) quinoxalone], R(-)isoproterenol, prazosin hydrochloride and yohimbine hydrochloride were all purchased from Research Biochemical International (Natick, MA). Tetrodotoxin (TTX), 1,2-bis(2aminophenoxy)ethane-N,N,N*,N*-tetraacetic acid (BAPTA) and Lucifer Yellow were obtained from Sigma Chemical Co. (St Louis, MO). RESULTS

Properties of median preoptic nucleus neurons Observations were based on data from 156 neurons (in 122 slices, maximum three cells/slice) located in the rostral or ventral MnPO immediately adjacent to the anterior commissure (Fig. 1). When visualized with Lucifer Yellow (Fig. 1B), most cells displayed a small (9–20 µm) ovoid soma and two or three main dendrites that lacked spines and extended for 200– 400 µm in all directions. A number of cells featured a clearly identified axon that arose from the soma (32 cells) or main dendrite (16 cells) and could be followed for several hundred micrometers usually in a ventral direction; an axon collateral was noted in 23 of 62 cells before the axon left the plane of section. MnPO neurons had a mean resting potential of "59&0.4 mV and resistance of 1.16&0.03 GÙ. All cells displayed spontaneous excitatory and/or inhibitory postsynaptic potentials and 53% of cells had spontaneous action potentials at or below 3.5 Hz. Action potentials had amplitudes exceeding 70 mV (measured from threshold), a duration of 1.18&0.03 ms (measured at half amplitude) and were

Noradrenergic receptors on rat MnPO neurons

907

Fig. 1. Location of recording sites and morphological profile of MnPO neurons. (A) A photomicrograph of a Neutral Red-stained midsagittal brain section to illustrate the location of the anterior commissure (AC) along the anterior wall of third ventricle (3V) and the position of the recording site in the anterior and ventral median preoptic nucleus, within the interrupted dotted line. Scale bar=100 µm. (B) Drawings depict the location and features of three Lucifer Yellow-stained neurons in the anterior and ventral MnPO; arrowheads depict axons that arise from the soma or a main dendrite.

followed by a prominent fast afterhyperpolarization with a mean amplitude of 29.0&2 mV (measured from threshold). A majority (86%) of MnPO neurons displayed a low threshold spike when membrane potentials were returned from a hyperpolarized level (Fig. 2A,C). Only 11/51 neurons demonstrated varying degrees of spike frequency adaptation during a depolarizing current pulse, possibly a reflection of the level of EGTA in the patch pipettes. MnPO neurons displayed varying degrees of voltage-activated inward rectification that allowed for their classification into type 1 cells (47%) that featured a prominent hyperpolarization-activated and time-dependent inward rectification (Fig. 2A,B) and type 2 cells (53%) that lacked the timedependent component and in 51/83 cells expressed varying degrees of time-independent (also called instantaneous) inward rectification (Fig. 2C,D). Responses to noradrenaline In recognition of the importance of noradrenaline in functions attributed to the MnPO, this investigation included an analysis of MnPO neuronal responses to bath applications of noradrenaline. In whole cell patch mode, 59% (n=92) of MnPO neurons were responsive to bath applied noradrenaline

(1–100 µM). Of these, a majority (n=62) featured membrane hyperpolarization, fewer (n=27) demonstrated membrane depolarization, and three cells showed a mixed response. These patterns of response were not an artifact of cell dialysis with the patchclamp approach since virtually identical changes in excitability were obtained from a population of 45 neurons recorded extracellularly in cellattached mode, prior to establishment of whole cell recordings: 58% of cells responded with 19 cells depressed, and seven cells excited by noradrenaline.

Noradrenaline-induced membrane hyperpolarization An estimated noradrenaline concentration of 50 µM resulted in the most consistent responses during current clamp recordings. At this concentration, responding neurons revealed a membrane hyperpolarization of 11.2&0.8 mV that peaked within 30–40 s, recovered gradually over 6–8 minutes and was accompanied by an apparent reduction in membrane resistance (Fig. 3A). Similar results obtained from 9/9 cells tested in the presence of 1 µM TTX (Fig. 3A) confirmed the involvement of postsynaptic receptors. Membrane hyperpolarizations were repeatable if applications were spaced at

908

D. Bai and L. P. Renaud

Fig. 2. Intrinsic properties of two categories of MnPO neurons recognized according to the presence or absence of a hyperpolarization-activated, time-dependent inward rectification. (A) When tested with intracellular current injections (500 ms, +8 pA to "80 pA in 8-pA increments), type 1 neurons displayed a strong hyperpolarization-activated, time-dependent inward rectification. (B) The voltage–current (V–I) plot for the neuron in A, obtained at either peak (,) or steady state (-). (C) Voltage responses to intracellular current pulses for a type 2 cell illustrate the lack of the time-dependent component, and revealed either no rectification or, in 51/83 cells, varied degrees of hyperpolarization-activated and time-independent inward rectification, as observed in this cell. (D) The I–V plot for the cell in C. Note that both cells displayed a low-threshold spike when membrane potentials were returned from a hyperpolarized level.

intervals of 15–20 m, suggesting minimal receptor desensitization under these conditions. In voltage-clamp mode, application of noradrenaline was associated with a small (5–45 pA) outward current whose amplitude increased as drug concentrations were increased with a maximum response seen at a concentration of 250 µM (Fig. 3B,C). Noradrenaline-induced membrane hyperpolarizations were mimicked by application of an á2 adrenoreceptor agonist UK14,304 (2 µM; 5/5 cells tested; Fig. 3D), an agent that induced prolonged responses. Partial recovery of response after blockade in the presence of an á2 adrenoreceptor antagonist yohimbine (5 µM; 5/5 cells tested; Fig. 3D) was observed after wash periods that lasted for 20–30 min. None of four cells that responded to noradrenaline with hyperpolarization were responsive to the á1 adrenoreceptor agonist, phenylephrine (50 µM), or the â adrenoreceptor agonist, isoproterenol (50 µM). An á1 adrenoreceptor antagonist, prazosin, appeared to be ineffective (3/3 cells tested) at concentrations (0.5 µM) that were effective in blocking

noradrenaline-induced membrane depolarizations (see below). Additional studies under voltage-clamp revealed that the noradrenaline-induced current reflected a 43&6% (n=17 cells) increase in membrane conductance (Fig. 4A,B). During a noradrenaline-induced response there was no apparent influence on the time-dependent relaxation (Fig. 4A, inset). The net noradrenaline-induced current (Fig. 4C), obtained by subtraction of control I–Vs from those obtained at the peak of the response (Fig. 4B), featured inward rectification with a mean reversal potential of "102&4 mV (n=11 cells). The latter approximated the calculated equilibrium potential for potassium ions in normal ACSF. In four cells, the reversal potential for the noradrenaline-induced current was shifted to "62&3 mV when the external potassium concentration was raised to 10 mM (Fig. 4C), as predicted by the Nernst equation. Noradrenaline-induced membrane hyperpolarization is not likely to involve calcium-activated potassium channels since, in 10 cells recorded with

Noradrenergic receptors on rat MnPO neurons

909

Fig. 3. Features of the noradrenaline-induced hyperpolarization in MnPO neurons. (A) The upper two voltage traces from the same neuron illustrate membrane hyperpolarization induced after bath application of 50 µM noradrenaline (NA, 20 s during the horizontal bar) in control ACSF (top) and after addition of tetrodotoxin (TTX, 1 µM, 6 min; below). Downward deflections, representing voltage transients resulting from intermittent injections of current pulses (lowest trace) used to monitor membrane resistance, suggest a reduction in membrane resistance during the noradrenaline-induced hyperpolarization. Membrane potential was "58 mV. (B) Voltage-clamp data from a MnPO neuron (holding potential of "65 mV) illustrate outward currents induced after 1 min bath applications of noradrenaline at concentrations between 1 and 50 µM. (C) Dose–response curve based on data from four MnPO neurons each tested for the complete range of concentrations. Points represent mean&S.E.M. (D) Traces in current clamp illustrate (top) a control noradrenaline-induced membrane hyperpolarization, loss of response after bath application of yohimbine (5 µM, 9 min) and partial recovery after a 21 min wash. Resting membrane potential "60 mV. Bottom trace from a different MnPO neuron illustrates a similar, although slower in onset and more prolonged, membrane hyperpolarization in response to bath applied UK14,304 (2 µM, 20 s). Resting membrane potential "57 mV.

micropipettes containing 10 mM BAPTA, measurements taken 20 min after the establishment of whole cell recording revealed that noradrenaline-

induced hyperpolarizations (seen in 5/10 cells) were not perceptably different (11.5&2.3 mV) from controls.

910

D. Bai and L. P. Renaud

Fig. 4. An inward rectifying potassium current underlies the noradrenaline-induced membrane hyperpolarization. (A) Current responses (upper trace) to a series of voltage pulses (middle trace; 500 ms duration, from "130 to "40 mV in 10-mV increments) obtained from a MnPO neuron held at "65 mV (rest "72 mV) during control and at the peak of a response (outward current of 45 pA) to bath-applied noradrenaline (NA, horizontal bar, 50 µM). Traces below are superimposed samples of the current responses to the "130 mV pulses during control (8) and after noradrenaline (9), to illustrate that the time-dependent rectification present under control conditions persists during the NA-induced response. (B) Corresponding I–V relationships (measured at the end of 500 ms pulse for the data in A) reveal a slope conductance increase at potentials near to resting membrane potentials ("60 to "80 mV) during the noradrenaline-induced response. (C) Subtraction of the two I–V curves in B yields a noradrenalineinduced net current (-) that is voltage dependent, shows inward rectification and is reversed at a potential of "110 mV. Note the shift of this reversal potential to "60 mV when extracellular potassium concentration is raised to 10 mM (,).

Cesium is a known blocker of voltage-activated time-dependent and time-independent inward rectification in CNS neurons.12 Cesium also blocked the time-independent (Fig. 5A) and time-dependent (Fig. 5B, middle panel) inward rectification in MnPO neurons displaying these features (n=7). However, at the same extracellular concentration (1 mM), cesium failed to block the noradrenaline-induced inwardly rectifying current over the whole voltage range tested (Fig. 5B right panel), suggesting that this conductance is mediated by a class of cesium-insensitive inwardly rectifying potassium channels. By contrast, the noradrenaline-induced current was reduced by 62&20% in 0.1 mM barium (n=3 cells) and 84&14% in 1 mM barium (n=4 cells) in cells held at "65 mV (Fig. 6B). Addition of barium (0.1–1 mM) also removed voltage-activated timeindependent (Fig. 6A), but not time-dependent, inward rectification (Fig. 6B, middle panel).

Noradrenaline-induced membrane depolarization This investigation encountered relatively few (17% or 27/156 cells tested) MnPO neurons that responded to noradrenaline with membrane depolarization. In the concentration range of 10–50 µM, responses typically peaked within 40 s, triggering action potentials in some quiescent cells (Fig. 7A) and inducing rhythmic membrane oscillations in others (Fig. 7B). Since these responses were usually prolonged and return to baseline was delayed, holding cells for repeated tests or manipulations were not always possible. Detection of similar responses in ACSF containing TTX (8/8 cells tested) did confirm a postsynaptic site of action (Fig. 7A). Moreover, two features supported an action mediated through á1 adrenoreceptors: responses were mimicked by phenylephrine (50 µM; 4/4 cells tested, Fig. 7B); responses appeared to be blocked (irreversibly over the period of recording)

Noradrenergic receptors on rat MnPO neurons

911

Fig. 5. Cesium blocks voltage-activated, time-independent and time-dependent inward rectifiers in MnPO neurons, but fails to block their noradrenaline-induced current. (A) Left and middle panels illustrate the current responses to voltage steps (inset), under control (-) and in the presence of 1 mM cesium (,) for 6 min. The panel on the right illustrates the instantaneous I–V relationships. Note that cesium blocked the voltage-activated, time-independent inward rectification. (B) Data from another MnPO neuron. A time-dependent inward rectification in a control current response (left panel) was blocked in ACSF containing 1 mM cesium (centre panel) although cesium did not block the noradrenaline-induced current (centre panel). The panel on the right illustrates the virtual overlap of the I–Vs of the noradrenalineinduced net current (see Fig. 4) in control conditions (-) and during application of cesium (,).

following addition of prazosin (0.5 µM, 7/7 cells tested, Fig. 7A). In two cells, the polarity of the noradrenaline-induced response was changed from depolarization to hyperpolarization after the application of prazosin, indicating that at least some MnPO neurons are likely to have both á1 and á2 adrenoreceptors; in both instances the hyperpolarizing response was blockable with yohimbine. In 3/3 cells, isoproterenol (50 µM) failed to mimic the noradrenaline-induced responses. For reasons cited above, less information was obtained on mechanisms of this pattern of response. Recordings in voltage-clamp mode revealed a noradrenaline-induced inward current that seldom exceeded 20 pA (Fig. 7C). In 6/10 neurons, this was accompanied by a 17&3% reduction in membrane conductance near resting membrane potential levels; in the remaining cells there was either no change (n=3), or an increase (13%, n=1) in membrane conductance. Mean net current calculated from the I–V relationships of five cells revealed a decrease in magnitude at more hyperpolarized levels (Fig. 7D), consistent with a reduction of a membrane conductance, although no reversal of the net current was observed in the tested voltage range of "50 to "130 mV.

Cell properties versus noradrenaline response pattern Whereas the position of 104 neurons tested for a response to noradrenaline revealed no apparent preferential distribution within MnPO (Fig. 8A), other comparisons suggested a correlation between neuronal properties and their responses to noradrenaline (Table 1). First, MnPO neurons responding with membrane hyperpolarization had a significantly less negative resting membrane potential and higher level of spontaneous activity when compared with cells demonstrating membrane depolarization, or lacking any response to this drug. In addition, type 1 MnPO neurons were more likely to reveal membrane hyperpolarization, whereas type 2 cells were more likely to be non-responsive to noradrenaline (Fig. 8B). DISCUSSION

Median preoptic nucleus contains a heterogenous population of neurons From a morphological viewpoint, the profiles of MnPO neurons revealed after filling with Lucifer Yellow suggests a relative homogeneity and similarity between cells. However, other properties imply considerable diversity among these cells. For example,

912

D. Bai and L. P. Renaud

Fig. 6. Barium blocks voltage-activated, time-independent inward rectification and the noradrenalineinduced current. (A) Similar to Fig. 6A, the left and middle panels are current responses under control (-) and in the presence of 1 mM barium (,) for 7 min. The panel at the right shows the instantaneous current-voltage relationships. Note that barium blocks the voltage-activated, time-independent inward rectification. (B) Data from another cell. Similar to A, barium is seen to block an inward current in this voltage pulse (left and centre panels). However, distinct from the action of cesium, barium failed to block the time-dependent inward rectification (centre panel, also shown in A). Furthermore, different from cesium, 1 mM barium for 7 min blocked the noradrenaline-induced current (centre panel). The panel on the right illustrates the noradrenaline-induced net current (see Fig. 4) for control (-) and in the presence of barium (,). Barium potently blocked noradrenaline-induced current over the whole tested voltage range.

heterogenous phenotypes can be recognized with immunocytochemistry.14 Moreover, data from extracellular recordings in vivo reveal diverse patterns of response between MnPO neurons in the same preparation to different cardiovascular perturbations,1,16,36 implying that afferent connectivity varies among individual MnPO cells. Similarly, there is diversity in the response patterns of different MnPO neurons tested with osmotic, thermal and/or peptidergic stimuli in vitro,38 indicating diversity among MnPO cells in terms of their intrinsic properties and neuropharmacology. The present investigation supports a functional heterogeneity among MnPO neurons, both in terms of their intrinsic membrane properties (inwardly rectifying properties) and their adrenoreceptors. The recognition of this diversity among MnPO neurons is an initial step towards a more detailed analysis of this area that is required at the cellular level. The relatively simple geometry of MnPO neurons deserves comment. Lucifer Yellow labelling, which is capable of revealing such details as dendritic spines, indicates that MnPO neurons do not possess an

extensive dendritic arborization. Assuming that most synaptic input takes place on dendrites, the question arises as to how effectively is this information transferred to the soma and integrated into a response, e.g., action potential initiation. Noting that the input resistance of MnPO neurons is high, one might predict that such cells are electrically compact, and that conductances associated with individual synaptic events, even if they are located remotely, can effect relatively large changes in membrane potential and thus alter cell excitability. As for the location of the noradrenergic input to these cells, one can surmise from an earlier anatomical study that a majority of the catecholaminergic synapses appear to be positioned on dendrites rather than on somata of MnPO neurons that project their axons to the hypothalamic paraventricular nucleus.15 The á2 adrenoreceptor-mediated current has distinct features The á2 adrenoreceptor-mediated membrane current in MnPO neurons features a time-independent

Noradrenergic receptors on rat MnPO neurons

913

Fig. 7. Characterization of noradrenaline-induced depolarization in MnPO neurons. (A) Traces under current clamp from the same MnPO neuron illustrate (upper) a 14 mV membrane depolarization in response to noradrenaline (NA, 50 µM, 20 s), sufficient to evoke spike discharges, persistence of the response in tetrodotoxin (TTX, 1 µM; middle trace) and blockade (irreversible) following addition of an á1 receptor antagonist, prazosin (0.5 µM for at least 8 min, bottom trace). Membrane potential: "61 mV. (B) Trace from another MnPO neuron illustrates a similar depolarizing action induced by an á1 receptor agonist, phenylephrine (50 µM, 20 s) in the presence of tetrodotoxin; oscillations, observed at the peak of this phenylephrine-induced depolarization, were also associated with its response to noradrenaline (not illustrated). Membrane potential: "66 mV. (C) Trace obtained under voltage-clamp (holding potential "65 mV; resting membrane potential "63 mV) illustrate a maximum inward current of 15 pA in response to bath-applied noradrenaline (NA). (D) Noradrenaline-induced net current (see Fig. 5) obtained for five MnPO neurons reveals a decrease in magnitude at negative membrane potentials, but failure to reverse within the tested voltage range ("50 to "130 mV).

inward rectification that is not blocked by cesium, although cesium does block the voltage-activated time-dependent and time-independent rectification in these neurons. Since barium blocked the noradrenaline-induced current, but had little effect on membrane conductance at resting membrane potentials (Fig. 6), these noradrenaline-activated channels are likely to be normally closed. The observation that a similar profile of responses to noradrenaline was obtained with cell-attached mode versus whole cell recording mode suggests that intracellular dialysis from the patch pipette did not alter the ability of the cells to respond to the ligand. While this does not necessarily rule out a role for secondary messengers in this response, it might be an indication that there is a close coupling between the G-protein(s) associated with the á2 receptor and the (potassium) effector channels. Similar inwardly rectifying currents follow activation of a variety of G-protein-coupled receptors on other CNS neurons, notably opioid and á2 adrenergic receptors in locus coeruleus neurons,27 5-HT1 receptors in dorsal raphe neurons40 and GABAB receptors in hippocampal CA3 neurons.32

Functional considerations: noradrenergic receptors The MnPO receives a prominent catecholamine (noradrenergic and adrenergic) and peptide (NPY) innervation15 from the ventrolateral and the dorsomedial medulla.29,30 Electrical stimulation in the ventrolateral medulla in vivo can alter the excitability of MnPO neurons as noted by Tanaka et al.35 who reported an increase the excitability of 14/21 MnPO neurons that had projections to the area of the hypothalamic paraventricular nucleus. In their study, the MnPO responses were blocked after microiontophoretic application of phentolamine, but not timolol, implying mediation through á adrenoreceptors. The present study clearly confirms the presence of depolarizing á1 adrenoreceptors on MnPO neurons, although these involve relatively fewer neurons than those demonstrating hyperpolarizing á2 adrenoreceptors. It is surprising that Tanaka et al.35 did not observe any depressant responses; it is possible that the input to MnPO from this area only engages cells with á1 receptors. Alternatively, MnPO neurons that are subject to a ‘‘depressant’’ and prevailing á2 adrenoreceptor-mediated input are likely to be silent, if this is a tonically active innervation, and such

914

D. Bai and L. P. Renaud

Fig. 8. (A) Location of MnPO neurons vs response to noradrenaline indicate no obvious clustering of neurons responding with hyperpolarization (,), depolarization (-), biphasic responses (0), or not responding (4). AC, anterior commissure. 3V, third ventricle. (B) Correlation between cell type and response to noradrenaline. Membrane hyperpolarization by noradrenaline is more frequently associated with a type 1 neuron (as defined in Fig. 2), while lack of response to noradrenaline is more often associated with a type 2 cell (see Fig. 2). Cells depolarized by noradrenaline are evenly distributed among the two cell types. Table 1. Membrane properties of median preoptic neurons that are depolarized, hyperpolarized or have no response to noradrenaline Response to noradrenaline Input resistance (MÙ) Resting potential (mV) Spontaneously active Action potential amplitude1 (mV) Action potential duration1 (ms) Afterhyperpolarization1 (mV)

Depolarization

n

Hyperpolarization

n

No response

n

1093&62 62&1.3 33% 92&2 1.2&0.06 29&2

21 21 21 11 11 11

1205&59 58&0.6* 67% 86&2 1.2&0.04 28&2

43 43 43 23 23 23

1149&61 62&1.0 42% 90&2 1.2&0.04 27&1

45 45 45 25 25 25

Data are presented as means&S.E.M. *P<0.01, 1Action potential amplitudes and afterhyperpolarizations (fast) were measured from threshold. The action potential duration was measured at 50% of the peak.

inputs are likely to remain undetected during any extracellular exploration. It is also possible that MnPO neurons with a prevalence of á2 adrenoreceptors may be more depolarized in a slice preparation because of removal of their noradrenergic input (see Table 1). In another recent in vivo extracellular investigation, Aradachi et al.1 reported that electrical stimulation in the dorsomedial medulla (nucleus tractus solitarius) increased neuronal excitability in 34% of MnPO cells, while 9% of cells showed a decrease in firing and 47% of cells were unresponsive. Interestingly, the latter authors also reported few depressant responses to electrical stimulation, but also noted that the direc-

tion of a specific cell’s response (i.e. increase or decrease in excitability) to electrical stimulation was often opposite to its response to hemodynamic or osmotic stimulation. The specific information conveyed through catecholaminergic inputs to MnPO has yet to be identified. However, ascending catecholaminergic fibres are known to convey visceral sensory information, including that from baroreceptors, to forebrain and hypothalamic neurosecretory neurons that participate in neuroendocrine and hydromineral regulation.10 The sensitivity of MnPO neurons to changes in arterial blood pressure is undisputed1,16,36 and indicates that they are also receiving information about

Noradrenergic receptors on rat MnPO neurons

peripheral baroreceptors. Since increases in arterial pressure are usually associated with a reduction in firing frequency of MnPO neurons,1,16,36 it is tempting to speculate that these ‘‘depressant’’ responses are mediated through á2 adrenoreceptors. Further studies should clarify this issue.

Functional considerations: neurosecretion An intact catecholamine innervation of the MnPO region is essential for both angiotensin-induced pressor and drinking responses, and for vasopressin release in the neurohypophysis.4,8 A recent study using focal electrical and chemical stimulation in the ventral MnPO indicates that MnPO’s output to the vasopressin- and oxytocin-secreting neurons in the supraoptic nucleus is predominantly inhibitory, and mediated through a direct GABAergic projection.26 If MnPO neurons do indeed contain a predominance of postsynaptic á2 (rather than á1) noradrenergic receptors, as indicated by the present data, any catecholaminergic inputs are likely to suppress, rather than enhance, their excitability. A hypothesis to be tested in future studies is that the noradrenergic innervation of MnPO (acting through á2 receptors) actually facilitates vasopressin release by ‘‘disinhibition’’, that is by reducing the GABAergic output of MnPO to the vasopressin-secreting neurons.

915

CONCLUSIONS

While our observations with Lucifer Yellow labelling indicate that there is a little morphological diversity among MnPO neurons, they do reveal a functional complexity and diversity among these neurons in terms of intrinsic membrane properties and the prevalence of hyperpolarizing á2 and, to a lesser extent, depolarizing á1 adrenoreceptors. The activation of their á2 adrenoreceptors can open potassium channels that are normally closed at rest, resulting in inwardly rectifying currents that resemble those observed in other CNS neurons following activation of G-protein coupled receptors. Evidently, noradrenergic inputs exert a prominent influence on the cellular excitability of MnPO neurons. It remains to be verified whether such inputs actually convey information to MnPO neurons about cardiovascular perturbations in vivo, and how these catecholaminegated conductances interact with both intrinsic conductances and those from other transmitters in cellular integration within this functionally important area of the lamina terminalis.

Acknowledgements—This work was supported by the Medical Research Council, and by the Heart and Stroke Foundation of Canada. DB is a Fellow of the Heart and Stroke Foundation, and LPR is a MRC Scientist. We thank Miloslav Kolaj for comments on an earlier version of the manuscript.

REFERENCES

1.

2. 3. 4. 5. 6. 7.

8.

9.

10. 11. 12. 13. 14.

Aradachi H., Honda K., Negoro H. and Kubota T. (1996) Median preoptic neurones projecting to the supraoptic nucleus are sensitive to haemodynamic changes as well as to rise in plasma osmolality in rats. J. Neuroendocr. 8, 35–43. Armstrong W. E., Tian M. and Wong H. (1996) Electron microscopic analysis of synaptic inputs from the median preoptic nucleus and regions to the supraoptic nucleus in the rat. J. comp. Neurol. 373, 228–239. Bai D. and Renaud L. P. (1995) Median preoptic nucleus (MnPO) neurons possess functional á1 and á2 adrenergic receptors. Soc. Neurosci. Abstr. 21, 1612. Bellin S. I., Landas S. K. and Johnson A. K. (1988) Selective catecholamine depletion of structures along the ventral lamina terminalis: effects on experimentally-induced drinking and pressor responses. Brain Res. 456, 9–16. Camacho A. and Phillips M. I. (1981) Horseradish peroxidase study in rat of the neural connections of the organum vasculosum of the lamina terminalis. Neurosci. Lett. 25, 201–204. Chiba T. and Murata Y. (1985) Afferent and efferent connections of the medial preoptic area in the rat: a WGA–HRP study. Brain Res. Bull. 14, 261–272. Cunningham J. T. and Johnson A. K. (1991) The effects of central norepinephrine infusions on drinking behavior induced by angiotensin after 6-hydroxydopamine injections into the anteroventral region of the third ventricle (AV3V). Brain Res. 558, 112–116. Cunningham J. T., Sullivan M. J., Edwards G. L., Farinpour R., Beltz T. G. and Johnson A. K. (1991) Dissociation of experimentally induced drinking behavior by ibotenate injection into the median preoptic nucleus. Brain Res. 554, 153–158. da Silva R. K., Saad W. A., Renzi A., Menani J. V. and Camargo L. A. (1995) Effect of lateral hypothalamus lesions on the water and salt intake, and sodium and urine excretion induced by activation of the median preoptic nucleus in conscious rats. J. auton. nerv. Syst. 53, 195–204. Dampney R. A. L. (1994) Functional organization of central pathways regulating the cardiovascular system. Physiol. Rev. 74, 323–364. Gardiner T. W., Verbalis J. G. and Stricker E. M. (1985) Impaired secretion of vasopressin and oxytocin in rats after lesions of nucleus medianus. Am. J. Physiol. 249, R681–8. Halliwell J. V. (1990) K+ channels in the central nervous system. In Potassium Channels: Structure, Classification, Function and Therapeutic Potential (ed. Cook N. S.), pp. 348–381. Ellis Horwood, Chichester. Johnson A. K., Cunningham J. T. and Thunhorst R. L. (1996) Integrative role of the lamina terminalis in the regulation of cardiovascular and body fluid homeostasis. Clin. exp. Pharmac. Physiol. 23, 183–191. Kawano H., Chiba T. and Masuko S. (1989) An immunohistochemical observation of polypeptides and monoamines in the nucleus preopticus medianus of the rat. Brain Res. 492, 139–148.

916 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

D. Bai and L. P. Renaud Kawano H. and Masuko S. (1993) Synaptic inputs of neuropeptide Y-immunoreactive noradrenergic nerve terminals to neurons in the nucleus preopticus medianus which project to the paraventricular nucleus of the hypothalamus of the rat: a combined immunohistochemical and retrograde tracing method. Brain Res. 600, 74–80. Knuepfer M. M., Gebhart G. F. and Brody M. J. (1985) Effects of baroreceptor activation on single unit activity of the anteroventral third ventricle region of the rat. Neurosci. Lett. 56, 79–85. Lind R. W. and Johnson A. K. (1982) Subfornical organ-median preoptic connections and drinking and pressor responses to angiotensin II. J. Neurosci. 2, 1043–1051. Lind R. W. and Swanson L. W. (1984) Evidence for corticotropin releasing factor and Leu-enkephalin in the neural projection from the lateral parabrachial nucleus to the median preoptic nucleus: a retrograde transport, immunohistochemical double labeling study in the rat. Brain Res. 321, 217–224. Mangiapane M. L. and Simpson J. B. (1980) Subfornical organ: forebrain site of pressor and dipsogenic action of angiotensin II. Am. J. Physiol. 239, R382–9. Mangiapane M. L., Thrasher T. N., Keil L. C., Simpson J. B. and Ganong W. F. (1983) Deficits in drinking and vasopressin secretion after lesions of the nucleus medianus. Neuroendocrinology 37, 73–77. McKinley M. J., Congiu M., Miselis R. R., Oldfield B. J. and Pennington G. L. (1988) The lamina terminalis and osmotically-stimulated vasopressin secretion. In Recent Progress in Posterior Pituitary Hormones (eds Yoshida S. and Share L.), pp. 117–124. Excerpta Medica, Amsterdam. McKinley M. J., Pennington G. L. and Oldfield B. J. (1996) Anteroventral wall of the third ventricle and dorsal lamina terminalis: headquarters for control of body fluid homeostasis. Clin. exp. Pharmac. Physiol. 23, 271–281. Miselis R. R. (1981) The efferent projections of the subfornical organ of the rat: a circumventricular organ within a neural network subserving water balance. Brain Res. 230, 1–23. Miselis R. R., Shapiro R. E. and Hand P. J. (1979) Subfornical organ efferents to neural systems for control of body water. Science 205, 1022–1025. Neher E. (1992) Correction for liquid junction potentials in patch clamp experiments. Meth. Enzymol. 207, 123–222. Nissen R. and Renaud L. P. (1994) GABA receptor mediation of median preoptic nucleus-evoked inhibition of supraoptic neurosecretory neurones in rat. J. Physiol., Lond. 479, 207–216. North R. A., Williams J. T., Surprenant A. and Christie M. J. (1987) Mu and delta receptors belong to a family of receptors that are coupled to potassium channels. Proc. natn. Acad. Sci. U.S.A. 84, 5487–5491. Oldfield B. J., Miselis R. R. and McKinley M. J. (1991) Median preoptic nucleus projections to vasopressin-containing neurones of the supraoptic nucleus in sheep. A light and electron microscopic study. Brain Res. 542, 193–200. Saper C. B. and Levisohn D. (1983) Afferent connections of the median preoptic nucleus in the rat: anatomical evidence for a cardiovascular integrative mechanism in the anteroventral third ventricular (AV3V) region. Brain Res. 288, 21–31. Saper C. B., Reis D. J. and Joh T. (1983) Medullary catecholamine inputs to the anteroventral third ventricular cardiovascular regulatory region in the rat. Neurosci. Lett. 42, 285–291. Silverman A. J., Hoffman D. L. and Zimmerman E. A. (1981) The descending afferent connections of the paraventricular nucleus of the hypothalamus (PVN). Brain Res. Bull. 6, 47–61. Sodickson D. L. and Bean B. P. (1996) GABAB receptor-activated inwardly rectifying potassium current in dissociated hippocampal CA3 neurons. J. Neurosci. 16, 6374–6385. Stitt J. T. (1985) Evidence for the involvement of the organum vasculosum laminae terminalis in the febrile response of rabbits and rats. J. Physiol., Lond. 368, 501–511. Tanaka J., Hori K., Koito H. and Nomura M. (1992) Enhanced monoamine release in the median preoptic area following reduced extracellular fluid volume in rats. Neurosci. Lett. 147, 110–113. Tanaka J., Nishimura J., Kimura F. and Nomura M. (1992) Noradrenergic excitatory inputs to median preoptic neurones in rats. NeuroReport. 3, 946–948. Tanaka J., Nomura M., Kariya K., Nishimura J. and Kimura F. (1993) Median preoptic neurons projecting to the hypothalamic paraventricular nucleus are sensitive to blood pressure changes. Brain Res. 605, 338–341. Tornøe C., Bai D., Holden-Dye L., Abramson S. N. and Sattelle D. B. (1995) Actions of neurotoxins (bungarotoxins, neosurugatoxin and lophotoxins) on insect and nematode nicotinic acetylcholine receptors. Toxicon 33, 411–424. Travis K. A. and Johnson A. K. (1993) In vitro sensitivity of median preoptic neurons to angiotensin II, osmotic pressure, and temperature. Am. J. Physiol. 264, R1200–5. Wilkin L. D., Patel K. P., Schmid P. G. and Johnson A. K. (1987) Increased norepinephrine turnover in the median preoptic nucleus following reduced extracellular fluid volume. Brain Res. 423, 369–372. Williams J. T., Colmers W. F. and Pan Z. Z. (1988) Voltage- and ligand-activated inwardly rectifying currents in dorsal raphe neurons in vitro. J. Neurosci. 8, 3499–3506. Zardetto-Smith A. M. and Johnson A. K. (1995) Chemical topography of efferent projections from the median preoptic nucleus to pontine monoaminergic cell groups in the rat. Neurosci. Lett. 199, 215–219. (Accepted 21 August 1997)