Intracellular recording of magnocellular preoptic neuron responses to olfactory brain

Pergamon

PII:

Neuroscience Vol. 78, No. 1, pp. 229–242, 1997 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/97 $17.00+0.00 S0306-4522(96)00566-0

INTRACELLULAR RECORDING OF MAGNOCELLULAR PREOPTIC NEURON RESPONSES TO OLFACTORY BRAIN A. G. PAOLINI*† and J. S. McKENZIE‡ *Department of Otolaryngology, University of Melbourne, 32 Gisborne Street, East Melbourne, 3002, Australia ‡Department of Physiology, University of Melbourne, Parkville, 3052, Australia Abstract––The magnocellular preoptic nucleus of the rat supplies centrifugal input to the olfactory bulb as well as projecting to other olfactory-related areas. The extent to which the piriform and entorhinal cortices can influence the activity of magnocellular preoptic neurons and hence that of the olfactory bulb were examined using intracellular in vivo recording. Stable recordings were obtained in 58 neurons impaled in the magnocellular preoptic nucleus. Antidromic responses occurred on stimulating olfactory bulb (15), piriform cortex (14), or entorhinal area (eight). Monosynaptic excitation was evoked by piriform (27 of 37 tested) and entorhinal cortex (15 of 32 tested) stimulation with polysynaptic inhibition occurring in seven and five neurons, respectively. Polysynaptic as well as antidromic excitation by olfactory bulb stimulation occurred in four; a further 28 tested responded polysynaptically. No response to olfactory bulb stimulation was monosynaptic. In stable impalements, 29 neurons discharged spontaneously in the absence of applied current. Lucifer Yellow and Neurobiotin were used to label 16 cells. All but one had smooth dendrites with soma diameters ranging from 8 to 24 µm. These results provide a framework in which magnocellular preoptic neurons can influence olfactory processing by direct action on the olfactory bulb, which action can be boosted by positive feedback from the bulb through the olfactory piriform and entorhinal cortices. ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: olfactory bulb, magnocellular preoptic nucleus, piriform cortex, entorhinal area, in vivo intracellular recording.

The basal magnocellular complex5 comprises a set of neurons extending from the medial septum to the caudal end of the corpus striatum. Central to this area is the human basal nucleus of Meynert. An assembly of magnocellular neurons in the same topographic area in many animals has been labelled as magnocellular preoptic nucleus (MCPO) by Loo,18 and nucleus of the diagonal band by Johnston14 or of its horizontal limb (HDB) by Price,25 Zaborsky et al.35 and others.1,17,19,28 The rat MCPO lies among fibres of the medial forebrain bundle35 and sends projections to the ipsilateral olfactory bulb (OB), the medial prefrontal cortex, entorhinal cortex, occipital lobe, limbic cortical structures, primary olfactory cortex, contralateral HDB, the caudal midbrain, pons and medulla.3,4,19,28 For afferent input the MCPO receives predominantly cholinergic projections from brainstem regions such as the pendunculopontine and laterodorsal tegmental nuclei, as well as serotonergic projections from the dorsal raphe nucleus.29 †To whom correspondence should be addressed. Abbreviations: EA, entorhinal area; EPSP, excitatory postsynaptic potential; HDB, horizontal limb of the diagonal band; IPSP, inhibitory postsynaptic potential; LY, Lucifer Yellow; MCPO, magnocellular preoptic nucleus; OB, olfactory bulb; PC, piriform cortex.

MCPO has a close association with the more medially located HDB, and the substantia innominata. Brauer et al.,2 using Golgi-stained material, have shown that MCPO is anatomically distinct from these areas, with three types of neurons. The largest cell type of the whole basal forebrain complex of the rat is found here, and generally lacks spines; in addition they describe two types of small cells, one with and the other without spines.2 Although these neurons differ from those of the HDB, the larger type is also found in the substantia innominata. As a bed nucleus of the medial forebrain bundle it has also been charted in a detailed atlas by Geeraedts et al.,7 using criteria of cell size and density to demarcate it and other nuclei in the basal region. In atlases of the rat brain, MCPO appears as a lateral and caudal extension of the horizontal diagonal band nucleus, but the demarcation is various and rather arbitrary.7,24,33 Most physiological experiments concerned with neuronal functioning of the basal forebrain have centred on the cholinergic neurons of the medial septum and the nucleus of the diagonal band,8–10,32 and the more posterior region associated with the globus pallidus.27 MCPO has been shown to contain an abundance of neurons which react positively for glutamate decarboxylase,1 and Zaborsky et al.35

229

230

A. G. Paolini and J. S. McKenzie

found that approximately 30% of GABAergic neurons in the MCPO/HDB complex send centrifugal input to the OB. Little is known, however, about the functioning of these mainly GABAergic cells located in MCPO and the part they play in influencing cells in other regions throughout the brain. Recently, electrical stimulation of this nucleus has been shown to inhibit OB granule cells,16 and appears thereby to facilitate mitral cell discharge.17 Behaviourally, lesions in the MCPO result in a decrease in olfactory investigation.23 Severe cell loss in the basal nucleus of Meynert is common in Alzheimer’s and other dementias34 with early cell loss in the rhinecephalon, together with decline of olfactory sense as a precocious symptom.6 By its projections to several olfactory related structures such as the OB, piriform and entorhinal cortices, MCPO may play an important part in olfactory functioning at several levels. In this investigation, MCPO neurons, particularly those which send centrifugal input to the OB, were recorded in vivo with intracellular micropipettes, to examine their intrinsic functional properties, and how they are influenced by excitation exerted from the piriform cortex (PC) and the entorhinal area (EA), which are established olfactory cortical structures.12 In addition, we investigated whether input from the OB is able indirectly to influence the firing of MCPO neurons. In a smaller sample of cells the morphology of impaled cell types was examined using intracellular dye filling. EXPERIMENTAL PROCEDURES

Preparation All experiments were performed on house-bred male hooded rats anaesthetized with intraperitoneal urethane in water (1.3 g/kg) and breathing spontaneously. All efforts were made to minimize animal suffering and to reduce the number of animals used, in accordance with University of Melbourne animal ethics guidelines. Surgery After a midline incision, the underlying muscle was cleared from the bone and the cranium and dura were removed over stimulating and recording positions. Any excessive bleeding was stopped with application of hydrogen peroxide (3%). Electrodes were inserted under stereotaxic control using coordinates according to the atlas of Paxinos and Watson.24 Intracellular recording electrodes for the magnocellular preoptic area were aimed at stereotaxic coordinates 8.7 mm anterior to the inter-aural plane and 2.5 mm lateral to the midline. Stimulating electrodes for EA, PC and substantia nigra were inserted at anterior 2.7, 5.7 and 3.2, lateral 5.0, 5.0 and 2.0, and 1.7, 0.0 and 1.8 mm above the interaural plane, respectively. Body temperature was maintained at approximately 37)C with a d.c. homoeothermic blanket and the heart beat monitored. Any brain swelling following surgery was reduced by 0.3 ml of dexomethasone phosphate (4 mg/ml Decadron, MSD) injected into the tail vein. Recording Microelectrodes, made from thick-walled (1.5 mm outside diameter) borosilicate glass (TW150F-15, WPI), were

filled with 1 M potassium acetate (70–120MÙ), 4% Lucifer Yellow in water (LY; 130–280 MÙ) or in some cases 4% Neurobiotin in 1 M potassium acetate (110–160 MÙ). Recordings made with LY- or Neurobiotin-containing electrodes are so indicated in figure legends. They were stepped from starting positions 7 mm below the dorsal brain surface and driven in 2-ìm steps using a Significat SCAT 01 microelectrode stepper (Digitimer Ltd), through MCPO. During recording, the position of the microelectrode tip was monitored using the field potential evoked by stimulation of the OB (Fig. 1). Upon stabilized cell impalement, stimuli to the OB, PC and EA were used to classify the cell response as antidromic, postsynaptic discharge, excitatory or inhibitory postsynaptic potentials (EPSP or IPSP) or absent. Stimuli (1 mA, delivered at 0.67 Hz) were delivered from isolated stimulators through coaxial bipolar stainless steel electrodes with the core negative. Antidromicity was determined by constant latency of action potential response with no discernible prepotential, ability to follow high-frequency stimulus bursts (up to 200 Hz), and collision with spontaneous or evoked synaptic discharge where available. Synaptic potentials were taken as monosynaptic if latency was constant over a range of stimulus strengths15 (approx. 0.5–2.0 mA). Cells were filled with LY by passing approximately 1.3 nA of hyperpolarizing current through the micropipette, for at least 3 min and optimally up to 15 min. In filling cells with Neurobiotin, 500 ms 1.1 nA positive current pulses (0.67 Hz) were used. When a number of cells (up to four) were considered to be filled, and could be easily distinguished by their electrophysiological responses based on cell depth and field potential profile, the experiment was terminated. Lesions were made at the tips of stimulating electrodes (10 µA, 60 s) in order to locate stimulation positions. Histology After anaesthetic overdose the rats were perfused transcardially with 10% formalin containing 30% sucrose. They were decapitated and the heads dissected free and preserved in sucrose–formalin solution. After four to five days the heads were returned to the stereotaxic apparatus, the cranial vaults removed and the brains blocked with a stereotaxically oriented blade. The brains, including the OBs, were then removed and sectioned on a freezing microtome. Sections of 120 µm through the MCPO region were coverslipped with phosphate-buffered glycerol and examined for LY-filled cells using a fluorescence microscope set up for fluorescein isothiocyanate. Sections containing stimulating electrode positions were cut at 60 µm and stained with Cresyl Violet. After electrophysiological experiments sections containing cells possibly filled with Neurobiotin13 were washed for 1 h in six changes of 0.1 M phosphate buffer. The excess buffered solution was then removed and the sections were bathed in a 1:50 dilution of streptavidin–Texas RedTM (Amersham, Australia) for approximately 2 h. The sections were then washed in three changes of phosphate buffer and coverslipped with 0.1 M phosphate-buffered glycerol. Sections were viewed under a fluorescence microscope (excitation 589 µm; emission, 615 µm). Imaging and cell location Cell body locations were also determined using a digitized video image of the sections at low-power magnification. Using fluorescence and light microscopy alternately, cell positions could be determined. Locations of cells not filled with dye-marker were determined by stereotaxic recording position, as belonging to MCPO if they were recorded within the boundaries defined by the atlas of Paxinos and Watson;24 this related consistently to changes in the field potential profile evoked by OB stimulation. During experimental recording, distance from the dorsal brain surface to basal pial surface was determined

Magnocellular preoptic neuron responses to olfactory brain

231

Fig. 1. Field potential profile evoked by olfactory bulb stimulation at selected depths through the ventral pallidum (VP), magnocellular preoptic nucleus (MCPO) and olfactory tubercle (OT). CP, caudate putamen. Upward deflection positive. by advancing the microelectrode until it touched the pial surface (as shown by a resistance drop). LY- and Neurobiotin-filled cell locations could thus be provisionally determined during the experiment. These predicted locations corresponded well with their actual locations after histological verification. RESULTS

Intracellular recordings Of 90 cells impaled in the MCPO, stable recordings were obtained from 58. Stable neurons were held for at least 5 min and had a mean (&S.E.) resting

membrane potential of "49.1&2.0 mV with a mean (&S.E.) action potential amplitude (with corresponding range) of 42.4&2.5 mV (20–80 mV). In a smaller sample of these cells the input resistance and time constant were also estimated (see below). Stimulating positions The stimulation positions in PC and EA are shown in Fig. 2A and B. In order to maximize stimulation of output cell axons, OB stimulation points were aimed towards the centre of the caudal bulb. Owing to space

232

A. G. Paolini and J. S. McKenzie

responded with polysynaptic EPSPs and one with an IPSP (Fig. 7A–F), of which two were spontaneously active. The antidromic spike occurred later than the polysynaptic response in one of these four cells which was not spontaneously active (Fig. 3). Three IPSPs (none monosynaptic) were seen in response to OB; one such cell was spontaneously active. The remaining nine gave no response to OB stimulation; two were spontaneously active. Responses to piriform cortex stimulation. EPSPs in response to PC stimulation were seen in 42 neurons; the majority of these responses were monosynaptic (27 out of 37 tested; e.g., Figs 3I, 4F, 5A,E, 7G–I) with 20 of the cells being spontaneously active. Of the 14 antidromically activated from PC, eight cells, six spontaneously active, also responded with monosynaptic EPSPs to PC stimulation. Three cells, all spontaneously active, failed to respond to PC. IPSP responses to PC stimulation, all polysynaptic, occurred in seven neurons (Fig. 7G–I), of which one was spontaneously active. Monosynaptic EPSP responses to PC stimulation showed a spread of latencies with a bimodal distribution, 11 with latencies of 3–5 ms and seven of 8–10 ms. The majority (60%) of EPSP responses to PC stimulation had durations between 20 and 50 ms (e.g., Figs 3I, 4D,E).

Fig. 2. Drawings of coronal sections through rat basal forebrain showing verified stimulation points in (A) caudal PC and (B) medial EA. M, medial; L, lateral. Scale bar=1 mm.

constraints, PC stimulation was directed to the most caudal parts and entorhinal stimulation to a medial region. Responses to stimulation Of the 58 stable cells recorded in MCPO, 29 were spontaneously active. All cells were tested for responses to stimulation of the PC, 56 tested with OB and 54 with EA stimulation. Output neurons were identified, 14 responding antidromically to PC, 15 to OB and eight to EA stimulation. Antidromic activation was not restricted to one stimulation site (see below). Spontaneous activity was seen in eight, two and four of those cells, respectively. The results are summarized in Table 1. Examples of intracellular MCPO responses are shown in Figs 3–7. Responses to olfactory bulb stimulation. OB stimulation evoked EPSPs in 33 cells (e.g., Figs 3A,E, 4B,C, 5B,G). None of 28 tested was shown to be monosynaptically driven. Of these synaptically activated cells 17 were spontaneously active. Of the 15 that were antidromically activated, four cells also

Responses to entorhinal area stimulation. EPSPs in response to EA stimulation were seen in 32 neurons, 20 of which were spontaneously active; 15 out of 32 were monosynaptic (e.g., Fig. 6D). Of the eight antidromically activated cells, four, of which three were spontaneously active, also responded with monosynaptic EPSPs to EA stimulation. Eleven cells, nine spontaneously active, gave no response to EA. Polysynaptic IPSP responses to EA stimulation (Fig. 4K) occurred in five neurons, of which one was spontaneously active. Monosynaptic EPSP responses to EA stimulation had a bimodal distribution of latencies, six between 5 and 7 ms and four between 13 and 15 ms (e.g., Figs 3J, 4K, 6D). Durations of EPSP also varied bimodally, six between 20 and 40 ms and four between 60 and 80 ms. Patterns of responsiveness. Of the 14 cells responding antidromically to PC, seven also responded antidromically to OB. Antidromic responses were also evoked by EA in four cells of 10 tested. Two of these cells were identified by dye filling as mediumsized (14–20 µm diameter) aspiny neurons (Fig. 8A,B). Only one neuron responded antidromically (together with EPSPs) to all three stimulation areas. In the remaining eight neurons that responded antidromically to OB, six EPSP and two IPSP responses were evoked by PC, EA evoking one antidromic, three EPSPs, and no response in three, with one cell not tested. Of the four cells with an EPSP in addition to an antidromic response to OB, all responded with monosynaptic EPSPs to PC, three with monosynaptic EPSPs to EA stimulation (Fig. 3).

Magnocellular preoptic neuron responses to olfactory brain

233

Table 1. Responses of magnocellular preoptic nucleus neurons to olfactory brain regions Response Antidromic Spontaneously active EPSP Monosynaptic Spontaneously active Antidromic/EPSP Monosynaptic Spontaneously active IPSP Monosynaptic Spontaneously active No response Spontaneously active

PC stimulation

OB stimulation

EA stimulation

(n=58) 14 8 42 27/37 20 8 8 6 7 0 1 3 3

(n=56) 15 2 33 0/28 17 4 0 2 3 0 1 9 2

(n=54) 8 4 34 15/32 20 4 4 3 5 0 1 11 9

Of the 41 neurons responding with EPSPs to PC stimulation, 27 also responded with polysynaptic EPSPs to OB stimulation and 28 with EPSPs to EA. Nineteen neurons were shown to respond with EPSPs to PC, OB and EA stimulation, this being the most predominant pattern of activation (e.g., Fig. 4A–H); two of these were identified as large aspiny neurons (24–30 µm; e.g., Fig. 8E). Of the seven neurons in which IPSP responses were evoked by PC, one responded to OB antidromically, one with an EPSP, one with an IPSP and four gave no response. EA stimulation in these cells evoked two antidromic responses, one EPSP, three IPSP, and one no response. Three of these neurons were identified with dye filling and shown to be small aspiny neurons (8–14 µm; e.g., Fig. 8D). In the nine cells that gave no response to OB stimulation, four EPSPs and four IPSPs were elicited by PC stimulation with one not responding. EA stimulation in these cells evoked one antidromic response, three EPSPs and three IPSPs, with two not responding. Intracellular dye filling Sixteen MCPO neurons were filled (e.g., Fig. 8), 11 with LY and five with Neurobiotin. All but one had smooth radial dendrites with few branches. The diameters of smooth cells ranged from 8–10 µm, 14–20 µm and 24–30 µm for small, medium and large cells, respectively. Three large, four medium and eight small neurons were recovered. One small cell (soma diameter 8 µm) excited by PC, and projecting (antidromic response) to OB, had dendrites with sparse spines. The three large neurons had polysynaptic EPSP responses to OB stimulation (e.g., Fig. 4B,C) and monosynaptic EPSP responses to PC (e.g., Fig. 4D– F). EA stimulation evoked monosynaptic EPSPs in two neurons and a polysynaptic response in one neuron (e.g., Fig. 4G,H). These large neurons did not respond antidromically to any of the stimulated areas. Of the medium aspiny neurons, the two tested

with OB stimulation had antidromic responses (e.g., Figs 4J, 8A), one also responding antidromically to PC. Responses only to PC and EA stimulation were obtained in three cells, two filled with Neurobiotin (Fig. 8C) and the other with LY, with antidromic response to PC seen in two cells and to EA in one cell, and monosynaptic EPSP responses to PC in one cell and to EA in two cells (Fig. 6). The small aspiny neurons (Fig. 8D) showed a mixture of responses to stimulated areas. In three neurons polysynaptic IPSPs were elicited by PC stimulation. In two of these cells OB stimulation elicited no response. Polysynaptic EPSP responses to OB stimulation were seen in the remaining cells. EA stimulation evoked antidromic responses in two, one with IPSP, and one did not respond; the remaining cells were not tested. Dendrites of some medium aspiny neurons in MCPO branched into the superficial olfactory tubercle (Fig. 8A–C). Small aspiny neurons were located throughout the MCPO and usually had few dendrites confined within the nucleus (Fig. 8D). Large aspiny neurons had few dendrites with branches extending towards the olfactory tubercle or dorsally into the ventral pallidum (Fig. 8E).

Electrophysiological properties In 26 MCPO neurons, the mean (&S.E.) input resistance (Ri) was 51.48&6.48 MÙ with a mean (&S.E.) time constant (ô) of 6.74&0.86 ms. No correlation was observed between resting membrane potential (mp), Ri and ô. Twenty-two cells were classified as responding either tonically or phasically in response to 500 ms current pulses ranging from "0.5 nA to +1.1 nA. Tonically responsive cells were defined as those responding with robust action potentials throughout the duration of the pulse, while phasically responsive cells fired either a single or a few spikes at the beginning of the pulse. A few cells did not respond to the depolarizing current pulse.

Fig. 3. Responses of a magnocellular preoptic neuron projecting to the olfactory bulb (OB). (A) EPSP response with spike to OB stimulation. (B) In the absence of synaptically activated spike an antidromic response (indicated by asterisk) is evident. (C) Response to double-shock stimulation showing collision extinction of antidromic response to first shock, not present in (D) in the absence of spike response on the evoked EPSP. (E) EPSP amplitude increases with more negative membrane potential. The antidromic spike response to OB (asterisk) shows no prepotential and has constant onset latency as indicated at arrow. (F–H) Antidromic response to OB on extracellular recording from this neuron. Again, note collision extinction of antidromic response to first stimulus in (F). (I) Depolarizing response to piriform cortex stimulation. (J) EPSP response to entorhinal area stimulation. S, stimulus artifact indicated. Resting membrane potential: "56 mV. Calibrations: (B–D) as in (A); (G-J) as in (F).

Fig. 4. Response of (A–H) large aspiny and (I–L) medium-sized aspiny magnocellular preoptic nucleus neurons recorded with LY-containing pipettes. These neurons are depicted in Fig. 8E and Fig. 8A, respectively. The large neuron is spontaneously active (A) responding with polysynaptic EPSP (B) but no spikes to olfactory bulb (OB) stimulation at 0.5 mA. At higher OB stimulus strength (1 mA) multiple firing of polysynaptic spikes is also evident (C). (D) EPSP response with spikes to piriform cortex (PC) stimulation at near threshold (0.5 mA) and (E) at three times threshold. (F) Multiple sweeps with different stimulus intensities demonstrating monosynaptic EPSP with spikes to PC stimulation (arrow indicates constant latency of EPSP generation). (G,H) Polysynaptic EPSP response to entorhinal area (EA) stimulation with delayed burst-type discharge at stimulus strength of 1.0 mA (H), not present at slightly lower stimulation strength of 0.9 mA (G). Resting membrane potential: "72 mV. (I) Spontaneous spike of medium-sized aspiny neuron. (J) Neuron responds antidromically [arrow (i), asterisk (ii) in expanded trace] to OB stimulation, followed by a polysynaptic IPSP response. IPSP amplitude decreases on membrane hyperpolarization. The apparent small depolarizing potentials preceding the IPSP were not always present. (K) Monosynaptic response (arrow indicates constant onset latency with varying stimulus intensity) to EA stimulation, followed by an IPSP. (L) IPSP response only was seen to PC stimulation in this neuron. Resting membrane potential: "70 mV. S, stimulus artifact. Calibrations: (A,F) 10 mV, 20 ms; (B–E,G,H) as in (A); (I) 20 mV, 20 ms; (Ji–L) as in (I); (Jii) 2 ms.

236

A. G. Paolini and J. S. McKenzie

Fig. 5. Magnocellular preoptic nucleus neurons responding to piriform cortex (PC) and olfactory bulb (OB) stimulation, and to current injection. Current injection allowed classification of the first neuron (A–D) as tonic and the second (E–J) as phasic. (A) PC stimulation at different membrane potentials, showing greater amplitude of monosynaptic EPSP with more negative potential. (B) Response to OB stimulation at different membrane potentials showing rise in amplitude of polysynaptic EPSP with more negative potential. (C) Tonic response from resting potential to +0.7 nA depolarizing current pulse lasting 250 ms. (D) Current–voltage relationship. Input resistance of 74 MÙ calculated using linear portion of curve between "0.2 and +0.2 nA. Resting membrane potential: "78 mV. (E) Monosynaptic EPSP response to PC stimulation in second neuron. (F) Monosynaptic EPSP to PC at spike threshold (1.2 mA). (G) Polysynaptic EPSP and spike response to OB stimulation. (H) Spontaneous spike showing no slow depolarization preceding spike, nor after hyperpolarization. (I) Single spike response (i), marked with asterisk, to 500 ms +0.7 nA depolarizing pulse, shown on slower time base in (ii). (J) Current–voltage relationship with input resistance of 42 MÙ calculated using the linear portion of curve between "0.2 and +0.2 nA. Resting membrane potential: "58 mV. S, stimulus artifact. Calibration: (B) as in (A); (E,F) as in (G); (Ii) 10 mV, 15 ms; (Iii) 10 mV, 100 ms.

Ten cells responded tonically (mean Ri=51.26&5.62, ô=8.49&1.2; mp="53.00&2.24) and nine phasically (mean Ri=47.14&7.07, ô=8.91&1.67; mp="53.7&5.10), while three did not respond (mean Ri=44.41&5.57, ô=8.69&1.25; mp="58.70&4.77). No significant differences were seen in Ri, ô or mp between these cell classes. In the cells responding tonically to current injection (e.g., Figs 5C, 6B,C), OB was stimulated in eight, eliciting antidromic responses in two, EPSPs in five and no response in one. For these cells stimulation of PC evoked antidromics in four cells, EPSPs in four, IPSP in one, and no response in one. EA stimulation evoked no response in five cells, EPSPs in four, and an antidromic response in one. Of the cells responding phasically to injected current (Fig. 5I), OB stimulation evoked EPSPs in eight and an IPSP in one. Six EPSPs were evoked by PC or EA stimulation, two antidromics and one IPSP by PC, and one antidromic response by EA stimulation, the remaining two cells not responding. In the cells that did not respond to injected current, OB stimulation evoked two EPSPs

with one not responding. PC and EA stimulation both evoked monosynaptic EPSP responses in these cells. Of the 10 cells responding tonically to current injection, nine were spontaneously active. In these spontaneously active neurons, an after-hyperpolarization followed the action potential (Fig. 6A). In addition to this after hyperpolarization, six such cells had a slow depolarization preceding spike generation (Fig. 6A); two were identified as mediumsized aspiny neurons (Fig. 8C). The neuron that was not spontaneously active had no after hyperpolarization following the current-induced spike. Five cells that responded phasically to current injection were spontaneously active, four with no postspike hyperpolarization (Fig. 5H). Two of these cells also showed a slow depolarization preceding spike generation. Of the four phasic neurons that were not spontaneously active, all showed hyperpolarization after the spikes elicited by current injection. Of the cells that did not respond to current injection none was spontaneously active, but all responded to stimulation of afferent sources.

Magnocellular preoptic neuron responses to olfactory brain

237

Fig. 6. Responses from two Neurobiotin-filled medium-sized aspiny neurons in the magnocellular preoptic nucleus depicted in Fig. 8C (1) (A–F) and Fig. 8C (2) (G–I). Although neurons were impaled successively, their responses to stimulation differ. Neuron 1 is spontaneously active (A) and responds with repetitive burst discharge to 500 ms depolarizing pulse (1.0 nA) from "60 mV membrane potential (B, C). At slower time base (C), a bursting pattern of discharge, already beginning in (B), continues throughout the 500 ms of depolarization. This neuron responds monosynaptically to entorhinal area (EA) stimulation (D) and polysynaptically to piriform cortex (PC) stimulation (E). (F) Current–voltage relationship. Input resistance of 35 MÙ calculated from linear portion of curve between "0.2 and +0.2 nA. (G) Neuron 2 responds antidromically to PC stimulation (shown as mean of eight traces). (H) Antidromic response evoked by EA stimulation collides with the spontaneous spike marked by an asterisk (I). S, stimulus artifact. Calibration: (A), 20 ms; (B,D,E) as in (A); (C) 200 ms; (G) 10 ms; (H,I) as in (G).

DISCUSSION

Olfactory influences on magnocellular preoptic nucleus neurons The OB sends extensive projections to both the piriform and entorhinal cortices, which in turn take part in the associational system of connections that link different olfactory cortical areas.11 Through these projections the bulb may indirectly influence the activity of other olfactory system neurons. These olfactory areas in turn have been shown in the present experiments to receive projections from MCPO, which may act as a mechanism to regulate olfactory information processing through the basal

forebrain. A majority (57%) of cells that responded antidromically to PC stimulation also responded with monosynaptic EPSPs, suggesting reciprocal connections. This was also seen for EA stimulation in half of the MCPO cells projecting to EA. In response to OB stimulation, only polysynaptic EPSP responses were seen in MCPO neurons sending centrifugal input to the bulb, suggesting that these cells are under only indirect influence from the bulb. A proportion of MCPO neurons appears to project to all three areas stimulated. Of the MCPO neurons that innervate PC, half project also to OB, suggesting that MCPO neurons have axons that branch out to a number of olfactory target areas. This is also true with EA, since 28% and 33% of MCPO neurons

238

A. G. Paolini and J. S. McKenzie

Fig. 7. Response of magnocellular preoptic neuron to stimulation of olfactory bulb (OB) and piriform cortex (PC). In A,C,D, resting potential "56 mV indicated by dotted line. (A) Antidromic response to OB stimulation (indicated by asterisk) followed by IPSP. Antidromicity verified using double shock stimulation (B) with constant latency of spike onset to OB stimulation without prepotential (C, arrow). Antidromic response followed by an IPSP which increases in amplitude with less negative membrane potential (arrows a1 compared to a2) (C, dashed line indicates membrane potential of neuron induced by constant depolarizing current). (D–F) Reduction of IPSP amplitude by injection of hyperpolarizing current. (D) IPSP response to OB stimulation shown with voltage and time scales expanded from (A). (E) Response to 0.3 nA hyperpolarizing pulse. (F) Combined response to OB stimulation and 0.3 nA hyperpolarizing pulse showing decrease in IPSP amplitude (dotted line shows passive membrane potential during current pulse). IPSP amplitude (indicated by a4) has decreased compared to (D, a3). (G) Spike response to PC stimulation shown not to be antidromic, because of prepotential (arrow) and failure to follow double shock stimulation (H, arrow). (I) Monosynaptic EPSP response to PC stimulation at different stimulus strengths showing constant onset latency (arrow), followed by IPSP. Calibration: (A) 20 mV, 20 ms; (C,D) 10 mV, 10 ms; (B,G,H) as in (A); (E,F) as in (D); (I) 15 mV, 15 ms.

which project to PC and OB, respectively, also project to this area. Three-way output was rare, however, with only one cell in MCPO activated antidromically from all three regions stimulated. The majority of MCPO neurons influenced by olfactory brain stimulation was spontaneously active, including two-thirds of those projecting to OB. It has been shown that stimulation of this nucleus can inhibit granule cell activity16 in the OB, an effect considered to disinhibit mitral cells30 thereby facilitating their firing.17 In addition, a large proportion of

MCPO output neurons to PC (79%) and EA (63%), and just under half of those receiving projections back from PC (49%) and EA (47%), were spontaneously active. This suggests that PC and EA may regulate the activity of OB output neurons indirectly through projections to MCPO as well as by direct31 centrifugal projections. Of cells responding with polysynaptic IPSPs to PC and EA stimulation, only a minority were spontaneously active (14% and 20%, respectively), suggesting that they are generally activated only by extrinsic input to MCPO.

Fig. 8. Neurons of the magnocellular preoptic nucleus (MCPO) filled with LY (A,B,D,E) or Neurobiotin (C). Medium-sized aspiny neurons (A,B) show multipolar arborization, dendrites extending into the olfactory tubercle (OT). Arrow in (A) indicates bleb, the possible site of impalement. Insert in (B) shows dendrites in adjacent section. (C) Reconstruction of two Neurobiotin-filled neurons impaled successively which had differing responses to stimulation (see Fig. 6). (D) Small aspiny neuron showing restricted dendritic arborization within MCPO. (E) Large reconstructed aspiny neuron. Calibration: (A) 50 µm; (B) as in (A) (insert 75 µm); (C,D) 20 µm; (E) 50 µm.

Magnocellular preoptic neuron responses to olfactory brain

239

240

A. G. Paolini and J. S. McKenzie

Morphology of magnocellular preoptic nucleus neurons Three cell types were identified in this investigation based on soma size. They were not consistent with the three types defined by Brauer et al.,2 who found large aspiny neurons and two medium-sized types, one with and the other without spines. In their investigation, six medium spiny and six medium aspiny neurons ranging from 13.2 to 21.6 µm in diameter were examined. We identified only one small-sized neuron with spines, which was excited by PC and projected to the bulb; we were also able to distinguish an additional aspiny class with small diameter soma. The large aspiny neuron defined by Brauer et al.2 is similar to that seen in this investigation. However, the possibility of recording from cells of the ventral pallidum and adjacent structures scattered within the MCPO area cannot be excluded. Not all neurons in MCPO actually belong to this nucleus, since it contains an admixture of neurons from surrounding areas,2 including the ventral pallidum, the ventral substantia innominata, the medially located HDB, and in addition displaced cells from the olfactory tubercle. Although Brauer et al.2 describe three distinct types of cells in the MCPO, cell types found in the HDB are also located around the border region with MCPO. The nucleus of the HDB contains mainly two cell types2 which are predominantly cholinergic.26 The possibility of our recording from such neurons cannot be dismissed. One distinguishing feature of cells in MCPO is the size of their cell soma. Large neurons are scattered throughout this nucleus,2 although our anatomical results appear to indicate that they are not the most abundant cell in this region. The function of these large neurons in olfactory processing is not clear. These large aspiny neurons were shown to respond monosynaptically to PC and EA stimulation. Multiple polysynaptic EPSP responses were seen following their monosynaptic response. These cells responded polysynaptically to OB, with increased amplitude and duration on increasing strength. But these large neurons failed to respond antidromically to OB or cortex stimulation, suggesting that they do not play a direct role in olfactory processing. Upon cellular reconstruction, they were shown to have large aspiny dendrites and the axon appeared to project in a caudal direction. Although dendrites of these large neurons appeared to extend into the olfactory tubercle they did not arborize to the same extent as those of identified medium-sized aspiny neurons. No IPSP response was seen in these cells, suggesting that they may not be influenced by GABAergic olfactory tubercle neurons. Medium aspiny neurons identified in this investigation responded antidromically to OB, PC and EA stimulation. GABAergic neurons have been shown to be predominant over cholinergic neurons in the MCPO region of the MCPO/HDB complex,1 with

30% of GABAergic neurons projecting from this complex to the OB.35 The size of OB projection neurons described by Zaborsky et al.35 is comparable with the medium-sized neurons seen in this investigation, suggesting that medium-sized neurons may directly inhibit OB granule cells.16 Small aspiny neurons identified in MCPO may act as local interneurons, for their dendrites and axon did not show extensive branching and appeared to be confined within the boundaries of this nucleus. A number of these small aspiny neurons responded with polysynaptic IPSP responses to olfactory cortical stimulation. Antidromic activation from OB and EA was also seen in a small number of them, however, indicating that not all small aspiny neurons have actions confined within MCPO. The dendritic morphology of certain identified neurons in MCPO suggests that they receive information from the olfactory tubercle. Although there is no direct projection reported from the tubercle to MCPO, filled cells often sent their dendrites into this area, extending to the base of the brain. If these dendrites receive synaptic contact from neuronal elements in the olfactory tubercle, the possibility exists for indirect action on MCPO neurons by olfactory information. The predominant cell type in the tubercle is the medium-sized spiny neuron;21 they are GABAergic22 and hence could have an inhibitory effect on MCPO neurons. Although only a very small proportion of MCPO cells (5%) responded to OB stimulation with polysynaptic IPSPs only, a number responded with an EPSP followed by an IPSP. This was seen in a filled neuron whose dendrites arborized extensively into the olfactory tubercle. There is also the possibility that direct stimulation of the bulb may activate excitatory pathways, such as those from the PC, which then activate MCPO neurons, thereby masking any inhibitory responses derived from the olfactory tubercle. As no MCPO neurons responded with monosynaptic EPSPs to OB stimulation, it appears that their dendrites extending into the tubercle do not receive synaptic contact from incoming OB fibres known to be located in the superficial molecular layer of the olfactory tubercle.21 Properties of magnocellular preoptic nucleus neurons At least three cell types could be distinguished in current injection experiments; two classes responded phasically and one responded tonically to depolarization. Of the phasically responding cells, one set was spontaneously active and showed no spike afterhyperpolarization, while the other was not spontaneously active but did show spike afterhyperpolarization when activated by current injection. The tonically responding cells were all spontaneously active and showed spike afterhyperpolarization. In general, the tonically responsive cells also showed a slow depolarization ramp preceding spike generation. A potassium A-current

Magnocellular preoptic neuron responses to olfactory brain

may underlie this slow depolarization.20 Two tonically activated cells were identified as medium-sized aspiny neurons, with one projecting to PC. OB was not tested in these cells. As it appears that mediumsized aspiny neurons provide the major projections of MCPO, such tonically activated neurons may play an important part in controlling the effectiveness of incoming olfactory information through their action on the OB. This is further supported by the number of antidromic responses seen in tonically activated cells as compared with phasically activated cells. No cells classified as phasic were antidromically activated by OB stimulation, suggesting that this class does not project to the bulb. Most of them responded with polysynaptic EPSPs to OB. Of neurons projecting to PC, more were identified as responding tonically than phasically to depolarizing current injection, although such bias was not seen with neurons projecting to EA. As with OB stimulation, synaptic activation of MCPO by PC and EA was seen more often in phasic neurons. This pattern of response suggests that olfactory information may trigger a synaptic response in phasic cells. If these act as excitatory interneurons in MCPO, they may in turn initiate a prolonged tonic burst in cells projecting to olfactory areas, thereby modifying olfactory processing.

241

CONCLUSIONS

These results demonstrate that many neurons in the MCPO are tonically active, projecting to the OB, PC and EA. The nucleus receives direct input from the olfactory cortical areas examined, and in conjunction with previous findings,16,17 the results provide a neurophysiological basis for a tonic, probably GABAergic, inhibitory action by spontaneously active MCPO neurons on granule cells. Boosting of this tonic MCPO activity may occur by positive feedback through the excitatory output of olfactory piriform and entorhinal cortices. This action, on OB neurons by MCPO, may be important in olfactory processing, as behaviourally in animals with MCPO lesions23 and in Alzheimer’s patients6 who have basal forebrain degeneration34 there is a loss of olfactory sensitivity. There is also anatomical evidence for possible indirect regulation of MCPO neuronal activity by the olfactory tubercle, which might be responsible for inhibiting MCPO output.

Acknowledgements—This work was conducted in and funded by the Department of Physiology, University of Melbourne.

REFERENCES

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Brashear H. R., Zaborszky L. and Heimer L. (1986) Distribution of GABAergic and cholinergic neurons in the rat diagonal band. Neuroscience 17, 439–451. Brauer K., Schober W., Werner L., Winkelmann E., Lungwitz W. and Hajdu F. (1988) Neurons in the basal forebrain complex of the rat: a golgi study. J. Hirnforsch. 29, 43–71. Carey R. G. and Rieck R. W. (1987) Topographic projections to the visual cortex from the basal forebrain in the rat. Brain Res. 424, 205–215. Dinopoulous A., Parnavelas J. G. and Eckenstein F. (1986) Morphological characterization of cholinergic neurons in the horizontal limb of the diagonal band of Broca in the basal forebrain of the rat. J. Neurocytol. 15, 619–628. Divac I. (1975) Magnocellular nuclei of the basal forebrain project to neocortex, brain stem, and olfactory bulb. Review of some functional correlates. Brain Res. 93, 385–398. Doty R. L., Reyes P. F. and Gregor T. (1987) Presence of both odor identification and detection deficits in Alzheimer’s disease. Brain Res. Bull. 18, 597–600. Geeraedts L. M. G., Nieuwenhuys R. and Veenig J. G. (1990) Medial forebrain bundle of the rat: III. Cytoarchitecture of the rostral (telencephalic) part of the medial forebrain bundle bed nucleus. J. comp. Neurol. 294, 507–536. Griffith W. H. (1988) Membrane properties of cell types within guinea pig basal forebrain nuclei in vitro. J. Neurophysiol. 59, 1590–1612. Griffith W. H. and Matthews R. T. (1986) Electrophysiology of AChE-positive neurons in basal forebrain slices. Neurosci. Lett. 71, 169–174. Griffith W. H., Sim J. A. and Matthews R. T. (1991) Electrophysiologic characteristics of basal forebrain neurons in vitro. In The Basal Forebrain Anatomy to Function (eds Napier T. C., Kalwas P. W. and Hanin I.), pp. 143–155. Plenum, New York. Haberly L. B. (1985) Neuronal circuitry in olfactory cortex: anatomy and functional implications. Chem. Sen. 10, 219–238. Haberly L. B. (1990) Olfactory cortex. In The Synaptic Organisation of the Brain (ed. Shepherd G. M.), pp. 317–345. Oxford University Press, New York. Jacquin M. F., Hu J. W., Sessle B. J., Renehan W. E. and White P. M. E. (1992) Intra-axonal Neurobiotin injection rapidly stains the long-range projections of identified trigeminal primary afferents in vivo: comparisons with HRP and PHA-L. J. Neurosci. Meth. 45, 71–86. Johnston J. B. (1923) Further contributions to the study of the forebrain. J. comp. Neurol. 35, 337–481. Kitai S. T. and Park M. R. (1990) Intracellular electrophysiological techniques. In Neuromethods, Vol 14, Neurophysiological Techniques: Basic Methods and Concepts (eds Boulton A. A., Baker G. B. and Vanderwolf C. H.). Humana Press, Clifton, NJ. Kunze W. A. A., Shafton A. D., Kemm R. E. and McKenzie J. S. (1992) Olfactory bulb output neurons excited from a basal forebrain magnocellular nucleus. Brain Res. 583, 327–331. Kunze W. A. A., Shafton A. D., Kemm R. E. and McKenzie J. S. (1992) Intracellular responses of olfactory bulb granule cells to stimulating the horizontal diagonal band nucleus. Neuroscience 48, 563–569. Loo Y. T. (1931) The forebrain of the opossum, Didelphis virginiana: Part II. Histology. J. comp. Neurol. 52, 1–48.

242 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.

A. G. Paolini and J. S. McKenzie Luiten P. G. M., Gaykema R. P. A., Traben J. and Spencer D. G. (1987) Cortical projection patterns of the magnocellular basal nucleus subdivisions as revealed by anterogradely transported Phaseolus vulgaris leucoaggultinin. Brain Res. 413, 229–250. McCormick D. A. (1990) Membrane properties and neurotransmitter actions. In The Synaptic Organisation of the Brain (ed. Shepherd G. M.), pp. 32–66. Oxford University Press, New York. Millhouse O. E. and Heimer L. (1984) Cell configuration in the olfactory tubercle of the rat. J. comp. Neurol. 228, 571–597. Mugnaini E. and Oertel W. H. (1985) An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry. In Handbook of Chemical Neuroanatomy, Vol. 4, GABA and Neuropeptides in the CNS, Part I (eds Bjorklund A. and Hokfelt T.), pp. 436–608. Elsevier, New York. Paolini A. G. and McKenzie J. S. (1996) Lesions in the magnocellular preoptic nucleus decrease olfactory investigation in rats. Behav. Brain Res. 83, 223–231. Paxinos G. and Watson C. (1986) The Rat Brain in Stereotaxic Coordinates, 2nd edn. Academic, Sydney. Price J. L. (1969) The origin of the centrifugal fibres to the olfactory bulb. Brain Res. 14, 542–545. Rye D. B., Wainer B. H., Mesulam M. M., Mufson E. J. and Saper C. B. (1984) Cortical projections arising from the basal forebrain: a study of cholinergic and noncholinergic components employing combined retrograde tracing and immunohistochemical localization of choline acetyltransferase. Neuroscience 13, 627–643. Saper C. B. (1984) Organization of cerebral cortical afferent systems in the rat. II. Magnocellular basal nucleus. J. comp. Neurol. 222, 313–342. Semba K., Reiner P. B., McGeer E. G. and Fibiger H. C. (1988) Non-cholinergic basal forebrain neurons project to contralateral basal forebrain in the rat. Neurosci. Lett. 84, 23–28. Semba K., Reiner P. B., McGeer E. G. and Fibiger H. C. (1988) Brainstem afferents studied by axonal transport, immunohistochemistry, and electrophsiology in the rat. J. comp. Neurol. 267, 433–453. Shepherd G. M. (1972) Synaptic organization of the mammalian olfactory bulb. Physiol. Rev. 52, 864–917. Shepherd G. M. and Greer C. A. (1990) Olfactory bulb. In The Synaptic Organisation of the Brain (ed. Shepherd G. M.), pp. 133–169. Oxford University Press, New York. Sim J. A. and Griffith W. H. (1991) Muscarinic agonists block a late-after hyperpolarization in medial septum/ diagonal band neurons in vitro. Neurosci. Lett. 129, 63–68. Swanson L. W. (1993) Brain Maps: Structure of the Rat Brain. Elsevier, Amsterdam. Whitehouse P. J., Price D. L., Stuble R. G., Clark A. W., Coyle J. T. and Delong M. R. (1982) Alzheimer’s disease and senile dementia: loss of neurons in the basal forebrain. Science 215, 1237–1239. Zaborszky L., Carlsen J., Brashear H. R. and Heimer L. (1986) Cholinergic and GABA-ergic afferents to the olfactory bulb in the rat with special emphasis on the projection neurons in the nucleus of the horizontal limb of the diagonal band. J. comp. Neurol. 243, 488–509. (Accepted 7 October 1996)