Brain Research, 559 (1991) 64-74 © 1991 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/91/$03.50 A DON1S 000689939116942G
64
BRES 16942
Extracellular characteristics of putative cholinergic neurons in the rat laterodorsal tegmental nucleus S.J. Grant and D.A. Highfield Department of Psychology and Program in Neuroscience, University of Delaware, Newark, DE 19716 (U.S.A.)
(Accepted 16 April 1991) Key words: Acetylcholine; Noradrenaline; Lateral-dorsal tegmental nucleus; Locus coeruleus; Anterior-ventral nucleus; Thalamus; Single unit activity; Rat
The extracellular electrophysiological properties of neurons in the laterodorsal tegmental nucleus (LDT), a major source of eholinergic afferents to the thalamus, were studied in chloral hydrate-anesthetized rats. A combination of antidromie activation from the thalamus and histological verification of recording sites was used to correlate the identity of extraeellular recordings in the rat LDT with eholinergic neurons in that region. All neurons antidromically activated by stimulation of the anteroventral thalamus were histologically verified to be within dusters of eholinergie (NADPH-d-positive) cells in the LDT or in the adjacent nucleus locus coerulens (LC). The thalamically projecting LDT neurons had a homogeneous neurophysioiogical profile consisting of long duration action potentials (mean = 2.5 ms), slow conduction velocities (mean = 0.78 m/s), and lengthy ehronaxie values (mean = 0.725 ms). The apl~aranee and axonal characteristics of these neurons resembled those of noradrenergic LC neurons, but the two populations exhibited substantially different spontaneous activity patterns and sensory responsiveness. These characteristics may be useful in the preliminary identification of putative eholinergie neurons in vivo, and thereby provide a foundation for exploring the neuropharmacology, afferent modulation, sensory responsiveness and behavioral correlates of the brainstem eholinergic system. INTRODUCTION There is great interest in the activity of brainstem cholinergic neurons because of their hypothesized relation to the classic reticular activating system z7'56'6°. Cholinergic receptors in the thalamus modulate thalamic rhythmicity and sensory responsiveness 39'56, and neurons along the mesopontine border are the major source of cholinergic innervation of the thalamus as well as other subcortical sites 24'5°'52'54. Like neighboring noradrenergic and serotonergic neurons, these brainstem cholinergic neurons are clustered in specific nuclei. The Ch6 group lies in the central gray within the laterodorsal tegmental nucleus (LDT) and the Ch5 group on the lateral edge of the brachium conjunctivum within the pedunculopontine tegmental nucleus (PPT) 42. Substantially less is known of the activity of cholinergic neurons compared to other systems with similar organization and functions, such as monoamine neurons. Neurophysiological studies of these cholinergic neurons in vivo have been hampered by the lack of extracellular criteria for their identification. The discovery of specific extracellular neurophysiological properties for other chemically defined systems, such as norpinephrine-, se-
rotonin- and dopamine-containing neurons, permitted extensive studies into the physiological, pharmacological and behavioral properties of these systems s'2°'41. Differentiation of monoamine neurons from surrounding populations was initially based on a combination of indirect criteria including location of recording sites, firing patterns, waveform, antidromic activation, and the absence of such activity after lesions with specific neurotoxic agents (e.g. 6-hydroxydopamine) l'5,tl'16,26'33,43,64This study uses a similar approach to identify putative cholinergic neurons in the brainstem. Antidromic activation from the thalamus allows a strong inference to be made about the cholinergic identity of extracellularly recorded neurons in the rat LDT. Retrograde tracing studies show that over 90% of thalamieally projecting neurons from the rat L D T and PPT are cholinergic 24's°'52'54. In particular, the anteroventral nucleus (AV) contains the densest plexus of cholinergic fibers in the thalamus, and the L D T is the major source of cholinergic afferents to the AV thalamus 24'36'52. Thus, stimulation of the A V thalamus should yield many antidromieally driven neurons in the LDT, and these neurons have a high probability of being cholinergic. Explicit histological localization of the extracellular
Correspondence: S.J. Grant, Department of Psychology, 220 Wolf Hall, University of Delaware, Newark, DE 19716, U.S.A. Fax: (1) (302) 292-3645.
65 r e c o r d i n g sites is necessary as a s e c o n d c r i t e r i o n since n e u r o n s in the adjacent locus c o e r u l e u s a n d dorsal rapile also project to t h e t h a l a m u s 5°. T h o u g h it is n o t possible with e x t r a c e l l u l a r t e c h n i q u e s to identify the specific n e u r o n r e c o r d e d , cells a n t i d r o m i c a l l y activated f r o m the t h a l a m u s are likely to b e cholinergic o n l y if t h e y lie
and a second overlying the LDT and LC (1.0-3.5 ram posterior to lambda, 0.5-1.3 mm lateral) 4s. The dura was reflected taking care to spare nearby sinuses. A co-axial bipolar stimulating electrode (Rhodes SNE-100 ) was advanced to the approximate stereotaxic depth of the AV thalamus. Multi-unit activity was recorded through the inner electrode to improve the accuracy of the placements 14. After placing the stimulating electrode the bone defect was covered with saline-soaked Gelfoam and a thin layer of bone wax.
w i t h i n clusters of n e u r o n s s t a i n i n g d e e p l y for a specific m a r k e r for Ch5 a n d Ch6 cholinergic n e u r o n s , such as nicotinamide adenine dinucleotide phosphate diaphorase ( N A D P H - d ) 6°. I n t e r p r e t a t i o n of histological d a t a is facilitated in the rat b y the spatial s e g r e g a t i o n of the Ch6 cholinergic n e u r o n s f r o m n o r a d r e n e r g i c a n d serotonergic n e u r o n s 5°. This contrasts with the cat w h e r e the interdigitation of n o r a d r e n e r g i c a n d cholinergic n e u r o n s a n d the g r e a t e r p r o p o r t i o n of n o n - c h o l i n e r g i c L D T a n d P P T n e u r o n s projecting to t h a l a m u s complicates b o t h the use of r e c o r d i n g site histology a n d a n t i d r o m i c activation from t h a l a m i c nuclei as m e a n s to identify cholinergic activity31,38,44,55,59. T h e r e f o r e , the c o m b i n a t i o n of a n t i d r o m i c activation f r o m the a n t e r o v e n t r a l t h a l a m u s a n d histological verific a t i o n o f r e c o r d i n g sites was used to infer the cholinergic i d e n t i t y of e x t r a c e l l u l a r recordings in the rat LDT. It was h y p o t h e s i z e d that these p u t a t i v e cholinergic n e u r o n s w o u l d h a v e o n e o r m o r e n e u r o p h y s i o l o g i c a l characteristics (e.g. c o n d u c t i o n velocity, w a v e f o r m , s p o n t a n e o u s activity) that w o u l d d i f f e r e n t i a t e t h e m f r o m s u r r o u n d i n g n e u r o n a l p o p u l a t i o n s . Specific e m p h a s i s was p l a c e d o n c o m p a r i s o n s with n e u r o n s in the a d j a c e n t locus coeru l e u s since cholinergic L D T n e u r o n s a n d n o r a d r e n e r i g c L C n e u r o n s have similar p r o j e c t i o n s to the t h a l a m u s , similar p o s t s y n a p t i c actions o n t h a l a m i c n e u r o n s 39, a n d s o m e m a i n t a i n that the e x t r a c e l l u l a r profile of L D T n e u r o n s r e s e m b l e s that of m o n o a m i n e n e u r o n s 32. P o r t i o n s of this d a t a have b e e n p u b l i s h e d in abstract f o r m TM. MATERIALS AND METHODS
Subjects Subjects were 40 male albino rats (250-350 g) obtained from commercial suppliers (Charles River). They were housed in group cages with ad libitum access to food and water. Lighting was controlled on a 12:12 h light-dark cycle and experiments were conducted during daylight hours.
Surgery Subjects were anesthetized with either chloral hydrate (400 mg/ kg, i.p.; n = 29) or urethane (1.8 g/kg, i.p.; n = 11) at a level sufficient to abolish corneal and hindlimb withdrawal reflexes. Chloral hydrate anesthesia was supplemented throughout the experiment with periodic injections via a lateral tail vein. Core temperature was monitored with a rectal probe and maintained at 3637 °C by a heating pad. Following induction of anesthesia the rat was placed in a small animal stereotaxic frame with the head level between bregma and lambda. The skull was exposed with a midline incision and two small holes in the bone were drilled: one overlying the AV thalamus (1.3-2.0 mm caudal to bregma, 1.5 mm lateral to the midline),
Electrophysiology Glass mieroelectrodes were used to record extracellular action potentials of single neurons. Micropipettes were pulled from single barreled omega dot tubing (1.5 mm o.d.), and the tips broken under microscopic control to a diameter of 1-3/~m. The micropipettes were backfilled with 2% Pontamine sky blue in 0.5 M Na acetate yielding electrodes with impedances of 3-7 Mfl at 1 kHz. The recording electrodes were slowly advanced through the brain at a 16° caudorostral angle via a hydraulic microdrive. ExtraceUular signals were routed through a single ended high impedance probe. Amplified and filtered (200 Hz to 8 kHz) signals were displayed on oscilloscopes and an audio monitor using standard electronics. Extracellular action potentials from well-isolated (> 3:1 signal/noise ratio) single neurons were convened to standard TI'L spike pulses by a window discriminator. The standardized pulses were directed to a microcomputer for construction of interspike interval or post-stimulus time interval histograms. Continuous time interval histograms of spontaneous activity were generated by a ratemeter and displayed on a strip chart recorder. Recording locations were estimated in vivo by a series of physiological landmarks. Passage through the overlying cerebellum was marked by appearance of climbing fiber activity while a sudden drop in baseline noise signaled entrance into the fourth ventricle. Jaw stretch was used to activate neurons of the mesencephalic sensory nucleus of the trigeminal nerve (Mes V). LC single unit activity was recognized in vivo using previously published physiological criteria including long duration action potentials with a notch on the ascending limb, slow spontaneous activity (250 Hz), and, for spontaneously active neurons, collision of stimulus evoked and spontaneous action potentials37. Evoked activity not meeting these criteria was considered to represent orthodromie (synaptic) responses. Threshold was defined as the current intensity that evoked antidromic responses on 50% of the trials. Latencies were determined at current intensities of 1.5 times threshold. Refractory periods were measured with twin pulse stimulation also at 1.5-times threshold. Chronaxie curves were derived by determining threshold at pulse widths varying from 0.05 to 2.0 ms 46.
Histology Dye spots were made to verify the anatomical location of each antidromicaUy activated neuron. Deposits of Pontamine sky blue were made by passing -10/~A through the microelectrode for 5 min via a constant-current unit (Fintronics). Successive penetrations
66 were separated by at least 300/~m so that individual dye spots could be unambiguously resolved on microscopic sections. Stimulation sites were also marked after the experiment by passing +20/~A for 30 s through the center electrode. At the end of each experiment the rat was rapidly killed by injecting chloral hydrate i.v. sufficient to abolish breathing. The rat was immediately perfused through the ascending aorta with saline followed by 10% formalin in 0.1 M phosphate buffer (pH 7.4). The brain was removed and cut into rostral and caudal blocks. The caudal block containing the recording sites was placed in formalin perfusate. The rostral block encompassing the stimulation site was placed into a vial containing 10% formalin plus 1% potassium ferricyanide. Both blocks were post-fixed for 12-18 h at 4 °C, then transferred to a 30% sucrose in 0.1 M phosphate buffer solution (pH 7.4) and stored at 4 °C until the blocks sank. Serial 50-/~m sections were cut on a freezing microtome. Alternate sections were stained for Nissl substance (neutral red) and either Perls (Prussian blue) iron reaction (stimulation sites) or reduced NADPH-d (recording sites) 53'6°. NADPH-d staining was used as a simple empirical histochemical marker for the Ch5 and Ch6 neurons since all neurons that stain deeply for NADPH-d activity in the LDT and PPT have been shown to contain choline acetyltransferase 6°. Neighboring noncholinergic neurons, including those in the LC and dorsal raphe and glia do not stain for NADPH-d activity. A modification of the
method of Scherer-Singler et al. 53 was used for NADPH-d staining. Alternate serial sections through the LDT were collected into ice-cold 0.1 M phosphate buffer (pH 7.4). Free floating sections were incubated immediately in a solution containing 15 mM sodium malate, 1 mM NADPH, 20 mM MgCI and 0.2 M nitro blue tetrazolium in 0.1 M Tris-HCl (pH 8.0) for 20-40 min at 37 °C in the dark. Following incubation, the sections were rinsed in 0.1 M Tris buffer, mounted, allowed to dry overnight, then dehydrated in alcohol, cleared in xylene, and cover slips were attached. The serial order of the NADPH-d sections was determined by comparisons with the corresponding Nissl sections. The locations of the recording and stimulation sites were reconstructed from the histological sections using the dye spots as reference points. A projected image of each histological section containing a spot was traced onto graph paper along with the outlines of prominent nuclear structures and the position of any NADPHd-positive neurons. The reconstructed recording sites were collapsed across all animals and plotted on tracings derived from a standard stereotaxic atlas 45.
Data analysis Single unit activity was classified according to their histological location and electrophysiologieal responses. The LDT was defined as co-extensive with the cluster of NADPH-d-stained neurons located lateral to the dorsal tegmental nucleus of Gudden, medial to
Fig. 1. Anatomical verification of representative recording (A,B) and stimulation (C,D) sites of antidromically driven LDT neurons. Adjacent Nissl (A) and NADPH-d (B) sections contain a dye spot marking a recording site (arrow) in proximity of cholinergic neurons. The neuron recorded at this site was antidromically driven by a stimulating electrode whose tip (arrowhead) was in the dorsal anterior-ventral thalamus shown in adjacent Nissl (C) and acetylcholinesterase (D) stained sections. DTg, dorsal tegmental nucleus of Gudden, V, ventricle, sm, stria medullaris, Rt, thalamic reticular nucleus. Calibration bar : A,B, 100 gm; C,D, 250/am.
67 the LC and Mes V and dorsal to the fibers of the mlf (Fig. 1). Thus, the adjacent NADPH-d-POor parvicellular Barrington's nucleus and the NADPH-d neurons in the parabrachial region lateral to Mes V were not considered part of the LDT. The LC, known to be a homogeneons duster of noradrenergic neurons, was easily recognized in Nissl sections as a compact collection of hyperehromic cells 2:'2s. LC neurons had to be located within the borders of the LC and conform to the standard physiological criteria described above. Most electrophysiological measurements were made from photographs of oscilloscope traces. Action potential duration was measured from the initial voltage deflection to the second zero (baseline) crossing. Latency was measured from stimulus onset to the initial action potential deflection. Conduction velocity was estimated using distances derived from published figures of the thalamic trajectory of LC and LDT axons 2s'52. Chronaxie values were derived by graphically estimating the stimulus duration at twice the rheobase (plateau) stimulus current 46. The interval at which the neuron failed to follow twin pulse stimulation at 1.5-times threshold intensity was considered the relative refractory period. Spontaneous firing rates were derived from interspike interval histograms or ratemeter generated continuous time histograms. Histograms were collected during a 5-15 min period in the absence of any stimulation. In the statistical analysis non-spontaneous neurons were assigned a firing rate of zero and included in the calculation of the average population firing rate. Data were analyzed by
parametric statistical tests (t-tests, ANOVA with post hoc comparisons) using a microcomputer program (SysStat).
RESULTS
Histological location Out of a total of 248 cells recorded from 40 rats, all neurons activated by antidromic stimulation were restricted to recording sites in either the LDT or LC. All recording sites in the LDT yielding antidromic activity were in proximity to neurons staining deeply for NADPH-d (Fig. 1); all other antidromic neurons were located in the rostral and middle portions of the LC (Fig. 2). All cells provisionally identified as noradrenergic by established physiological criteria were histologically located within the borders of the LC. The remaining nonantidromically activated cells were found within a region encompassing the LDT, Barrington's nucleus, dorsal tegmental nucleus of Gudden and the reticular formation immediately ventral to the mlf (Fig. 2).
D A:
I
•
20 msec
B:
v•
•
v•
•
•
o LC -AD
I~
LDT-AD LDT-ND
Fig. 2. Reconstruction of recording sites. Sections are redrawn from the atlas of Paxinos and Watson. Borders of the LDT were based on the distribution of NADPH-d neurons. Note that antidromically activated neurons were confined to either the LC (LC-AD) or the LDT (LDT-AD) whereas neurons not antidromically activated (LDT-ND) were also found beyond the LDT. For clarity only antidromically activated locus coeruleus neurons (LC-AD) were plotted, sc, sub coeruleus.
lO msec
Fig. 3. Antidromic activation of two LDT neuroas. A: collision test. In the top trace, a stimulation pulse (triangle) triggered by a spontaneous action potential (left solid circle) after a critical interval eficited a constant latency action potential (right solid circle). In the bottom trace a slight decrease in the interval blocked the evoked action potential. B: high frequency follo~ng. In the top trace dual pulse stimulation (triangles) at 300 Hz elicits two constant latency action potentials (solid circles). In the bottom trace a small decrease in the interpulse interval fails to elicit the second action potential. All figures are 5 overlapping traces.
68 Effective stimulation sites were located predominantly in the anterior-ventral thalamic nucleus, but since stimulation currents could spread into adjacent anterior nuclei (e.g. anterior-medial, anterior-dorsal and lateral-dorsal nuclei, effective stimulation sites will be collectively called the anterior thalamus. A typical stimulation site in the anterior-ventral thalamus is shown in Fig. 1. No antidromically driven neurons were recorded when the stimulating electrodes were located outside the thalamus, nor were antidromic L D T neurons recorded when the stimulation sites were in the reticular or ventrolateral thalamic nuclei. Stimulation electrodes were histologically located in the anteroventral (29/40), reticular (5/40), ventrolateral (1/40), and central lateral/medial dorsal (2/40) thalamic nuclei. In 3 rats the stimulation electrodes were located outside the thalamus, in the hippocampus (2/40) and the bed nucleus of the stria terminalis (1/40).
A:
== o ~e
5-9.9
0-4.9
10-14.9
15-19.9 LATENCY
20-24.9
25-29.9
30-349
35-39,9
l
>40
(reset)
B:
Antidromic activation Out of the total population, 60 cells (24%) exhibited constant latency spikes following stimulation of the anterior thalamus, and 44 (17%) satisfied at least one other criterion for antidromic activation. Of all L D T neurons, 16% (20/126) were antidromically driven, while 19% (24/ 122) of all LC neurons were antidromically driven. Recordings of 16 neurons were not sufficiently stable to permit testing for additional criteria, with the LC neurons constituting most of this population (13/16). Since the antidromic status of these cells is ambiguous, they were excluded from further analysis. The antidromically driven L D T and LC neurons will be hereafter referred to as LDT-AD and LC-AD, respectively. The neurons that were not antidromically driven are designated as LDT-ND and LC-ND. Fig. 3 shows the responses of two representative LDT=AD neurons meeting the criteria for antidromic activation. Collision testing was precluded for many LDT-AD neurons due to absence of spontaneous activity. Therefore, non-spontaneous LDT neurons that exhibited constant latency and followed high frequency stimulation (>250 Hz) were considered antidromic 37. These criteria were considered reasonable as we have previously found that the orthodromic activation of LDT neurons produced by single pulse stimulation of the medial prefrontal cortex fails to exhibit constant latency or follow stimulation frequencies greater than 50 Hz (Grant and Highfield, unpublished results). In addition, all LC-AD neurons that were spontaneously active and satisfied the collision criterion also exhibited high frequency following in excess of 250 Hz. As seen in Fig. 4A, the antidromic response latency for L D T neurons ranged from 6 to 30 ms with an average latency of 16.5 - 3.7 (S.E.M.) ms. By contrast, LC neurons had a longer average latency (25.2 -+ 2.3 ms), and a slightly wider, but overlapping range (6-52 ms). These latency differences were statistically significant (t = 2.87, df = 42, P < 0.01) (Fig. 4B), but the calculated conduction velocities for L D T (0.78 --- 0.09 m/s) and LC (0.62 - 0.08 s) neurons were not (t = 1.29, df = 42, P < 0.075). This was probably because the shorter latency
Z
A LDT
LC
B
LDT
LC
Fig. 4. Distribution of antidromic latencies for LDT and LC neurons. The distributions have overlapping ranges (A), but LC neurons had a significantly longer mean latency due to a few neurons with especially long latencies (** P < 0.01, t.test). Ordinate labels indicate the range of the histogram bins (B). There was no significant difference between LC and LDT neurons in calculated conduction velocities because of differences in axonal pathway lengths (see text).
I msec
I msec
1 msec
Fig, 5. Representative waveforms for antidromically activated LC (A) and LDT (B) neurons in filtered recordings (200-8 kHz). The major distinguishing feature is the prominent notch (IS-SD break)
on the ascending limb of the LC action potential.
69 of L D T neurons was offset by its more rostral location. No significant differences were found between L D T - A D and L C - A D neurons with respect to stimulation threshold (LDT, 489 -+ 88 # A ; LC neurons, 476 __ 57/zA), relative refractory period (LDT, 2.1 +- 0.2 ms; LC, 2.2 -- 0.2 ms), or chronaxie (LDT, 0.725 -+ 0.275 ms; LC, 0.470 -+ 0.20 ms). The conduction velocity, refractory period, and antidromic stimulation thresholds of L C neurons are within the range of previous descriptions for this population 11'13'43.
single unit activity. This notch is usually interpreted as marking the spread of the action potential from the axon hillock/initial segment to the soma-dendritic compartment (IS-SD break). Most of the non-driven L D T cells also had initially positive biphasic waveforms. Only a few non-driven neurons (12/126) were initially negative, and they were all located ventral to the mlf at the border of the reticular formation. Although there was substantial overlap in the range of values, there were reliable quantitative differences among the action potential duration widths of 3 populations (Fig. 6). L D T - A D neurons (2.5 - 0.2 ms, n = 14) had shorter duration action potentials than L C - A D neurons (3.1 - 0.4 ms, n = 13), but both were longer than L D T - N D neurons (1.6 - 0.1 ms, n = 82). All of these differences were found to be significant by A N O V A (F = 40.2, df = 2, 106, P < 0.001) and post-hoc comparisons (Newman-Keuls, P
64
BRES 16942
Extracellular characteristics of putative cholinergic neurons in the rat laterodorsal tegmental nucleus S.J. Grant and D.A. Highfield Department of Psychology and Program in Neuroscience, University of Delaware, Newark, DE 19716 (U.S.A.)
(Accepted 16 April 1991) Key words: Acetylcholine; Noradrenaline; Lateral-dorsal tegmental nucleus; Locus coeruleus; Anterior-ventral nucleus; Thalamus; Single unit activity; Rat
The extracellular electrophysiological properties of neurons in the laterodorsal tegmental nucleus (LDT), a major source of eholinergic afferents to the thalamus, were studied in chloral hydrate-anesthetized rats. A combination of antidromie activation from the thalamus and histological verification of recording sites was used to correlate the identity of extraeellular recordings in the rat LDT with eholinergic neurons in that region. All neurons antidromically activated by stimulation of the anteroventral thalamus were histologically verified to be within dusters of eholinergie (NADPH-d-positive) cells in the LDT or in the adjacent nucleus locus coerulens (LC). The thalamically projecting LDT neurons had a homogeneous neurophysioiogical profile consisting of long duration action potentials (mean = 2.5 ms), slow conduction velocities (mean = 0.78 m/s), and lengthy ehronaxie values (mean = 0.725 ms). The apl~aranee and axonal characteristics of these neurons resembled those of noradrenergic LC neurons, but the two populations exhibited substantially different spontaneous activity patterns and sensory responsiveness. These characteristics may be useful in the preliminary identification of putative eholinergie neurons in vivo, and thereby provide a foundation for exploring the neuropharmacology, afferent modulation, sensory responsiveness and behavioral correlates of the brainstem eholinergic system. INTRODUCTION There is great interest in the activity of brainstem cholinergic neurons because of their hypothesized relation to the classic reticular activating system z7'56'6°. Cholinergic receptors in the thalamus modulate thalamic rhythmicity and sensory responsiveness 39'56, and neurons along the mesopontine border are the major source of cholinergic innervation of the thalamus as well as other subcortical sites 24'5°'52'54. Like neighboring noradrenergic and serotonergic neurons, these brainstem cholinergic neurons are clustered in specific nuclei. The Ch6 group lies in the central gray within the laterodorsal tegmental nucleus (LDT) and the Ch5 group on the lateral edge of the brachium conjunctivum within the pedunculopontine tegmental nucleus (PPT) 42. Substantially less is known of the activity of cholinergic neurons compared to other systems with similar organization and functions, such as monoamine neurons. Neurophysiological studies of these cholinergic neurons in vivo have been hampered by the lack of extracellular criteria for their identification. The discovery of specific extracellular neurophysiological properties for other chemically defined systems, such as norpinephrine-, se-
rotonin- and dopamine-containing neurons, permitted extensive studies into the physiological, pharmacological and behavioral properties of these systems s'2°'41. Differentiation of monoamine neurons from surrounding populations was initially based on a combination of indirect criteria including location of recording sites, firing patterns, waveform, antidromic activation, and the absence of such activity after lesions with specific neurotoxic agents (e.g. 6-hydroxydopamine) l'5,tl'16,26'33,43,64This study uses a similar approach to identify putative cholinergic neurons in the brainstem. Antidromic activation from the thalamus allows a strong inference to be made about the cholinergic identity of extracellularly recorded neurons in the rat LDT. Retrograde tracing studies show that over 90% of thalamieally projecting neurons from the rat L D T and PPT are cholinergic 24's°'52'54. In particular, the anteroventral nucleus (AV) contains the densest plexus of cholinergic fibers in the thalamus, and the L D T is the major source of cholinergic afferents to the AV thalamus 24'36'52. Thus, stimulation of the A V thalamus should yield many antidromieally driven neurons in the LDT, and these neurons have a high probability of being cholinergic. Explicit histological localization of the extracellular
Correspondence: S.J. Grant, Department of Psychology, 220 Wolf Hall, University of Delaware, Newark, DE 19716, U.S.A. Fax: (1) (302) 292-3645.
65 r e c o r d i n g sites is necessary as a s e c o n d c r i t e r i o n since n e u r o n s in the adjacent locus c o e r u l e u s a n d dorsal rapile also project to t h e t h a l a m u s 5°. T h o u g h it is n o t possible with e x t r a c e l l u l a r t e c h n i q u e s to identify the specific n e u r o n r e c o r d e d , cells a n t i d r o m i c a l l y activated f r o m the t h a l a m u s are likely to b e cholinergic o n l y if t h e y lie
and a second overlying the LDT and LC (1.0-3.5 ram posterior to lambda, 0.5-1.3 mm lateral) 4s. The dura was reflected taking care to spare nearby sinuses. A co-axial bipolar stimulating electrode (Rhodes SNE-100 ) was advanced to the approximate stereotaxic depth of the AV thalamus. Multi-unit activity was recorded through the inner electrode to improve the accuracy of the placements 14. After placing the stimulating electrode the bone defect was covered with saline-soaked Gelfoam and a thin layer of bone wax.
w i t h i n clusters of n e u r o n s s t a i n i n g d e e p l y for a specific m a r k e r for Ch5 a n d Ch6 cholinergic n e u r o n s , such as nicotinamide adenine dinucleotide phosphate diaphorase ( N A D P H - d ) 6°. I n t e r p r e t a t i o n of histological d a t a is facilitated in the rat b y the spatial s e g r e g a t i o n of the Ch6 cholinergic n e u r o n s f r o m n o r a d r e n e r g i c a n d serotonergic n e u r o n s 5°. This contrasts with the cat w h e r e the interdigitation of n o r a d r e n e r g i c a n d cholinergic n e u r o n s a n d the g r e a t e r p r o p o r t i o n of n o n - c h o l i n e r g i c L D T a n d P P T n e u r o n s projecting to t h a l a m u s complicates b o t h the use of r e c o r d i n g site histology a n d a n t i d r o m i c activation from t h a l a m i c nuclei as m e a n s to identify cholinergic activity31,38,44,55,59. T h e r e f o r e , the c o m b i n a t i o n of a n t i d r o m i c activation f r o m the a n t e r o v e n t r a l t h a l a m u s a n d histological verific a t i o n o f r e c o r d i n g sites was used to infer the cholinergic i d e n t i t y of e x t r a c e l l u l a r recordings in the rat LDT. It was h y p o t h e s i z e d that these p u t a t i v e cholinergic n e u r o n s w o u l d h a v e o n e o r m o r e n e u r o p h y s i o l o g i c a l characteristics (e.g. c o n d u c t i o n velocity, w a v e f o r m , s p o n t a n e o u s activity) that w o u l d d i f f e r e n t i a t e t h e m f r o m s u r r o u n d i n g n e u r o n a l p o p u l a t i o n s . Specific e m p h a s i s was p l a c e d o n c o m p a r i s o n s with n e u r o n s in the a d j a c e n t locus coeru l e u s since cholinergic L D T n e u r o n s a n d n o r a d r e n e r i g c L C n e u r o n s have similar p r o j e c t i o n s to the t h a l a m u s , similar p o s t s y n a p t i c actions o n t h a l a m i c n e u r o n s 39, a n d s o m e m a i n t a i n that the e x t r a c e l l u l a r profile of L D T n e u r o n s r e s e m b l e s that of m o n o a m i n e n e u r o n s 32. P o r t i o n s of this d a t a have b e e n p u b l i s h e d in abstract f o r m TM. MATERIALS AND METHODS
Subjects Subjects were 40 male albino rats (250-350 g) obtained from commercial suppliers (Charles River). They were housed in group cages with ad libitum access to food and water. Lighting was controlled on a 12:12 h light-dark cycle and experiments were conducted during daylight hours.
Surgery Subjects were anesthetized with either chloral hydrate (400 mg/ kg, i.p.; n = 29) or urethane (1.8 g/kg, i.p.; n = 11) at a level sufficient to abolish corneal and hindlimb withdrawal reflexes. Chloral hydrate anesthesia was supplemented throughout the experiment with periodic injections via a lateral tail vein. Core temperature was monitored with a rectal probe and maintained at 3637 °C by a heating pad. Following induction of anesthesia the rat was placed in a small animal stereotaxic frame with the head level between bregma and lambda. The skull was exposed with a midline incision and two small holes in the bone were drilled: one overlying the AV thalamus (1.3-2.0 mm caudal to bregma, 1.5 mm lateral to the midline),
Electrophysiology Glass mieroelectrodes were used to record extracellular action potentials of single neurons. Micropipettes were pulled from single barreled omega dot tubing (1.5 mm o.d.), and the tips broken under microscopic control to a diameter of 1-3/~m. The micropipettes were backfilled with 2% Pontamine sky blue in 0.5 M Na acetate yielding electrodes with impedances of 3-7 Mfl at 1 kHz. The recording electrodes were slowly advanced through the brain at a 16° caudorostral angle via a hydraulic microdrive. ExtraceUular signals were routed through a single ended high impedance probe. Amplified and filtered (200 Hz to 8 kHz) signals were displayed on oscilloscopes and an audio monitor using standard electronics. Extracellular action potentials from well-isolated (> 3:1 signal/noise ratio) single neurons were convened to standard TI'L spike pulses by a window discriminator. The standardized pulses were directed to a microcomputer for construction of interspike interval or post-stimulus time interval histograms. Continuous time interval histograms of spontaneous activity were generated by a ratemeter and displayed on a strip chart recorder. Recording locations were estimated in vivo by a series of physiological landmarks. Passage through the overlying cerebellum was marked by appearance of climbing fiber activity while a sudden drop in baseline noise signaled entrance into the fourth ventricle. Jaw stretch was used to activate neurons of the mesencephalic sensory nucleus of the trigeminal nerve (Mes V). LC single unit activity was recognized in vivo using previously published physiological criteria including long duration action potentials with a notch on the ascending limb, slow spontaneous activity (250 Hz), and, for spontaneously active neurons, collision of stimulus evoked and spontaneous action potentials37. Evoked activity not meeting these criteria was considered to represent orthodromie (synaptic) responses. Threshold was defined as the current intensity that evoked antidromic responses on 50% of the trials. Latencies were determined at current intensities of 1.5 times threshold. Refractory periods were measured with twin pulse stimulation also at 1.5-times threshold. Chronaxie curves were derived by determining threshold at pulse widths varying from 0.05 to 2.0 ms 46.
Histology Dye spots were made to verify the anatomical location of each antidromicaUy activated neuron. Deposits of Pontamine sky blue were made by passing -10/~A through the microelectrode for 5 min via a constant-current unit (Fintronics). Successive penetrations
66 were separated by at least 300/~m so that individual dye spots could be unambiguously resolved on microscopic sections. Stimulation sites were also marked after the experiment by passing +20/~A for 30 s through the center electrode. At the end of each experiment the rat was rapidly killed by injecting chloral hydrate i.v. sufficient to abolish breathing. The rat was immediately perfused through the ascending aorta with saline followed by 10% formalin in 0.1 M phosphate buffer (pH 7.4). The brain was removed and cut into rostral and caudal blocks. The caudal block containing the recording sites was placed in formalin perfusate. The rostral block encompassing the stimulation site was placed into a vial containing 10% formalin plus 1% potassium ferricyanide. Both blocks were post-fixed for 12-18 h at 4 °C, then transferred to a 30% sucrose in 0.1 M phosphate buffer solution (pH 7.4) and stored at 4 °C until the blocks sank. Serial 50-/~m sections were cut on a freezing microtome. Alternate sections were stained for Nissl substance (neutral red) and either Perls (Prussian blue) iron reaction (stimulation sites) or reduced NADPH-d (recording sites) 53'6°. NADPH-d staining was used as a simple empirical histochemical marker for the Ch5 and Ch6 neurons since all neurons that stain deeply for NADPH-d activity in the LDT and PPT have been shown to contain choline acetyltransferase 6°. Neighboring noncholinergic neurons, including those in the LC and dorsal raphe and glia do not stain for NADPH-d activity. A modification of the
method of Scherer-Singler et al. 53 was used for NADPH-d staining. Alternate serial sections through the LDT were collected into ice-cold 0.1 M phosphate buffer (pH 7.4). Free floating sections were incubated immediately in a solution containing 15 mM sodium malate, 1 mM NADPH, 20 mM MgCI and 0.2 M nitro blue tetrazolium in 0.1 M Tris-HCl (pH 8.0) for 20-40 min at 37 °C in the dark. Following incubation, the sections were rinsed in 0.1 M Tris buffer, mounted, allowed to dry overnight, then dehydrated in alcohol, cleared in xylene, and cover slips were attached. The serial order of the NADPH-d sections was determined by comparisons with the corresponding Nissl sections. The locations of the recording and stimulation sites were reconstructed from the histological sections using the dye spots as reference points. A projected image of each histological section containing a spot was traced onto graph paper along with the outlines of prominent nuclear structures and the position of any NADPHd-positive neurons. The reconstructed recording sites were collapsed across all animals and plotted on tracings derived from a standard stereotaxic atlas 45.
Data analysis Single unit activity was classified according to their histological location and electrophysiologieal responses. The LDT was defined as co-extensive with the cluster of NADPH-d-stained neurons located lateral to the dorsal tegmental nucleus of Gudden, medial to
Fig. 1. Anatomical verification of representative recording (A,B) and stimulation (C,D) sites of antidromically driven LDT neurons. Adjacent Nissl (A) and NADPH-d (B) sections contain a dye spot marking a recording site (arrow) in proximity of cholinergic neurons. The neuron recorded at this site was antidromically driven by a stimulating electrode whose tip (arrowhead) was in the dorsal anterior-ventral thalamus shown in adjacent Nissl (C) and acetylcholinesterase (D) stained sections. DTg, dorsal tegmental nucleus of Gudden, V, ventricle, sm, stria medullaris, Rt, thalamic reticular nucleus. Calibration bar : A,B, 100 gm; C,D, 250/am.
67 the LC and Mes V and dorsal to the fibers of the mlf (Fig. 1). Thus, the adjacent NADPH-d-POor parvicellular Barrington's nucleus and the NADPH-d neurons in the parabrachial region lateral to Mes V were not considered part of the LDT. The LC, known to be a homogeneons duster of noradrenergic neurons, was easily recognized in Nissl sections as a compact collection of hyperehromic cells 2:'2s. LC neurons had to be located within the borders of the LC and conform to the standard physiological criteria described above. Most electrophysiological measurements were made from photographs of oscilloscope traces. Action potential duration was measured from the initial voltage deflection to the second zero (baseline) crossing. Latency was measured from stimulus onset to the initial action potential deflection. Conduction velocity was estimated using distances derived from published figures of the thalamic trajectory of LC and LDT axons 2s'52. Chronaxie values were derived by graphically estimating the stimulus duration at twice the rheobase (plateau) stimulus current 46. The interval at which the neuron failed to follow twin pulse stimulation at 1.5-times threshold intensity was considered the relative refractory period. Spontaneous firing rates were derived from interspike interval histograms or ratemeter generated continuous time histograms. Histograms were collected during a 5-15 min period in the absence of any stimulation. In the statistical analysis non-spontaneous neurons were assigned a firing rate of zero and included in the calculation of the average population firing rate. Data were analyzed by
parametric statistical tests (t-tests, ANOVA with post hoc comparisons) using a microcomputer program (SysStat).
RESULTS
Histological location Out of a total of 248 cells recorded from 40 rats, all neurons activated by antidromic stimulation were restricted to recording sites in either the LDT or LC. All recording sites in the LDT yielding antidromic activity were in proximity to neurons staining deeply for NADPH-d (Fig. 1); all other antidromic neurons were located in the rostral and middle portions of the LC (Fig. 2). All cells provisionally identified as noradrenergic by established physiological criteria were histologically located within the borders of the LC. The remaining nonantidromically activated cells were found within a region encompassing the LDT, Barrington's nucleus, dorsal tegmental nucleus of Gudden and the reticular formation immediately ventral to the mlf (Fig. 2).
D A:
I
•
20 msec
B:
v•
•
v•
•
•
o LC -AD
I~
LDT-AD LDT-ND
Fig. 2. Reconstruction of recording sites. Sections are redrawn from the atlas of Paxinos and Watson. Borders of the LDT were based on the distribution of NADPH-d neurons. Note that antidromically activated neurons were confined to either the LC (LC-AD) or the LDT (LDT-AD) whereas neurons not antidromically activated (LDT-ND) were also found beyond the LDT. For clarity only antidromically activated locus coeruleus neurons (LC-AD) were plotted, sc, sub coeruleus.
lO msec
Fig. 3. Antidromic activation of two LDT neuroas. A: collision test. In the top trace, a stimulation pulse (triangle) triggered by a spontaneous action potential (left solid circle) after a critical interval eficited a constant latency action potential (right solid circle). In the bottom trace a slight decrease in the interval blocked the evoked action potential. B: high frequency follo~ng. In the top trace dual pulse stimulation (triangles) at 300 Hz elicits two constant latency action potentials (solid circles). In the bottom trace a small decrease in the interpulse interval fails to elicit the second action potential. All figures are 5 overlapping traces.
68 Effective stimulation sites were located predominantly in the anterior-ventral thalamic nucleus, but since stimulation currents could spread into adjacent anterior nuclei (e.g. anterior-medial, anterior-dorsal and lateral-dorsal nuclei, effective stimulation sites will be collectively called the anterior thalamus. A typical stimulation site in the anterior-ventral thalamus is shown in Fig. 1. No antidromically driven neurons were recorded when the stimulating electrodes were located outside the thalamus, nor were antidromic L D T neurons recorded when the stimulation sites were in the reticular or ventrolateral thalamic nuclei. Stimulation electrodes were histologically located in the anteroventral (29/40), reticular (5/40), ventrolateral (1/40), and central lateral/medial dorsal (2/40) thalamic nuclei. In 3 rats the stimulation electrodes were located outside the thalamus, in the hippocampus (2/40) and the bed nucleus of the stria terminalis (1/40).
A:
== o ~e
5-9.9
0-4.9
10-14.9
15-19.9 LATENCY
20-24.9
25-29.9
30-349
35-39,9
l
>40
(reset)
B:
Antidromic activation Out of the total population, 60 cells (24%) exhibited constant latency spikes following stimulation of the anterior thalamus, and 44 (17%) satisfied at least one other criterion for antidromic activation. Of all L D T neurons, 16% (20/126) were antidromically driven, while 19% (24/ 122) of all LC neurons were antidromically driven. Recordings of 16 neurons were not sufficiently stable to permit testing for additional criteria, with the LC neurons constituting most of this population (13/16). Since the antidromic status of these cells is ambiguous, they were excluded from further analysis. The antidromically driven L D T and LC neurons will be hereafter referred to as LDT-AD and LC-AD, respectively. The neurons that were not antidromically driven are designated as LDT-ND and LC-ND. Fig. 3 shows the responses of two representative LDT=AD neurons meeting the criteria for antidromic activation. Collision testing was precluded for many LDT-AD neurons due to absence of spontaneous activity. Therefore, non-spontaneous LDT neurons that exhibited constant latency and followed high frequency stimulation (>250 Hz) were considered antidromic 37. These criteria were considered reasonable as we have previously found that the orthodromic activation of LDT neurons produced by single pulse stimulation of the medial prefrontal cortex fails to exhibit constant latency or follow stimulation frequencies greater than 50 Hz (Grant and Highfield, unpublished results). In addition, all LC-AD neurons that were spontaneously active and satisfied the collision criterion also exhibited high frequency following in excess of 250 Hz. As seen in Fig. 4A, the antidromic response latency for L D T neurons ranged from 6 to 30 ms with an average latency of 16.5 - 3.7 (S.E.M.) ms. By contrast, LC neurons had a longer average latency (25.2 -+ 2.3 ms), and a slightly wider, but overlapping range (6-52 ms). These latency differences were statistically significant (t = 2.87, df = 42, P < 0.01) (Fig. 4B), but the calculated conduction velocities for L D T (0.78 --- 0.09 m/s) and LC (0.62 - 0.08 s) neurons were not (t = 1.29, df = 42, P < 0.075). This was probably because the shorter latency
Z
A LDT
LC
B
LDT
LC
Fig. 4. Distribution of antidromic latencies for LDT and LC neurons. The distributions have overlapping ranges (A), but LC neurons had a significantly longer mean latency due to a few neurons with especially long latencies (** P < 0.01, t.test). Ordinate labels indicate the range of the histogram bins (B). There was no significant difference between LC and LDT neurons in calculated conduction velocities because of differences in axonal pathway lengths (see text).
I msec
I msec
1 msec
Fig, 5. Representative waveforms for antidromically activated LC (A) and LDT (B) neurons in filtered recordings (200-8 kHz). The major distinguishing feature is the prominent notch (IS-SD break)
on the ascending limb of the LC action potential.
69 of L D T neurons was offset by its more rostral location. No significant differences were found between L D T - A D and L C - A D neurons with respect to stimulation threshold (LDT, 489 -+ 88 # A ; LC neurons, 476 __ 57/zA), relative refractory period (LDT, 2.1 +- 0.2 ms; LC, 2.2 -- 0.2 ms), or chronaxie (LDT, 0.725 -+ 0.275 ms; LC, 0.470 -+ 0.20 ms). The conduction velocity, refractory period, and antidromic stimulation thresholds of L C neurons are within the range of previous descriptions for this population 11'13'43.
single unit activity. This notch is usually interpreted as marking the spread of the action potential from the axon hillock/initial segment to the soma-dendritic compartment (IS-SD break). Most of the non-driven L D T cells also had initially positive biphasic waveforms. Only a few non-driven neurons (12/126) were initially negative, and they were all located ventral to the mlf at the border of the reticular formation. Although there was substantial overlap in the range of values, there were reliable quantitative differences among the action potential duration widths of 3 populations (Fig. 6). L D T - A D neurons (2.5 - 0.2 ms, n = 14) had shorter duration action potentials than L C - A D neurons (3.1 - 0.4 ms, n = 13), but both were longer than L D T - N D neurons (1.6 - 0.1 ms, n = 82). All of these differences were found to be significant by A N O V A (F = 40.2, df = 2, 106, P < 0.001) and post-hoc comparisons (Newman-Keuls, P
