Brain Research, 562 (1991) 126-135 © 1991 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/91/$03.50 ADONIS 0006899391170750
Rostral ventrolateral medulla" a source of the glutamatergic innervation of the sympathetic intermediolateral nucleus Shaun E Morrison, Janie Callaway, Teresa A. Milner and Donald J. Reis Division of Neurobiology, Department of Neurology and Neuroscience, Cornell University Medical College, New York, NY 10021 (U.S.A.) (Accepted 28 May 1991) Key words: Sympathetic preganglionic neurons; Sympathetic nerve discharge; Phaseolus vulgaris-leucoagglutinin; Spinal cord; Excitatory amino acid; Autoradiography; Immunocytochemistry; Electron microscopy
To determine whether the sympathoexcitatory projection from the rostral ventrolateral medulla (RVL) to the sympathetic intermediolateral nucleus (IML) of the spinal cord might use glutamate as an excitatory transmitter, we performed a dual-label, transport and immunocytochemical ultrastructural study. Axon terminals within the IML were examined to determine whether anterogradely transported Phaseolus vulgaris-leucoagglutinin (PHA-L) following injections into the RVL, was coloealized with glutamate immunoreactivity using an antibody to hemocyanin-conjugated L-glutamate (Hepler et al., J. Histochem. Cytochem., 36 (1988) 13-22). Transported PHA-L was visualized with the peroxidase-antiperoxidase technique while glutamate-like immunoreactivity was localized within the same section of the thoracic spinal cord with immunoautoradiography. By light microscopy, PHA-L immunoreactivity was found within a plexus of fine fibers and varicose processes localized to the IML. Silver grains indicative of glutamate immunoreactivity were concentrated over the IML and also over the superficial layers of the dorsal horn. Electron microscopic analysis revealed PHA-L immunoreactivity in axons and axon terminals within the IML. They ranged in diameter from 0.5 to 2.0/,m, contained numerous small clear and 0-3 large, dense-core vesicles, and formed primarily asymmetric synaptic contacts on small dendrites of IML neurons. Some of the PHA-L immunoreactive terminals making asymmetric (excitatory) synaptic contacts on the small dendrites of IML neurons also contained glutamate-like immunoreactivity. We conclude that at least a portion of the input to the IML from the RVL uses glutamate as its transmitter. These findings provide ultrastructural support for the hypothesis that the release of glutamate onto neurons in the IML plays a key role in the regulation of arterial pressure by reticulospinal, sympathoexcitatory pathways from the RVL.
INTRODUCTION Neurons in the rostral ventrolateral m e d u l l a ( R V L ) that project to the spinal intermediolateral nucleus (IML) 13'29'34 p r o d u c e a tonic excitation of sympathetic preganglionic neurons 1'4'17'23 that maintains resting sympathetic tone and arterial pressure ~°'3°. The excitatory neurotransmitter released within the I M L by the terminals of RVL-spinal neurons is not known. Two lines of evidence suggest that the transmitter is an excitatory amino acid, possibly L-glutamate 25. Physiologically, spinal application of the b r o a d - s p e c t r u m glutamate r e c e p t o r antagonist, kynurenic acid, blocks the excitatory responses of sympathetic preganglionic neurons 24, the increases in sympathetic nerve activity ~2 and the pressor responses 3'19 e v o k e d by R V L stimulation. A n a t o m i c a l l y , glutamate-like immunoreactivity has been localized within the I M L in axon terminals making asymmetric (excitatory) synapses on dendritic processes 21. The
glutamate-like immunoreactivity in the I M L appears to originate from a supraspinal source since it is reduced caudal to spinal cord transection 21. W h e t h e r it arises from neurons in the R V L or from other brainstem nuclei innervating the I M L has not been established. To d e m o n s t r a t e directly that neurons in the R V L contribute to the gilutamatergic innervation of the I M L , we c o m b i n e d p e r o x i d a s e - a n t i p e r o x i d a s e ( P A P ) immunocytochemistry to detect a n t e r o g r a d e labeling for Phaseolus vulgaris leucoagglutinin ( P H A - L ) 11 following injections into the R V L with i m m u n o a u t o r a d i o g r a p h i c labeling for glutamate-like immunoreactivity in the I M L . We found that m a n y P H A - L immunoreactive terminals in the I M L also contained glutamate-like immunoreactivity. These results provide anatomic support for the hypothesis that regulation of sympathetic nerve discharge by reticulospinal neurons of the R V L is m e d i a t e d , at least in part, by the release of glutamate in the I M L . A preliminary r e p o r t of this work has a p p e a r e d 22.
Correspondence: S.E Morrison. Present address: Department of Physiology (Tarry 5-703), Northwestern University Medical School, 303 East Chicago Avenue, Chicago, IL 60611, U.S.A.
127 MATERIALS AND METHODS
Animal preparation Eight male Spragne-Dawley rats were anesthetized with chloral hydrate (420 mg/kg, i.p.). The occipital bone and the caudal apex of the fourth ventricle (calamus scriptorius) were exposed by retraction of the overlying muscle and cutting t h e atlanto-oeeipital membrane. A hole was drilled in the occipital bone to allow stereotaxic positioning of a mieropipette (12-15/~m tip) containing the anterograde tracer, PHA-L (2.4% in 0.1 M phosphate-buffered saline, pH 7.4, Vector Laboratories) n. Relative to the caudal apex of the fourth ventricle, iontophoretic injections of PHA-L (7/~A anodal DC; 7 s on, 7 s off for 20 rain) were made at AP + 2.6 mm, ML 1.9 mm, DV -2.2 ram. This site corresponds to the highest density of RVL-spinal sympathoexcitatory neurons identified with the antidromic activation technique 4"23.
Tissue preparation Following a 14-21 day survival period, animals were anesthetized with sodium pentobarbital (75 mg/kg, i.p.). Six rats were transeardially perfused with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) and two were perfused with 0.2% glutaraldehyde and 4% paraformaldehyde for light microscopic examination of PHA-L transport only. The medulla and the upper thoracic spinal cord were removed and postfixed for 12-24 h. For light microscopic examination of the PHA-L injection site, frozen coronal sections (40 Hm) were cut through the RVL. For examination of the single and dual labeling for PHA-L and glutamate in the IML, vibratome sections (40 /~m) were cut coronally through thoracic segments T1-T 3 and collected in 0.1 M phosphate buffer. The dual labeling procedure has been described 26'27. Briefly, all sections were rinsed in Tris-buffered saline (TBS, 0.1 M, pH 7.6), incubated in H20 2 (1% in 0.1 M TBS, 30 min), rinsed in 0.1 M TBS several times until the bubbles disappeared and incubated (18 h at room temperature) in a solution (1% bovine serum albumin and 0.05% Triton X-100 in 0.1 M TBS) containing: (1) antibody to PHA-L raised in goat (1:2000, Vector Laboratories); and (2) antibody to hemocyanin-conjugated glutamate raised in rabbit (1:4000). The glutamate antibody was obtained from Drs. P. Petrusz and A. Rustioni, University of North Carolina. The specificity of this antibody has been extensively tested 15 and its labeling characterized in the IML 21, the somatosensory cortex 6 and sensory ganglia 2. Spinal sections for dual labeling were next incubated (2 h) in 12SI-labeled anti-rabbit IgG raised in donkey (1:100, Amersham, 100 Ci/ ml) and then washed (15 rain) in TBS until negligible radioactivity was detected in the wash solution. All sections were then sequentially incubated in anti-goat IgG raised in sheep (1:100, 2 h), peroxidase-antiperoxidase (PAP) 33 raised in goat (1:200, 30 rain) and 3,3"-diaminobenzidine (0.025 and 0.003% H 2 0 2 in TBS, 6 rain). All incubations were carried out at room temperature with continuous agitation. Sections were rinsed (3 x 15 rain) between each incubation. For light microscopic examination of PHA-L injection sites and PHA-L-labeled processes in the spinal cord, sections were rinsed in phosphate buffer, mounted on glass slides, air-dried, dehydrated and coverslipped with Histoclad. For light microscopic localization of glutamate-like immunoreactivity using autoradiography, the procedure described by Cowan et al. s and Piekel et a l Y was used. Sections were mounted on acid-cleaned slides previously coated with 0.5% gelatin, air-dried and defatted (chloroform-100% ethanol (1:1,2 × 15 rain) and graded ethanol to water). Slides were dipped in Ilford LA emulsion (1:1 with water, 50 °C), dried and stored in light-proof boxes for 6-12 days at 4 °C. Slides were developed (Kodak D-19, 2 rain, 16 °C), fixed (Kodak Ektaflo, 8 rain), washed, dehydrated, cleared and coverslipped using Histoclad. Sections for light microscopy were examined with a Nikon Microphot microscope using DIC and bright field optics. For electron microscopy, labeled sections were fixed further in glutaraldehyde (1% in 0.1 M phosphate buffer, 20 rain) and osmium tetroxide (2%, 1 h), dehydrated through a graded series of
alcohols and propylene oxide, and embedded in Epon 812. U1trathin sections (50 nm) were cut from the outer surface of the flat embedded sections and deposited on slides previously coated with 2% parlodion in amyl acetate. Sections were counterstained with uranyl acetate and Reynold's lead citrate 2s, coated with a silvergray layer of carbon (Varian vacuum evaporator), dipped in Ilford L4 emulsion (1:4 with water, 50 °C) and stored in light-proof boxes for 8-12 months. Slides were developed (Kodak Microdol-X, 1-3 rain, 16 °C), fixed in sodium thiosulfate (30%, 6 min) and washed. The ultrathin sections on the parlodion coating were collected on grids and immersed in amyl acetate (3 min) prior to examination with a Philips 201 electron microscope. Electron microscopic analysis was conducted on 3-6 plastic-embedded sections (at least 6 grids from each with 3-6 thin sections/ grid) from the 4 animals with optimal PHA-L injection sites and anterograde labeling in the IML. Single- or double-labeled axon terminals were photographed at lower magnification to assess the level of background autoradiographic labeling and to determine the size of the postsynaptic target and at higher magnification to ascertain the nature of the synaptic specializations. Regarding the autoradiographic labeling of axon terminals for glutamate immunoreactivity, only those axon terminals containing one or more silver grains in at least two serial sections were photographed for further analysis. To distinguish positive, autoradiographic labeling of axon terminals for glutamate from the random positioning of grains arising from the exposure of the emulsion by background radiation, we employed the following statistical analysis based on the assumption that silver grains arising from background exposure of the emulsion are randomly distributed over any given section. For each terminal over which grains were observed in at least two serial sections, its area and that of a region of its surrounding neuropil (about 20 times greater than the area of the terminal) were measured from each of the serial electron micrographs. The size of the area of the surrounding neuropil (20-30 #m 2) was selected to include some background grains in the cases of lowest grain density. This area was then maintained in probability determinations for the other terminals examined. The ratio of the area of the terminal to that of the region of surrounding neuropil represents the probability (Pt) that a single, randomly-positioned grain would be found over the terminal. The mean value for Pt was 0.049 -+ 0.003, obtained from measurements of 92 micrographs of 38 terminals. The total number (13 of grains over the neuropil region (including those over the terminal) were counted, as were the number (t) overlying the terminal. In 63% of the eases, two or more silver grains were located over the terminal profile in at least one of the serial sections. The probability that t out of T grains would be located over the terminal by chance alone is Pr.t = k (Pt) t (1 - Pt)T-t, where k is the coefficient of the appropriate term in the binomial expansion. For the individual micrographs, Pr, t values ranged from 0.0001 for a terminal with 4 out of 6 grains positioned over it, to 0.21 for a terminal with 1 out of 7 grains over it. The mean background grain density was 12 - 2% of that over labeled terminals. The power of the serial section approach is that the probability of observing random background grains over the same terminal in serial sections is the product of those probabilities derived from measurements on each individual section. Only those terminals for which this product was less than 0.05 were considered to have a statistically significant indication of glutamate immunoreactivity and were included in this report.
P H A - L single labeling Light microscopy. I n j e c t i o n s o f P H A - L w e r e c e n t e r e d in t h e r e g i o n b e t w e e n t h e nucleus a m b i g u u s a n d t h e v e n tral surface o f t h e m e d u l l a o b l o n g a t a , i m m e d i a t e l y cau-
Fig. 1. Schematic diagrams of coronal sections through the medulla (left) approximately 2.5 mm rostral to the calamus seriptorius and through the third thoracic spinal segment (right). The outline in the medullary diagram indicates the region containing the rostral ventrolateral medulla (RVL) into which PHA-L was iontophoretically applied and which is pictured in B. The outline in the spinal diagram indicates the region of the intermediolateral nucleus (IML) which receives projections from the RVL and which is pictured in C and D. B: PHA-L injection site in the RVL. Note that the center of the injection is surrounded by darkly labeled neurons (arrows) that have taken up the PHA-L. C: immunoautoradiographic demonstration of glutamate-like immunoreactivity within the IML of the upper thoracic spinal cord. Autoradiographic exposure was 3 days. D: fine axonal processes and varicosities of the IML terminal field of RVL-spinal neurons labeled with PHA-L following the injection illustrated in B. A, nucleus ambiguus; DH, dorsal horn; LF, lateral funiculus; MV, medial vestibular nucleus; NTS, nucleus of the solitary tract; RO, raphe obscurus; RP, raphe pallidus; SV, vestibular nucleus; $5, spinal nucleus of the trigeminal; VH, ventral horn. Calibration bar is 150/tm in B and 100/~m in C and D. dal to the facial nerve nucleus (Fig. 1A,B). U p t a k e of P H A - L by several neurons within and surrounding each injection site was d e m o n s t r a t e d by their darkly-labeled p e r i k a r y a (Fig. 1B). L a b e l e d neurons were contained within the C 1 area, the region of the R V L containing neurons that project to the I M L , m a n y of which are immunoreactive for the epinephrine-synthesizing enzyme, phenylethanolamine N-methyltransferase (PNMT)30. Injections in this position, but not those outside of it, produced a dense plexus of PHA-L-labeled axons and bouton-like varicosities in the I M L of the upper thoracic spinal cord (Fig. 1D). In general, P H A - L labeling was stronger: (1) in the I M L ipsilateral to the injection site; and (2) in the rostral portions of the I M L (Cs-T3). The conditions necessary for electron microscopic colocalization of P H A - L with glutamate (i.e. fixation with paraformaldehyde only 21 and processing of the spinal sections with a low amount of Triton)
decreased the observed P H A - L labeling in the IML. Electron microscopy. Immunoreactivity for P H A - L was localized exclusively within axons and axon terminals in the I M L (Figs. 2, 4, 5). Table I describes the relationship between the type of synaptic contact m a d e by
TABLE I Dendritic size VS type of association of terminals with PHA-L immunoreactivity in IML Type of dendrite
Number of associations (n = 86) Asymmetric synapses
Large (1.5-3.5/~m) Small (0.5-1.5 pm) Spine
10 31 8
10 16 2
Fig. 2. Axon terminals in the IML containing PHA-L immunoreactivity following injections in the rostral ventrolateral medulla in sections processed for both peroxidase-antiperoxidase (PAP) labeling of PHA-L and autoradiographic labeling for glutamate. A: axon terminal (PHA-T) containing PHA-L, visualized with the PAP technique, makes an apparent asymmetric synapse (curved arrow) on a small unlabeled dendrite (uD) in the IML. Note the fine unlabeled glial processes (asterisks) enveloping the pre- and postsynaptic elements and the unmyelinated axon containing glutamate-like immunoreactivity (arrowhead) indicated by the duster of silver grains. B: PHA-L-labeled axon terminal (PHA-T) makes synaptie contacts (curves arrows) on the head and neck portions of a spinous process of a uD in the IML. Autoradiographic exposure was 9 months. Calibration bar is 0.5 #m.
PHA-L-labeled axon terminals and size of their postsynaptic dendritic targets. Labeled axon terminals made primarily asymmetric synaptic contacts on small dendrites (0.5-1.5/~m in diameter) of neurons in the IML (Figs. 2A and 4A). PHA-L-containing terminals of RVL neurons also synapsed on the head or neck portions of spinous dendritic processes (Fig. 2B) and on larger (1.53.5 #m) dendrites (Fig. 5) of IML neurons. We also observed a number of instances in which the membranes of PHA-L-containing axon terminals and dendritic processes in IML were apposed but did not exhibit synaptic specializations in the plane of section studied (Table I). Labeled terminals ranged from 0.5 to 1.5 #m in diameter, contained many small clear vesicles and in addition, in some instances of light labeling, 1-3 dense-core vesicles were seen. The population of terminals labeled in IML following P H A - L injections in the R V L is likely to include those containing PNM'T 9'14'20'29. The differences between the relative distribution of the terminals of RVL neurons (Table I) and that of PNMT immunoreactive
terminals in the IML (Table I in ref. 20) suggests, however, the existence of at least one, non-PNMT-containing terminal population labeled following P H A - L injections in the RVL.
Glutamate single-labeling Light microscopy. Silver grains indicating the presence of glutamate-like immunoreactivity were concentrated over the region of the IML (Fig. 1C) and over the outer laminae of the dorsal horn of the thoracic spinal cord. This localization corresponds to that seen using the PAP technique 21.
Electron microscopy. Only those neuronal profiles over which silver grains were found in at least two serial sections and for which statistical analysis yielded a significant indication of glutamate immunoreactivity were considered to be positively labeled for glutamate (Figs. 3-5). Although only 1-4 grains might be found over a labeled terminal in each section, positively labeled terminals were usually followed through 3-6 serial sections,
Fig. 3. Immunoautoradiographic identification of a glutamate-immunoreactive terminal in the IML of the upper thoracic spinal cord. A,B: serial sections through the same axon terminal (GLU-T) making an asymmetric (excitatory) synapse with a small uD in the IML. Autoradiographic exposure was 10 months. Calibration bar is 0.4/~m.
particularly near the surface of the e m b e d d e d tissue with more access to i m m u n o r e a g e n t s and consequently m o r e intense labeling. A total of 38 axon terminals with statistically significant ( P < 0.05) autoradiographic indication of glutamate immunoreactivity were identified, of which 13 also contained the P H A - L label. Within the I M L , silver grains indicating glutamatelike immunoreactivity were primarily found over axon terminals that m a d e asymmetric synapses (Figs. 3 A , B , 4 A and 5) with small ( 0 . 5 - 1 . 5 / ~ m ) dendritic processes. Glutamate-containing terminals were p o p u l a t e d by small clear vesicles and several mitochondria (Figs. 3 and 4C, D), but rarely contained dense core vesicles. The localization and morphological characteristics of the axon terminals labeled for glutamate using immunoautoradiography paralleled those we have described 21 using the peroxidase technique.
Glutamate and PHA-L colocalization In optimally labeled sections through the I M L , colocalization of glutamate-like and P H A - L immunoreactivities was found in 13 axon terminals (Figs. 4 and 5). As expected from the single-labeling results, dual-labeled terminals were most often in contact with small (0.5-1.5 ~tm) dendrites in the I M L (Fig. 4). While occasional synapses on larger ( 1 . 5 - 3 . 5 / ~ m ) dendrites were seen (Fig. 5), no axosomatic contacts were observed. Most of the synaptic contacts could be classified as asymmetric (Fig. 4 A ) , however, the morphological preservation with p a r a f o r m a l d e h y d e fixation sometimes p r e v e n t e d distinct characterization of m e m b r a n e specializations (Fig. 5A, B). A s shown in Fig. 4 A , B , an additional factor in the identification of synaptic association was the potential for variations in the intensity of the postsynaptic density b e t w e e n different planes of section. In coronal sections, most dendritic processes were
Fig. 4. Colocalization of PHA-L and glutamate immunoreactivities in the same axon terminals in the IML. A,B: serial sections through an axon terminal (GLU + PHA-T) in the IML labeled with PHA-L following an injection in the RVL. Note the silver grains (arrowheads) over the terminal indicating that it is also immunoreactive for glutamate. This terminal makes an asymmetric synapse (curved arrow in A) with a small uD in the IML. The random silver grain over the uD in A is considered background since the same dendrite was not labeled in other sections (B). C,D: serial sections through an axon terminal (GLU + PHA-T) in the IML which is lightly labeled with PHA-L following an injection in the RVL. Silver grains (arrowheads), indicating glutamate-like immunoreactivity, were consistently found to overlie this terminal in two additional serial sections. The surrounding neuropil and postsynaptic dendrite were devoid of immunolabeling. Autoradiographic exposure was 11 months. Calibration bar is 0.4/~m.
133 Fig. 5. Serial sections through a large, glutamate-containing axon terminal (GLU + PHA-T) of a reticulospinal neuron in the RVL which forms a synapse on a medium-size uD in the IML. In each section, this terminal is heavily labeled with the peroxidase reaction product indicating PHA-L immunoreactivity. The localization of silver grains (arrowheads) over these and several additional serial sections demonstrated that this terminal also contained glutamate. The synaptic contact (curved arrow) which this terminal makes with the medium-size dendrite in the IML, although not clearly defined, appears to be of the asymmetric type. Note that an additional PHA-L-labeled terminal (PHA-T) is apposed to the same dendrite (small arrows in A), but is not labeled for glutamate since grains were not consistently seen overlying this terminal or in the surrounding neuropil (B and C). Autoradiographic exposure was 11 months. Calibration bar is 0.4/~m for A, and 0.5/~m for B and C.
contacted by only one glutamate-containing axon terminal (Figs. 4 and 5). Although multiple PHA-L labeled terminals could be apposed to larger dendrites, only one would also contain detectable immunoautoradiographic labeling for glutamate (Fig. 5). DISCUSSION In the present study, we used an anatomical approach to test the hypothesis that the reticulospinal projection from the R V L to the IML is glutamatergic. By combining P H A - L anterograde labeling of the RVL-spinal projection with immunocytochemical detection of glutamate, we have demonstrated that the axon terminals of RVL neurons that project to the IML contain glutamate-like immunoreactivity and make excitatory (asymmetric) contacts upon local dendrites. While, as with most immunocytochemical studies, the nature of the antigen should be interpreted cautiously (conceivably the antibody could be identifying an unspecified glutamate-like molecule or a larger molecule with a terminal glutamate), the glutamate antibody used here has been extensively examined for its specificity as. Moreover, axon terminals labeled in the dorsal horn TM and in the hippocampus 7 with different antibodies to L-glutamate conjugates exhibit similar characteristics to those we have found in the IML. Given this caveat, our results provide direct evidence that reticulospinal neurons in the R V L contribute to the glutamatergic innervation of the sympathetic nucleus in the IML. They also provide f u r t h e r support 3'12"t9'24'25 for the view that glutamate is the excitatory neurotransmitter released by RVL-spinal neurons to increase sympathetic nerve activity and arterial pressure. The results of the present study demonstrate that P H A - L anterograde transport and immunoautoradiographic techniques can be combined to obtain information on the potential transmitters within specific projection pathways. In the present case, however, our ability to observe dual-labeled terminals was likely compromised since the fixation conditions necessary to obtain glutamate labeling were not optimal for immunocytochemical detection of the P H A - L anterograde tracer. As emphasized in this and other studies using 125I-1a-
beled IgG 2°'27, only those processes over which silver grains were observed in more than one serial section were considered to be labeled for the antigen. The low sensitivity of the technique and the potential for spread of radiation away from the source 31'32 make this criterion difficult to achieve in the case of smaller axons and axon terminals. In biological tissue fixed with osmium tetroxide and stained with uranyl acetate, the resolution or half-distances for 125I-labeled specimens using I1ford-L4 emulsion is approximately 80 nm 32. Thus, even small (0.1 /~m) glutamate-containing axon terminals would be sufficiently large to permit tentative identification of the source of the radiation in single sections. Indeed, statistical analysis confirmed the glutamate labeling in axon terminals tentatively identified through examination of serial sections. The colocalization within axon terminals in the IML of glutamate-like immunoreactivity and P H A - L anterogradely transported from neurons in the RVL, extends our previous observation zl from spinal cord transection, that glutamate-like immunoreactivity in the IML is derived primarily from supraspinal sources. In contrast, glutamate labeling in the dorsal horn appears to have a significant component of segmental origin 21, possibly arising from glutamate-containing neurons in dorsal root ganglia 2. While the present study demonstrates directly that some of the suprasegmental glutamatergic input to the IML arises from neurons of the RVL, it does not eliminate the possibility that other brainstem nuclei may also contribute to the glutamatergic innervation of the IML. The finding that P H A - L injections into the RVL labeled many axon terminals in the IML which made asymmetric synaptic contacts with local dendrites is consistent with the physiological data indicating that RVL neurons projecting directly to the IML are sympathoexcitatory 1'4'14'17'23 and contribute substantially to the maintenance of tonic sympathetic tone and resting arterial pressure 1°'3°. In this regard, synapses with asymmetric (Gray type I) specializations have been proposed to mediate excitation 5'35. The demonstration that some of these PHA-L labeled terminals also contained glutamate-like immunoreactivity strongly suggests that gluta-
134 mate mediates the sympathoexcitatory effects of at least a population of RVL-spinal neurons. In support of this view: (1) glutamate produces an exclusively excitatory action on m a m m a l i a n neurons16; (2) glutamate receptor antagonists block the excitation of sympathetic preganglionic neurons 24 and the increases in arterial pressure 3" 12.19 evoked by R V L stimulation; (3) glutamate receptors of the kainate subclass are concentrated in the I M L 2s. Since the majority of the neurons in the R V L that project to the I M L contain the epinephrine-synthesizing enzyme, P N M T 9'29, and are presumably sympathoexcitatory 14'23, the question arises whether glutamate and catecholamines coexist within the same terminals or are derived from different populations of RVL-spinal neurons. The characteristics of glutamate-containing terminals in the I M L are similar to those of the majority of terminals containing PNMT: they form asymmetric synapses on small dendrites 2°'21. The principal differences arise from the fact that a significant n u m b e r of P N M T containing terminals: (1) m a k e symmetric synapses on proximal dendrites and on cell bodies of I M L neurons; and (2) contain dense-core vesicles 2°. In the present study, P H A - L labeled terminals making symmetric contacts with dendrites in the I M L (see Table I) may have arisen from PNMT-containing neurons in the R V L .
REFERENCES 1 Barman, S.M. and Gebber, G.L., Axonal projection patterns of ventrolateral meduliospinal sympathoexcitatory neurons, J. Neurophysiol,, 53 (1985) 1551-1566. 2 Battaglia, G. and Rustioni, A., Coexistence of glutamate and substance P in dorsal root ganglion neurons of the rat and monkey, J. Comp. Neurol., 227 (1988) 302-312. 3 Bazil, M.K. and Gordon, F.J., Spinal N-methyl-o-aspartic acid (NMDA) receptors mediate pressor and tachycardic responses evoked by excitation of the rostral ventrolateral medulla (RVM), Soc. Neurosci. Abstr., 15 (1989) 965. 4 Brown, D.L. and Guyenet, P.G., Cardiovascular neurons of brain stem with projections to spinal cord, Am. J. Physiol., 247 (1984) R1009-R1016. 5 Cohen, R.S., Carlin, R.K., Grab, D.J. and Sukevitz, P., Phosphoproteins in postsynaptic densities, Prog. Brain Res., 56 (1982) 49-76. 6 Conti, E, Rustioni, A., Petrusz, P. and Towle, A.C., Glutamate-positive neurons in the somatic sensory cortex of rats and monkeys, J. Neurosci., 7 (1987) 1887-1901. 7 Cotman, C.W., Monaghan, D.T., Ottersen, O.P. and StormMathisen, J., Anatomical organization of excitatory amino acid receptors and their pathways, Trends Neurosci., 10 (1987) 273279. 8 Cowan, W.M., Gottlieb, D.I., Hendrickson, A., Price, J.L. and Woolsey, T.A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 37 (1972) 21-51. 9 Cravo, S., Ruggiero, D.A., Anwar, M. and Reis, D.J., Quantitative-topographic analysis of adrenergic and non-adrenergic spinal projections of cardiovascular area of RVL, Soc. Neurosci. Abstr., 14 (1988) 328.
Thus, while PNMT-containing axon terminals making asymmetric (excitatory) synapses in the I M L cannot be readily distinguished morphologically from those containing glutamate, it is unlikely that glutamate is colocalized with P N M T in that subset which makes inhibitory synapses on I M L neurons. In conclusion, a fraction of the glutamate-like immunoreactivity in the I M L is localized within the axon terminals of reticulospinal neurons of the R V L which may use this excitatory amino acid in mediating the excitation of sympathetic preganglionic neurons that maintains basal sympathetic tone and arterial pressure. The extent to which glutamate is colocalized with other neurotransmitters in the RVL-spinal pathway and whether glutamate-containing terminals synapse directly on sympathetic preganglionic neurons or on local interneurons in the I M L remain to be determined.
Acknowledgements. The authors wish to thank Drs. P. Petrusz (Department of Cell Biology and Anatomy, University of North Carolina at Chapel Hill) and A. Rustioni (Department of Cell Biology and Anatomy and Department of Physiology, University of North Carolina at Chapel Hill) for generously providing their glutamate antibody and Dr. Virginia M. Pickel for helpful comments on the manuscript. This research was supported by National Institutes of Health Grant HL 18974.
10 Dampney, R.A.L. and Moon, E.A., Role of ventrolateral medulla in vasomotor response to cerebral ischemia, Am. J. Physiol., 139 (1980) H349-H358. 11 Gerfen, C.R. and Sawchenko, P.E., An anterograde neuroanatomical tracing method that shows the detailed morphology of neurons, their axons and terminals. Immunohistochemical localization of an axonally transported plant lectin, Phaseolus vulgaris, Leucoagglutin (PHA-L), Brain Research, 290 (1984) 219-238. 12 Guyenet, P.G., Sun, M.-K. and Les Brown, D., Role of GABA and excitatory amino acids in medullary baroreflex pathways. In J. Ciriello, ER. Calaresu, L.P. Renaud and C. Polosa (Eds.), Organization of the Autonomic Nervous System: Central and Peripheral Mechanisms, Liss, New York, 1987, pp. 215-225. 13 Guyenet, P.G. and Young, B.S., Projections of nucleus paragigantocellularis lateralis to locus coeruleus and other structures in rat, Brain Research, 406 (1987) 171-184. 14 Haselton, J.R. and Guyenet, EG., Electrophysiological characterization of putative C1 adrenergic neurons in the rat, Neuroscience, 30 (1989) 199-214. 15 Hepler, J.R., Toomim, C.S., McCarthy, K.D., Conti, E, Battaglia, G., Rustioni, A. and Petrusz, P., Characterization of antisera to glutamate and aspartate, J. Histochem. Cytochem., 36 (1988) 13-22. 16 Mayer, M.L. and Westbrook, G.L., The physiology of excitatory amino acids in the vertebrate central nervous system, Progr. Neurobiol., 28 (1987) 197-276. 17 McAllen, R.M., Identification and properties of sub-retrofaciai bulbospinal neurones: a descending cardiovascular pathway in the cat, J. Auton. Nerv. Sys., 17 (1986) 151-164. 18 Miller, K.E., Clements, J.R., Larson, A.A. and Beitz, A.J., Organization of glutamate-like immunoreactivity in the rat superficial dorsal horn: light and electron microscopic observa-
135 tions, Synapse, 2 (1988) 28-36. 19 Mills, E.H., Minson, J.B., Pilowsky, EM. and Chaimers, J.P., N-methyl-D-aspartate receptors in the spinal cord mediate pressor responses to stimulation of the rostral ventrolateral medulla in the rat, Clin. Exp. Pharmacol. Physiol., 15 (1988) 147-155. 20 Milner, T.A., Morrison, S.E, Abate, C. and Reis, D.J., Phenylethanolamine N-methyltransferase-containing terminals synapse directly on sympathetic preganglionic neurons in the rat, Brain Research, 448 (1988) 205-222. 21 Morrison, S.E, CaUaway, J., Milner, T.A. and Reis, D.J., Glutamate in the spinal sympathetic intermediolateral nucleus: localization by light and electron microscopy, Brain Research, 503 (1989) 5-15. 22 Morrison, S.E, Callaway, J., Milner, T.A. and Reis, D.J., Glutamate is contained in axon terminals of neurons in the rostral ventrolateral medulla that project to the spinal sympathetic intermediolateral nucleus, Soc. Neurosci. Abstr., 15 (1989) 965. 23 Morrison, S.E, Milner, T.A. and Reis, D.J., Reticulospinal vasomotor neurons of the rat rostral ventrolateral medulla: relationship to sympathetic nerve activity and the C1 adrenergic cell group, J. Neurosci., 8 (1988) 1286-1301. 24 Morrison, S.E and Reis, D.J., Glutamate receptor antagonist blocks the responses of sympathetic preganglionic neurons (SPN) to stimulation in the C1 area of the rostrai ventrolateral medulla (RVL), Soc. Neurosci. Abstr., 13 (1987) 808. 25 Morrison, S.E, Ernsberger, E, Milner, T.A., Callaway, J., Gong, A. and Reis, D.J., A glutamate mechanism in the intermediolateral nucleus mediates sympathoexcitatory responses to stimulation of the rostral ventrolateral medulla. In J. Ciriello, M.M. Caverson and C. Polosa (Eds.), The Central Neural Organization of Cardiovascular Control, Progress in Brain Research, Vol. 81, Elsevier, Amsterdam, 1989, pp. 159-169. 26 Pickel, V.M. and Beaudet, A., Combined use of autoradiography and immunocytochemical methods to show synaptic interactions between chemically defined neurons. In J. Polak and
31 32 33 34
Varnell (Eds.), lmmunolabeling for Electron Microscopy, Elsevier, Amsterdam, 1984, pp. 259-265. Pickel, V.M., Chan, J. and Milner, T.A., Autoradiographic detection of 125I-secondary antiserum: a sensitive light and electron microscopic method compatible with peroxidase immunocytochemistry for dual localization of neuronal antigens, J. Histochem. Cytochem., 34 (1986) 707-718. Reynolds, E.S., The use of lead citrate at high pH as an electron-opaque stain in electron microscopy, J. Cell Biol., 17 (1963) 208. Ross, C.A., Ruggicro, D.A., Joh, T.H., Park, D.H. and Reis, D.J., Rostral ventrolatcral medulla: selective projections to the thoracic autonomic cell column from the region containing C1 adrenaline neurons, J. Comp. Neurol., 228 (1984) 168-185. Ross, C.A., Ruggiero, D.A., Park, D.H., Joh, T.H., Sved, A.E, Fernandez-Pardal, J., Saavedra, J.M. and Reis, D.J., Tonic vasomotor control by the rostral ventrolateral medulla: effect of electrical or chemical stimulation of the area containing C1 adrenaline neurons on arterial pressure, heart rate, and plasma catecholamines and vasopressin, J. Neurosci., 4 (1984) 474-494. Salpeter, M.M., Backman, L. and Salpeter, E.E., Resolution in electron microscopic autoradiography, J. Cell Biol., 41 (1969) 1-20. Salpcter, M.M., McHcnry, EA. and Salpeter, E.E., Resolution in electron microscopic autoradiography IV. Application to analysis of autoradiographs, J. Cell Biol., 76 (1978) 127-145. Sternberger, L.A., Immunocytochemistry, Wiley, New York, 1979. Strack, A.M., Sawyer, W.B., Hughes, J.H., Platt, K.B. and Loewy, A.D., A general pattern of CNS innervation of the sympathetic outflow demonstrated by transneuronal pseudorabies viral infections, Brain Research, 491 (1989) 156-162. Uchizono, K., Characterization of excitatory and inhibitory synapses in the CNS of the cat, Nature, 207 (1965) 642-643.