Revelation in the Rat Brainstem. BARBARA E. JONES. Department of Neurology and Neurosurgery, McGill University, Neuroanatomy Laboratory. Montreal ...
Brain Research Bulletin, Vol. 9, pp. 33-44, 1982.Printed in the U.S.A.
Relationship Between Catecholamine Neurons and Cerebral Blood Vessels Studied by Their Simultaneous Fluorescent Revelation in the Rat Brainstem BARBARA E. JONES Department
of Neurology and Neurosurgery, McGill University, Neuroanatomy Laboratory Montreal Neurological Institute, Montreal, Quebec, Canada H3A 2B4
JONES, B. E. Relationship revelation
between catecholamine neurons and cerebral blood vessels by their simultaneousjluorescent in the rut brainstem. BRAIN RES. BULL. 9(1-6) 33-44, 1982.-In order to study the relationship between
catecholamine neurons and cerebral blood vessels, a technique was developed which permitted the simultaneous visualization of blue-green fluorescent catecholamine neurons and red fluorescent stained blood vessels in the brain of the normal rat. Sympathetic nerve fibers were found on the major arteries and in the pia-arachnoid at the base of the brainstem and also along paramedial and lateral perforating arteries as small as l&12 pm within the brain. Running within the parenchyma, central catecholamine nerve fibers occasionally approached and intersected smaller blood vessels, either arterioles or venules of 8-12 pm, and infrequently climbed along or encircled these vessels for a limited distance, particularly within the lateral tegmentum. Across the nuclei of the brainstem, no overall contiguity of catecholamine terminals with capillaries was apparent, and no correlation between the density of catecholamine varicosities and that of capillaries existed. Only in regions with a high density of both catecholamine varicosities and capillaries, such as in the principal olivary nucleus, did a significant overlap of the two occur. But in most cases of moderately to densely innervated and vascularized regions, such as the solitary tract nuclei, the greatest concentration of terminals appeared over the parenchyma. Regarding the blood supply to the catecholamine neurons, their perikarya did not receive a particularly dense capillary supply relative to other nuclei. However, a special relationship of catecholamine cells to blood vessels was suggested, particularly in the case of dopamine neurons in the substantia nigra by the close apposition of cellular processes to adjacent small vessels. This morphological study was undertaken to determine whether central catecholamine neurons may significantly innervate cerebral blood vessels and accordingly, may function analogously to the peripheral sympathetic adrenergic neurons in the regulation of the vascular system. Although a limited number of associations between central catecholamine nerve tenninals and small blood vessels suggested the possibility of an innervation in a few regions, the lack of an overah correspondence and correlation between the two across brainstem nuclei indicated that the analogy of central catecholamine neurons to the sympathetic nervous system was inappropriate. On the other hand, evidence of contact with vessels by presumed dendrites of the catecholamine neurons suggested a possible vascular sensory function. Catecholamine
neurons
Cerebral blood vessels
Morphology
THE catecholamine neurons of the brain, and particularly the noradrenaline locus coeruleus neurons, project widely throughout the central nervous system [26,49] (see [32,33] for review) and modulate multiple and diverse functions, including the sleep-waking cycle, motor activity, eating and drinking, temperature control and cardiovascular regulation [5, 25, 27, 501. Such an ubiquitous structural and functional influence suggests that the catecholamine neurons serve a very general role in the central nervous system. Indeed, Raichle, Hartman, Eichling and Sharpe [39] suggested that central adrenergic neurons may be analogous to the peripheral sympathetic nervous system, and accordingly may produce regional changes in cerebral blood flow and cerebral vascular permeability. This primarily metabolic function of catecholamine neurons would explain their ubiquitous pres-
Copyright
0 1982 ANKHO
International
Brainstem
ence and generalized function in diverse systems of the brain and spinal cord. The original suggestion that central noradrenaline neurons may represent a sympathetic system originated from the demonstration by Hartman, Zide and Udenfriend [24] of an apparent innervation of small arteries by dopamine-Bhydroxylase immunoreactive fibers within the hypothalamus. The proximity to and overlap by central adrenergic fibers of small blood vessels in the hypothalamus had been and was subsequently noted by other investigators, who however questioned the significance of this structural association [16,18]. Only by ultrastructural proof of the close apposition of dense-core vesicle-containing varicosities to capillary endothelial cells and pericytes in the hypothalamus of Shydroxydopamine pretreated rats [47] came acceptance
Inc.-
0361-9230/82/070033-12$03.0010
JONES
34
of the idea that small in&a-parenchymal blood vessels, particularly capillaries, may be innervated by central adrenergic neurons (see [37,40]). Whereas it was originally believed that this innervation of blood vessels in the hypothalamus derived from noradrenaline locus coeruleus neurons [39], it was subsequently shown that the perivascular nerves originated from other brainstem catecholamine cell groups as well [29]. If the central catecholamine neurons were analogous to the sympathetic nervous system in the regulation of the cerebral vascular system, they would be expected to innervate blood vessels and capillaries throughout the brain and not only in the hypothalamus. Such a far reaching analysis to investigate this principle is not practicable at the electron microscopic level. Due to the difftculty of visualizing unstained vessels and capillaries in hlstofluorescent material, a general appreciation of the relationship between catecholamine neurons and blood vessels, particularly capillaries, has not been possible with the fluorescent technique up to this time. For this reason, a method was developed which permitted the simultaneous visualization of blue-green catecholamine neurons and red stained blood vessels by histofluorescence. With this technique, a morphological evaluation of the relationship between catecholamine neurons and cerebral blood vessels was undertaken through the cervical spinal cord and entire brainstem of the normal rat. METHOD
The histochemical procedure was developed by combining the technique for catecholamine histofluorescence [4, 17, 301 with the perfusion or injection of Evans blue for the staining of blood vessels [8, 23, 361. A similar approach has been developed independently by McGinty, Koda and Bloom [3 11. The histochemical procedure was performed on normal male Sprague Dawley rats (125-150 g) which were anesthetized with Sodium pentobarbital(60 mg/kg) for transcardiac perfusion or decapitation. For catecholamine fluorescence, a two step perfusion of ice cold solutions was performed at an initial rate of 100 ml per minute by placing the perfusion bottles at approximately 3 feet above the animal. In accordance with a procedure established for the cat sympathetic ganglion by Birks [2], the perfusion solutions contained a high concentration of Mg++ which has been shown to prevent transmitter release caused by perfusion with aldehyde fixatives. The rats were perfused
for -60 sets with the first solution that consisted of 20 mM MgC&, 4 mM KC1 in 10 mM Hepes Buffer (Sigma Chemical Co.) with 2% Glyoxylic Acid (Sigma Chemical Co.) (pH adjusted to 7.5 with sodium hydroxide) and perfused for 120 seconds with the second solution that contained 4% paraformaldehyde in addition to the constituents of the first solution (pH adjusted to 7.5 with sodium hydroxide). The brain was rapidly removed and dissected, frozen on a cryostat chuck by dry CO* gas and placed in powdered dry ice for l-2 days. Twenty pm thick sections were cut at -30°C on an American Optical Cryostat and mounted by contact onto clean glass slides held at room temperature. The sections were immediately dipped into the first perfusion solution containing 2% Glyoxylic Acid (cooled to 4”C), dried under a stream of 37°C air with a hair dryer, and reacted under paraffin oil at 100°C in an oven for 5 minutes before coverslipping. Adjacent sections were dipped in 0.5% thionine stain (in 20% ethanol), rinsed in water, dried and coverslipped with pan&in oil. For the simultaneous visualization of blood vessels, Evans blue stain which fluoresced red under ultraviolet light was found to be compatible with catecholamine fluorescence. The stain could also be carried through the same histochemical procedure without loss or diffusion of the dye. Three different techniques for staining vessels with Evans blue were tested in this study. First, the free dye was added to the second perfusion solution in a concentration of 2%. A volume of 100-150 ml of the solution was delivered in one minute. Second, the dye was coupled to bovine serum albumin (Cohn, Fraction II, Sigma Chemical Co.) and added in concentrations of 1% dye and 5% albumin to the second perfusion solution. In this case, not more than 50 ml of the solution were perfused in one minute. Third, the Evans blue albumin complex was dissolved in saline and injected (0.1 ml per 10 g body weight) into the tail vein of the rat (125-150 g) 30 min prior to decapitation [36]. In this case, the animals were not perfused and the unperfused brains were processed for histofluorescence in the same manner as the perfused brain described above. Fluorescent slides were viewed under a Leitz Dialux 20 microscope equipped with incident fluorescence. Leitz filter block D for Ploemopak was employed for excitation with ultraviolet and violet light (355-425 Band pass filter) and emission of blue-green and longer wavelength light (460 long-pass suppression filter). The slides were photographed
FIG. 1. Catecholamine histofluorescence in a coronal section through the medulla. Sympathetic fibers surround the basilar artery and penetrate the pia mater where they form a plexus around the base of the brainstem. Medially, central catecholamine fibers arrive from the lateral tegmentum and ramify within the raphe pallidus nucleus. Calibration bar: 100 pm. B-F. Combined fluorescence of catecholamine neurons and vessels stained by perfusion with Evans blue dye. B. Saggital section through the base of the medulla at the level of the raphe pallidus nucleus. A dense plexus of adrenergic fibers within the pia-arachnoid and surrounding the perforating arteries is evident. The large penetrating artery measures approximately 60 pm across. The basilar artery borders the section ventrally. Calibration bar: 100 pm. C. Coronal section through the raphe pallidus nucleus showing paramedian perforating vessels at the base of the brain. Catechohunine varicose fibers follow along perforating blood vessels into the brain. Calibration bar: 100 pm. D. Horizontal section through the lateral medulla in the region of the lateral funiculus and lateral reticular nucleus. Sympathetic fibers form a plexus at the pial surface of the brain and appear to follow along perforating blood vessels into the brain. Calibration bar: 50 pm. E. Catecholamine varicose fibers surrounding a vessel of approximately 7-9 pm in width within the ventrolateral tegmentum of the medulla. Calibration bar: 25 pm. F. Catecholamine varicose fibers within the dorsolateral pontine tegmentum, near the locus coeruleus. Some fibers and/or varicosities intersect and/or overlap blood vessels (- 6 pm thick) but most are concentrated over the brain parenchyma.
CATECHOLAMINE
NEURONS/CEREBRAL
BLOOD VESSELS
35
FIG. 2. Combined fluorescence of catecholamine neurons and Evans blue stained blood vessels. A. Coronal section through the inferior olivary nuclei characterized by a dense catecholamine terminal plexus. Although the capillary supply is homogeneously dense throughout the olivary complex, the catecholamine varicosities are most concentrated over the principal nucleus. Calibration bar: 100 pm. B. High power photomicrograph through the principal inferior olivary nucleus (pars dorsalis) where catecholamine varicosities are superimposed though not concentrated over capillaries (-6 pm thick). Calibration bar: 25 pm. C. Coronal section through the locus coeruleus nucleus, adjacent to the IV ventricle, and the mesencephalic trigeminal nucleus located laterally, visible as a collection of large yellow autofluorescent soma. It is apparent that the capillary density is high in the locus coeruleus but as high in the mesencephalic trigeminal nucleus. A slight diffusion of dye is evident surrounding the ventricle. Calibration bar: 100 pm. D. High power photomicrograph of the autofluorescent mesencephalic trigeminal perikarya which are surrounded by capillaries. In this case, dye has diffused slightly from the vessels to enshroud the cell soma of this nucleus. Calibration bar: 2.5 pm.
with the Wild Photoautomat MS 55 utilizing Kodak Ectachrome 160 Professional Film (Tungsten). The analysis of the association of catecholamine neurons with blood vessels was based on material from 20 rats perfused with the free dye, 4 rats perfused with the conjugated dye and 4 rats injected with the conjugated dye. The photomicrographs were all taken from material perfused with 2% Evans blue dye. RESULTS
Sympathetic nerve fibers associated with the major cerebral arteries, pial vessels and large penetrating intraparenchymal arteries were evident in material treated uniquely for catecholamine fluorescence (Fig. ]A). The relationship between central catecholamine fibers and small intra-parenchymal vessels could only be appreicated in material treated for combined fluorescence of catecholamine neurons and blood vessels (Fig. IB-F).
The most intense and complete staining of blood vessels was achieved by perfusion with free Evans blue dye. In most regions, the Evans blue stained the vessel walls without any diffusion into the surrounding parenchyma (see Fig. 2B and 2C). The meninges, the stroma and ependyma of the choroid plexus and the area postrema were usually moderately to intensely stained. The ependyma of the fourth ventricle was lightly to moderately stained. Occasionally with cases of ubiquitously intense staining, the dye extended into the periventricular gray (Fig. 2C). Finally in one group of cells, the mesencephalic trigeminal neurons, evidence of extravasation of the dye appeared in certain animals with very intense staining (Fig. 2C and D). Perfusion with conjugated Evans blue produced similar staining of the vessel walls but with less intensity. In this case, the meninges were also darkly stained but the ependyma of the choroid plexus and fourth ventricle and the area postrema were very lightly stained. Many of the mesencephalic trigeminal cells appeared to be partially encircled by capillaries, the walls of which were sharply defined, in con-
CATECHOLAMINE
NEURONS/CEREBRAL
BLOOD VESSELS
FIG. 3. Combined fluorescence of catecholamine
neurons and Evans blue stained blood vessels in the region of the substantia nigra and the neostriatum. A. Horizontal section through the dopamine perikarya whose processes appear to surround small blood vessels in between myelinated fiber bundles. Calibration bar: 25 pm. B. Varicose catecholamine fibers envelop a small blood vessels (-6 pm) near the dopamine perikarya seen in sagittal section. Calibration bar: 25 pm. C. High power photomicrograph of coronal section showing close association of dopamine perikarya and their processes with a small blood vessel (-110 pm). The cell soma measures approximately 15x30 pm. Calibration
bar: 10 pm. D. Horizontal section through the neostriatum showing dopamine varicosities overlying blood vessels (-6 pm) between the myelinated fascicles of the internal capsule. Calibration bar: 25 Frn
trast to those visualized in some cases involving free dye perfusion. Injection of Evans blue albumin resulted in a light red coloration visible within the lumen of the larger vessels. The meninges were lightly stained. The dye complex did not color the stroma or ependyma of the choroid plexus, the ventricular ependyma or the area postrema. Vessels containing dye were visible surrounding many of the mesencephalic trigeminal neurons. In general, however, the staining was not sufficiently intense for a clear histological picture of vessels. Association
of Catecholamine
Nerves
with Blood Vessels
Catecholamine fluorescent fibers with thick varicosities were evident surrounding the vertebra1 and basilar arteries (Fig. 1A). Continuous with these sympathetic nerves, fibers extended into the pial membrane surrounding the base of the medulla where they formed a plexus (Fig. 1A and IB). Along the midline fibers with similarly large varicosities followed along the paramedial perforant arteries into the brain through the raphe pallidus and raphe obscurrus nut. (Fig. 1C).
Within these nuclei smaller varicosities derived from central catecholamine neurons were densely distributed over raphe neurons. Thick varicose axons were also observed directly alongside paramedial perforating arteries running dorsally up to the periventricular gray, and also along smaller lateral branches of the arteries (2&10 pm). Ventrolaterally, thick varicose fibers emerged from the pial surface and followed arteries into the brain in the region of the lateral funiculus where within the interstitial plexus catecholamine fibers were present in close association with vessels (Fig. 1B). Within the parenchyma radially oriented perforating arteries (measuring from 35 pm down to 10 pm) were accompanied by thick varicose fibers running directly alongside. These thick varicose fibers were believed to be of sympathetic origin. Within the tegmentum, particularly in the lateral portion, smaller vessels of 6-12 pm in diameter were often observed in association with thin varicose catecholamine fibers of presumed central origin, running in the same direction (Fig. 1E and F). Occasionally a fiber appeared to climb like a vine along a small blood vessel for a certain distance (up to 0.8
38
JONES TABLE I DENSITY
ESTIMATES
OF CAPILLARIES AND CATECHOLAMINE BRAINSTEM NUCLEf*
Capillaries Sensory Nuclei Dorsal Horn V Spinal Nut. (Trigeminal) V Principal Sensory Nut. (Trigeminal) V Mesencephalic Nut. (Trigeminal) Gracile Nut. Cuneate Nut. Vestibular Nut., Med. Vestibular Nut., Lat. Vestibular NW., Sup. Cochlear NW., Dor. Cochlear NW., Vent. Trapezoid NW. Superior Olivary Complex Lateral Lemniscus Nut. Inferior Colliculus Superior Colliculus
3 2 2 2 1 I 0.25 0.5 0.5 3 2.5 0.1 0.25 0.5 2
i 1.35 k 1.00 4 5 5 5
3.5 4 4 2 3.38 I*I0.95
4.75 t 0.503
Motor Nuclei Ventral Horn XII Nut. (Hypoglossal) VII Nut. (Facial) VI Nut. (Abducens) V Motor Nut. (Trigeminal) IV Nut. (Troclearf III Nut. (Ocuiomotor)
3 3 3 1 3 1 0.5 2.86 -r 0.38t
Reticular Formation Lateral Tegmental Field (Medulla) Gigantocellular Tegmental Field (Medulla) Magnocellular Tegmental Field (Medulla) Lateral Tegmental Field (Pans) Gigantocellular Tegmental Field (Pans) Lateral Tegmental Field (Midbrain) Central Tegmental Field (Midbrain) Cuneiform Nut.
3 ? 3 3 2 3 2 2 2.50 I 0.53*
Cerebellar Associated Nuclei Inferior Olivary Nut. Principal Dors. Inferior Olivary Nut. Principal Vent. Inferior Olivary NW. Accessory Dors. Inferior Olivary NW. Accessory Med. Lateral Reticular Nut. Tegmental Reticular NW. Pontine Nut. Red Nut.
IN
Varicosities
3 3 4 5 3 3 5 4 4 5 5 5 5 5 4 4 4.19 + 0.83
Visceral Nuclei X Nut. (Dorsal Motor Nut. Vagus) Solitary Tract Nut. Commissural NW. Ambiguous Nut.
VARICOSITIES
5 4 4 4 3.5 3 4 4 3.94 t 0.56
2.07
5
I.17
3
1 3 4 I 4 3
I 2.50 i
1.31$
5 4
2
I 2 2 3
I 2.50 2 1.411
CATECHOLAMINE
NEURONS/CEREBRAL
BLOOD VESSELS TABLE
39
1
CONTINUED
Capillaries Other Nuclei Periventricular Gray (Pans) Dorsal Tegmental NW. Ventral Tegmental NW. Parabrachial Nut. Med. Periaqueductal Gray Interpeduncular Nut.
Varicosities
0.5 0.1 4 4 5 2 4.17 f 0.98
2.60 lr 2.03
2.67 2 0.82t
3.00 + 0.89$
3.60 * 0.89
3.60 r 1.14$
Raphe Nuclei Raphe pallidus Raphe obscurus
Raphe magnus Raphe pontis Raphe dorsalis Central superior
Catecholamine Cell Groups Al A2 A6 (LBCUSCoeruleus)
A9 (Substantia Nigra, pars compacta) A10
*For all tracts (e.g., white columns of the spinal cord, pyramidal tract, spinal trigeminal tract) the capillary density was rated as 1.0 and the catecholamine varicosity density as 0. Bignifkantly less than that of sensory nuclei (Students t-test psO.001). SSignificantly more than that of sensory nuclei (Students t-test ~~0.05).
mm). Such a profile has been noted in the lateral tegmentum
and also at the medial edge of the dorsal cochlear nucleus. But in most cases the overlap of catecholamine varicosities with blood vessels appeared almost incidental along the ramifying course of the catecholamine axons (Fig. 1F). In no case was there a perfect contiguity of central catecholamine nerves and blood vessels, such that their fibers were uniquely visible along the vessels and not over the surrounding parenchyma. In fact, only a very small proportion of central catecholamine fibers were associated in any manner with blood vessels. Association
of Catecholamine
Varicosities
with Capillaries
The densities of catecholamine varicosities and capillaries were estimated (on a scale of &5) for each nucleus in the brainstem (Table 1) and their association was tested by Pearson’s correlation coefficient. The highest density of capillaries was apparent within the sensory nuclei, cerebellar associated nuclei and other nuclei which have been linked, albeit indirectly, with the limbic system. Within these categories, the nuclei of the auditory and vestibular systems were salient for their commonly rich capillary supply. The visceral nuclei had a moderate capillary density; the densi-
ties for the motor nuclei, raphe nuclei, and the reticular formation were significantly less than those of the other categories. Myelinated fiber tracts, such as the pyramidal tract had the lowest density of capillaries. There was an obvious correlation between the density of capillaries and that of neuronal perikarya, which was reflected in the decreasing density of capillaries respectively present in sensory nuclei, motor nuclei, the reticular formation and white matter. Similarly to that of capillaries, the density of catecholamine terminals was visibly greater in gray matter than in white matter. But across the nuclei of the brainstem, there was no correlation between the densities of catecholamine varicosities and capillaries (r=O.lO). Whereas the highest concentration of capillaries was apparent in sensory nuclei, and particularly within those of the auditory and vestibular systems, the catecholamine innervation was sparse to moderate in these structures. By category, the most dense catecholamine innervation in the brainstem was found within the visceral nuclei. Within areas, such as the sensory nuclei, where catecholamine varicosities were of low density, the few terminals visible were found over the parenchyma and rarely superimposed on vessels. No association of capillaries and catechol-
JONES
30
amine varicosities was evident within the cochlear nuclei. Only in those regions where a moderate to high density of varicosities was apparent, were the catecholamine terminals found overlying capillaries. However even in these regions, including the nucleus of the solitary tract and the commissural nucleus, the majority of the varicosities were concentrated over the surrounding parenchyma. The nucleus which demonstrated the greatest overlap of catecholamine varicosities and blood vessels was the principal inferior olivary nucleus, which receives both a dense catecholamine innervation and a rich capillary supply (Fig. 2A and B). In this case noradrenaline varicosities which are of known central origin were profusely distributed over small vessels as well as over the parenchyma within the nucleus. Association
of Cutecholamine
Perikarya Mith Capillaries
The association of catecholamine perikarya with blood vessels was assessed according to two measures. First the density of capillaries in the region of the catecholamine cell bodies, as well as the serotonin neurons, was compared to that of other nuclei in the brainstem (Table 1). All the catecholamine cell groups had a moderate to high density of capillaries. Particularly dense was the blood supply to the noradrenaline locus coeruleus cell bodies and the adrenaline and noradrenaline cells in the regions of the commissural and solitary tract nuclei, both areas of very dense cell packing. Yet even in these regions, the capillary supply to catecholamine cell groups was not as dense as that to other nuclei of the brainstem (Table 1). The serotonin raphe cells, most of which are not very tightly grouped, appeared to receive a less dense capillary supply, although the neurons were found lying close to large perforant vessels. Second, the actual proportion of cell bodies which were adjacent to blood vessels was estimated. In those cell groups of the lateral tegmentum (Al and AS) only 35% of the cells counted appeared to be apposed to blood vessels. In A2 within the commissural nucleus approximately 50% of the cell soma appeared in apposition to small blood vessels. The locus coeruleus perikarya were too closely packed to allow counting of cells and capillaries, but these neurons did appear to have a rich capillary supply (Fig. 2C). The density of the capillary supply, however, was as high in the adjacent periventricular gray (Fig. 2C) and in the mesencephalic trigeminal nucleus (Fig. 2C and 2D). These large trigeminal neurons appeared to have one capillary per cell body. They were often encircled by blood vessels and in cases of intense staining showed evidence of extravasation of the dye (Fig. 2D). Some other nuclei which appeared to have a 1: 1 relationship of cells to capillaries were the cochlear nuclei, the superior olivary complex, the trapezoid nucleus, and the inferior olivary complex. Although the density of capillaries in the A9 and Al0 ventral tegmental area was not very high and the percentage of the A9 and Al0 perikarya which were adjacent to blood vessels was low (-25%), the dopamine neurons in this region revealed a special relationship of catecholamine neurons to blood vessels (Fig. 3A, B, and C). The thick processes, presumed to be dendrites, which emanated from the substantia nigra cell bodies appeared closely associated with nearby blood vessels of 46 pm (Fig. 3C), and catecholamine varicosities were visible over these vessels (Fig. 3B and C). In the caudate nucleus, the terminal region of the dopamine nigrostriatal neurons, fine dopamine terminals appeared overlying and surrounding blood vessels located between the
myelinated tracts of the internal capsule fibers (Fig. 3D). Although dendritic processes were not easily visible in noradrenaline neurons of the brainstem, particularly in the A2 and A6 neurons, faintly fluorescent extensions of the A 1, A5 and A7 noradrenaline neurons appeared to run near blood vessels in a manner similar to that observed for the dopamine neurons in the ventral tegmentum. DISCUSSION
Application
c$ the Technique
The combination of catecholamine histofluorescence with supravital staining of blood vessels by a fluorescent dye provides a rapid and reliable technique for the simultaneous visualization of catecholamine neurons and blood vessels at the light microscopic level. Evans blue has been used for years as a stain for blood vessels but most especially as an in viva and post-mortem marker for the pathological opening of the blood brain barrier [8, 9, 10, 36, 54, 551. In the present study, it was found that perfusion of the free dye in the presence of paraformaldehyde fixative produced intense staining of the vessels for fluorescence microscopy with no significant extravasation of the dye. Generally it is the serum albumin complex of the dye, formed in the blood after in ~*ivo injection or by conjugation prior to injection as practiced in the present study, that is believed to be impermeable to the blood brain barrier [55]. However the previous use of free dyes in post-mortem applications [8] and the present demonstration of such an application suggest that the cerebral blood vessels are relatively impermeable to free Evans blue. The anionic composition, lipid insolubility and relative high molecular weight of Evans blue may account for this impermeability [6]. In addition the use of Evans blue in association with paraformaldehyde fixative as practiced with the present technique would further prevent intracerebral diffusion of the free dye [7,9]. Nonetheless the blood brain barrier is not an all-or-none phenomenon and differing degrees of permeability in different regions of the brain have been described [7, 52, 531. In the present study, passage of the free or conjugated dye through the vessels was not seen in most regions deep within the parenchyma but was apparent to varying degrees in particular regions, some known to have no barrier [54]. The extravascular accumulation of dye was directly correlated with the amount of dye perfused and with the resulting overall intensity of the staining. Accordingly increased volumes of perfused dye resulted in increased staining of area postrema and of the endothelial cells, stroma and epithelial cells of the choroid plexus. With maximal staining the Evans blue colored the ependyma of the ventricle and diffused through the ependyma into the periventricular gray, a phenomena mentioned with intravitam staining in the older literature [9,52]. Such ventricular diffusion probably arises from a “functional leak” discovered with horseradish peroxidase at the root of the choroid plexus [7]. The one area deep in the parenchyma where slight extravasation of dye occurred in some animals with an intense staining was within the mesencephalic trigeminal nucleus. In cases where no extravasation was seen it was apparent that many of these cells were partially encircled by blood vessels. It is of interest that this relationship of the perikarya to blood vessels and the partial permeability of these vessels to the dye are similar characteristics to those features described for the gasserian ganglion [ 11. It is possible that this central sensory ganglion may have special vascular characteristics. On the other hand free
CATECHOLAMINE
NEURONS/CEREBRAL
BLOOD VESSELS
Evans blue dye perfused in a large volume may diffuse beyond the vessel wall when saturation occurs in certain regions. It is nonetheless the case that perfusion with the free dye yields the most intense and complete staining of the cerebral blood vessels. Brains in which there was a slight diffusion of the free dye beyond the vessel wall within certain regions did not affect the interpretation of the material in the present study. This histofluorescent technique was used to investigate the relationship of central catecholamine neurons to cerebral blood vessels in the brainstem. Three possible associations were studied: (1) the innervation of large vessels, either arterioles or veins (2) the innervation of smaller vessels, including capillaries and (3) the density and proximity of capillaries to catecholamine perikarya. Such an analysis in histofluorescent material depends upon the interpretation of respective contiguity, overlap or proximity of fibers, varicosities or soma with blood vessels. Of course, these concepts have limited meaning, particularly in 20 pm thick sections, at the light microscopic level and even pose problems at the ultrastructural level in the interpretation of what constitutes the innervation of a blood vessel [37]. However if the central catecholamine neurons do innervate arteries, veins and/or capillaries, such as to function in an analogous manner to the peripheral sympathetic nervous system, a significant association of the two should be evident at the light microscopic level. It was the investigation of this principle which stimulated the present study. Innervation of Large Blood Vessels b> Catecholumine Neurons In the present study, it appeared chat large blood vessels (50 pm-10 pm), which may include both arteries and veins, were innervated by sympathetic nerves which accompanied the vessels into the brain. The distinction between sympathetic and central catecholamine neurons was made according to the morphology and origin of the fluorescent fibers. In the medulla, there was a certain continuity of fibers with a similar morphology, that is having relatively large varicosities, encircling the basilar artery, crossing through the pia arachnoid, and penetrating the brain a short distance along the perforating arteries. These thick varicose fibers were seen running directly alongside the artery with no deviation into the parenchyma. Within the brain a similar relationship of thick varicose fibers to large blood vessels was presumed to be the continuation of the sympathetic innervation. However, in the latter case, absolute distinction was not always possible in the normal animal. These observations confirm earlier reports regarding the sympathetic innervation of the penetrating arteries in various regions of the brain. Pentield [38] and Clark [12] demonstrated by specially applied silver techniques that the small intracerebral arteries were innervated in a similar manner to the pial vessels and that their respective nerve plexuses were continuous. Many years later, investigations with the histofluorescent technique [ 19,341 and the electron microscope [ 11,411, confirmed these earlier observations and showed that penetrating arteries as small as 10-20 pm in diameter were accompanied by peripheral nerves. Central catecholamine fibers were distinguished by smaller varicosities and by their origin in the parenchyma. Such fibers rarely appeared to be closely associated with blood vessels although they frequently intersected or overlapped vessels in their path. Within the lateral tegmentum through the entire brainstem, small radially coursing vessels
of 8-12 pm diameter frequently appeared to be encircled or embraced by central catecholamine fibers which arrived from the parenchyma. This configuration in this particular region represents the one morphological picture which is suggestive of a vascular innervation. It is, however, within this region of the tegmental radiations that blood vessels and catecholamine fibers travel in the same radial orientation and may therefore simply overlap in a parallel course. Such contiguity may be explained by the underlying structural correspondence of cytoarchitecture and angioarchitecture in the brain [ 131. Similar conclusions and reservations were extended by Edvinsson and his colleagues [18] regarding the overlap of central catecholamine fibers with vessels in the hypothalamus. Innervation
ofCapillaries by Catecholamine
Neurons
In an effort to determine if central catecholamine terminals may be associated with capillaries, the density of each of these was estimated in the nuclei of the brainstem. The subjective method of rating was deemed valid and reliable since the estimates of capillary density correlated very highly (r=0.93) with the original measures of Craigie [14,15] and those of catecholamine varicosities correlated well (r=0.70) with similar estimates by Levitt and Moore [28] for the commonly studied nuclei. As Craigie [14] originally noted, gray matter has a higher density of blood vessels than white matter and that within the brainstem nuclei, sensory regions have a richer capillary supply than motor regions and than the reticular formation. The capillary supply was closely correlated with the neuronal density in a region. But in addition, in comparing nuclei with similarly dense cell packing, it became apparent that some nuclei, such as the auditory and vestibular, have even richer capillary supplies than others, such as the solitary tract nuclei. It is of interest that in comparing regional brain glucose consumption of published autoradiographic maps [45] with capillary density in the present study, regional metabolism appeared highly correlated with capillary supply, being the highest in sensory and particularly in the vestibular and auditory regions. Since it is known that regional metabolism is related to local cerebral blood flow (see [46]), these intercorrelations render even more interesting the potential association of catecholamine terminals with capillaries. Indeed the underlying structure for the proposed roles of these neurons in both glucose metabolism [43,44] and cerebral blood flow [39] might be explained by such an association. However, capillary density was not correlated with that of catecholamine varicosities in the gray matter of the brainstem and cervical spinal cord. In fact as a category, the sensory nuclei which have the highest density of capillaries have the lowest density of catecholamine terminals. This negative finding indicates that catecholamine neurons would not play an ubiquitous role in cerebral blood flow and capillary permeability. On the other hand, these results do not exclude the possibility that they may be involved in vascular regulation in certain nuclei. A clearcut contiguity or correspondence of catecholamine terminals and capillaries was not observed in any region of the brainstem. In regions such as the sensory nuclei where a low concentration of terminals was observed, virtually no terminals were seen in close proximity to blood vessels. In nuclei in which a moderate to dense catecholamine innervation coincided with a rich capillary supply, catecholamine varicosities overlapped the vessels, but a much greater con-
JONES centration of terminals was present over the surrounding parenchyma. These results indicate that only in a small number of densely innervated nuclei, such as the principal inferior olivary nucleus, might catecholamine varicosities contact capillaries and that even in these nuclei, only a small proportion of the catecholamine terminals might contact the vessel. Indeed, in the paraventricular nucleus where the richest capillary bed in the brain [22] is coincident with one of the densest adrenergic innervations, only 5% of 5-hydroxydopamine-tagged terminal varicosities were associated with capillaries in the sympathectomized rat; the majority were apposed to other neuronal processes [48]. In fact in a recent study in which noradrenaline terminals were identified by dopamine+-hydroxylase immunoreactivity in normal rats, no such associations were found [35]. Although further studies with the electron microscope in the normal animal are imperative, particularly in such areas as the principal inferior olivary nucleus, the present sum of results indicates that innervation of blood vessels by central catecholamine neurons is probably a rare phenomenon and does not constitute a basic principle of their organization and function. A.ssociation
oj’Catrcholaminr
Prrikuryu
und Blood
Vrs,sds
In 1940 Finley and Cobb [22] reported that in the monkey, the locus coeruleus had one of the richest capillary beds in the brain. They found that its supply was much greater than that of the mesencephalic trigeminal nucleus. The observations of the present study do not support these findings in the rat, in which the locus coeruleus had a moderate to dense capillary supply but no more than the periventricular gray or the mesencephalic trigeminal nucleus. Other catecholamine groups, the cells of which are much less densely packed than those of the locus coeruleus, have a less dense capillary supply. Only one-third to one-half of the catecholamine perikarya appear closely apposed to capillaries. These observations therefore indicate that catecholamine neurons are not special in the density of their vascular supply, a fact which probably means that they do not have extraordinary metabolic requirements. Although catecholamine perikarya were not frequently apposed to blood vessels, thick processes very near to the soma appeared to contact nearby vessels in the substantia nigra. These processes have previously been identified as dendrites of the dopamine neurons [3,20]. Similar observations were made by Felten and Crutcher [21] in histofluorescent material and correlated with ultrastructural evidence of appositions between unidentified dendrites and vessels in the
substantia nigra, locus coeruleus and raphe. Although with the present fluorescent technique, the noradrenaline and adrenaline perikarya and dendritic processes are not as clearly visible as the dopamine ones, some large cells of Al. AS and A7 also appeared to send processes to nearby blood vessels. A similar relationship was originally noted for the serotonin raphe cell bodies and the paramedian perforating arteries in Golgi material by Scheibel, Tomiyasu and Scheibel 1421. With evidence of apparent specialized en pc~ssugc~ or terminal expansions of dendritic processes, the authors speculated that raphe neurons may fulfill a chemosensor or mechanoreceptor function which might be of importance for their role in slow wave sleep. The present observations would suggest that a vascular sensory modality of central catecholamine neurons might underlie their putative roles in such functions as eating and drinking, waking and arousal, and central cardiovascular regulation [S, 2.5, 49, 501. Indeed recent physiological evidence would support a vascular sensor function for central noradrenaline neurons and an indirect role in vascular regulation through neuroendocrine and cardiovascular reflexes, as well as arousal, instead of a direct cerebral vasomotor function [5 I]. The sum of these results would indicate that morphological evidence from histofluorescence material does not support the concept that the central catecholamine neurons are analogous to the peripheral sympathetic nervous system. Their predominant mode of action on the cardiovascular system, as on other central systems, is probably through neural intermediaries rather than directly on cerebral microvasculature. Such a modulatory role may depend on sensory information received through dendritic processes which may contact nearby blood vessels.
ACKNOWLEDGEMENTS 1 would like to thank Lynda Mainville for her excellent techmcal assistance, Lee Friedman, Mitchell Schiller. and Linda Lewis for their respective participation as students in various aspects of the research, and Elisabeth Mullin for preparation of the manuscript. I am grateful to Alain Beaudet for his consultation and stimulating discussion. Finally, I greatly appreciate the advice and assistance of Robert LeMarche for all aspects of the photomicroscopy which was so important in this work. Dr. Barbara E. Jones is a Canadian Medical Research Council Scholar. Her research was supported by the Canadian Medical Research Council (Grant No. 6464). Le Conseil de Recherche en SantC du QuCbec. the Killam Memorial Fund of the Montreal Neurological Institute, and the George W. Stairs Memorial Fund of McGill University.
REFERENCES Arvidsson, B., K. Kristensson and Y. Olsson. Vascular permeability to fluorescent protein tracer in trigeminal nerve and gasserian ganglion. Acta nruroputh. 26: 199-205, 1973. Birks, R. I. The relationship of transmitter release and storage to tine structure in a sympathetic ganglion. .I. Nauroqvtol. 3: 133-160, 1974. Bjiirklund, A. and 0. Lindvall. Dopamine in dendrites of substantia nigra neurons: suggestions for a role in dendritic terminals. Brain Rrs. 83: 531-537, 1975. Bloom, F. E. and E. L. F. Battenberg. A rapid, simple, and sensitive method for the demonstration of central catecholamine-containing neurons and axons by glyoxylic acidinduced fluorescence. II. A detailed description of methodology. .I. Histochrm. C’ytochern. 24: 561-571. 1976.
5. Bolme, P., K. Fuxe and P. Lidbrink. On the function of central catecholamine neurons-their role in cardiovascular and arousal mechanisms. Rcs. C‘ommun. chum. puth. Pharmtrc 4: 657-697, 1972. 6. Bradbury, M. The Concept of‘u Blood-Btain Burrkr. Chichester, England: John Wiley & Sons, 1979. 7. Brightman, M. W. and T. S. Reese. Junctions between intimately apposed cell membranes in the vertebrate brain. .I. (‘cl// Bid. 40: 648-677, 1969. 8. Broman. T., 0. Grontoft and 0. Steinwall. Comparative intravital and postmortem studies on chemically induced blood-brain barrier damage tested with trypan blue. Acrrr nrrrrol. .rcurtd. 41: 527-538. 1965.
CATECHOLAMINE
NEURONS/CEREBRAL
BLOOD VESSELS
9. Cappell, D. F. Intravitam and supravital staining. I. The principles and general results. J. Path. Bact. 32: 595-628, 1929. 10. Cappell, D. F. Intravitam and supravital staining. II. Blood and organs. .I. Path. Bact. 32: 629-674, 1929. 11. Cervos-Navarro, J. and F. Matakas. Electron microscopic evidence for innervation of intracerebral arterioles in the cat. Neurology 24: 282-286, 1974. 12. Clark, S. L, inne~ation of the blood vessels of the medulla and spinal cord. .I. camp. Neural. 48: 247-266, 1929. 13. Cobb, S. The cerebrospinal blood vessels. In: Cytology and Cellular Pathology ofthe Nervous System. vol. 2, edited by W. Penfield. New York: Hoeber, 1932, pp. 577-610. 14. Craigie, E. H. On the relative vascularity of various parts of the central nervous system of the albino rat. J. cojnp. Neural. 31: 429-464, 1920. 15. Craigie, E. Ii. The vascularity of the vagus nuclei and of the adjacent reticular formation in the albino rat. J. camp. Neural. 42: 57-68, 1926. 16. Csillik, B., G. Jancs6, L. Tbth, M. Kozma, G. Krilman and S. Karcsu. Adrenergic innervation of hypothalamic blood vessels. A contribution to the problem of central thermodetectors. Acta anat. 80: 142-151, 1971. 17. De la Torre, J. C. and J. W. Surgeon. A meth~ological approach to rapid and sensitive monoamine histolluorescence using a modified glyoxylic acid technique: the SPG method. Histochemistry 49: 81-93, 1976. 18. Edvinsson, L., M. Lindvall, K. C. Nielsen and Ch. Owman. Are brain vessels innervated also by central (non-sympathetic) adrenergic neurones? Brain Res. 63: 496-499, 1973. 19. Falck, B., G. I. Mchedlishvili and Ch. Owman. Histochemical demonstration of adrenergic nerves in cortex-pia of rabbit. Atta pharmat. fox. 53: 133-142, 1965. 20. Felten, D. I,. Dendritic monoamines in the squirrel monkey brain. Brain Res. 120: 553-558, 1977. 21. Felten, D. I,. and K. A. Crutcher. Neuronal-vascular relationships in the raphe nuclei, locus coeruleus, and substantia nigra in primates. Am. J. Anat. 15% 467-482, 1979. 22. Finley, K. H. and S. Cobb. The capillary bed of the locus coeruleus. J. camp. Neural. 73: 49-58, 1940. 23. Gershon, M. D. and S. Bursztajn. Properties of the enteric nervous system: limitation of access of intravascular macromolecules to the myenteric and muscularis externa. J. camp. Neurol. 180: 467-488, 1978. 24. Hartman, B., D. Zide and S. Udenfriend. The use of dopamine-~-hydroxylase as a marker for the central noradrenergic nervous system in the rat brain. Pror. natn. Acad. Sci. U.S.A. 69: 2722-2726, 1972. 25. Jacobs, B. L. and B. E. Jones. The role of central monoamine acetylcholine systems in sleep-wakefulness states: mediation or modulation? In: Cholinergic-Monoaminergic Interactions in the Brain. edited by L. L. Butcher, New York: Academic Press, 1978, pp. 271-290. 26. Jones, B. E. and R. Y. Moore. Ascending projections of the locus coeruieus in the rat. II. Autoradiog~a~h~ study. Brain Res. 127: 2FS3, 1977. 27. Jouvet, M. The role of monoamines and acetvlcholinecontaining neurons in the regulation of the sleep-waking cycle. Ernebn. Phvsiol. 64: 166-307. 1972. 28. Ld&tt, P. .and R. Y. Moore. Origin and organization of brainstem catecholamine innervation in the rat. J. camp. Nearol. 186: 505-528, 1979. 29. Lindvall, M., J. Cervos-Navarro, L. Edvinsson, Ch. Owman and U. Stenevi. Non-sympathetic perivascular nerves in the brain: origin and mode of innervation studied by fluorescence and electron microscopy combined with stereotaxic lesions and sympathectomy. In: Bbod Flow and Metabolism in the Brain, edited by M. Harper, B. Jennett. D. Miller and J. Roman. Edinburgh: Churchill iivingstone, 197.5,pp. 1.7-1.9. 30. Lindvall, 0. and A. Bj~rklund. The glyoxylic acid fluorescence histochemical method: a detailed account of the methodology for the visualization of central catecholamine neurons. Histochemistry 39: 97-127, 1974.
43
31. McGinty, J. F., L. Y. Koda and F. E. Bloom. A novel fluorescent marker of CNS vasculature used in combination with monoamine histofluorescence. Sot. Neurosci. Abstr. 5: 344, 1979. 32. Moore, R. Y. and F. E. Bloom. Central catecholamine neuron systems: anatomy and physiology of the dopamine systems. A. Rev. Neurosci. 1: 129-169, 1978. 33. Moore, R. Y. and F. E. Bloom. Central catecholamine neuron systems: anatomy and physiology of the norepineprhine and epinephrine systems. A. Rev. Neurusci. 2: 113-168, 1979. 34. Nielsen, K. C. and Ch. Owman. Adrenergic innervation of pial arteries related to the circle of Willis in the cat. Brain Res. 6: 773-776, 1967. 35. Olschowka, J. A., M. E. Molliver, R. Grzanna, F. L. Rice and J. T. Coyle. UItrastructural demonstration of noradrenergic synapses in the rat central nervous system by dopamine-~hydroxylase immunocytochemistry. J. Histochem. Cytochem. 29: 271-280, 1981. 36. Olsson, Y. and T. S. Reese. Permeability of vasa nervorum and perineurium in mouse sciatic nerve studied by fluorescence and electron microscopy. J. Neuropath. exp. Nrurul. 30: 105-I 19, 1971. 37. Owman, Ch. and L. Edvinsson. Histochemic~ and pharmacological approach to the investigation of neurotransmitters, with particular regard to the cerebra-vascular bed. WennerGren Center, Int. Symp. Ser. 30: 15-38, 1977. 38. Penlield, W. Intracerebral vascular nerves. Archs Neural. Psych&. 27: 30-44, 1932. 39. Raichle, M. E., B. K. Hartman, J. 0. Eichling and L. G. Sharpe. Central noradrenergic regulation of cerebral blood flow and vascular permeability. Proc. natn. Acad. Sci. U.S.A. 72: 3726-3730, 1975. 40. Rennels, M. L., M. S. Forbes, J. J. Anders and E. Nelson. Innervation of the microcirculation in the central nervous system and other tissues. Wenner-Gren Center. Int. Symp. Ser. 20: 91-104, 1977. 41. Sato, So. An electron microscopic study on the innervation of the intracranial artery of the rat. Am. J. Anat. 118~873-890, 1966. 42. Scheibel, M. E., U. Tomiyasu and A. B. Scheibel. Do raphe nuclei of the reticular formation have a neurosecretory or vascular sensor function? Expl Neural. 47: 316-329, 1975. 43. Schwartz, W. J. A role for the dopaminergic nigrostriatal bundle in the pathogenesis of altered brain glucose consumption after lateral hypothalamic lesions. Evidence using the ‘%-labeled deoxyglucose technique. Brain Res. 15% 129-147, 1978. 44 Schwartz, W. J. ~Hydroxydopamine lesions of rat locus coeruleus alter brain glucose consumption, as measured by the 2-deoxy-d-(I%) glucose tracer technique. Nrurosci. Lett. 7: 141-150. 1978. 45. Schwartz, W. J. and F. R. Sharp. Autoradiographic maps of regional brain glucose consumption in resting, awake rats using ti4Cf 2-deoxyglucose. J. camp. Neural. 177: 335-359, 1978. 46, Sokoloff, L. Local cerebral energy metabolism: its reIationships to local functional activity and blood flow. In: Cerebral Vascular Smooth Muscle and its Control, CIBA Foundation Symposium 56 (new series). Amsterdam: Elsevier/North-Holland, 1978, pp. 171-197. 47 Swanson, L. W., M. A. Connelly and B. K. Hartman. Ultrastructural evidence for central monoaminergic innervation of blood vessels in the paravent~cular nucleus of the hypothalamus. Brain Res. 136: 166-173, 1977. 48. Swanson, L. W., M. A. Connelly and B. K. Hartman. Further studies on the fine structure of the adrenergic innervation of the hypothalamus. Brain Res. 151: 165-174, 1978. 49. Gngerstedt, U. Stereotaxic mapping of the monoamine pathways in the rat brain. Acta physiol. srand. Suppl. 367: l-48, 1971. 50. Ungerstedt, U. Adipsia and aphagia after ~hydroxydopamine induced degeneration of the nigro-striatal dopamine system. Acta physiol. stand. Suppl. 367: 95-122, 1971.
44
51. Ward, D. G., A. M. Lefcourt and D. S. Gann. Neurons in the dorsal rostra1 pons process information about changes in venous return and in arterial pressure. Brain Rcs. 181: 75-88, 1980. S2. Wells. A. Q. and E. A. Carmichael. Microglia: an experimental study by means of tissue culture and vital staining. Brain 53: I-IO, 1930. 53. Westergaard, E. and M. W. Brightman. Transport of proteins across normal cerebral arterioles. J. clomp. Neural. 152: 17-44. 1973.
JON
ES
54. Wislocki, G. B. and E. H. Leduc. Vital staining of the hematoencephalic barrier by silver nitrate and trypan blue. and cytological comparisons of the neurohypophysis, pineal body. area postrema, intercolumnar tubercle and supraoptic crest. ./. c’omp. Neural. 96: 371-413, 1952. 55. Woman, M., I. Klatzo, E. Chui, F. Wilmes, K. Nishimoto. K. Fujiwara and M. Spatz. Evaluation of the dye-protein tracers in pathophysiology of blood-brain barrier. Actrr rwrr’opdr. 54: 55-61. 1981.
