Fluorescence photomicrograph of whole mount of rat mesentery showing perivascular sympathetic nerves demonstrated by Falck‐Hillarp formaldehyde technique. Very dense plexus of green‐fluorescent, noradrenergic fibers supplies a small artery, whereas corresponding vein has less well‐developed innervation. Nerve terminals are seen to accompany small vessels of arteriolar caliber. × 155.

Arrangement of vascular neuroeffector apparatus. Postganglionic autonomic nerves ramify into small bundles forming a primary plexus, which is located in loose adventitia. Bundles consist of axons with smooth appearance characteristic of preterminal portion of nerve fiber. Bundles give rise to beaded (varicosed) fibers forming terminal effector plexus, located on surface of media layer. A few terminals may sometimes extend, but only for a short distance, into outermost layer of the media. Thus only smooth muscle cells located in adventi‐tia‐media junction receive direct autonomic innervation.

Location of pressor (crosshatching) and depressor (horizontal lines) centers in brain stem of cat. A‐C: cross sections through medulla at levels indicated by dotted lines to D. D: semidiagrammatic projection of pressor and depressor regions onto dorsal surface of brain stem, viewed with cerebellar peduncles cut across and cerebellum removed. AT, auditory tubercle; BC, brachium conjunctivum; BP, brachium pontis; C₁, 1st cervical nerve; CN, cuneate nucleus; FG, facial genu; GN, gracile n.; IC, inferior colliculus: IO, inferior olivary n.; LN, lateral reticular n.; RB, restiform body; SO, superior olivary n.; SPV, spinal trigeminal tract; TB, trapezoid body; TC, tuberculum cinereum; TS, tractus solitarius; V‐VII and X, cranial nerves; and I‐III, levels of transection.

Representation of some cerebral circulatory studies in which either stimulation or lesions of brain stem have been examined. To allow comparison, studies have been transferred to midsagittal section of human brain. Sections I and II, positions of ascending noradrenergic system and ascending reticular activating system, respectively. Section III, filled areas show lesioned sites that abolish cerebral CO₂ reactivity in dogs. Crosshatching in this and remaining sections show lesions where no effect was noted. Sections IV and V, brain stem transections and pontine cold lesions, respectively, that abolish CO₂ reactivity in cats. Sections VI‐VIII, studies in rhesus monkeys and baboons that demonstrate brain stem sites where stimulation results in increased cerebral blood flow and metabolism.

Cardiovascular reflex arc involving nucleus of solitary tract (NST). 1: Afferent cardiovascular neuron, from periphery to NST (cell bodies in nodose ganglion). 2: Central cardiovascular neuron (cell bodies in NST); a: axons to efferent cardiovascular neurons; b: axons to medullary centers; and c: axons running back from medullary centers to efferent neurons. 3: Efferent cardiovascular neuron, running from medulla or spinal cord to periphery; a: preganglionic cells in dorsal vagal nucleus and intermediolateral nucleus; b: postganglionic cells in peripheral autonomic ganglia. Filled circles with open centers, reflex arc neurons; open circle with filled center, modulatory neuron; solid lines, central axons; dashed lines, peripheral nerve connections.

a: Electron micrograph of innervated capillary in cat anterior hypothalamus, lined by endothelium (E) with junctional complexes (J) at cellular margins. Endothelial nucleus is not in plane of section. Pericyte cell nucleus (Pn) appears at top left. Arrows, pericyte cytoplasmic processes. Axon terminal is in outlined area. Vertical bar, 1.0 μm. b: Outlined area in a at higher magnification, showing axon terminal (Ax) surrounded by basement membrane and containing synaptic vesicles. Major contact in this section is between terminal and pericyte, although a small portion of axon abuts directly on endothelium. Pp, pericyte process; A, astrocyte process. Vertical bar, 0.5 μm. c: Axon terminal in direct apposition to endothelium of a different capillary. Vertical bar, 0.5 μm. d: Portion of large capillary (∼9.3 μm luminal diam) of heart ventricle from mouse treated with 5‐hydroxydopamine. Adrenergic terminals contain small, dense‐core vesicles (arrows), 1 of which is closely apposed both to a pericyte (PP) and to endothelium. No Schwann cell covering is present. Small circular profile between fibroblast (F) and axon terminal is intervaricose axon. Vertical bar, 0.5 μm.

Fluorescence photomicrograph showing myosin immunoreactivity in wall of rat brain capillaries. In separate tests the same structures were shown to contain immunoreactive actin. Strong myosin reaction is also seen in smooth muscle media layer of arterioles (A). × 250.

A: regenerative sprouting of catecholaminergic nerve fibers 7 wk after damage of central neurons by electrolytic lesion in rat mesencephalon. Entire thickness of wall of cross‐sectioned small vessel close to the lesion at base of the brain is invaded by abnormally distributed, catecholaminergic fibers. B: small artery at base of the brain of normal rat showing characteristic distribution of sympathetic nerve fibers in adventitia, superimposed on smooth muscle media layer. Nonspecific autofluorescence of ruffled inner elastic membrane is clearly visible. × 180.

After release of norepinephrine (NE) from sympathetic nerve terminals in the adventitia‐media junction of vessel, receptors of the directly innervated smooth muscle cells in outermost layer of media are activated first. Deeper muscle cell layers are activated by diffusion of transmitter substance, which reaches receptors in decreasing concentrations, and by propagation between adjacent smooth muscle cells via low‐resistance connections. Part of released transmitter is removed from effector region through neuronal uptake and part of it is metabolized extraneurally.

A: Falck‐Hillarp formaldehyde technique for fluorescence‐histochemical visualization of monoamine transmitters. Proposed reaction sequence between catecholamines and formaldehyde to form products visible in fluorescence microscope. Amines are condensed with formaldehyde to 6,7‐dihydroxy‐1,2,3,4‐tetrahydroisoquinoline (I) and 4,6,7‐trihydroxy‐1,2,3,4‐tetrahydroisoquinoline (IV). Formation of fluorescent products (fluorophores) can proceed in 2 ways: either through autoxidation to 6,7‐dihydroxy‐3,4‐dihydroisoquinolines (R‐H) or by a 2nd, acid‐catalyzed reaction with formaldehyde to yield 2‐methyl‐6,7‐dihydroxy‐3,4‐dihydroisoquinolinium compounds (R‐CH₃). At neutral pH fluorophores are in their tautomeric quinoidal forms (III and VI); they are converted to their nonquinoidal forms (II and V) after brief treatment with HCl. On prolonged treatment with HCl, the fluorophores with a 4‐hydroxy group (V) are transformed into fully aromatic compounds (VII). B: histochemical visualization of acetylcholinesterase (AChE), which is primarily localized in cholinergic nerves. Because the substrate used, acetylthiocholine (AThCh⁺), is hydrolyzed both by AChE and the nonspecific enzyme butyrylcholinesterase (BuChE), BuChE must be efficiently inhibited in order to show cholinergic nerves selectively. Tissue sections are incubated with AThCh⁺ in a medium saturated with copper thiocholine in order to effect its precipitation as rapidly as that formed from the enzymatic liberation of thiocholine. After rinsing, sections are immersed in yellow ammonium sulfide solution to convert precipitated copper sulfide. Sites of enzymatic activity are then visualized as dark‐brown sulfide deposits.

Indirect immunofluorescence technique (top) of Coons et al. 57 and peroxidase‐antiperoxidase (PAP) technique (bottom) of Sternberger 263 are the most commonly used methods for immunohistochemistry. In indirect immunofluorescence technique, sections are exposed to specific peptide antiserum in appropriate dilution for various lengths of time (usually 3 h) at room temperature. After thorough rinsing in phosphate buffer, they are incubated in fluorescein‐labeled antirabbit (or anti‐guinea pig) immunoglobulin G (IgG) for 30 min at room temperature. Sections are mounted in phosphate‐buffered glycerin after another rinsing in phosphate buffer and are examined with fluorescence microscope equipped with filters selected to give peak excitation at 490 nm. In the PAP procedure, sections are exposed to peptide antiserum overnight at 4°C. After rinsing in phosphate buffer, sections are incubated in unlabeled goat (or swine or sheep) antirabbit or anti‐guinea pig IgG for 30 min at room temperature, followed by incubation for 30 min with PAP complex [i.e., peroxidase coupled to rabbit (or guinea pig) antibody raised against horseradish peroxidase]. This is followed by incubation with a solution of 0.05% diaminobenzidine hydrochloride and 0.01% hydrogen peroxide in Tris buffer (pH 7.6) for 1 h. Sections are then examined under a light microscope.

Anterior cerebral arteries obtained from human fetuses at 21 wk of gestation. Vessels were cut open and mounted flat on microscope slides. A: fluorescence photomicrograph after using Falck‐Hillarp formaldehyde technique to demonstrate adrenergic nerves. Vessel is supplied by dense plexus of sympathetic fibers. × 100. B: AChE‐containing nerves visualized by incubation with AThCh⁺ after inhibition of pseudocholinesterase with Mipafox. Artery receives equally dense supply of cholinergic fibers. × 100.

Cat's posterior cerebral artery, cut open longitudinally and mounted flat on microscope slide. Vessel was first treated with formaldehyde gas under mild conditions for visualization of norepinephrine (a) and was subsequently incubated for staining of AChE (b). There is close correspondence between the adrenergic and cholinergic nerve plexuses. × 235. c: Electron micrograph of adrenergic nerve varicosity (A) containing electron‐dense synaptic vesicles, located in close apposition (membrane‐to‐membrane distance, 20 nm) to cholinergic varicosity (Ch). Both terminals are located at innervation distance (∼100 nm) from membrane of smooth muscle cell (M). × 73,500.

Interaction between adrenergic and cholinergic mechanisms in brain circulation. Cerebral blood flow (CBF) was measured in caudate n. (CN) by heat clearance, using a thermistor heated a few tenths of a degree above cerebral temperature (TEMP) and measured by a contralateral reference thermistor. Temperature difference is maintained at constant level despite blood flow changes, with heating current required varying in direct proportion to alterations in flow. Left part of figure shows effects on CN blood flow of postganglionic sympathetic nerve stimulation (St; 6 V, 2‐ms duration, and 20‐Hz freq) or infusion of 2.0 μg·kg⁻¹·min⁻¹ of norepinephrine [N‐Ep (NE)] before or during intracarotid carbachol (CARB) infusion (2.5 μg·kg⁻¹·min⁻¹). Administration of atropine (At; 0.5 mg/kg) inhibited CARB‐induced increase in CN blood flow but did not affect antagonistic action of CARB during sympathetic nerve stimulation. Also, NE response was unchanged. There was no significant alteration in any other parameters recorded, except for small rise in systemic blood pressure (BP) during NE infusion. Slight rebound in flow after cessation of sympathetic nerve stimulation was potentiated by CARB. Diagram to right summarizes measurements in CN of 13 unanesthetized rabbits. Open bars, mean increases (± SE) in CN blood flow during intracarotid infusion with 2.5 or 1.3 μg·kg⁻¹·min⁻¹ of CARB. Stippled bars, mean reduction (± SE) in CN blood flow during postganglionic cervical sympathetic nerve stimulation (CS), either alone or during intracarotid infusion of 2.5 μg·kg⁻¹·min⁻¹ of CARB. , partial pressure of CO₂ in arterial blood; , partial pressure of O₂ in arterial blood.

Schematic drawing of ultrastructure of different types of autonomic nerve endings. Left: cholinergic; center: noradrenergic; right: “P type.” Cholinergic nerve profile contains vesicles, most of which are small and electron lucent, whereas vesicles of adrenergic nerve profile are predominantly small and electron dense. The P‐type nerve profile is characterized by relative abundance of fairly large granules of varying electron density; they are sometimes referred to as large opaque vesicles.

Schematic representation of how synthesis, storage, release, and inactivation of ATP is assumed to take place in a purinergic autonomic nerve terminal.

Vasoactive intestinal polypeptide (VIP) immunoreactivity in anterior cerebral arteries of cat. A: picture is composed of 2 adjacent cross sections and shows many VIPergic nerves enclosing vessel. × 200. B: oblique section through vessel wall. Several delicate VIPergic nerve fibers with beaded appearance run in wall and also enclose adjacent small arteriole in bottom right corner. × 450. C: cross section of vessel, demonstrating that VIPergic nerves run on surface of smooth muscle layer in a characteristic manner. × 450. Internal elastic membrane is autofluorescent in all 3 pictures.

Sympathetic nerve fibers visualized with Falck‐Hillarp formaldehyde histofluorescence technique. A: section of cat's brain, showing intracerebral arteriole branching into brain parenchyma from pial arteries visible in top lefthand margin of picture. Resistance vessel is supplied with a well‐developed plexus of sympathetic nerve terminals. × 250. B: whole mount of cat pial vein, cut open and mounted flat on microscope slide. Capacitance vessel receives substantial number of sympathetic fibers forming a network in wall. × 100.

Superior cervical ganglion of rat showing green‐fluorescent, noradrenergic nerve cell bodies as demonstrated by Falck‐Hillarp formaldehyde technique. Cell nucleus appears dark. Note that ganglion also contains nonadrenergic cells. × 400.

Fluorometric determinations of NE content in rabbit pial vessels at various times after pre‐ or postganglionic sympathectomy. Values indistinguishable from 0 are reached within 2 days after ganglionectomy. Decentralization leads to slightly elevated tissue NE level, due to reduced activity in postganglionic neuron devoid of its preganglionic input. Values are means ± SE, n = 4 determinations (10 at time 0), with tissues from 3 animals being pooled for each determination. Student's t test: ^*, 0.01 < P < 0.05; ^***, P < 0.001.

Fluorescence microscopic estimation of sympathetic innervation (formaldehyde reaction) of intracerebral resistance vessels in baboons. In each region, number of arterioles accompanied by sympathetic fibers was related to total number of arterioles in 6‐μm‐thick sections.

Continuous measurement of blood flow in caudate nucleus (CN) and lateral geniculate body (LGB) in unanesthetized rabbits, with thermoclearance technique (cf. Fig. 14). After unilateral transection of sympathetic trunk below superior cervical ganglion, there was an immediate increase in flow, followed by prompt reduction during electrical stimulation (30 Hz) of distal nerve stump. In reference to flow seen during inhalation of 7% CO₂, effect of sympathetic denervation or stimulation was always greater in CN, which is better innervated by sympathetic perivascular nerves, than in LGB (cf. Fig. 21). ECoG, electrocorticogram; BP, systemic blood pressure; and , arterial gas tensions, measured by mass spectrometry; and TEMP, contralateral brain temperature.

Schematic representation of modulatory effects of cranial sympathetics on cerebral blood flow autoregulation. A, lower limit of autoregulation during hemorrhage, which is accompanied by marked sympathoadrenal activation. B, upper limit of autoregulation during hypertension. C, lower limit of autoregulation can be moved toward lower blood pressure levels, provided perivascular sympathetic innervation of brain vessels is inactivated. D, upper limit of autoregulation can be moved toward higher blood pressure levels during enhanced activation of cerebrovascular sympathetic nerves.

Possible pathogenesis of stroke after sympathetic denervation in stroke‐prone, spontaneously hypertensive rats.

Effect of sympathomimetic and sympathetic stimulation on cerebral blood volume (CBV) in mice. A: tyramine (which acts by uptake into adrenergic nerves, followed by displacement of transmitter) reduced CBV in intact, but not to any significant degree in sympathectomized, animals. B: bilateral electrical stimulation of sympathetic trunk in the neck significantly lowered CBV in comparison with untreated or sham‐operated animals, with the effect being blocked by injection of the α‐receptor antagonist phenoxybenzamine (PBZ; 2 mg/kg, iv). Values are means ± SE; number of animals in parentheses.

Sympathetic denervation leads to a variety of changes in cerebral hemodynamics, as illustrated by measurements of CBV in mice with a non‐diffusible tracer ([¹³¹I]‐labeled serum albumin), depending on type of denervation and postoperative time lapse. After postganglionic denervation (bilateral excision of superior cervical ganglia), CBV decreases during transmitter leakage, then increases when the degenerating vasoconstrictor nerves lose their transmitter, and finally normalizes when denervation supersensitivity ensues. After preganglionic denervation (decentralization of superior cervical ganglia), CBV is enhanced because of loss of vascular tone (cf. Fig. 20).

Circadian rhythm of CBV in mice subjected to 12‐h light/12‐h dark periods. Values are means ± SE; number of animals in parentheses.

Both the choroid plexus, responsible for cerebrospinal fluid (CSF) production, and the cerebrovascular bed, where most of blood volume is held in capacitance vessels, are innervated by sympathetic nerves. CSF dynamics and cerebral blood volume (CBV) are major factors in regulation of intracranial pressure. Postganglionic sympathetic denervation (SyX) of rabbits leads to approximately equal elevations in CSF production and CBV compared with unoperated controls (C), and to a marked increase in intracranial pressure. Values are means ± SE; number of animals in parentheses; differences between mean values are indicated according to Student's t test.

Whole mount preparations of longitudinally opened rabbit anterior cerebral arteries, processed according to Falck‐Hillarp formaldehyde technique for demonstration of perivascular adrenergic nerves. A: vessel from untreated control animal showing dense sympathetic nerve plexus; B: artery from animal treated with subarachnoid blood injection 3 days before sacrifice. Both the number of nerve terminals and their fluorescence intensity are markedly reduced (cf. Fig. 30). × 300.

Experimental subarachnoid hemorrhage in rabbit, achieved by injection of 1–2 ml of autologous blood into chiasmatic and basal cisterns. Blood deposited in subarachnoid space results in reduced in vitro uptake of [³H]NE into perivascular adrenergic nerves, concomitant with reduction in number and intensity of nerves with formaldehyde‐induced fluorescence (cf. Fig. 29), reflecting decrease in concentration of NE in blood vessels. Those changes are followed by “vasospasm,” as measured on arteriograms in terms of basiliar artery caliber. All parameters show similar pattern of normalization in course of 3 wk after maximum changes.

Cisternal blood injections in rabbits, used in an attempt to mimic subarachnoid hemorrhage, within 3 days results in 5‐fold increase in in vitro reactivity of basilar artery to NE (○‐ ‐ ‐○) in comparison with vessels from control animals (•–•). Values are means ± SE; n = 10.

Schematic representation of brain capillary. Endothelial cells are connected by continuous belts of tight junctions at several levels of opposing membranes, thereby restricting intercellular diffusion. Together with paucity of transendothelial pinocytosis in brain vessels, morphological blood‐brain barrier efficiently impedes entry of, e.g., transmitter amines and peptides from circulation into brain.

Fluorescence‐microscopic demonstration of enzymatic blood‐brain barrier. Animal received L‐dihydroxyphenylalanine, which is taken up into capillary endothelial cells and pericytes by facilitated transport system for amino acids. Within cells, the amino acid is decarboxylated to dopamine, which is trapped locally because the degrading enzyme, monoamine oxidase (also present within endothelial cells and pericytes), has been inhibited. Highest fluorescence intensity is in nuclear region, where cell is most voluminous. × 225.

Incidence and degree of blood‐brain barrier lesions, as evidenced by extravasation of Evans blue‐albumin (EBA) complex or [¹⁴C]inulin in right hemisphere of rat brain after hypertensive insult (rapid blood injection into right carotid artery), hypercapnia or papaverine, or combined treatment. Vasodilatation by hypercapnia or papaverine significantly enhanced leakage induced by hypertensive insult (Student's t test: P < 0.001). Values are means ± SE.

Regional cerebral blood flow (rCBF), measured by [¹⁴C]ethanol method in untreated control rats (C), as well as after infusion of catecholamines [NE, epinephrine (E), and isoproterenol (Iso)], alone at rate of 1 mg·kg⁻¹·min⁻¹, or after pretreatment of animals with competitive antagonists [phentolamine (Phent), propranolol (Prop)]. Values are means ± SE; n = 10 controls, 4–5 treated. Examples from frontal cortex (marked reduction in flow after NE and E, no effect by Iso; effects blocked or slightly reversed by Phent) and thalamus (little or no reduction by NE and E, enhanced flow after α‐receptor blockade by Phent; increased flow by Iso, counteracted by Prop).

rCBF in rats, measured by [¹⁴C]ethanol method, exemplified by cortex (open bars) and caudate nucleus (stippled bars). Values are means ± SE; number of animals indicated. In comparison with untreated controls, hypercapnia induced expected flow increase, whereas NE (1 μg·kg⁻¹·min⁻¹, iv) reduced rCBF (cf. Fig. 35). When the same dose of NE was infused after opening of morphological blood‐brain barrier by hypertonic shock (2 M urea injected into internal carotid), flow was significantly elevated secondary to NE‐induced increase in brain metabolism. Comparison between mean values (Student's t test): *, 0.01 < P < 0.05; **, 0.001 < P < 0.01; ***, P < 0.001.

Parasympathomimetic effects on brain circulation (CBF) of rats studied with venous outflow method. A: with intact blood‐brain barrier, intracarotid infusion of carbacholine induced small and short‐lasting changes in flow via direct effects on cholinergic receptors of vessel walls. B: after hypertensive opening of blood‐brain barrier with intracarotid bolus of autologous blood, biphasic response was more pronounced and long lasting. Augmented effect was due to added parasympathomimetic stimulation of cerebral metabolic rate of O₂, which was enhanced by ∼20% when carbacholine was allowed to enter brain. MABP, mean arterial blood pressure; ICP, intracranial pressure.

Composite picture of representative curves comparing effects of histamine on feline intra‐ and extracranial arteries in vitro. Contractile responses were obtained with histamine H₂ receptors blocked. Dilatation effects were recorded during blockade of H₁ receptors at an active tonic contraction of 250 dyns by 3 × 10⁻⁶ M serotonin. Histamine contracted extracranial arteries more efficiently than intracranial arteries. Dilatory response was slightly more pronounced intracranially. Contractile effect on intracranial vessels was, in contrast to other responses, nonspecific in the sense that the specific histamine antagonist tested failed to produce a parallel shift of dose‐response curve toward higher agonist concentrations. ED₅₀, effective dose, 50%; E_AM, maximum response.

Experiments on 3 different strips of canine saphenous vein showing that concentrations of 3 vasoactive amines, which do not cause changes in tension of unstimulated preparations or during contractions with exogenous NE, do cause pronounced relaxation during sympathetic nerve stimulation (top graphs). Relaxation is due to decrease in evoked release of adrenergic neurotransmitter (bottom graphs), measured in terms of tritium overflow after preincubation of vessel preparation in presence of [³H]NE. ES, electrical stimulation.

Schematic illustration of events mediated by substance P (SP) occurring during peripheral activation of small‐diameter sensory nerves. Intradermal injection of SP, for example, releases histamine (H), which acts on sensory nerves to produce an itching sensation as well as release of SP from perivascular nerves containing it, resulting in antidromic vasodilatation. BV, blood vessel; M, mast cell; filled circles, mast cell granules; and P1 and P2, primary sensory neurons 1 and 2.

Increase in caliber of cat's pial arteries (ranging in size from 40 to 180 μm) and veins, expressed as percent change in size from preinjection values, during administration of various concentrations of VIP into perivascular space. Values are means ± SE. Significant changes (Student's t test) in caliber in presence of VIP in comparison with injection of vehicle (mock CSF): *, 0.01 < P < 0.02; **, 0.001 < P < 0.01; and ***, P < 0.001.

Time course of changes in CBF and cerebral O₂ consumption in baboons during intracarotid infusion of VIP (10⁻¹¹ mol/min for 20 min). Open circles, with intact blood‐brain barrier, the 2 parameters are not significantly altered. Filled circles, significant and parallel increase in both flow and metabolism after opening of barrier by intracarotid bolus injection of hypertonic urea at time 0. Values are means ± SE and are expressed as percent changes from their respective levels prior to administration of peptide. Significant differences from base line values (Student's t test): *, 0.01 < P < 0.05.

Vasodilatory effects of morphine (A) and Leu‐enkephalin (B) on isolated feline middle cerebral artery. Vessel had been given tonic contraction with prostaglandin F_2α (2.5 × 10⁻⁶ M) beforehand. Dose‐response curves are shifted in parallel to right, indicating competitive inhibition, in presence of 2 concentrations of naloxone.