The peptidergic innervation of the human superficial temporal artery: Immunohistochemistry, ultrastructure, and vasomotility

Peptides. Vol. 16. No. 2. pp. 275-287, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in the USA. All rights reserved 0196-9781/95 $9.50 + .OO

Pergamon 0196-9781(94)00165-O

The Peptidergic Innervation of the Human Superficial Temporal Artery: Immunohistochemistry, Ultrastructure, and Vasomotility I. JANSEN

OLESEN, *’ S. GULBENKIAN,? J. WHARTON,$ J. M. POLAK$

A. VALENCA,? J. L. ANTUNES,.+ AND L. EDVINSSON*l

*Department of Experimental Research, Lund University, Malmii General Hospital, S-214 01 Malmii, Sweden, fDepartment of Cell Biology, Gulbenkian Institute of Science, 2781 Oeiras, Portugal, $University Clinic of Neurosurgery, University of Lisbon Medical School, Santa Maria Hospital, 1600 Lisbon, Portugal, jDepartment of Histochemistry, Royal Postgraduate Medical School, Hammersmith Hospital, London WI2 ONN, UK, and TDepartment of Internal Medicine, University Hospital of Lund, S-221 85 Lund, Sweden Received

12 January

1993

OLESEN, I. J.. S. GULBENKIAN, A. VALENCA, J. L. ANTUNES, J. WHARTON, J. M. POLAK AND L. EDVINSSON. The peptidergic innervation of the human superjkial temporal artery: Immunocytochemistry, ultrastructure, and vasomotili@. PEPTIDES 16(2) 275-287, 1995.-The peptidergic innervation of the human superficial temporal artery was investigated by means of immunohistochemical, ultrastructural, and in vitro pharmacological techniques. A dense network of nerve fibers was found in the adventitia. The majority of the nerve fibers displayed immunoreactivity for tyrosine hydroxylase and neuropeptide Y (NPY). A moderate supply of perivascular nerve fibers displayed either acetylcholinesterase activity or immunoreactivity for vasoactive intestinal peptide (VIP), peptide histidine methionine-27 (PHM), and calcitonin gene-related peptide (CGRP). Only a few nerve fibers displayed substance P (SP). neurokinin A (NKA), and neuropeptide K (NPK) immunoreactivity. In double immunostained preparations, SP immunoreactivity was co-localized with NPK and CGRP in the same nerve fibers. Ultrastructural studies revealed the presence of numerous axon variocosities at the adventitial-medial border. NPY, VIP, and CGRP immunoreactivities occurred in the same type of large granular vesicles. but in morphological distinct nerve profiles. NPY had. in general, no direct vasoconstrictor effect. However. at a low concentration of NPY contractile response induced by NA (IO-‘- 10mh M) was 9- 15 times enhanced. The NPY-induced potentiation of the NA-induced contraction was not dependent on the presence of an intact endothelium. No significant difference was found between acetylcholine, VIP. and PHM in either potency or degree of relaxation. SP, NKA, and CGRP also acted as vasodilatory agents, with CGRP being more potent than the tachykinins. The response to SP, but not CGRP. was dependent on an intact endothelium. Pretreatment of the vessels with a low concentration of NPY did not change the responses to ACh, VIP. SP, or CGRP. Neuropeptide Y Superficial temporal

Vasoactive intestinal artery innervation

polypeptide Calcitonin Immunohistochemistry

studies on the pathophysiology of migraine have revealed increases in pulsations of the superior temporal artery at the beginning of the headache on the affected side. Following treatment with ergotamine, the pulsations were reduced in amplitude with the decline in the intensity of headache (7). In acute migraine attacks, we have demonstrated an activation of the trigeminovascular system, together with release of stored sensory neuropeptides (22). Furthermore, the demonstration that the release of neuropeptides occurs in parallel with headache and vasodilatation (18) suggests that perivascular nerves are involved

for reprints

should be addressed

Tachykinins

in the dynamic changes occurring in conjunction with attacks of migraine. The perivascular innervation of the superficial temporal artery has been studied by immunohistochemistry and retrograde tracing in the rat (47). It was observed that the superficial temporal artery is supplied with sympathetic nerves storing neuropeptide Y (NPY) and noradrenaline (NA), which emanate from the superior cervical ganglion; parasympathetic nerves arising from the sphenopalatine and otic ganglia containing vasoactive intestinal peptide (VIP) and peptide histidine isoleucine (PHI); sensory fi-

CLINICAL

I Requests

gene-related peptide In vitro pharmacology

to Inger Jansen Olesen. Ph.D.

275

276

JANSEN

bets, storing substance P (SP) and calcitonin gene-related peptide (CGRP), which originate from the first division of the trigeminal ganglion (47). Apart from a brief preliminary study (36), a detailed analysis of the peptidergic innervation of the human superficial temporal artery has not been performed. In the present study, we have further examined the peptidergic innervation (autonomic and sensory) of the human superficial temporal artery by means of immunohistochemical and ultrastructural techniques. Vasomotor responses were also examined in artery preparations both in the presence and absence of intact endothelium. Thus, by using a combination of immunohistochemical, ultrastructural, and pharmacological techniques we have attempted to further our understanding of neuroeffector mechanisms in the human superficial temporal artery. METHOD

Immunojluorescence

Staining

Superficial temporal arteries were obtained from six patients during neurosurgical tumor resections. Immediately after excision, vessel segments were fixed by immersion for 16-24 h at 4°C in Zamboni’s fixative (43). The tissue was rinsed in several changes of phosphate-buffered saline (PBS, 0.01 M, pH 7.2) containing 15% (w/v) sucrose and 0.1% (w/v) sodium azide, then processed as whole-mount preparations for indirect immunofluorescence staining as previously described (23). Briefly, after pretreatment with a solution containing 0.2% Triton X-100 for 2 h at room temperature and impregnation with the dye pontamine sky blue (6) for 30 min, blood vessels were incubated in diluted primary antisera (Table 1) overnight at room temperature. The preparations were then washed in PBS and incubated with fluorescein isothiocyanate-conjugated (FITC) goat anti-rabbit IgG (1:lOO dilution; Sigma) for 1 h at room temperature. For the simultaneous localization of two antigens, preparations were first exposed to a primary antiserum raised in rabbit that was visualized by a rhodamine-labeled goat anti-rabbit IgG (1: 100 dilution; Sigma) and then to a second primary antiserum raised in rat that was visualized by a FITC-labeled goat anti-rat IgG (1: 100 dilution; Sigma). The preparations were finally examined using an Olympus BH-2 microscope equipped for epi-illumination with filters selective for fluorescein and rhodamine fluorescence. Acetylcholinesterase

(AChE) Staining

The histochemical demonstration of acetylcholinesterase (AChE) activity in whole-mount preparation was performed as previously described (24). Briefly, whole-mount preparations were immersed in incubation medium at a 20-fold higher dilution for 30 min at 37°C (16). After rinsing in distilled water, the AChE activity was visualized by immersing the preparations for 5 min in Tris-HCl buffer (50 mM, pH 7.6) containing 0.04% 3,3’-diaminobenzidine tetrahydrochloride and 0.3% nickel ammonium sulfate and then for a further 5 - 10 min with the addition of 0.003% hydrogen peroxide (46). After a brief wash in distilled water, whole-mount preparations were mounted on glass slides, dehydrated in an ascending series of acetone concentrations, cleared with xylene, and mounted in DPX. The preparations were finally observed and photographed with an Olympus BH-2 microscope using the transmitted bright field illumination. Conventional Electron Microscopy Immunogold Staining

and Postembedding

Vessel segments were immersed in 2.5% glutaraldehyde (v/v) in 0.1 M phosphate buffer (pH 7.2) for 2 h at 4°C. Arteries were then washed in buffer containing 0.1 M sucrose, postfixed

TABLE CHARACTERISTICS

OF

OLESEN

ET AL.

1 ANTIGENES

USED

Dilutions Antigen

PGP 9 .5 TH NPY VIP PHM SP SP NPK CGRP

Species

Code No.

LM

Rabbit Rabbit Rabbit

Ra95 103 Tel01 IO86

I:1600 I:100 I :400

Rabbit Rabbit Rabbit Rat Rabbit Rabbit

652 1655 910 MASO 35b 15.36R2 I208

I :2000 I:800 I:500 I:100 I:600 I:200

EM

SOlIKe

I :2000 I :4000

Ultraclone, UK Eugene Tech, USA Hammersmith Hosp. Hammersmith Hosp.

Hammersmith Hosp. Hammersmith Hosp. Sera Lab, UK I :2000

Dr. Valentino. Hammersmith

USA Hosp.

PGP 9.5, protein gene product 9.5; TH. tyrosine hydroxylase; NPY, neuropeptide tyrosine; VIP, vasoactive intestinal polypeptide; PHM. peptide histidine methionine; SP, substance P; NPK, neuropeptide K; CGRP, calcitonin gene-related peptide. References for characterization of the antisera (23,25,26,50).

in 1% osmium tetroxide for 1 h at 4°C rinsed in buffer containing 0.1 M sucrose, dehydrated in a graded series of ethanol concentrations, cleared in propylene oxide, and infiltrated with Epon resin. Ultrathin sections of silver interference color (70-90 nm) were collected onto 300-mesh formvar-coated grids. For the localization of peptide-immunoreactive sites at the ultrastructural level, an on-grid immunogold staining method was employed (48). Briefly, the sections were treated for IO-40 min at room temperature with a saturated aqueous solution of sodium metaperiodate (Sigma), a strong oxidizing agent that unmasks antigenie sites on osmium-fixed tissue (2). After rinsing in 0.05 M Tris buffer, sections were incubated in normal goat serum ( I:30 dilution) for 15 min at room temperature and then in diluted primary antisera (Table 1) for 16 h at 4°C. After washing thoroughly in Tris buffer, the sections were incubated in a goat antirabbit IgG antiserum labeled with 15-nm gold particles (1: 15 dilution; Amersham) for 1 h at room temperature. After washing in Tris buffer, sections were finally counterstained with uranyl acetate and lead citrate, and were examined using a Jeol 100 CX electron microscope operating at 60 kV. Antisera The antisera used in this study are listed in Table 1. The polyclonal (code 910) and monoclonal (code MAS 035b) antisera raised against SP showed partial cross-reactivity with neurokinin A and neurokinin B; the antiserum raised against neuropeptide K (code 15-36 R2) cross-reacted with neurokinin A and B, but did not cross-react with SP. Immunohistochemical

and Histochemical

Controls

In control experiments, no immunostaining was observed when one primary antiserum (Table 1) was omitted, replaced with nonimmune serum, or preabsorbed with the corresponding antigens ( 10m5- 10d6 M) for 24 h at 4°C. Labeled secondary antisera also exhibited no cross-reactivity with IgG from inappropriate species. For the demonstration of AChE activity, the following control incubation media were used: 1) 10-4-10-6 M tetraisopropyl-pyrophospharamine (iso-OMPA; Sigma) was added to the incubation medium as an inhibitor for nonspecific cholinesterases. (2) preparations were incubated in a substratefree medium omitting acetylthiocholine.

INNERVATION

OF HUMAN

SUPERFICIAL

TEMPORAL

Vasomotor Responses In Vitro Superficial temporal arteries were obtained from 18 patients during neurosurgical tumor resections. Immediately after excition, vessel segments were immersed in aerated (5% COP in 0,) buffer solution (4’C) and transported to the laboratory for studies of vasomotor reactivity. Two to three millimeter-long ring segments of the superficial temporal artery were suspended between two L-shaped metal prongs (0.2 mm) in small tissue baths containing a buffer solution aerated with 5% CO2 and 95% OZ. The vessels were given a passive load of 2-4 mN depending on the vessel size, and were allowed to stabilize at this tension for an equilibration period of 1.5 h before isometric circular contractions were recorded. The contractile capacity of the preparations was first tested by exposure to a buffer solution containing 60 mM potassium; this resulted in strong contractions of temporal arteries: 23.0 -C 2.0 mN (n = 49). Only vessels showing reasonably strong (> 1 mN) and reproducible (< 10% variation between two tests) responses were used in this study. Prostaglandin Fti (PGF,,) 3 x 10e6 M was used as a contractile agent and evoked a stable level of tension: 11.6 ? 1.1 mN. During this precontraction, relaxant responses to acetylcholine and the peptides were examined. In experiments with NPY, the peptide was applied alone in increasing concentrations. When the potentiating capacity of NPY was examined it was given in a concentration of 3 X 10m8 A4 5 min prior to a single concentration of noradrenaline. A modulating effect of NPY (3 X lo-’ M, 5 min before test) on relaxations induced by VIP, SP, and CGRP, NPY was performed in PGFti-precontracted vessels, In all studies controls were run in parallel. To remove the endothelium, arterial segments were perfused during 20 s with buffer solution containing 0.1% Triton X-100. Successful removal of the endothelium was always demonstrated by the absence of relaxation to acetylcholine. The in vitro data are expressed as EC5,, or I&, values (concentration of agonist eliciting half-maximum contraction or relaxation, respectively) and as E,,,,, or I,,,,, (maximum contraction or relaxation, respectively). Data are presented as the means -C SEM of the responses obtained in a given number (n) of vessels segments. one or two from each patient. Solutions and Drugs The buffer solution used was of the following composition (mM): NaCl, 119; NaHCO,, 15; KCl, 4.6; CaC&, 1.5; MgCl*, 1.2; NaH2P04, 1.2; and glucose, 11. The potassium buffer was obtained by an equimolar substitution of NaCl for KCl, resulting in a potassium concentration of 60 mM. Pharmacological

Agents

Acetylcholine hydrochloride, noradrenaline hydrochloride (Sigma, USA), propranolol (Inderal@, ICI, UK), and prostaglandin Fti (Amoglandin@; Astra, Sweden) were diluted in 0.9% saline. Human (u-CGRP, NKA, PHM-27 (CRB, Cambridge, UK), NPY, SP, and VIP (Sigma, USA) were dissolved in 0.9% saline containing 1% bovine serum albumin (BSA) and 0.1 mM ascorbic acid. Prazosin HCl (Fermion, Orion Corp. Ltd., Finland) was dissolved in methanol and 1 mM HCl, giving a final concentration of 0.1 r&f, and then further diluted in 0.9% saline. Statistics Statistical analysis of the results was performed by using Kruskal-Wallis test (two-tailed) followed by Mann-Whitney IItest for comparison between each individual group. In the experiments of NPY potentiation of NA-induced contraction in

217

ARTERY

rubbed vessels, statistics were calculated by paired Student’s ttest. A probability value of 0.05 was accepted as significant for differences between groups of data. RESULTS

Light Microscopical

Immunohistochemistty

and Histochemistry

Immunofluorescence staining with the antiserum to the general neuronal marker protein gene product (PGP) 9.5 demonstrated that the human superficial temporal artery possesses a very dense supply of nerves fibers in the adventitia (Fig. 1). The distribution of this perivascular innervation was arranged in two layers, consisting of an outer layer of nonvaricose nerve fibers and fascicles and an inner layer of varicose and nonvaricose nerve fibers that mainly run parallel to the long axis of the vessel at the adventitial-medial border (Fig. 1). The majority of the nerve fibers displayed immunoreactivity for tyrosine hydroxylase (TH) (Fig. 2) and NPY (Fig. 3). It was also observed that the number and the distribution of TH-immunoreactive nerves fibers paralleled that of nerve fibers containing NPY immunoreactivity (Figs. 2, 3). The superficial temporal artery was also found to be innervated by a moderate supply of nerve fibers containing AChE activity (Fig. 4) and VIP (Fig. 5), PHM (Fig. 6) and CGRP (Fig. 7) immunoreactivity. In contrast, relatively few nerve fibers displaying SP (Fig. 8) and neuropeptide K (Fig. 9) immunoreactivity were observed. The use of a double immunofluorescence staining technique revealed that SP immunoreactivity was colocalized with neuropeptide K and CGRP immunoreactivities in the same varicose nerve fibers (Figs. 10, 11). Conventional

Electron Microscopy

Electron microscopical examination of the human superficial temporal artery revealed the presence of numerous unmyelinated axons in the adventitia. At the adventitial-medial border, axon varicosities were separated from adjacent smooth muscle cell membranes by a cleft 100 to several hundred nanometers wide (Fig. 12). At sites where the neuromuscular gap was 100 nm wide, fusion of the basal laminae that surrounds the axon varicosity and the smooth muscle cell was observed (Fig. 12). Although axon varicosities represent a spectrum of profiles containing a heterogeneous vesicle population, seven principal types were distinguishable. 1. Varicosities containing a heterogeneous population of small round (40-60 nm in diameter) and flattened agranular (2040 nm in diameter and extending up to 80 nm) vesicles, as well as a few occasional large granular ones (80- 150 nm in diameter) (Fig. 12). 2. Varicosities containing a large number of small granular vesicles (40-60 in diameter) and a few large granular vesicles (80-150 nm in diameter). These varicosities also possessed varying numbers of small vesicles with indistinct granular cores as well as agranular vesicles (Figs. 13, 14). 3. Varicosities containing numerous large granular vesicles (80150 nm in diameter) together with a varying number of small agranular and/or granular vesicles (Fig. 13). 4. Varicosities containing large amounts of glycogen-like granules (Fig. 14). 5. Varicosities containing numerous osmiophilic bodies, broken mitochondria, and autophagic vacuoles together with vesicles of varied size, shape, and electron density (Fig. 15). 6. Varicosities containing tightly packed mitochondria as well as aggregates of glycogen-like granules (Fig. 16).

278

JANSEN

OLESEN

FIGS. l-9. Whole-mount preparations of human superficial temporal artery immunostained for PGP 9.5 (Fig. I), TH (Fig. 2). NPY (Fig. 3). VIP (Fig. 5). PHM (Fig. 6). CGRP (Fig. 7). substance P (SP) (Fig. 8), and neuropeptide K (NPK) (Fig. 9). Figure 4 shows perivascular nerve fibers displaying acetylcholinesterase (AChE) activity. Bar = 50 nm.

ET AL

INNERVATION

FIGS.

IO-

OF HUMAN

11.Whole-mount

SUPERFICIAL

TEMPORAL

ARTERY

279

preparations of human superficial temporal artery. The colocalization of substance P (SF’)with CGRP [Fig. [Fig. 1l(a.b)] was determined by double immunostaining of the same preparations. Bar

IO(a,b)] as well as with neuropeptide K (NPK) =50nm

7. Varicosities (Fig. 17).

containing

Electron Microscopicul

numerous lamellated or myelin bodies Immunocytochemist~

Postembedding immunogold staining demonstrated that NPY immunoreactivity consistently occurred in large granular vesicles (80- 150 nm in diameter) in varicosities that also contained numerous small granular vesicles (40-60 nm in diameter) (Fig. 18). VIP and CGRP immunoreactivities were found in the same type of large electron-dense vesicles. VIP-positive large granular vesicles were usually observed in varicosities that contained a large number of small, round, and flattened agranular vesicles (Fig. 19), whereas CGRP-immunoreactive vesicles were localized in varicosities that displayed only a few scattered small, round, agranular vesicles (Fig. 20). In Vitro Pharmacology Noradrenaline caused a concentration-dependent contraction of the superficial temporal artery (Table 2). The a,-adrenoceptor

antagonist prazosin ( 10m9- 10e7 kf) caused a significant shift of the concentration-response curve of noradrenaline towards higher noradrenaline concentrations with a pA2 value of 8.87 (Fig. 21a). NPY (3 x lo-” M) induced a slight but nonsignificant contraction in 9 out of 32 of the arterial segments examined. The vessel segments responding lo NPY were obtained from two of seven patients. The contraction induced by NPY was 4.0 k 1.4 mN or 15.7 2 4.7% of the contraction induced by 60 mkJ K+. However, in arteries from five patients NPY (3 X lo-’ M) significantly shifted the noradrenaline-induced concentration-response curve towards lower noradrenaline concentrations (i.e., potentiation) (Table 2) (p < 0.05). In arteries from two patients, NA-induced contractions were not potentiated in the presence of NPY (3 x IO-’ kf). When NPY (3 X lo-’ kf) was given 5 min prior to a single concentration of NA, a IO- 15 times stronger contraction was observed at the lower NA (lo-‘- 10m6kf) concentrations (Fig. 21 b). The response to noradrenaline was not changed by endothelium removal (Fig. 21~). Furthermore, the

JANSEN

OLESEN

FIGS. I2- 17. Electron micrographs of nerve varicosities at the adventitial-medial border of the human superficial temporal artery Axon varicosities are found in close apposition to smooth muscle cells (SM) with only basal lamina (BL) material between them (Fig. 12). The minimum space between axon varicosities and muscle cell membranes (double headed arrows) is in the order of 100 nm (Fig. 12). Several axon varicosities may be demonstrated containing numerous small, round (arrowheads), and flattened (arrows) agranular vesicles (Fig. 12). small granular vesicles (Figs. 13- 14, arrows), or mainly large granular vesicles (Fig. 13. asterisk). Figure 14 contains an axon bundle and displays a varicosity possessing a large amount of glycogen-like granules (asterisk). The axon varicosity in Fig. 15 shows degenerative features, including autophagic vacuoles (large arrows) and osmiophilic bodies (arrowheads). Large vesicles (thin arrows) are also present in this varicosity. Figure 16 displays an axon varicosity containing tightly packed mitochondria (M) together with aggregates of glycogen-like granules (arrowheads). Numerous lamellated or myelin bodies are also present in some varicosities (Fig. 17). SC, Schwann cell. Bar = 200 nm.

ET AL

INNERVATION

OF

HUMAN

SUPERFICIAL

TEMPORAL

ARTERY

281

FIGS. 18-20. Electron micrographs demonstrating the ultrastructural localization of NPY (Fig. 18). VIP (Fig. 19). and CGRP (Fig. 20) immunoreactivity in nerve varicosities of the human superficial temporal artery. In Fig. 18. the large granular vesicles (arrowheads), but not the small granular vesicles (arrows), display NPY immunogold labeling (I 5-nm gold particles). In Fig. 19. VIP immunogold labeling (arrowheads) is localized over large granular vesicles in varicosities that also contain numerous unlabeled small, round (large arrows), and flattened (thin arrows) agranular vesicles. Figure 20: CGRP immunogold labeling (arrowheads) is concentrated over large granular vesicles in axon varicosities that contain only occasional small agranular vesicles (arrows). SC. Schwann cell. Bar = 150 nm

effect of NPY on NA-induced contractions was also observed in vessels without endothelium (Fig. 2la). ACh, VIP, PHM, SP, NKA, and CGRP did not alter the resting level of tension but caused relaxation of PGF?,-precontracted arteries (Fig. 22, Table 2). The order of potency for the parasympathetic-associated agents was VIP > PHM = ACh, with no difference in maximum effects (Table 2). In vessel segments with the endothelium removed, ACh only induced a weak relaxation whereas the response to VIP and PHM was unchanged. Furthermore, in vessel segments with intact endothelium, ACh caused a rapid relaxation that was quickly reversed. In vessels without endothelium, the relaxation slowly developed and occurred at higher ACh concentrations: 3 X lo-* M of NPY had no modulatory effect on ACh- or VIP-induced relaxations (Fig. 23).

potentiating

Peptides associated with sensory nerves were found to induce potent vasodilatory responses. The order of potency was CGRP > SP > NKA. CGRP was a stronger relaxant agent compared to SP and NKA (Table 2). The response induced by CGRP occurred in a wide range of concentrations (10ei3- 10m7M), and in some of the experiments a clear biphasic relaxation was observed. The first phase was seen with CGRP at concentrations of 10-‘5- lo-” A4whereas the second phase occurred between lo-” and 10m7M (Fig. 22). The response to CGRP was unchanged in vessels without endothelium (Fig. 22). In contrast, SP and NKA failed to induce relaxation in vessels where the endothelium had been removed beforehand. NPY given 5 min before the addition of SP or CGRP did not alter the responses to these relaxant agents (Fig. 23).

282

JANSEN

TABLE

2

CHARACTERISTCS OF THE CONCENTRATION-DEPENDENT CONTRACTILE AND RELAXANT RESPONSES OF THE HUMAN SUPERFICIAL TEMPORAL ARTERY. EXPLAINED AS pDL OR E,,x (I,,,, VALUES Contractile Responses Agonm

NA NA + NPY

PDZ

6.2 -c 0.1 7.0 t 0.1*

95 2 5 97 t- 5

32 (7) 32 (7)

ACh

7.0 5 0.1

712

VIP PHM-27

7.3 t 0.2 6.9 2 0.3

81 25 76 5 8

11

6 (3) 11 (6) 6 (3)

Relaxant Responses PD:

SP NKA CGRP

8.7 t 0.2 8.4 5 0.3 9.4 z 0.3

I,,,(R)

II

70 t 4 65 -c 9 85 -c 4

13 (9) 9 (6) 19 (IO)

II = number of experiments from the number parentheses. For further details see text. * p < 0.05 Mann-Whitney U-test.

of patients

shown

in

DISCUSSION

The present study indicates that the human superficial temporal artery is supplied by a very rich innervation. It was also demonstrated that nerve varicosities were often found in close association with the smooth muscle cells, with lOO-nm-wide neuromuscular junctions and fused basal lamina. Although we did not find pre- or postsynaptic membrane specializations, our observations provide morphological evidence that some nerve terminals may be capable of affecting the artery tone directly. A spectrum of axon varicosity profiles was identified at the ultrastructural level. Seven principal types appeared to be distinguishable that are similar to axon varicosities previously identified in the mammalian vascular system (4,23,32). They include nerve profiles that contain populations of small, round, and flattened agranular vesicles together with some large granular vesicles that correspond to terminals described as being cholinergic. Other varicosities may be distinguished on the basis of the numerous small granular vesicles they contain and that are thought to represent noradrenergic axons. A further axon profile contained a heterogenous population of large granular or so called p-type (peptidergic-containing) vesicles (1). Four other axons profiles were interpreted as sensory because they displayed ultrastructural features similar to the ones described for presumed sensory or baroreceptor nerve terminals (5,23,32,37). These distinguishing features include an unusual abundance of mitochondria, autophagic vacuoles, pleomorphic dense bodies, lamellated or myelin figures, and large amounts of glycogen-like granules. Distribution

of NPY Fibers

NPY-immunoreactive nerve fibers were the most abundant of all peptide-containing nerve populations identified in the superficial temporal artery. These nerve fibers exhibited a similar dis-

ET AL.

tribution to that of nerves containing the catecholamine-synthesizing enzyme TH and are thought to represent noradrenergic sympathetic neurones (1.539). Using immunogold staining procedures, it was possible to demonstrate that NPY immunoreactivity is localized in large granular vesicles in nerve varicosities that also contain numerous small granular vesicles. These observations are in agreement with several biochemical studies showing that in sympathetic nerves NA occurs mainly in the small granular vesicles, whereas NPY is contained, together with some NA, in the large granular vesicles (8,19). Vasomotor

Relaxant Responses

OLESEN

Responses

to Noradrenaline

and NPY

When exposed to NA, the superficial temporal artery displayed a strong contractile response. The contraction was blocked by the a,-adrenoceptor antagonist prazosin. The pA2 value correlated with that reported for other human vascular tissues with cY,-adrenoceptors (42). In general, NPY did not act as a vasoconstrictor; however, a weak contractile response was observed in two out of seven subjects. This is different from studies performed on human cerebral and middle meningeal arteries where NPY per se was found to be a stronger and more potent constrictor than NA (12,35). In most superficial temporal arteries examined, NPY caused potentiation of the NA-induced contractile responses. Arteries from two patients, however, did not respond either with contraction or NA-enhancing effects. A NAenhancing effect has been noted in human omental arteries (14) and in peripheral (femoral, mesenteric) arteries of various laboratory animals (1549). The potentiating effect of NPY on NAinduced contraction is in favor of a frequency-dependent differential release of NPY and NA (38); for example, NA is released from small vesicles at low stimulation frequences, whereas the neuropeptide is preferentially released from large vesicles at high stimulation frequencies to potentiate the effect of NA. The NPYinduced potentiation (lo- 15.fold increase) of the NA-induced contractile response was noticeable at lower NA concentrations. It appears that NPY may have an important role in the regulation of transmitter economy at the neuroeffector junction in the human superficial temporal artery (11). In some vascular tissues. NPY may also downregulate vasodilator responses to, for example, acetylcholine, SP and VIP (17). However, such an effect was not observed in the present study. Thus, the main effect of NPY in the human superficial temporal artery may be to potentiate the action of the sympathetic transmittor NA. Distribution

of VIP and PHh4 Fibers

A moderate supply of nerve fibers positive for VIP and PHM, two peptides derived from the same precursor molecule (3 1), was observed in the superficial adventitia of the temporal artery. It was also found that the relative number and distribution of VIPand PHM-immunoreactive nerve fibers was similar to that of AChEcontaining nerve fibers. These observations correlate well with the results of previous studies showing that VIP-immunoreactive nerve fibers supplying the cerebral vasculature contain AChE activity and choline acetyltransferase immunoreactivity and arise from parasympathetic cholinergic neurons (24,28,33,45). This is further supported by our ultrastructural observations showing that VIP-immunoreactive large granular vesicles are found within nerve varicosities that also contain small, round, and flattened agranular vesicles and are presumed to represent cholinergic nerves. It should be noted, however, that not all cholinergic nerve fibers may be positive for VIP, immunoreactivity, or vice versa (40).

INNERVATION

OF HUMAN

H

SUPERFICIAL

TEMPORAL

283

ARTERY

Control

l Prazosin 10m9M +

Prazosin

10m8M

A

Prazosin

10m7M

_;

-j

_;

Noradrenaline A

-;

-i

-6.5

-7

-6

-5.5

NORADRENALINE

log cont. (M)

-5

LOG CONC.

-4.5 (M)

B

loo w 8.

.

l Control n +NPY 3xdM

G @

Log [noradrenaline) (M)

C FIG. 21. (a) Concentration-dependent contractions of human superficial temporal artery to noradrenaline in the presence of prazosin 10m9- 10m7 M. Mean values k SEM. n = 6-8. (b) Shows the ratio of contraction induced by noradrenalin in the presence of 3 X IO-"M NPY and without NPY at different noradrenaline concentrations. (c) Shows concentration-dependent contractions of human superficial temporal arteries to noradrenaline (0) and noradrenaline in the presence of 3 x lo-” M NPY (W) in vessel segments without endothelium. Mean values -t SEM, n = 4-6. Statistical analysis using Mann-Whitney U-tests. In (c) paired Student’s t-tests was used. *p < 0.05.

Vasomotor Responses to ACh. VIP, and PHM-27 AC4 VIP, and PHM-27 induced relaxation to the same degree and potency in the superficial temporal artery. The mechanism of action of VIP and PHM involves an increase in the adenylate cyclase activity within the vessel wall (10,30,44). Acetylcholine, on the other hand, induces relaxation through the release of an endothelium-derived relaxing factor (20) that has now been identified as nitric oxide (41). This agent stimulates smooth muscle guanylate cyclase to elicit dilatation (3). Thus, the different

agents stored in parasympathetic nerve terminals may modulate tone via different postjunctional mechanisms. We have now demonstrated in human superficial temporal artery that ACh, but not VIP, induces relaxation via an endothelium-dependent mechanism. Furthermore, it was shown that the response to ACh was faster in onset than the one induced by VIP.

Distribution of CGRP, SP, and Neuropeptide K Fibers The superficial temporal artery received a moderate CGRPimmunoreactive nerve supply. The relative density of CGRP-

284

JANSEN OLESEN ET AL

With endothelium

Without endothelium -9 8

Log cont. (M) -7 -9.58 -5 4

Log cont. (M) -3.5

-9 -9 -7 -5.5 -5

-5

-3.5

4

0 20 40 60 60 100

I

ACh

Log cont. (M)

Log cont. (M) -12 -11 -10 -8

-6

-7 -6.5 -6

-12 -11 -10 -9

-8

-7 8.5

0

8

20 40 60 60 100

Log cont. (M) -11 -10 -9

Log cont. (M)

-6 -7.5 -7 8.5

-11 -10 -9

-6

0

0

20

20

40

40

60

60

80

80

100

100

-9

-8

-6

Log cont. (M)

Log cont. (M) -13 -12 -11 -10

-8 -7.5 -7 6.5

-7

-13 -12 -11 -10

J

-9

-6

-7

-6

20 40 60 80 100 FIG. 22. Dilatory response to acetylcholine, SP, VIP, and CGRP of human superficial temporal arteries that have been precontracted by 3 x 10m6M PGF,,. The responses of vessels with intact endothelium are shown to the right and the responses of arteries where the endothelium had been removed beforehand are shown to the left. Mean values 2 SEM. n = 6-8.

INNERVATION

OF HUMAN

SUPERFICIAL

TEMPORAL

285

ARTERY

Log cont. (M)

Log cont. (M)

-9

-8

-7 -6.5 -6

-5

-4 -3.5

g

-10 -9

-13

-12

-8 -7.5 -7 -6.5 -6

20

.g 40 ‘r;;

.5 40 t;i

i d

i d

60

-11

60 80

80

-12

Log cont. (M) -8

-11 -10 -9

Log cont. (M)

-7 -685 -6

-11 -10 -9

-8

-7

-6

80

FIG. 23. Dilatory response to ACh, VIP, SP. and CGRP with (m) and without (0) pretreatment with 3 X lo-” M NPY. The experiments were performed on human superficial temporal arteries precontracted with 3 X 10m6M PGF?,. Mean values -t SEM, n = 6-8.

immunoreactive perivascular nerves appears to be between that of nerves displaying NPY and VIPiPHM immunoreactivities. In contrast, relatively few nerve fibers containing SP and neuropeptide K immunoreactivity were identified in the superficial temporal artery. It was also shown that SP and neuropeptide K coexisted with CGRP in the same nerve fibers. These observations correlate well with the results of previous studies showing that tachykinins and CGRP coexist in capsaicin-sensitive neurons (21,29,51) and are costored in the same p-type vesicles in both sensory ganglia and perivascular nerve terminals of the guinea pig ( 13,27). Vasomotor Responses

to SP. Neurokinin

A, and CGRP

Substance P, neurokinin A, and CGRP acted as potent relaxant agents on the superficial temporal artery. CGRP induced a more potent vasodilator response than that elicited by the two tachykinins. CGRP also exhibited a biphasic pattern of response. Thus, there may exist both low- and high-affinity binding sites for a-CGRP in the human superficial temporal artery. However, more detailed analysis of these will be required in the future. In laboratory animals, the vasomotor responses to a-CGRP occur concomitantly with the activation of adenylate cyclase, whereas SP and neurokinin A require an intact endothelium to induce an adequate response (10,34). Similar results were obtained for the human superficial temporal artery where SP and NKA required the presence of an intact endothelium to exert their relaxant effect and CGRP did

not. In the guinea pig basilar, coronary, and uterine arteries, the relaxant responses induced by ACh, VIP, and SP were shown to be blocked by pretreatment with a single concentration of NPY (9). However, in the human superficial temporal artery, NPY did not modify the relaxant responses to ACh, VIP, SP, or CGRP when given in the same concentrations as potentiation occurs (i.e., pharmacologically active). It is known that NPY in the guinea pig basilar and uterine arteries induces contraction; thus, the mechanisms of action appear to be different from that found in our preparation. In conclusion, we have morphological evidence that the superficial temporal artery is richly innervated by nerve fibers containing vasoactive substances. The stored agents have strong vasomotor effects. The potentiating effect of NPY on NA-induced contractions was not dependent on an endothelium while relaxation induced by ACh and SP was dependent on an intact endothelium. The biphasic response induced by CGRP suggests the pretense of low- and high-affinity binding sites to this peptide. ACKNOWLEDGEMENTS Antisera raised to regulatory peptides at the Hammersmith Hospital were produced in conjunction with Prof. S. R. Bloom. The authors are grateful to Dr. K. Valentino for providing antiserum to neuropeptide K. We thank Mrs. M. R. Alpiaqa for excellent technical assistance. This work was supported in part by the Swedish Medical Research Council (grant No. 05958) Faculty of Medicine, University of Lund, and the Swedish Society of Medicine.

286

JANSEN

OLESEN

ET

AL.

REFERENCES 1. Baumgarten, H. G.; Holstein, A. F.; Owman, C. Auerbach’s plexus of mammals and man: Electron microscopic identification of three different types of neuronal processes in myenteric ganglia of large intestine from rhesus monkeys, guinea pig and man. Z. Zellforsch. 106:376-397; 1970. 2. Bendayan, M.; Zollinger, M. Ultrastructural localization of antigenic sites in osmium-fixed tissues applying the protein A-gold technique. J. Histochem. Cytochem. 3 1: IOl- 109; 1983. 3. Boheme, E.; Grossmann, G.; Herz. J.; et al. Regulation of cyclic GMP formation by soluble guanylate cyclase: Stimulation by NOcontaining compounds. In: Greengard, P.. ed. Advances in cyclic nucleotide and protein phosphorylation research. New York: Raven; 1984:259-266. 4. Burnstock, G.; Charnley, J. H.; Campbell, G. R. The innervation of arteries. In: Schwartz, J.; Werhessen, N. T.; Wolf, S.. eds. Structure and function of the circulation, vol. 1. New York: Plenum Publishing Corp.; 1980:729-767. 5. Chiba, T. Fine structure of the baroreceptor nerve terminals in the carotid sinus of the dog, J. Electron Microsc. 21: 139- 148; 1972. 6. Cowen, T.; Haven, A. J.; Bumstock, G. Pontamine sky blue: A counterstain for background autofluorescence in fluorescence and immunofluorescence histochemistry. Histochemistry 82:205-208: 1985. 7. Dalessio, D. J. Wolff s headache and other head pain. New York: Oxford University Press: 1980. 8. De Potter, W. P.; Dillen, L.; Annaert, W.; Tombeur, K.; Bergmans, R.; Coen, E. P. Noradrenaline from large dense cored vesicles in sympathetic nerves of the bovine vas deferens. Synapse 2: 157- 162; 1988. 9. Edvinsson, L.; Adamsson, M. Neuropeptide Y inhibits relaxation of guinea pig cerebral, coronary and uterine arteries: Blockade by D-myo-inositol-1.2,6-trisphosphate. J. Cardiovasc. Pharmacol. 20:466-472; 1992. 10. Edvinsson, L.; Fredholm, B. B.; Hamel, E.; Jansen, I.; Verrecchia. C. Perivascular peptides relax cerebral arteries concomitant with stimulation of cyclic adenosine monophosphate accummulation or release of an endothelium-derived relaxing factor in the cat. Neurosci Lett. 58:213-217; 1985. 11. Edvinsson. L.; Ekman, R.; Jansen, I.; Ottosson, A.: Uddman, R. Peptide-containing nerve fibers in human cerebral arteries: Immunocytochemistry, radioimmunoassay, and in vitro pharmacology. Ann. Neurol. 21:431-437; 1987. 12. Edvinsson, L.; H?rkanson, R.; Wahlestedt C.; Uddman, R. Effects of neuropeptide Y on the cardiovascular system. Trends Pharmacol. Sci. 8:231-235; 1987. S.; Jansen, I.; Wharton, J.; Cervantes, 13. Edvinsson, L.; Gulbenkian. C.; Polak, J. M. Comparison of peptidergic mechanisms in different parts of the guinea pig superior mesenteric artery: Immunocytochemistry at the light and ultrastructural levels and responses in vitro of large and small arteries. J. Auton. Nerv. Syst. 28: 141- 154; 1989. 14. Edvinsson. L.; H&anson, R.; Steen, S.; Sundler, F.; Uddman, R.; Wahlestedt. C. Innervation of human omental arteries and veins and vasomotor responses to noradrenaline, neuropeptide Y, substance P and vasoactive intestinal peptide. Regul. Pept. 12:67-79; 1985. 15. Ekblad, E.; Edvinsson, L.; Wahlestedt, C.; Uddman, R.; H%kanson, R.; Sundler, F. Neuropeptide Y co-exists and co-operates with noradrenaline in perivascular nerve fibers. Regul. Pept. 8:225-235; 1984. A.; Shenk, E. A. Histochemical methods for separate, 16. El-Badawi, consecutive and simultaneous demonstration of acetylcholinesterase and norepinephrine in cryostat sections, J. Histochem. Cytochem. 1:580-588; 1967. 17. Fallgren, B.; Ekblad, E.; Edvinsson, L. Co-existence of neuropeptides and differential inhibition of vasodilator responses by neuropeptide Y in guinea pig uterine arteries. Neurosci. Lett. 100:71-76; 1990. 18. Friberg, L.; Olesen, J.; Iversen, H. K.; Sperling, B. Migraine pain associated with middle cerebral artery dilatation: Reversal by sumatriptan. Lancet 338:13- 17; 1991. 19. Fried, G.; Terenius, L.; Hokfelt, T.: Goldstein, M. Evidence for differential localization of noradrenaline and neuropeptide Y in neu-

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36

37 38

ronal storage vesicles isolated from bovine vas deferens, J. Neurosci. 5:450-458: 1985. Furchgott. R. F.; Zawadzki, J. V. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 288:373-376; 1980. Gibbins, I. L.; Furness, J. B.; Costa, M.; Maclntyre, 1.; Hillyard, C. J.; Girgis. S. Co-localization of calcitonin gene-related peptidelike immunoreactivity with substance P in cutaneous vascular and visceral sensory neurons of guinea pigs. Neurosci. Lett. 57:125130; 1985. Goadsby, P. J.; Edvinsson, L.: Ekman, R. Vasoactive peptide release in the extracerebral circulation of humans during migraine headache. Ann. Neurol. 28: 183- 187; 1990. Gulbenkian. S.; Edvinsson. L.; Saetrum Opgaard, 0.; Wharton, J.; Polak, J. M.; David-Ferreira, J. F. Peptide-containing nerve fibers in guinea-pig coronary arteries: Immunohistochemistry, ultrastructure and vasomotility, J. Auton. Nerv. Syst. 31:153- 168; 1990. Gulbenkian, S.; Valenca. A.; Wharton, J.: Polak, J. M.; David-Ferreira, J. F. Simultaneous visualization of neuropeptide and acetylcholinesterase nerve subpopulations in the perivascular plexus. Histochem. J. 23:553-558; 1991. Gulbenkian. S.; Wharton, J.; Polak, J. M. The visualization of cardiovascular innervation in the guinea pig using an antiserum to protein gene product 9.5 (PGP 9.5). J. Auton. Nerv. Syst. 18:235-247; 1987. Gulbenkian, S.; Wharton, J.; Hacker. G. W.; Varndell, 1. M.; Bloom, S. R.; Polak, J. M. Co-localization of neuropeptide tyrosine (NPY) and its C-terminal flanking peptide (C-PON). Peptides 6:12371243; 1985. Gulbenkian, S.: Merighi, A.; Wharton, S.; Vandell, I M.; Polak, J. M. Ultrastructural evidence for the coexistence of calcitonin generelated peptide and substance P in sensory vesicles of peripheral nerves in the guinea pig. J. Neurocytol. 15:535-542; 1986. Hara, H.; Hamill. G. S.; Jacobowitz, D. M. Origin of cholinergic nerves to the rat major cerebral arteries: Coexistence with vasoactive intestinal polypeptide. Brain Res. 14:179- 188; 1985. Hua, X.-Y.; Theodorsson-Norheim, E.; Brodin, E.: Lundberg, J. M.; Hokfelt, T. Multiple tachykinins (neurokinin A. neuropeptide K and substance P) in capsaicin-sensitive neurons in the guinea pig. Regul. Pept. 13:1-19; 1985. Huang. N.; Rorstad, 0. P. Cerebral vascular adenylate cyclase: Evidence for coupling to receptors for vasoactive intestinal peptide and parathyroid hormone. J. Neurochem. 43:849-856; 1984. Itoh, N.; Obata. K.-I.; Yanaihara. N.: Okamoto, H. Human preprovasoactive intestinal polypeptide contains a novel PHI-27 like peptide, PHM-27. Nature 304:547-549; 1983. Iwayama, T.; Furness, J. B.; Bumstock, G. Dual adrenergic and cholinergic innervation of the cerebral artery of the rat. An ultrastructural study. Circ. Res. 26:653-646: 1970. Jacobowitz. D. M.; Hara, H.; Kobayashi. S. Autonomic innervation of the major blood vessels of the brain. In: Edvinsson, L.; McCulloch. J., eds. Peptidergic mechanisms in the cerebral circulation Weinheim: VCH Verlagsgesellschaft; 1987:48-63. Jansen, I.; Alafaci, C.; McCulloch, J.; Uddman. R.; Edvinsson, L. Tachykinins (substance P, neurokinin A, neuropeptide K and neurokinin B) in the cerebral circulation: Vasomotor responses in vitro and in situ. J. Cereb. Blood Flow Metab. 11:567-575; 1991. Jansen, I.; Uddman, R.; Ekman. R.; Olesen. J.; Ottosson. A.; Edvinsson, L. Distribution and effects of neuropeptide Y. vasoactive intestinal polypeptide, substance P and calcitonin gene-related peptide in human middle meningeal artery: Comparison with cerebral and temporal arteries. Peptides 13:527-536; 1992. Jansen, I.; Uddman, R.; Hocherman, M.; et al. Localization and effects of neuropeptide Y. vasoactive intestinal polypeptide. substance P, and calcitonin gene-related peptide in human temporal arteries. Ann. Neural. 20:496-501; 1986. Kraus, J. M. Structure of rat aortic baroreceptors and their relationship to connective tissue. J. Neurocytol. 8:401-414; 1979. Lundberg, J. M.; Rudehill, A.; Sollevi, A.; Fried, G.; Wallin, G. Corelease of neuropeptide Y and noradrenaline from pig spleen in viva;

INNERVATION OF HUMAN SUPERFICIAL TEMPORAL ARTERY

39.

40.

41.

42.

43. 44.

45.

importance of subcellular storage, nerve impulse frequency and pattern, feedback regulation and resupply by axonal transport. Neuroscience 28:475-486; 1989. Lundberg. J. M.; Terenius, L.; Hokfelt, T.; et al. Neuropeptide Y (NPY)-like immunoreactivity in peripheral noradrenergic neurons and effects of NPY on sympathetic function. Acta Physiol. Stand. 116:477-480; 1982. Miao. F.J.-P.; Lee. T.J.-F. Cholinergic and VIPergic innervation in cerebral arteries: A sequential double-labelling immunohistochemical study. J. Cereb. Blood Flow Metab. 10:32-37: 1990. Palmer. R. M. J.; Ferrige. A. G.; Moncada, S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 3275244526; 1987. Skiby. T. V. C.: Andersson, K.-E.; Edvinsson, L. Pharmacological characterization of postjunctional a-adrenoceptors in isolated feline cerebral and peripheral arteries. Acta Physiol. Stand. 117:63-73; 1983. Steffanini, M.; De Martino, C.; Zamboni, I. Fixation of ejaculated spermatozoa for electron microscopy. Nature 2 16: 173- 174: 1967. Suzuki. Y.; McMaster, D.; Lederis. K.; Rorstad, 0. P. Characterization of the relaxant effects of vasoactive intestinal peptide (VIP) and PHI on isolated brain arteries. Brain Res. 322:9- 16; 1984. Suzuki. N.; Hardebo, J. E.; Owman, C. Origins and pathways of choline-acetyltransferase-positive parasympathetic nerve fibers to

46.

47.

48.

49.

50.

51.

cerebra1 blood vessels in rat. J. Cereb. Blood Flow Metab. 10:399408; 1990. Tago. H.; Kimura, H.; Maeda, T. Visualization of detailed acetylcholinesterase fiber and neuron staining in rat brain by a sensitive histochemical procedure. J. Histochem. Cytochem. 11:143 1- 1438; 1986. Uddman, R.; Edvinsson, L.; Hara, H. Axonal tracing of autonomic nerve fibers to the superficial temporal artery in the rat. Cell Tissue Res. 256:559-565; 1989. Varndell. I. M.; Tapia, F. J.; Probert, L.; et al. Imunogold staining procedure for the localization of regulatory peptides. Peptides 31259-272; 1982. Wahlestedt, C.; Edvinsson. L.; Ekblad, E.; Hakanson, R. Neuropeptide Y potentiates noradrenaline-evoked vasoconstriction: Mode of action. J. Pharmacol. Exp. Ther. 234:735-741; 1985. Wharton. J.; Gulbcnkian, S.; Merighi, A.; et al. Immunohistochemical and ultrastructural localisation of peptide-containing nerves and myocardial cells in the human atrial appendage. Cell Tissue Res. 254:155-166: 1988. Wharton. J.; Gulbenkian, S.; Mulderry. P. K.; MacGregor, G. P.; Bloom, S. R.; Polak. J. M. Capsaicin induces a depletion of calcitonin gene-related peptide (CGRP)-immunoreactive nerves in the cardiovascular system of the guinea pig and rat. J. Auton. Nerv. Syst. 16:535-542; 1986.