- Email: [email protected]
Calcitonin gene-related peptide in cardiovascular tissues of the rat
CALCITUNIN GENE-RELATED PEPTIDE fN CARDIOVASCULAR TISSUES OF THE RAT P. K. MULDBRRY,M. A. GWTEI, J. RODRIGO, J. M. ALLEN, M. G. ROSEWELD*,J. M. POLAK and S. R. BL.~oM~ Departments of Medicine and Histochemistry,
Royal Postgraduate London Wf2 OHS, U.K.
Medical School, Du Cane Road,
Abstract-The distribution of calcitonin gene-related peptide immuuoreactivity in the cardiovascular system of the rat was investigated by radioimmunoassay and immunocytochemistry. The nature of the immunoreactivity was studied by gel permeation and high performance liquid chromatography. Immunocytochemistry demonstrated the existence of calcitonin gene-related peptide-containing nerve fibres throughout the cardiovascular system. These were present in all regions of the heart, particularly in association with the coronary arteries, within the pap&try mu&es and within the sinoatrial and at~oven~cu~ar nodes. Cakitonin gene-related ~ptid~~~in~ng flbres were found mainly in the adventitia of the arteries and veins. Calcitonin gene-related peptide coucentrations were high br major arteries and veins but comparatively low in the heart, aortic arch and thoracic aorta. Chromatography showed that approximately 70% of the total immunoreactivity was identical to synthetic calcitonin gene-related peptide. Calcitonin gene-related peptide concentrations in the blood vessels of rats treated neonatally with capsaicin were not found to be significantly different from those in control animals ahhougb capsaicin caused significant reductions of cabitonin gene-related peptide levels in certain other tissues. The rest&s of this study suggest that calcitunin gene-related ~tid~“c#nt~n~ng fibres are likely to be of importance in the innervation of vascular tissues and raise the possibility that these fibres are different in character from calcitonin gene-related peptide-containing fibres found In other tissues.
in the rat 5f a novel putative neuropeptide resulting from alternative RNA splicing processes during caicitonin gene expression has recently been described.‘*‘* This 37 amino acid calcitonin gene-related peptide (CGRP), which has no sequence homology with calcitonin, has subsequently been shown to have potent effects on the rat cardiovascular system, causing increased heart rate and a pressure following intrablood rise in cerebroventricular administration, but causing hypotension with an accompanying increase in heart rate when given intravenously.’ Immunohistochemical localisation studies have demonstrated an abundance of CGRP~ontaiRing nerve iibres in the rat brain,‘* spinal sensory ganglia and spinal cord.‘* CGRPstaining was also observed in fibres throughout many parts of the body, often associated with the smooth muscle of blood vessels. Is The calcium dependent release of CGRP from cultured rat trigeminal ganglion celIs has been demonstrat~‘~ and a human CGRP whose amino acid sequence differs from that of the rat peptide in only four positions has been isolated from human medullary thyroid carcinoma tissue.‘7
The presence of ~p~d~o~t~ni~~ nerve gbres in vascular smooth muscle and in the heart has already been described, notably for substance P,2-3*68-9+‘2,2’n23 vasoactive intestinal polypeptide (VIP),2,4a,5.62’*2’a neurotensin’7”~2’”and neuropeptide Y”s”~ and there is now considerable evidence for the existence of nonnon-adrenergic neurotransmitters in chohnergic, nerves suppIying vascular tissue~.~*‘“‘~It is possible that neuropeptides play an important role in neural mechanisms controlling blood flow either as neurotransmitters or as modulators of synaptic transmission. In this study we have investigated the distribution of CGRP in the heart and major blood vessels of the rat using radioimmunoassay to determine CGRP concentrations in the various tissues and immunocytochemistry to observe the precise localisation of the peptide. In addition, an attempt was made to gain some information on the nature of CGRF-containing gbres by measuring CGRF concentrations in the blood vessels of rats treated neonatally with cap saicin.
The ocm-rence
EXPERIMENTALPROCEDURES Tissue prepuration
*Division of ~doc~no~ogy,
SchooI of Medicine, University of Cahfornia, San Diego, CA 92093, U.S.A. ?To whom correspondence shouid be addressed. Abbreviations: CGW, calcitonin gene-related peptide; BPLC, high performance liquid chromatography; PBS, phosphate-buffered saline; VIP, vasoactive intestinal polypeptide. 947 N.S.C. 14i3-L
Male W&tar rats were used throughout. Normal animals were aged 6 months to I year and weighed 13%250 g. For neonatal capsaicin studies, animals were injected subcutaneausly on the second day of life either with capsaicin (50 mg/kg body wt) emulsified in isotonic saline containing lOuS, (v/v) ethanol and 10% (v/v) Tween 80 or with s&e-ethanol-Tween vehicle only (control group). They
948
P. K. Mulderry et al.
were then killed at the age of 3 months. For radioimmunoassay the rats were killed by a sharp blow to the head and were immediately dissected. Tissues taken are listed in Tables 1 and 2. After trimming of fat and connective tissues, each specimen was weighed and immediately extracted into boiling 0.5 mol/l acetic acid as described previously by Bryant and Bloom.4 After cooling, the extracts were stored frozen at -20°C until assay. For immunocytochemistry, animals were anaesthetised and afterwards perfused via the ascending aorta with pbenzoquinone 0.4% (w/v) in 0.01 mol/l phosphate-buffered saline (PBS), pH 7.0, at room temperature. The three regions of aorta (arch, thoracic and abdominal), common carotid, superior and inferior mesenteric, renal and femoral arteries, superior and inferior vena cava and heart were removed. The tissues were immersed in a fresh solution of p-benzoquinone for a period of 20-60 min to complete the fixation and were then washed in PBS containing sucrose (7%) and sodium azide (0.01%) at 4°C. Specimens of right and left ventricle and septum were then prepared by cutting 60pm cryostat sections. Arteries, veins and atria were studied in whole mount preparations. Radioimmunoassay Tissue extracts were assayed for CGRP in duplicate 20 pl aliquots in 0.8 ml 0.06 mol/l phosphate buffer, pH 7.4, containing 0.01 mol/l ethylenediaminetetra-acetate, 0.05% (w/v) sodium azide and 3% (w/v) bovine serum albumin. The assay used an antiserum previously raised in rabbits immunised with a synthetic C-terminal CGRP fragment as described by Rosenfeld et al.” at a final dilution of 1:120.000. Radiolabelled CGRP was prepared by a chloramine-T iodination of the histidine residue in synthetic CGRP (Peninsula Laboratories) and purified by reversephase high performance liquid chromatography (HPLC) on a pBondapak C-18 column (Waters Associates). Following 5 days incubation at 4°C bound and free fractions of the peptide were separated by charcoal adsorption of the free fraction using 8mg charcoal (Norit GSX) suspended in 250~1 phosphate buffer, containing 0.25% (w/v) gelatine, added to each assay tube. The tubes were then centrifuged at 4°C and the supematant immediately separated from the charcoal pellet. CGRP concentrations were measured against a synthetic CGRP standard (Peninsula) with which the assay could detect 2 fmol with 95% confidence. Chromatography The identity of CGRP-like immunoreactivity was verified by gel filtration and HPLC. For gel filtration, pooled tissue extracts (0.5 ml) were loaded directly onto Sephadex G50 superfine columns (60 x 0.9 cm) eluted with 0.06 mol/l phosphate buffer pH 7.4 containing 0.3% bovine serum albumin and 0.2 M sodium chloride at a flow rate of 3 ml/h. Fractions were collected to a volume of 0.6 ml. The columns were calibrated with dextran blue (mol. wt = 2,ooO,ooO) and NalZSI as markers of void volume and total volume respectively and with horse heart cytochrome c (mol. wt = 12,384). HPLC was carried out on a Techsil C-18 reverse-phase column, particle size 5 pm, (HPLC Technology) eluted with an acetonitrile 0.1% (v/v) aqueous trifluroacetic acid solvent system at 1 ml/min. Prior to loading each sample was purified on Sep-Pak C- 18 reverse ohase cartridges (Waters) and injected onto the column in 2 ml 12”/, (v/v)ac&onitrile, 0.1% (v/v) aqueous trifluoroacetic acid. Fractions were coliected to a volume of 1 ml. The concentration of CGRP-Iike immunomaotivity in all column fractions was determined by assaying duplicate 100 ~1 aliquots.
Immunohistochemisq The same antiserum as for iYidiOimmUnOaS%+y was used in a peroxidase-anti-peroxidase technique, modilied for whole mount preparations as previously described by ‘Terenghi et al.” Free floating fragments were dehydrated with ethanol, cleared in xylol and rehydrated. They were then rinsed in PBS and incubated overnight in non-immune goat serum. diluted I:IO in PBS containing 0.2% Triton X-100. The samples were then incubated with CG’RP antiserum at a dilution of 1:I000 for 48 h at 4’C in a humid chamber. Following an overnight wash in PBS, 5-6 h incubation with unconjugated goat anti-rabbit immunoglobulin serum (Miles Laboratories) at a 1:50 dilution in PBS and a further overnight wash in PBS, the sections were incubated with peroxidase-anti-peroxidase complex (UCB Bioproducts) at a dilution of 1: 100 for 5-6 h at room temperature. Following this, they were again washed overnight in PBS. The sections were finally incubated for 5-20 min in PBS containing 0.05% 3,3_diaminobenzidine tetrahydrochloride and 0.3% hydrogen peroxide and were then washed. The sections were mounted on poly-L-lysine-coated slides as described by Huang et al.” dehydrated through graded alcohols, cleared in xylene and covered with DPX mounting medium (Raymond A. Lamb). Conlrol.~ The stability of CGRP to extraction in boiling acetic acid was assessed by addition of synthetic CGRP to tissue extracts which were then reboiled. These extracts were then assayed alongside aliquots from the same extracts before addition of the synthetic peptide and before reboiling. The recovery of exogenous CGRP-like immunoreactivity after boiling was 111 k 9%, n = 4. The extracts were also andlysed by gel filtration on Sephadex G50 columns. All the exogenous immunoreactivity after boiling co-eluted with unboiled synthetic CGRP as a single peak. The specificity of the immunostaining was established by absorption experiments in which the diluted antiserum was treated prior to use with various peptides. CGRP-staining could be completely eliminated with 0.01 pmol/l CGRP, but was unaffected by incubation with substance P, somatostatin-14, calcitonin or bombesin at concentrations of lOpmol/l. Other sections were incubated with nonimmune rabbit serum in place of the primary antiserum. In radioimmunoassay the antiserum showed no cross reactivity (less than 0.02%) with calcitonin, somatostatin-28, substance P, VIP or gastrin-releasing peptide.
RESULTS Distribution and characterisation of calcitonin generelated peptide-like immunoreactivity CGRP concentrations in cardiovascular tissues of normal rats are given in Table 1. CGRP-like
immunoreactivity was found to be widespread, occurring in both arteries and veins. Particularly high concentrations were found in the renal and superior mesenteric arteries and in the inferior vena cava and femoral vein. By contrast, CGRP levels in the heart, aortic arch and thoracic aorta were comparatively low. The results for capsaicin treated animals and controls are given in Table 2. Comparison of capsaicin treated rats with vehicle injected controls showed that they had slightly higher CGRP concentrations in the abdominal aorta, abdominal bifurcation and renal artery and slightly lower levels in the inferior vena cava. None of these differences reached statistical
949
CGRP in cardiovascular tissues of the rat Table 1. Calcitonin gene-related peptide concentrations in cardiovascular tissues of normal rats as determined by radioimmunoassay CGRI’ concentration Tissue Heart
(pmol/g tissue)
(Whole) (Left atrium) (Right atrium) (Left ventricle) (Right ventricle) (Inter-ventricular septum) Aortic arch Thoracic aorta Abdominal aorta Abdominal bifurcation Carotid artery Carotid bifurcation Superior mesenteric artery Renal artery Femoral artery Renal vein Femoral vein Inferior vena cava Results are given as mean + SEM.
Renal
0.5 & 0.2 0.8 * 0.3 1.5 kO.5 0.3 + 0.1 0.7 + 0.2
60-
O-
0.5 f 0.2 1.220.2 0.6kO.l 20.6 +_3.8 13.6 f 2.3 2.8 f 1.1 12.4 + 4.2 50.5 * 14.3 27.7 + 6.3 9.2 + 1.2 8.0 + 1.6 24.8 f 4.2 25.6 + 7.5
5 6 4 9 11 5 5 4 10 5 5 4 5
(P < 0.005).
Aorta
60Vena
Cava
30-
O-
In both gel f&ration and HPLC analysis of vascular tissue extracts, the major peak of CGRP-like immunoreactivity, comprising more than 70% of the total, had the same elution characteristics as synthetic CGRP. On Sephadex G50 columns, the fraction of the gel phase volume (Kay) available to this material was 0.37 f 0.02, n = 3, whereas the synthetic CGRP standard was eluted with a KV of 0.36 + 0.01, n = 4. The extracts also presented two smaller peaks with K,,s of 0.53 + 0.03, n = 3 and 0.70 f 0.02, n = 3 (Fig. 1). On HPLC, the major peak was co-eluted with synthetic CGRP after 53 min (Fig. 2). Locahation of calcitonin gene-related immunoreactivity
peptide-like
Superior and inferior vena cava. At the junction of the superior vena cava with the right atrium, abunTable 2. Calcitonin gene-related peptide concentrations (mean f SEM) in blood vessels and other tissues of rats treated neonatally with capsaicin (capsaicin group) and rats injected with vehicle only (control group) as described under Experimental Procedures
Abdominal aorta Abdominal bifurcation Renal artery Inferior vena cava Bladder Oesophagus
Abdominal
Inferior
significance according to Student’s unpaired t-test with a critical P value of 0.05. In contrast, CGRP levels in other tissues taken from the capsaicin treated rats were drastically and significantly reduced
Tissue
Artery
n
CGRP concentration pmol/g tissue (n) Control Capsaicin group group 12.8 f 2.2 (5) 11.9 &-2.6 (5) 10.9 * 4.1 (5) 9.2 + I .2 (5) 37.6 + 6.6 (4) 32.1 + 5.3 (5) 8.2 + 2.1 (5) 13.8 f 3.1 (5) 4.0 + 1.5 (5) 41.9 f 4.3 (5) 1.5 f 0.5 (5) 9.4 f 1.9 (5)
t 0
1
0.5
I
1.0
Kav
Fig. 1. Gel permeation profiles of CGRP-like immunoreactivity per 100~1 aliquot of fractions collected from columns loaded with pooled blood vessel extracts from normal rats as described under Experimental Procedures. K, was calculated as K,, = (V, - V&/(V, - V,) where V, = elution volume, V, = void volume, V, = total volume. Synthetic CGRP standard was eluted as a single peak at the position shown.
dant varicose and smooth CGRP-immunoreactive nerve fibres were observed. Nerve fibres ran along the adventitia forming parallel meshes at different levels. Large fascicles were situated on the periphery and small fascicles or isolated nerve fibres within the walls of the veins. The number of stained nerve fibres decreased as the veins were followed distally from the junction with the right atrium. No neuronal cell bodies were found in this region. Heart. In the right and left ventricle walls and in the interventricular septum, CGRPcontaining nerve fibres were seen running parallel to the coronary arteries and their collaterals. From this perivascular plexus some nerve fibres extended into the myocardium and ran beside muscle fibres forming a nerve plexus which was particularly well developed within the papillary muscles. Both atria contained immunoreactive nerve fibres and these were more numerous in the right atrium than in the left. CGRP-containing fibres were present within the sinoatrial and atrioventricular nodes and
950
P. K. Mulderry et al.
CGRP pmol/fraction 1.6
%ACN 0.8
0'
1.6
0.8
0 I
0
I
30 uiinutes
1
60
Fig. 2. Typical HPLC elution profiles of CGRP-like immunoreactivity (solid line): (A) synthetic CGRP and (B) pooled rat abdominal aorta extracts. The column was eluted with a gradient (hatched line) between two a&on&rile (ACN) 0.1% aqueous trifluoroacetic acid solvent system: (1) 21% (v/v) ACN; (2) 40% (v/v) ACN. Recovery of CGRP-like immunoreactivity was greater than 8V/
around the conduction system. Prominent CGRPcontaining nerves surrounding unreactive ganglion cells were found in these regions, particularly in the sinoatrial node (Figs 3-6). Common carotid artery. This artery was sparsely innervated by CGRP immunoreactive nerve fibres. These formed a diffuse plexus consisting of isolated varicose nerve fibres and nerve fascicles containing up to five fibres running along the wall of the artery. Aorta. The abdominal aorta was richly innervated by CGRP-immunoreactive nerve fibres. A dense mesh was seen in the adventitia with branches penetrating the wall of the vessel. This innervation was particularly marked close to the origins of the superior and inferior mesenteric and renal arteries. The aortic arch was quite clearly innervated, in particular close to the origin of the carotid artery. The thortic aorta showed the least density of innervation. Superior and inferior mesenteric arteries; renal and femoral arteries. Abundant networks of CGRP-
immunoreactive nerve fibres were observed around the walls of these arteries, the fibres being found mostly in the adventitia and within the muscle layer (Figs 7 and 8). DISCUSSION
The results of this study demonstrate that CGRP is present in nerve fibres throughout the cardiovascular system of the rat. A provisional co-son with the known distributions of substance P- and VIP-containing fibres in the blood vessels of various species reveals some interesting similar&&s and differences. The distribution of substance Pcontaining nerves has already been described in Mood vessels of the rat3 and guinea-pig9 and in both species was densest in the visceral arteries including the mesenteric arteries. In this respect, the distribution is perhaps similar to that found for C!GRP where the superior mesenteric artery contained the highest
Fig. 3. CGRP-immunoreactive
nerve fibres (nf) around unreactive ganglion cells (g) in the sinoatrial node ( x 800). Fig. 4. A bundle of CGRP-immunoreactive nerve fibres (nf) in the wall of the right atrium ( x 600). Fig. 5. A perivascular plexus of CGRP-immunoreactive nerve fibres (nf) around the right coronary artery (bv) ( x 525).
951
Fig. 6. A dense plexus of CGRP-immunoreactive Fig. 7. Varicose
CGRP-containing
Fig. 8. A perivascular
nerve fibres (nf) in a papillary (m) (x 550).
nerve fibres (nf) on the wall of the superior
muscle of the myocardium mesenteric
artery
( x 450).
plexus of CGRP-containing nerve fibres (nf) in the adventitia of the renal artery. Single fibres and small nerve fascicles are distinguished ( x 450).
952
CGRF’ in cardiovascular tissues of the rat levels. By contrast, substance P fibres were very sparse or completely absent in the renal artery and were relatively sparse in the veins, all of which contained considerable levels of CGRP and in which numerous CGRP-containing fibres could be stained. A survey of the occurrence of VIP-containing fibres in blood vessels of the cat” showed that although they, could be stained in almost all other arteries, none could be found in the renal artery and that they were almost completely absent from veins. In the guinea pig the veins contained only relatively few VIP fibres.& It appears, therefore, that CGRP-containing fibres may be of particular importance in the innervation of the renal arteries and large veins, as CGRP is relatively more abundant there than other peptides so far studied. The localisation of CGRP fibres in the heart was similar to that already described for substance Pz3but contrasts sharply with the distribution of neuropeptide Y. ” Radioimmunoassay detected only low levels of CGRP in the heart, presumably because although CGRP-containing fibres are present in localised regions, the density of the innervation overall is not great. However, the potent effects of exogenous CGRP on heart rate’ suggest that this innervation may still be of importance. The high levels of CGRP in the carotid bifurcation, which includes the baro- and chemoreceptors of the carotid sinus, are of particular interest. The presence of peptide-containing nerves in the carotid sinus has already been reported for VIP, substance P and enkephalin in the cat22 and substance P in the rat.12 These findings suggest a role for peptidergic sensory systems in central cardiovascular control. Neonatal administration of capsaicin to rats causes a permanent degeneration of unmyelinated sensory nerve fibres.” The depletion of substance Pcontaining fibres in vascular tissues has already been described in the rat’ and guinea-pigas and this phenomenon has been taken to indicate a sensory role for these fibres. Our results for the effect of neonatal capsaicin treatment on CGRP levels in the blood vessels show no significant reduction although CGRP levels in other tissues were depleted by capsaicin, and the distribution of CGRP in the rat spinal cord indicates that a sensory role is likely.‘O It should be noted that there are differences between the CGRP
953
levels found in the vascular tissues of normal rats (Table 1) and control animals in the capsaicin experiment (Table 2, right hand column). This apparent discrepancy could result from the fact that the two experiments used quite separate groups of rats which differed in age and were killed, dissected and the tissues extracted at different times. However, although the absolute levels are different, the pattern of relative levels between different tissues is the same in both groups. No conclusions can therefore be drawn on the possible function of CGRP-containing nerves in the blood vessels but it would appear that they may be in a distinct class from those in some other tissues which is either not vulnerable to the effects of capsaicin or is capable of regeneration after capsaicin treatment. Further studies are obviously needed to elucidate the function of vascular CGRP, including the tracing of pathways of CGRP-containing fibres from blood vessels to regulatory centres in the central nervous system and measuring the local release of CGRP from fibres in the blood vessels following various forms of stimulation. Fisher et al.’ have reported that CGRP causes a generalised vasodilation when given intravenously to rats. If the peptide has a vasomotor role it should be possible to correlate local concentrations of endogenous CGRP with the magnitude of vasodilatory response resulting from direct application of the peptide to isolated blood vessel preparations. Conclusion
It is not yet possible to say what function(s) CGRP may have in the cardiovascular system, but the presence of such high concentrations of the peptide and the extent of the innervation make an important physiological role seem likely and add weight to the evidence for the importance of peptidergic nerves in cardiovascular regulation. Acknowledgements-Support
for this work was gratefully received from the Tobacco Advisory Council. J. Rodrigo is a visiting professor from the Institute Santiago Ram& Y Cabal, Madrid, Snain and has a fellowshin from the Roval Society. J. M. Allen is in receipt of a’ Wellcome T&t Training Fellowship. We thank P. Marwaha and J. P. Chatterton for their help in preparing the manuscript.
REFERENCES Amara S. G., Jonas V., Roscnfeld M. G., Ong E. S. and Evans R. M. (1982) Alternative RNA processing in calcitonin gene expression generates mRNAs encoding different polypeptide products. Nature 298, 2&244. Baja F. and Mathison R. (1982) Adrenergic and peptidergic (substance P and vasoactwe intestinal polypeptide) innervation of the rat portal vein. Blood Vess. 19, 263-272. Baja F., Mathison R. and Huggel H. (1983) Substance P-containing nerve fibres in peripheral blood vessels of the rat. Cell Thue Res. 229, 411-422. Bryant M. G. and Bloom S. R. (1982) In Radioimmunoassay of Gut Regulatory Peptides (eds Bloom S. R. and Long R. G.), pp. 36-41. W. B. Saunders, London. 4a. Della N. G., Papka R. E., Fumess J. B. and Costa M. (1983) Vasoactive intestinal polypeptide-like immunoreactivity in nerves associated with the cardiovascular system of guinea-pigs. Neuroscience 9, 605419.
954
P. K. Mulderry er al
5. Duckles S. P. and Said S. I. (1982) Vasoactive intestinal polypeptide as a neurotransmitter
in the cerebral circulation.
Eur. J. Pharmac. 70, 311-374. 6. Edvinsson L., McCullough .I. K. and Uddman R. (1982) Feline cerebral veins and arteries: comparison of autonomic innervation and vasomotor responses. J. Physiol., Lond. 325, 161-173. 7. Fisher L. A., Kikkawa D. O., Rivier J. E., Amara S. G., Evans R. M., Rosenfeld M. G., Vale W. W. and Brown M. R. (1983) Stimulation of noradrenergic sympathetic outflow by calcitonin gene-related peptide. Nature 305,534536.
8. Furness J. B., Elliot J. M., Murphy R., Costa M. and Chalmers J. P. (1982) Baroreceptor reflexes in conscious guinea-pigs are unaffected by depletion of cardio-vascular substance P nerves. Neurosci. Left. 32, 285-290. 9. Furness J. B., Papka R. E., Della N. G., Costa M. and Eskay R. L. (1982) Substance P-like immunoreactivity in nerves associated with the vascular system of guinea-pigs. Neuroscience 7, 447459. 10. Gibson S. J., Polak J. M., Bloom S. R., Sabate I. M., Mulderry P. K., Ghatei M. A., McGregor G. P., Morrison J. F. B., Kelly J. S. and Rosenfeld M. G. (1984) Calcitonin gene-related peptide immunoreactivity in the spinal cord of man and of eight other species. J. Neurosci. In press. 11. Gu J., Polak J. M., Adrian T. E., Allen J. M., Tatemoto K. and Bloom S. R. (1983) Neuropeptide Y (NPYta major cardiac neuropeptide. Lancer 1, 1008-1010. lla. Gu J., Polak J. M., Allen J. M., Huang W. M., Sheppard M. N., Tatemoto K. and Bloom S. R. (1984) High concentrations of a novel peptide, neuropeptide Y, in the innervation of mouse and rat heart. J. Histochem. Cyrochem. 32, 467-472. 12. Helke C. J., O’Donohue T. L. and Jacobowitz D. M. (1980) Substance P as a baro- and chemoreceptor alferent neurotransmitter: immunocytochemical and nemochemical evidence in the rat. Peptides 1, 1-9. 13. Huang K., Grossman M. I. and Yvy A. C. (1948) Nervous control of the cervical portion of the oesophagus. Am. .J. Physiol. 154, 343-357. 14. Hughes J. and Vane J. R. (1970) Relaxations of the isolated portal vein of the rabbit induced by nicotine and electrical stimulation Br. J. Pharmac. 39, 476-489. 15. Lee R. J-F., Wyatt R. A., Che S. and Bevan J. A. (1978) Neurogenic vasodilation of cat cerebral arteries. Circulation Res. 42, 535-542. 16. Lee T. J-F. (1980) Direct evidence against acetylcholine as the dilator transmitter Pharmac. 68, 393-394.
in the cat cerebral artery. Eur. J.
16a. Mason R. T., Peterfreund R. A., Sawchenko P. E., Corrigan A. Z., Rivier J. E. and Vale W. W. (1984) Release of the predicted calcitonin gene-related peptide from cultured rat trigeminal ganglion cells. Nature 308, 653-655. 17. Morris H. R., Panic0 M., Etienne T., Tippins J., Girgis S. I. and McIntyre I. (1984) Isolation and characterisation of human calcitonin gene-related peptide. Nature 30?!, 746-748. 17a. Reinecke M., Weihe E., Carraway R. E., Leeman S. E. and Forssmann W. G. (1982) Localisation of neurotensin immunoreactive nerve fibres in the guinea pig heart. Neuroscience 7, 1785-1795. 18. Rosenfeld M. G., Mennod J. J., Amara S. G., Swanson L. W., Sawchenko P. E., Rivier J., Vale W. W. and Evans R. M. (1983) Production of a novel neuropeptide encoded by the calcitonin gene via tissue-specific RNA processing. Nature 304, 129-135. 19. Scadding J. W. (1980) The permanent anatomical effect of neonatal capsaicin on somatosensory nerves. J. Anar. 131, 473-484. 20. Terenghi G., Polak J. M., Probet L., McGregor G. P., Ferri G. L., Blank M. A., Butler J. M., Unger W. G., Zhang
S., Cole D. F. and Bloom S. R. (1982) Mapping, quantitative distribution and origin of substance P- and VIP-containing nerves in the uvea of guinea-pig eye. Htstochemistry 75, 399-417. 21. Uddman R., Alumets J., Edvinsson L., Hakanson R. and Sundler F. (1981) VIP-nerve fibres around peripheral blood vessels. Acta physiol. stand. 112, 65-70.
21a. Weihe E.. Reinecke M.. (1981) Pentidernic innervation of the mammalian sinus nodes: vasoactive intestinal polypeptide, neurotensin, &&stance P.* Net&i. Lett. 26, 283-288. 22. Wharton J., Polak J. M., Pearse A. G. E., McGregor G. P., Bryant M. G., Bloom S. R., Emson P. C., Bisgard E. E. and Will J. A. (1980) Enkephalin-, VIP- and substance P-like immunoreactivity in the carotid body. Nature 2M, 269-27 1. 23. Wharton J., Polak J. M., McGregor G. P., Bishop A. E. and Bloom S. R. (1981) The distribution immunoreactive nerves in the guinea-pig heart. Neuroscience 6, 2193-2204. (Accepted 2 October 1984)
of substance P-like
Royal Postgraduate London Wf2 OHS, U.K.
Medical School, Du Cane Road,
Abstract-The distribution of calcitonin gene-related peptide immuuoreactivity in the cardiovascular system of the rat was investigated by radioimmunoassay and immunocytochemistry. The nature of the immunoreactivity was studied by gel permeation and high performance liquid chromatography. Immunocytochemistry demonstrated the existence of calcitonin gene-related peptide-containing nerve fibres throughout the cardiovascular system. These were present in all regions of the heart, particularly in association with the coronary arteries, within the pap&try mu&es and within the sinoatrial and at~oven~cu~ar nodes. Cakitonin gene-related ~ptid~~~in~ng flbres were found mainly in the adventitia of the arteries and veins. Calcitonin gene-related peptide coucentrations were high br major arteries and veins but comparatively low in the heart, aortic arch and thoracic aorta. Chromatography showed that approximately 70% of the total immunoreactivity was identical to synthetic calcitonin gene-related peptide. Calcitonin gene-related peptide concentrations in the blood vessels of rats treated neonatally with capsaicin were not found to be significantly different from those in control animals ahhougb capsaicin caused significant reductions of cabitonin gene-related peptide levels in certain other tissues. The rest&s of this study suggest that calcitunin gene-related ~tid~“c#nt~n~ng fibres are likely to be of importance in the innervation of vascular tissues and raise the possibility that these fibres are different in character from calcitonin gene-related peptide-containing fibres found In other tissues.
in the rat 5f a novel putative neuropeptide resulting from alternative RNA splicing processes during caicitonin gene expression has recently been described.‘*‘* This 37 amino acid calcitonin gene-related peptide (CGRP), which has no sequence homology with calcitonin, has subsequently been shown to have potent effects on the rat cardiovascular system, causing increased heart rate and a pressure following intrablood rise in cerebroventricular administration, but causing hypotension with an accompanying increase in heart rate when given intravenously.’ Immunohistochemical localisation studies have demonstrated an abundance of CGRP~ontaiRing nerve iibres in the rat brain,‘* spinal sensory ganglia and spinal cord.‘* CGRPstaining was also observed in fibres throughout many parts of the body, often associated with the smooth muscle of blood vessels. Is The calcium dependent release of CGRP from cultured rat trigeminal ganglion celIs has been demonstrat~‘~ and a human CGRP whose amino acid sequence differs from that of the rat peptide in only four positions has been isolated from human medullary thyroid carcinoma tissue.‘7
The presence of ~p~d~o~t~ni~~ nerve gbres in vascular smooth muscle and in the heart has already been described, notably for substance P,2-3*68-9+‘2,2’n23 vasoactive intestinal polypeptide (VIP),2,4a,5.62’*2’a neurotensin’7”~2’”and neuropeptide Y”s”~ and there is now considerable evidence for the existence of nonnon-adrenergic neurotransmitters in chohnergic, nerves suppIying vascular tissue~.~*‘“‘~It is possible that neuropeptides play an important role in neural mechanisms controlling blood flow either as neurotransmitters or as modulators of synaptic transmission. In this study we have investigated the distribution of CGRP in the heart and major blood vessels of the rat using radioimmunoassay to determine CGRP concentrations in the various tissues and immunocytochemistry to observe the precise localisation of the peptide. In addition, an attempt was made to gain some information on the nature of CGRF-containing gbres by measuring CGRF concentrations in the blood vessels of rats treated neonatally with cap saicin.
The ocm-rence
EXPERIMENTALPROCEDURES Tissue prepuration
*Division of ~doc~no~ogy,
SchooI of Medicine, University of Cahfornia, San Diego, CA 92093, U.S.A. ?To whom correspondence shouid be addressed. Abbreviations: CGW, calcitonin gene-related peptide; BPLC, high performance liquid chromatography; PBS, phosphate-buffered saline; VIP, vasoactive intestinal polypeptide. 947 N.S.C. 14i3-L
Male W&tar rats were used throughout. Normal animals were aged 6 months to I year and weighed 13%250 g. For neonatal capsaicin studies, animals were injected subcutaneausly on the second day of life either with capsaicin (50 mg/kg body wt) emulsified in isotonic saline containing lOuS, (v/v) ethanol and 10% (v/v) Tween 80 or with s&e-ethanol-Tween vehicle only (control group). They
948
P. K. Mulderry et al.
were then killed at the age of 3 months. For radioimmunoassay the rats were killed by a sharp blow to the head and were immediately dissected. Tissues taken are listed in Tables 1 and 2. After trimming of fat and connective tissues, each specimen was weighed and immediately extracted into boiling 0.5 mol/l acetic acid as described previously by Bryant and Bloom.4 After cooling, the extracts were stored frozen at -20°C until assay. For immunocytochemistry, animals were anaesthetised and afterwards perfused via the ascending aorta with pbenzoquinone 0.4% (w/v) in 0.01 mol/l phosphate-buffered saline (PBS), pH 7.0, at room temperature. The three regions of aorta (arch, thoracic and abdominal), common carotid, superior and inferior mesenteric, renal and femoral arteries, superior and inferior vena cava and heart were removed. The tissues were immersed in a fresh solution of p-benzoquinone for a period of 20-60 min to complete the fixation and were then washed in PBS containing sucrose (7%) and sodium azide (0.01%) at 4°C. Specimens of right and left ventricle and septum were then prepared by cutting 60pm cryostat sections. Arteries, veins and atria were studied in whole mount preparations. Radioimmunoassay Tissue extracts were assayed for CGRP in duplicate 20 pl aliquots in 0.8 ml 0.06 mol/l phosphate buffer, pH 7.4, containing 0.01 mol/l ethylenediaminetetra-acetate, 0.05% (w/v) sodium azide and 3% (w/v) bovine serum albumin. The assay used an antiserum previously raised in rabbits immunised with a synthetic C-terminal CGRP fragment as described by Rosenfeld et al.” at a final dilution of 1:120.000. Radiolabelled CGRP was prepared by a chloramine-T iodination of the histidine residue in synthetic CGRP (Peninsula Laboratories) and purified by reversephase high performance liquid chromatography (HPLC) on a pBondapak C-18 column (Waters Associates). Following 5 days incubation at 4°C bound and free fractions of the peptide were separated by charcoal adsorption of the free fraction using 8mg charcoal (Norit GSX) suspended in 250~1 phosphate buffer, containing 0.25% (w/v) gelatine, added to each assay tube. The tubes were then centrifuged at 4°C and the supematant immediately separated from the charcoal pellet. CGRP concentrations were measured against a synthetic CGRP standard (Peninsula) with which the assay could detect 2 fmol with 95% confidence. Chromatography The identity of CGRP-like immunoreactivity was verified by gel filtration and HPLC. For gel filtration, pooled tissue extracts (0.5 ml) were loaded directly onto Sephadex G50 superfine columns (60 x 0.9 cm) eluted with 0.06 mol/l phosphate buffer pH 7.4 containing 0.3% bovine serum albumin and 0.2 M sodium chloride at a flow rate of 3 ml/h. Fractions were collected to a volume of 0.6 ml. The columns were calibrated with dextran blue (mol. wt = 2,ooO,ooO) and NalZSI as markers of void volume and total volume respectively and with horse heart cytochrome c (mol. wt = 12,384). HPLC was carried out on a Techsil C-18 reverse-phase column, particle size 5 pm, (HPLC Technology) eluted with an acetonitrile 0.1% (v/v) aqueous trifluroacetic acid solvent system at 1 ml/min. Prior to loading each sample was purified on Sep-Pak C- 18 reverse ohase cartridges (Waters) and injected onto the column in 2 ml 12”/, (v/v)ac&onitrile, 0.1% (v/v) aqueous trifluoroacetic acid. Fractions were coliected to a volume of 1 ml. The concentration of CGRP-Iike immunomaotivity in all column fractions was determined by assaying duplicate 100 ~1 aliquots.
Immunohistochemisq The same antiserum as for iYidiOimmUnOaS%+y was used in a peroxidase-anti-peroxidase technique, modilied for whole mount preparations as previously described by ‘Terenghi et al.” Free floating fragments were dehydrated with ethanol, cleared in xylol and rehydrated. They were then rinsed in PBS and incubated overnight in non-immune goat serum. diluted I:IO in PBS containing 0.2% Triton X-100. The samples were then incubated with CG’RP antiserum at a dilution of 1:I000 for 48 h at 4’C in a humid chamber. Following an overnight wash in PBS, 5-6 h incubation with unconjugated goat anti-rabbit immunoglobulin serum (Miles Laboratories) at a 1:50 dilution in PBS and a further overnight wash in PBS, the sections were incubated with peroxidase-anti-peroxidase complex (UCB Bioproducts) at a dilution of 1: 100 for 5-6 h at room temperature. Following this, they were again washed overnight in PBS. The sections were finally incubated for 5-20 min in PBS containing 0.05% 3,3_diaminobenzidine tetrahydrochloride and 0.3% hydrogen peroxide and were then washed. The sections were mounted on poly-L-lysine-coated slides as described by Huang et al.” dehydrated through graded alcohols, cleared in xylene and covered with DPX mounting medium (Raymond A. Lamb). Conlrol.~ The stability of CGRP to extraction in boiling acetic acid was assessed by addition of synthetic CGRP to tissue extracts which were then reboiled. These extracts were then assayed alongside aliquots from the same extracts before addition of the synthetic peptide and before reboiling. The recovery of exogenous CGRP-like immunoreactivity after boiling was 111 k 9%, n = 4. The extracts were also andlysed by gel filtration on Sephadex G50 columns. All the exogenous immunoreactivity after boiling co-eluted with unboiled synthetic CGRP as a single peak. The specificity of the immunostaining was established by absorption experiments in which the diluted antiserum was treated prior to use with various peptides. CGRP-staining could be completely eliminated with 0.01 pmol/l CGRP, but was unaffected by incubation with substance P, somatostatin-14, calcitonin or bombesin at concentrations of lOpmol/l. Other sections were incubated with nonimmune rabbit serum in place of the primary antiserum. In radioimmunoassay the antiserum showed no cross reactivity (less than 0.02%) with calcitonin, somatostatin-28, substance P, VIP or gastrin-releasing peptide.
RESULTS Distribution and characterisation of calcitonin generelated peptide-like immunoreactivity CGRP concentrations in cardiovascular tissues of normal rats are given in Table 1. CGRP-like
immunoreactivity was found to be widespread, occurring in both arteries and veins. Particularly high concentrations were found in the renal and superior mesenteric arteries and in the inferior vena cava and femoral vein. By contrast, CGRP levels in the heart, aortic arch and thoracic aorta were comparatively low. The results for capsaicin treated animals and controls are given in Table 2. Comparison of capsaicin treated rats with vehicle injected controls showed that they had slightly higher CGRP concentrations in the abdominal aorta, abdominal bifurcation and renal artery and slightly lower levels in the inferior vena cava. None of these differences reached statistical
949
CGRP in cardiovascular tissues of the rat Table 1. Calcitonin gene-related peptide concentrations in cardiovascular tissues of normal rats as determined by radioimmunoassay CGRI’ concentration Tissue Heart
(pmol/g tissue)
(Whole) (Left atrium) (Right atrium) (Left ventricle) (Right ventricle) (Inter-ventricular septum) Aortic arch Thoracic aorta Abdominal aorta Abdominal bifurcation Carotid artery Carotid bifurcation Superior mesenteric artery Renal artery Femoral artery Renal vein Femoral vein Inferior vena cava Results are given as mean + SEM.
Renal
0.5 & 0.2 0.8 * 0.3 1.5 kO.5 0.3 + 0.1 0.7 + 0.2
60-
O-
0.5 f 0.2 1.220.2 0.6kO.l 20.6 +_3.8 13.6 f 2.3 2.8 f 1.1 12.4 + 4.2 50.5 * 14.3 27.7 + 6.3 9.2 + 1.2 8.0 + 1.6 24.8 f 4.2 25.6 + 7.5
5 6 4 9 11 5 5 4 10 5 5 4 5
(P < 0.005).
Aorta
60Vena
Cava
30-
O-
In both gel f&ration and HPLC analysis of vascular tissue extracts, the major peak of CGRP-like immunoreactivity, comprising more than 70% of the total, had the same elution characteristics as synthetic CGRP. On Sephadex G50 columns, the fraction of the gel phase volume (Kay) available to this material was 0.37 f 0.02, n = 3, whereas the synthetic CGRP standard was eluted with a KV of 0.36 + 0.01, n = 4. The extracts also presented two smaller peaks with K,,s of 0.53 + 0.03, n = 3 and 0.70 f 0.02, n = 3 (Fig. 1). On HPLC, the major peak was co-eluted with synthetic CGRP after 53 min (Fig. 2). Locahation of calcitonin gene-related immunoreactivity
peptide-like
Superior and inferior vena cava. At the junction of the superior vena cava with the right atrium, abunTable 2. Calcitonin gene-related peptide concentrations (mean f SEM) in blood vessels and other tissues of rats treated neonatally with capsaicin (capsaicin group) and rats injected with vehicle only (control group) as described under Experimental Procedures
Abdominal aorta Abdominal bifurcation Renal artery Inferior vena cava Bladder Oesophagus
Abdominal
Inferior
significance according to Student’s unpaired t-test with a critical P value of 0.05. In contrast, CGRP levels in other tissues taken from the capsaicin treated rats were drastically and significantly reduced
Tissue
Artery
n
CGRP concentration pmol/g tissue (n) Control Capsaicin group group 12.8 f 2.2 (5) 11.9 &-2.6 (5) 10.9 * 4.1 (5) 9.2 + I .2 (5) 37.6 + 6.6 (4) 32.1 + 5.3 (5) 8.2 + 2.1 (5) 13.8 f 3.1 (5) 4.0 + 1.5 (5) 41.9 f 4.3 (5) 1.5 f 0.5 (5) 9.4 f 1.9 (5)
t 0
1
0.5
I
1.0
Kav
Fig. 1. Gel permeation profiles of CGRP-like immunoreactivity per 100~1 aliquot of fractions collected from columns loaded with pooled blood vessel extracts from normal rats as described under Experimental Procedures. K, was calculated as K,, = (V, - V&/(V, - V,) where V, = elution volume, V, = void volume, V, = total volume. Synthetic CGRP standard was eluted as a single peak at the position shown.
dant varicose and smooth CGRP-immunoreactive nerve fibres were observed. Nerve fibres ran along the adventitia forming parallel meshes at different levels. Large fascicles were situated on the periphery and small fascicles or isolated nerve fibres within the walls of the veins. The number of stained nerve fibres decreased as the veins were followed distally from the junction with the right atrium. No neuronal cell bodies were found in this region. Heart. In the right and left ventricle walls and in the interventricular septum, CGRPcontaining nerve fibres were seen running parallel to the coronary arteries and their collaterals. From this perivascular plexus some nerve fibres extended into the myocardium and ran beside muscle fibres forming a nerve plexus which was particularly well developed within the papillary muscles. Both atria contained immunoreactive nerve fibres and these were more numerous in the right atrium than in the left. CGRP-containing fibres were present within the sinoatrial and atrioventricular nodes and
950
P. K. Mulderry et al.
CGRP pmol/fraction 1.6
%ACN 0.8
0'
1.6
0.8
0 I
0
I
30 uiinutes
1
60
Fig. 2. Typical HPLC elution profiles of CGRP-like immunoreactivity (solid line): (A) synthetic CGRP and (B) pooled rat abdominal aorta extracts. The column was eluted with a gradient (hatched line) between two a&on&rile (ACN) 0.1% aqueous trifluoroacetic acid solvent system: (1) 21% (v/v) ACN; (2) 40% (v/v) ACN. Recovery of CGRP-like immunoreactivity was greater than 8V/
around the conduction system. Prominent CGRPcontaining nerves surrounding unreactive ganglion cells were found in these regions, particularly in the sinoatrial node (Figs 3-6). Common carotid artery. This artery was sparsely innervated by CGRP immunoreactive nerve fibres. These formed a diffuse plexus consisting of isolated varicose nerve fibres and nerve fascicles containing up to five fibres running along the wall of the artery. Aorta. The abdominal aorta was richly innervated by CGRP-immunoreactive nerve fibres. A dense mesh was seen in the adventitia with branches penetrating the wall of the vessel. This innervation was particularly marked close to the origins of the superior and inferior mesenteric and renal arteries. The aortic arch was quite clearly innervated, in particular close to the origin of the carotid artery. The thortic aorta showed the least density of innervation. Superior and inferior mesenteric arteries; renal and femoral arteries. Abundant networks of CGRP-
immunoreactive nerve fibres were observed around the walls of these arteries, the fibres being found mostly in the adventitia and within the muscle layer (Figs 7 and 8). DISCUSSION
The results of this study demonstrate that CGRP is present in nerve fibres throughout the cardiovascular system of the rat. A provisional co-son with the known distributions of substance P- and VIP-containing fibres in the blood vessels of various species reveals some interesting similar&&s and differences. The distribution of substance Pcontaining nerves has already been described in Mood vessels of the rat3 and guinea-pig9 and in both species was densest in the visceral arteries including the mesenteric arteries. In this respect, the distribution is perhaps similar to that found for C!GRP where the superior mesenteric artery contained the highest
Fig. 3. CGRP-immunoreactive
nerve fibres (nf) around unreactive ganglion cells (g) in the sinoatrial node ( x 800). Fig. 4. A bundle of CGRP-immunoreactive nerve fibres (nf) in the wall of the right atrium ( x 600). Fig. 5. A perivascular plexus of CGRP-immunoreactive nerve fibres (nf) around the right coronary artery (bv) ( x 525).
951
Fig. 6. A dense plexus of CGRP-immunoreactive Fig. 7. Varicose
CGRP-containing
Fig. 8. A perivascular
nerve fibres (nf) in a papillary (m) (x 550).
nerve fibres (nf) on the wall of the superior
muscle of the myocardium mesenteric
artery
( x 450).
plexus of CGRP-containing nerve fibres (nf) in the adventitia of the renal artery. Single fibres and small nerve fascicles are distinguished ( x 450).
952
CGRF’ in cardiovascular tissues of the rat levels. By contrast, substance P fibres were very sparse or completely absent in the renal artery and were relatively sparse in the veins, all of which contained considerable levels of CGRP and in which numerous CGRP-containing fibres could be stained. A survey of the occurrence of VIP-containing fibres in blood vessels of the cat” showed that although they, could be stained in almost all other arteries, none could be found in the renal artery and that they were almost completely absent from veins. In the guinea pig the veins contained only relatively few VIP fibres.& It appears, therefore, that CGRP-containing fibres may be of particular importance in the innervation of the renal arteries and large veins, as CGRP is relatively more abundant there than other peptides so far studied. The localisation of CGRP fibres in the heart was similar to that already described for substance Pz3but contrasts sharply with the distribution of neuropeptide Y. ” Radioimmunoassay detected only low levels of CGRP in the heart, presumably because although CGRP-containing fibres are present in localised regions, the density of the innervation overall is not great. However, the potent effects of exogenous CGRP on heart rate’ suggest that this innervation may still be of importance. The high levels of CGRP in the carotid bifurcation, which includes the baro- and chemoreceptors of the carotid sinus, are of particular interest. The presence of peptide-containing nerves in the carotid sinus has already been reported for VIP, substance P and enkephalin in the cat22 and substance P in the rat.12 These findings suggest a role for peptidergic sensory systems in central cardiovascular control. Neonatal administration of capsaicin to rats causes a permanent degeneration of unmyelinated sensory nerve fibres.” The depletion of substance Pcontaining fibres in vascular tissues has already been described in the rat’ and guinea-pigas and this phenomenon has been taken to indicate a sensory role for these fibres. Our results for the effect of neonatal capsaicin treatment on CGRP levels in the blood vessels show no significant reduction although CGRP levels in other tissues were depleted by capsaicin, and the distribution of CGRP in the rat spinal cord indicates that a sensory role is likely.‘O It should be noted that there are differences between the CGRP
953
levels found in the vascular tissues of normal rats (Table 1) and control animals in the capsaicin experiment (Table 2, right hand column). This apparent discrepancy could result from the fact that the two experiments used quite separate groups of rats which differed in age and were killed, dissected and the tissues extracted at different times. However, although the absolute levels are different, the pattern of relative levels between different tissues is the same in both groups. No conclusions can therefore be drawn on the possible function of CGRP-containing nerves in the blood vessels but it would appear that they may be in a distinct class from those in some other tissues which is either not vulnerable to the effects of capsaicin or is capable of regeneration after capsaicin treatment. Further studies are obviously needed to elucidate the function of vascular CGRP, including the tracing of pathways of CGRP-containing fibres from blood vessels to regulatory centres in the central nervous system and measuring the local release of CGRP from fibres in the blood vessels following various forms of stimulation. Fisher et al.’ have reported that CGRP causes a generalised vasodilation when given intravenously to rats. If the peptide has a vasomotor role it should be possible to correlate local concentrations of endogenous CGRP with the magnitude of vasodilatory response resulting from direct application of the peptide to isolated blood vessel preparations. Conclusion
It is not yet possible to say what function(s) CGRP may have in the cardiovascular system, but the presence of such high concentrations of the peptide and the extent of the innervation make an important physiological role seem likely and add weight to the evidence for the importance of peptidergic nerves in cardiovascular regulation. Acknowledgements-Support
for this work was gratefully received from the Tobacco Advisory Council. J. Rodrigo is a visiting professor from the Institute Santiago Ram& Y Cabal, Madrid, Snain and has a fellowshin from the Roval Society. J. M. Allen is in receipt of a’ Wellcome T&t Training Fellowship. We thank P. Marwaha and J. P. Chatterton for their help in preparing the manuscript.
REFERENCES Amara S. G., Jonas V., Roscnfeld M. G., Ong E. S. and Evans R. M. (1982) Alternative RNA processing in calcitonin gene expression generates mRNAs encoding different polypeptide products. Nature 298, 2&244. Baja F. and Mathison R. (1982) Adrenergic and peptidergic (substance P and vasoactwe intestinal polypeptide) innervation of the rat portal vein. Blood Vess. 19, 263-272. Baja F., Mathison R. and Huggel H. (1983) Substance P-containing nerve fibres in peripheral blood vessels of the rat. Cell Thue Res. 229, 411-422. Bryant M. G. and Bloom S. R. (1982) In Radioimmunoassay of Gut Regulatory Peptides (eds Bloom S. R. and Long R. G.), pp. 36-41. W. B. Saunders, London. 4a. Della N. G., Papka R. E., Fumess J. B. and Costa M. (1983) Vasoactive intestinal polypeptide-like immunoreactivity in nerves associated with the cardiovascular system of guinea-pigs. Neuroscience 9, 605419.
954
P. K. Mulderry er al
5. Duckles S. P. and Said S. I. (1982) Vasoactive intestinal polypeptide as a neurotransmitter
in the cerebral circulation.
Eur. J. Pharmac. 70, 311-374. 6. Edvinsson L., McCullough .I. K. and Uddman R. (1982) Feline cerebral veins and arteries: comparison of autonomic innervation and vasomotor responses. J. Physiol., Lond. 325, 161-173. 7. Fisher L. A., Kikkawa D. O., Rivier J. E., Amara S. G., Evans R. M., Rosenfeld M. G., Vale W. W. and Brown M. R. (1983) Stimulation of noradrenergic sympathetic outflow by calcitonin gene-related peptide. Nature 305,534536.
8. Furness J. B., Elliot J. M., Murphy R., Costa M. and Chalmers J. P. (1982) Baroreceptor reflexes in conscious guinea-pigs are unaffected by depletion of cardio-vascular substance P nerves. Neurosci. Left. 32, 285-290. 9. Furness J. B., Papka R. E., Della N. G., Costa M. and Eskay R. L. (1982) Substance P-like immunoreactivity in nerves associated with the vascular system of guinea-pigs. Neuroscience 7, 447459. 10. Gibson S. J., Polak J. M., Bloom S. R., Sabate I. M., Mulderry P. K., Ghatei M. A., McGregor G. P., Morrison J. F. B., Kelly J. S. and Rosenfeld M. G. (1984) Calcitonin gene-related peptide immunoreactivity in the spinal cord of man and of eight other species. J. Neurosci. In press. 11. Gu J., Polak J. M., Adrian T. E., Allen J. M., Tatemoto K. and Bloom S. R. (1983) Neuropeptide Y (NPYta major cardiac neuropeptide. Lancer 1, 1008-1010. lla. Gu J., Polak J. M., Allen J. M., Huang W. M., Sheppard M. N., Tatemoto K. and Bloom S. R. (1984) High concentrations of a novel peptide, neuropeptide Y, in the innervation of mouse and rat heart. J. Histochem. Cyrochem. 32, 467-472. 12. Helke C. J., O’Donohue T. L. and Jacobowitz D. M. (1980) Substance P as a baro- and chemoreceptor alferent neurotransmitter: immunocytochemical and nemochemical evidence in the rat. Peptides 1, 1-9. 13. Huang K., Grossman M. I. and Yvy A. C. (1948) Nervous control of the cervical portion of the oesophagus. Am. .J. Physiol. 154, 343-357. 14. Hughes J. and Vane J. R. (1970) Relaxations of the isolated portal vein of the rabbit induced by nicotine and electrical stimulation Br. J. Pharmac. 39, 476-489. 15. Lee R. J-F., Wyatt R. A., Che S. and Bevan J. A. (1978) Neurogenic vasodilation of cat cerebral arteries. Circulation Res. 42, 535-542. 16. Lee T. J-F. (1980) Direct evidence against acetylcholine as the dilator transmitter Pharmac. 68, 393-394.
in the cat cerebral artery. Eur. J.
16a. Mason R. T., Peterfreund R. A., Sawchenko P. E., Corrigan A. Z., Rivier J. E. and Vale W. W. (1984) Release of the predicted calcitonin gene-related peptide from cultured rat trigeminal ganglion cells. Nature 308, 653-655. 17. Morris H. R., Panic0 M., Etienne T., Tippins J., Girgis S. I. and McIntyre I. (1984) Isolation and characterisation of human calcitonin gene-related peptide. Nature 30?!, 746-748. 17a. Reinecke M., Weihe E., Carraway R. E., Leeman S. E. and Forssmann W. G. (1982) Localisation of neurotensin immunoreactive nerve fibres in the guinea pig heart. Neuroscience 7, 1785-1795. 18. Rosenfeld M. G., Mennod J. J., Amara S. G., Swanson L. W., Sawchenko P. E., Rivier J., Vale W. W. and Evans R. M. (1983) Production of a novel neuropeptide encoded by the calcitonin gene via tissue-specific RNA processing. Nature 304, 129-135. 19. Scadding J. W. (1980) The permanent anatomical effect of neonatal capsaicin on somatosensory nerves. J. Anar. 131, 473-484. 20. Terenghi G., Polak J. M., Probet L., McGregor G. P., Ferri G. L., Blank M. A., Butler J. M., Unger W. G., Zhang
S., Cole D. F. and Bloom S. R. (1982) Mapping, quantitative distribution and origin of substance P- and VIP-containing nerves in the uvea of guinea-pig eye. Htstochemistry 75, 399-417. 21. Uddman R., Alumets J., Edvinsson L., Hakanson R. and Sundler F. (1981) VIP-nerve fibres around peripheral blood vessels. Acta physiol. stand. 112, 65-70.
21a. Weihe E.. Reinecke M.. (1981) Pentidernic innervation of the mammalian sinus nodes: vasoactive intestinal polypeptide, neurotensin, &&stance P.* Net&i. Lett. 26, 283-288. 22. Wharton J., Polak J. M., Pearse A. G. E., McGregor G. P., Bryant M. G., Bloom S. R., Emson P. C., Bisgard E. E. and Will J. A. (1980) Enkephalin-, VIP- and substance P-like immunoreactivity in the carotid body. Nature 2M, 269-27 1. 23. Wharton J., Polak J. M., McGregor G. P., Bishop A. E. and Bloom S. R. (1981) The distribution immunoreactive nerves in the guinea-pig heart. Neuroscience 6, 2193-2204. (Accepted 2 October 1984)
of substance P-like