Potent Vasodilator Activity of Calcitonin Gene-Related Peptide in Human Skin

Susan D. Brain, Ph.D., * John R. Tippins, Ph.D., Howard R. Morris, Ph.D., lain Macintyre, Ph.D., D.Sc., F.R.C.P., and Timothy J. Williams, Ph.D. Section of Vascular Biology, Clinica l Resea rch Centre (SOB, TJW), Harrow, Department of Biochemistry, Imperial College ORT, HRM), London, and Department of Chemical Pathology, Roya l Postgraduate Medical School (1M), London. U.K.

We have recently shown that the novel neuropeptide cal­citonin gene-related peptide, CGRP, is a potent vasodila­tor. In this paper we report a detailed study of the effects of CGRP in human skin . CGRP induces a clearly defined, long-lasting erythema. We have measured the effect ofCGRP on blood flow in human skin using a laser Doppler tech­nique and have demonstrated increased local blood flow that persists for a number of hours . We com pared the re­sponse of CGRP with other known vasodilators [hista­mine, prostaglandin (PG) E 2 , PGI2 , substance P, and vaso­active intestinal peptide (VIP)] in the skin, and in all subj ects the erythema induced by CGRP was more persistent than that induced by the other mediators tes ted. Except at high

Calcitonin gene-related peptide (CGRP) is a 37 amino acid peptide which is predicted to occur as a result of alternative processing of mRNA from the calci­tonin gene [1,2] . The calcitonin-secreting cells of the thyroid gland are derived embryonically from the

neural crest and although calcitonin is the major product, it has been possible to isolate and characterize a human CGRP from thyroid tissue taken from patients with n1.edullary thyroid ca r­cinoma [3] . This human CGRP consists of37 amino acid residues but differs at 4 amino acid sites from the previously reported predicted structure of rat CGRP [1] . Immunocytochemical studies in the rat have revealed the distribution of immunoreactive CGRP in the central and peripheral nervous systems [1]. In the periphery immunoreactive CGRP was found in thin beaded nerve fibers that were associated with the smooth muscle of blood vessels [lJ . We have found that both natural human CGRP and synthetic rat CGRP increase microvascular blood flow in rabbit skin, measured with a xenon clearance technique [4]. By directly visualizing the microvascular bed of the hamster cheek pouch we found that CGRP acts on arterioles. In human skin intradermal injection of low doses ofCGRP induces only local erythema, unlike the triple response of skin to agents such as histamine [4] . It has been

Manuscript received February 5, 1986; accepted for publication May 8, 1986.

Supported by the Wellcome Trust, British Heart Foundation, and Med­ical Research Council.

*Dr. Brain is a Halley Stewart research fellow. Reprint requests to: Susan D. Brain, Ph.D., Section of Vascular Bi­

ology, MRC Clinical Research Centre, Watford Road , Harrow, HAl 3U] , U.K.

Abbreviations: CGRP: calcitonin gene-related peptide HPLC: high-performance liquid chromatography PG: prostaglandin VIP: vasoactive intestinal peptide

doses the local vasodilatation induced by CGRP was not associated with a wheal and flare as seen with histamine, substance P, and VIP. CGRP is an extremely potent vaso­dilator and if released into the circulation, or locall y from peripheral nerve endings, it could have a role in the reg­ulation of blood flow in both physiologic and pathologic conditions; CGRP may be the endogenous m ediator of the flare in the triple response. A deficiency in CGRP secretion or action could be an important component of peripheral vascular disease. Some flushin g reactions (e.g., those as­sociated with medullary thyroid carcinoma) may result from circulating CGRP.) Invest Dermato/ 87:533-536, 1986

sugges ted that the flushing of the skin which is observed in some patients with medullary thyroid carcinoma [5] cou ld be due to circulating CGRP [4]. More recently, immunoreactive CGRP has been detected in the plasma ofnorm31 individuals [6]. The potency ofCGRP in inducing erythema and its locali zation in perivascular nerves suggest that CGRP could have an important regulatory role in the con trol of blood fl ow in humans.

In this study we have examined the effect of CGRP in human skin and report that CGRP has potent and persistent effects in stimulating local blood flow . We have compared the effect of CGRP in human skin with that of other known vasodilators and give ev idence to suggest that CGRP is one of the most potent and persistent vasodilators yet discovered.

SUBJECTS AND METHODS

Pep tides were purified by high-perfo rm ance liquid chromatog­raphy (HPLC). Human CGRP (Sandoz) was eluted from a Waters J-L Bondapak C18 semipreparative column with a gradient of 0-40% n-propanol in 5% acetic acid in water. Substance P (Sigma) and vasoactive intestinal peptide (VIP) (Sigma) were eluted from the same column with a gradient of10-40% n-propanol in 5% acetic acid in water. Elution positions were confirmed by fast atom bombardment mass spectrometry and the peptides were quan­titated by amino acid analysis.

Experiments were carried out on health y volunteers using a total of 9 individuals (6 male and 3 fem ale, aged 21-40 years). The subjects gave informed consent and ethi cal approval was obtained.

The test substances were prepared in 0.9% sterile pyrogen-free saline (saline) and injected in 50 J-LI volumes at predetermined sites into the volar surface of the forearm. C hanges in blood flow were recorded in 4 volunteers usin g a Perimed II laser Doppler flow meter [7]. The probe was secured in a holder 0 11 the arm for each reading. Readings (each taken continuously for 2 min) were taken at 3 sites on the forearm, just before and 10 min , 1 h, 2 h, 3 h, and 4 h after intraderm al injection. One site received 10 pmol

0022-202X/86/S03.50 Copyright © 1986 by The Society for Investigative Dermatology, Inc.

533

534 BRAIN ET AL

CGRP in sa line, the second site sa line only, and the third site was not injected. Results were recorded as mean % flux val ue (number of red blood cells moving in the volume measured by the path of the laser beam, 1-2 cu mm x mean cell velocity).

CGRP (1-100 pmol) was compared w ith 5 other established vasod ila tors: VIP (10 pmol), substance P (10 pmol), prostaglandin (PG)E2' (10 pmol), PGI2, (10 pmol), and histamine (500 pmol) in 4 individuals. T he effect of loca l and system ic HI antihistamines on CG RP-indu ced erythema was exa mined by injectin g chlor­pheniramine (Piriton, 1 J.L g) intradermal ly with CGRP (10 pmol) and hi stamine (500 pmol) in 3 individuals or by ad ministering terfenad ine (Tri ludan, 60 m g p.o . in 4 indi viduals) or chlo rphen­iramine (Piriton 4-8 mg p.o. in 3 individuals) 2-3 h before in­tradermal injection of CGRP (1-10 pmol) and histamine (500 pmo l). Res ponses obtained in the presence of the antihistamines were compared with control responses o btained by injecting hi s­tamine and CGRP at alternate sites on the sa me forearm in the case of loca l anti hi stamine, o r by obta ining responses to CGR P and hi sta mine on the other forearm 24 h before ad ministration of systemic antihistam ine. In 1 individual (TJW) the stability of the vasodil ator activity of CGRP was tested in autologous plasma and blood . Periphera l venous blood was co ll ected into heparin (10 units / ml). The GH.P (800 pmol/m l) was incubated for 1 h at 37°C in plasma, blood, or sa line. The samples were centrifuged and supernatants diluted 4-fold in sa line before intradermal in­j ection of 50 J.LI in'to the forea rm. Each test sample was compared with a control injecti on where CGRP had been added to the plasma , blood, or sa line after incubation .

The area of the res ponses was measured by placing a clear plastic sheet over the skin and tracing the edges of the wheal and ery­thema at 10 min , 1 h , 2 h, 3 h, and 4 h after injection. All results are expressed as area (sq mm) and ca lculated by use of an Apple digitizer board attached to an Apple li e computer. The responses were usually photographed.

RESULTS

Fig 1 shows that CGRP increased blood flow, assessed using a laser Doppler flow meter. Ten minutes after injection , the saline­injected site exhibited a raised blood flow , but this returned to basal values by 1 h . By this time, clearly defined local reddenjng (141 ± 20 sq mm, mean ± SE) was observed at sites injected with CGRP in all 4 individuals and blood flow was high. Blood flow decreased from 1 h, a decrease that correlated with a grad ual return to normal skin color of the erythematous area. The re­sponses of human skin to certain inflammatory mediators is well documented: histamine induces the classical triple response (wheal, loca l reddenin g, and surrounding fl are) [8] which is also observed with substan ce P [9] and VIP [10], while PGE2 and PGh are potent in inducing loca l reddenin g [11 ,12]. In the results shown in Fig 2 histamine, substance P, VIP, and PGE2 (in 1 individual) induced whea l and flare responses which had faded by 1 h after injection . Prostagland in E2 in the other 3 volunteers and PGI2 in all subjects caused only local reddening. The loca l reddening response to PGh faded by 1 h and the response to PGE2 disappeared by 3 h. The CGRP (1 and 10 pmol), in every individual, induced local red­denin g which appeared slower in onset than that induced by PGE2

or PGI2, because it spread slowly in the first hour as indicated by the increasing area (see Fig 2b) . The color of the skin was bright pink 1-2 h afier injection; then the response slowly faded and the area of erythema became smaller until it had disappeared in most subj ects by 3 h (1 pmol) or 5-6 h (10 pmol). However a higher dose of CGRP (100 pmol) did initially induce a wheal and flare followed by persistent local reddening. The loca l reddening in­du ced by CGRP was similar to the persistent local reddening of skin to VIP . T his response to VIP consisted of a small area of erythema wh ich became evident as the wheal and flare to VIP subsided. As shown in Fig 2, the persistent erythematous response to VIP was smaller than that due to CGRP and faded more quickly in each individual. The persis tent local reddening by VIP was not

THE JOURNAL OF INV ESTIGATIVE DERMATO LOGY

90

70

30

10 i . pre-. 0 injection 2 3 4

TIME (h)

Figure 1. Local increased blood flow after intradermal injection of CGRP. Blood flow was measured at 3 sites on the volar surface of the same forearm in 4 individuals. Solid circle, sites that received intradermal CGRP (10 pmol/SO J.Ll) ; so lid square, sites that received intrade rmal sa line (50 J.Ll) ; ope" square, llninjected sites . B lood fl ow was measured before and at the times indica ted after injections. Results are expressed as mean ± SEM, n = 4.

reported in a previous study of the response of human skin to VIP [10] . The CGRP induced a similar persistent local reddening whcn injected into skin after its incubation in vitro in autologous blood and plasma for 1 h at 37°C. This suggests either that CGRP is stable in blood, or that during incubation a similarly potent metabolite is formed; the finding is supported by our studies in the rabbit [13].

Thc efficacy of HI antihistamines was tested by their effect on the histamine-induced triplc response. As prcviously reported [14], H I antihistamine's had little effect on the histaminc-induccd wheal but caused an almost total suppression of the histamine­induced flare when administered either locally, (chlorphenira­mine, 10 min: 96.0 ± 2.0% reduction in area; mcan ± SEM, n = 3) or systcmica lly , (Terfenadine, 10 min: 91.3 ± 8.0% re­duction, n = 4). Interestingly, the histamine-induced local red­dening was abolished by local antihistamines but unaffected by systemic antihistamines. Local chlorpheniramine had little initial effect on CGRP-induced local reddening (10 min: 14.3 ± 2.4% reduction, n = 3) but, by 2 h a reduction in thc intensity oflocal reddening was apparent. By 3 h after injection this inhibition was more obvious and associated with a reduction in the area of local reddening (56.7 ± 21.3% reduction). Another antihistamine (promethazine hydrochloride, Phenergan, 1 J.Lg) had a similar inhibitory cffect when given locally. In contrast, thc persistence of the CGRP-induced local reddening was unaffected by systemic terfenadine up to 4 h after injcction (4 h: 1.7 ± 18.0% difference, n = 4) . Systcmic administration of chlorpheniramine, similarly had no effect on CGRP-induccd local reddening. It is unlikcly that the inhibitory effects oflocal antihistamines were due to local anesthetic effects as lignocaine hydrochloride (Xylocaine, 0.02%) did not affect the CGRP-induced local reddcning.

VOL. 87, N O.4 OCTOBER 1986

E E 0-VI

~ 400 a: <t

150

_ 100

E E

:t '" w a: ~

50

1 ------1' )/'[......... . .....

[

\.

2 TIME (h)

fI

b

1

--, -'l.

-- ---. -. - .----- .I

3 4

Figure 2. Skin responses after intradermal injections of vasodilators. fI ,

Immediate wheal and flare. The wheal (opell coil/mIlS) and flare (shaded co /umll s) responses to intradermal inj ection of histamine (HA, 500 pmol) substance P (SP, 10 pmol), vasoactive intes tinal peptide (VIP, 10 pmol). and PCE2 (10 pmol) are shown. Responses were calculated as area (sq rnm) and measured 10 min after injection . They are expressed as mean ± SEM . n = 4, except for PGE2 where resul ts are fro m 1 volun teer. The other 3 volunteers exhibi ted only loca l reddening in response to PGE2.

No wheal and fl are was seen after injection of CCRP (10 pmoD . PCh (10 pmol) or sa line (SAL) in any volunteers. b, Local reddenil~g. The area (sq mm) and duration of erythema in response to CGRP (circle, 10 pmol), solid sg'ltlre, PGE2 (10 pl11ol), PGI2 (opell sql/are, 10 pmol), and VIP (Iriml,!!le, 10 pmol) is shown. r~csponses arc expresscd as mcan ± SE~ , n = 4, except for *, PGEz 10 min reading where n = 3. At 10 min I"volunteer responded to PGE2 and all volunteers responded to VIP with a wheal and flare (see a) ; it was not possible to accurately mcasure loca l reddening in the presence of a wheal and flare.

D[SCUSSION

T hese observat ions support ou r previous guantitative studies in animals and preliminary studies in humans [4] w hich showed that CGRP is a vasodilator. In this paper we have shown by visual observation and using laser Qoppler blood flow measurements that intraderma lly inj ected CGRP has potent local vasodilator acti vity in humans which exceeds , in duration of action, that of any vasodilator yet tested. T he characteristics of other potent vasodilators , PGEz, PGI2 , substa nce P, VIP, and histamine, were compared w ith those of CGRP. At the doses used, responses were characterist ica ll y different, The PGs induced np inly loca l erythema, ;'ast in o nset, and pers istin g fo r no longer than 2-3 h . Su bstance P induced loca l erythem a and a whea l and fl are which

VASOD ILATO R ACTIVITY OF cent> 535

had faded by 1 h . VIP induced a similar initia l response to that of substance P, but with a loca l erythem a persisting for 3 h. In con trast, CGRP induced on ly intense local erythema, slow in onset but very persi stent, up to 10-12 h duration at high doses [41· .

We have suggested that CGRP may be the endogenous vaso­di lator released fro m the endings of branches of afferent nerves in the skin , mediating the axon reflex flare that surroun ds the loca l response to inflammatory stimuli [4]. T his is supported by the recent finding of immunoreactive CGRP in cutaneous nerves associated w ith blood vessels in humans (Dr SpringaU and Pro­fessor Polak , personal communica tion) . However, immunoreac­tive substance P and VIP have also been detected in human skin [1 5] and these substances have been claimed to be the endogenous med iators of th e fl are reaction [1 0, 14,16]. At present, the evidence for and aga inst any of these ca ndidates is circumstantial; never­theless substance P and VIP themselves induce a fl are, probably by releasing histamine w hich stimulates pain receptors and trig­gers the axon refl ex. If either of these peptides is the endogenous m ediator, it is diffi cult to see how the area of fl are wouJd be limited-release of histam ine in the area surroundin g the 'ini tia l stimulus inducing a secondary axon reflex, and so on-unless other mechanisms exist to lin1it the flare area. T hus CGRP is a better cand idate as the mediator of the flare because it di lates onl y the arterioles to w hich it is exposed, without itself triggering an axon reflex fl are. On the other hand, the flare reaction is fast in onset and can be (dependin g on the intensity of the inflammatory stimulus) of short duration. From th is point of v iew CGRP is apparently not a good candidate. However, the slow onset of responses to intradermall y injected CGRP ma y be because it is a relatively large m olecule (when compared to released histamine) and m ay take time to diffuse to superficial mi crovessels . Our previous experim ents in rabbit 1'131 together w ith the res ult of the present experiment in humans sugges t that the vasodi lator activity of CGRP is stable in plas m a, and therefore presum ably in extravascul ar tissue fluid . T his m ay part.1 y exp lain the long duration of action of in'traderma ll y injected CGRP. [n contrast, CG RP released from cutaneous nerve endin gs may be metabo­li zed ncar its site of release (e.g., b y enzy mes on arterio les) or there m ay be a reuptake mechanism in nerves .. Thus, CGRP released at low levels may ca use a transient flare, whereas high levels rel eased by an intense inflam m atory stimu lus may induce a persistent fl are.

T he inhibitory effect of loca l H, antih istamines on the persis­tence of loca l ery thema induced by CGRP was un expected. If CGRP, at the doses used , releases histamine, one wou ld expect a w hea l and flare reaction, w hich was not seen with CG RP at these concentrations . Further, o ral H, antihistamines suppressed res ponses to histam ine itself, but had no effect on erythem a in­duced by CGRP. Hence, the phenomen on may represent an ac­tivity of loca l antihistamines independent of hista mine receptor antagonism , but apparentl y not associated with loca l anesthetic activity of antihistamines (as li gnoca ine did not suppress the re­sponse to CGRP). T he paradoxica l o bservation Illa y relate to the effectiveness o flo call y appli ed anti hista mines (w hen applied ea rly) .in suppressing certain local allergic reactions in the ski n .

Locall y released CGRP m ay be in volved in the ph ysiologic contro l of tissu e blood fl ow and a defici~n cy in release or action may be an important contributory fac tor in peripheral vascular diseases. [n some patients w ith Rayna ud 's phenomenon or dia­betic polyneuropath y, electrica l transcutaneous nerve stimulation induces vasodilation and relief from ischemic pain. It has been suggested that the release of an endogenous vasodilator is partly responsible for the beneficial effects [17]. Transcutaneous nerve stimulation has been associated with a rise in plasma VIP in these patients and normal indi viduals [18]. altho ugh further studies r 19] have suggested tlpt this endogenous vasodi.1ato r is not V[ P nor any other known substance. It is possible that this unidentified vasodilato r is CGRP, w hich we show here to retai n its vasodilator ~ctivity in plasm a. T hus, exogenous CGRP, or vasodilator 3n-

536 I3I(A IN ET AL

alogues, should be considered as a possib le therapeuti c agent in peripheral vascular diseases. We demonstrate here that the un­stable PG I2 induces a ve ry acute vasodi lator response il1 the sk in w hen compared w ith the response to CGRP, and yet PG I2 has been reportcd to have long-term beneficial effects in periph era l ischemic diseases [20,21] . In support of the loca l effects we have demonstrated, intravenous CGRP has been shown to lower sys­temic arterial blood pressure by reducing peripheral res istance [22].

Recen tly a rad ioimmunoassay for CG RP has been developed [6'1. Levels of immunoreactive CGRP in normal plas ma have been reported, but it is not clea r w hether this refl ects the level of the intact, biologicall y active molecu le. A second hum an calci­tonin/CGRP. gene has beel) described that encodes ano ther CGRP [23]. This second CGRP differs in structure at 3 amino acid s it~s from the CGRP described and used in these experim ents. H ow­ever, preliminary resuJts suggest that bo th of th ese forms of CGRP have sim ilar potencies as vasodi lators 124]. As CGRP has been extracted from medulJary thyroid ca rci no ma tissue, and elevated levels of immunoreactive CGR P have been detected in patients w ith such tum ors [5]. it is possible that flu shing reactions seen in these individua ls (and possib ly those affected w ith other tum ors) may resu lt from hig l; circul atin g CGRP level s. ' . We consider that the stud y of the roles ofCGRP in ph ysio logy

and pathology, and its possible use as a th erapeutic agent, may prove to be fruitfu l areas for future resea rch.

We 1IJ01lid like to thallk Dr. Fiolla C II/'l/Iillghall' , IlIStilllte oj DenllalOlogy, jor her " ell' lIJith laser Doppler blood flollJ lII eaSll remellts alld the jo llollJillg lII elllbers ojlhe Departlll el'll ojBioehclllistry, IlII perial College: TOllY Etie'lII ejorl-/PLC, Dr Maria Partieo jar lII ass spectrollletry , attd DalJid Fea t/terbe jor alllillo acid al1alysis. We are also illdeb/cd to the (Jo /lllll eers (./)" 0 took parI ill t"is slIIdy .

REFERE N CES

1. Rosenfe ld MG, M ermodJ-J, Amara SG, Swanson LW , Sawchenko PE, Rivier J , Va le WW, Evans RM : Production of a novel neu­ropeptide encoded by the calcitonin gene via tissue-speci fi c RN A process ing . N atu re 304:129-135, 1983

2. Edbrooke MR, Parker 0, M cVey JH , Riley JH, Sorenson GO, Pet­tengell OS, Craig RK : Expression of the hum an calcitonin /CG RP gene in lung and thyroid carcinoma . E M BO J 4:715-724, 1985

3. Morris HR, Panico M , Etienlle T, T ippins JR, Gi rg is SI, Macintyre 1: Iso lation and characterization of human calcitonin gene- related peptide. N ature 308:746-748, 1984

4. Brain SD, Williams TJ , T ipp ins In., Morris HR , MacIntyre I: Cal­citonin gene- related peptide is a potent vasodi lator. N ature 313:54-56, 1985 .

5. Cohen SL, M acinty re I, Grahame-Smith D , Walker J G: Alcohol­stimulated calcitonin release in medullary ca rcinoma of the thy­roid. Lancet 2:1 172- 11 74, 1973

6. Girgis SI, M acdonald DWR, Stevenson J C, Bevis PJR, Lynch C, Wimalawansa 'SJ, Self C H , M o rri s HR , M acintyre I: Calci tonin

T H E JOU I(NAL OF INVESTIGATIVE DERMATOLOGY

gene- related peptide: potent vasod il ato r and maj o r product of ca l­citonin gene. Lancet 2: 14-16, 1985

7. H oll oway AG: Laser Doppler m easure m ent of cutaneous blood fl ow. J Invest D ermato l 69:306-309, 1977

8. Lewis T: The Blood Vessels of the H uman Ski n and T heir l~esponses. London, Shaw and Sons, 1927

9. Ha germ ark 0, H oHelt T, Pern ow 13 : Flare and itch induced by substance P in hum an skin . J In ves t D ermato l 71 :233-235, 1977

10. An and A, Bloom SR, M cG rego r G P: Topica l capsaicin pre trea tmen t inhibits axon re fl ex vasod ilatation caused by som atos tatin and vasoactive inrestinal polypeptide in human skin . Br J Pharmacol 78:665-669, 1983

11 . So lo man LM , Juhl in L, Kirschbaum MM: Prostag landins on cuta­neous vascubture. J In vest Dermatol 51 :280-282, 1968

12. Higgs EA, O'Grady J , T hrower PA , M oncada S: Prostacyclin: in­fl ammatory effects in hum an sk in . Fourth Inrerna\iona l Prosta­g landin Conference, Washington DC, M ay 1979, Meeting Ab­stracts p 48

13. Brain SD, Williams TJ: In flammatory oedem a induced by synergiS111 between ca lcitonin gene- related peptide and m ediators of increased vascula r permeabi lity. Br J Pharm aco l 86:855-860, 1985

14. Forem an J C, Jo rdan CC, Oehme P, Renner H: Structure-acti vi ty relationships for 'some substance P-relatcd peptides that cause w heal and fl are reactions in hunpn skin. J Physiol 335:449-465, 1983

15. B loo m SR, Po l ~ k JM : I~egul ato r y pcp tides and the skin. C lin Exp Dcrmatol 8:3- 18, 1983

16. BUTllstock : Autonomi c neuroeffector junctions-relex vasodila­tation of the sk in. J In ves t Dermatol 69:47-57, 1977

17. Kaada 13 : Vasodi latation induced by transcutaneous nerve stim u la­tion in periphera l ischem ia (Ibynaud's phenomenon and diabetic polyneu ropath y). E ur H ea rt J 3:303-314, 1982

18. Kaada B, O lsen E, Eiclscn 0: In sea rch of mediato rs of skin vaso­di lation induced by transcutancou ~ nerve stimul ation: III. Increase in plasm a VIP in no rm al subj ects and in Raynaud 's disease. G~n Pharmacol 15 :1 07-11 3, 1984

19. Kaada B, H elle KB : In sea rch of mediators of skin vasodilation indu ced by transcutaneous nerve stimulat io n : IV. In vit ro bioassay of the vasoinhibi tory activi ty of sera from patients suffering fr0111 peripheral ischaemia. Gen Pharma col 15:115-122, 1984

20. Szczeklik A, Nizankowski R , Shawinski S, Szczeklik J , G luszko P, Gryglewski RJ: Successful therapy of advanced arte rioscle rosis obli terans with prostacycl in. Lancet 1 :1111 - 111 4, 1979

21. Belch JF, Drury JK , Capell H , Forkes C D, N ewman P, M cK enzie F, Leiberman P , Prentice C RM: In termittent epoprostenoI (pros­tacyclin) ipfusion in patients with Haynaud's syndrome; a doubIe­bli nd contro lied tri al. Lance t 1:313-315, 1983

22. Struthers AD , 'I~rown MJ , Bea cham JL, Mo~ris HR , M acIn tyre I, Stevenson J C: T he acute effect of hu man calcitonin genc- rcbtcd peptide in m an. J Endocrin o l 104(suppI): 129, 1985

23. Stecnbergh PH, H oppencr JWM , Zandberg J , Lips CJ M , J ansz HS : A second human calcitonin/CG I~P gene. FEBS Lett 183:403-407 , 1985

24. Brain SD, M acI ntyre I, Wi ll ia ms TJ: A secon d form of human ca l­cito nin gene-rebted peptide w hich is a potenr vasodilator. Eur J Pharm aco l 124:349-352, 1986