Vernonia spp. (Asteraceae) are reported in Mexican herbo-laria for used in menstrual disorders, hair regeneration,dysentery, and as a coagulant.1) Brazilian folk medicine hasshown that the crude extract prepared from leaves of V. con-densata prevents stomach and liver disturbances (sic).2) Sev-eral of these applications suggest presence of compounds ac-tive on smooth muscle. Sesquiterpene lactones, mainly glau-colide- and hirsutinolide-type, have been isolated from Ver-nonia spp. –sensu latu–.3—6) However, several hirsutinolidesmight be artifacts formed from glaucolides on exposure toacidic adsorbents during chromatography.7—9) Glaucolideshave been reported to elicit molluscicidal, antitumoral, andantimicrobial properties5,10,11); but their effects on smoothmuscle have not been investigated previously. Parthenolide, astructurally related sesquiterpene lactone (Fig. 1D) bearingalso a germacra-1(10),4-diene-4-epoxide skeleton is knownto inhibit smooth muscle contractility likely due to the a-methylene g-lactone moiety.12) Therefore, the effects of glau-colides D (Fig. 1A) and E (Fig. 1B) on vascular and uterinesmooth muscle contraction were studied. The effect of a hir-sutinolide type sesquiterpene lactone (Fig. 1C) lacking bothgermacrane skeleton and a-methylene g-lactone was also in-vestigated. We have assayed the action of these compoundson smooth muscle pre-contracted with either depolarizinghigh-KCl solution or an agonist (noradrenaline for aorta andoxytocin for uterus). This model was selected as mean to in-dicate whether glaucolides D and E are able to relax the con-traction dependent on different pathways of calcium influx.

MATERIALS AND METHODS

Compounds Glaucolides D (Fig. 1A) and E (Fig. 1B)were isolated from Vernonia liatroides by chromatographyand identified by spectroscopic methods as previously de-scribed.3) The hirsutinolide [1R,4S,5R,6S,8S,10R]-8,10,13-triacetoxy-1(4)-epoxy-1,5-dihydroxygermacr-7(11)-en-6(12)-olide (Fig. 1C) was obtained from Vernonia paniculata astransformation product of glaucolide B treated with bentonite

earth.7) Vernonia liatroides was collected in Ixtapan de la Sal,State of Mexico, while V. paniculata was collected inMatatlán, State of Oaxaca, in Mexico. Herbarium specimensare deposited in IMMS and MEXU herbaria, respectively.

Animals Sprague-Dawley rats, female for uterine exper-iments and male for aorta experiments, body weight approxi-mately 300 g, were used. Virgin female rats received a subcu-taneous injection of estradiol benzoate (40 mg/kg) 24 h priorto the experiment. Animal experiments were performed fol-lowing the recommendations of the Policies on Animal Ex-perimentation of the Scientific Committee of the InstitutoMexicano del Seguro Social; animals were decapitated, anduterus or descending thoracic aorta were isolated, fat andconnective tissue were removed, and tissues were placed inKrebs–Ringer bicarbonate solution.

Uterus Experiments Uterine rings (5—7 mm long)were placed in a 5 ml organ bath containing Krebs–Ringerbicarbonate (KRb) solution with the following composition(mM): NaCl, 120; KCl, 4.6; CaCl2, 1.5; NaHCO3, 20;KH2PO4, 1.2; MgSO4, 1.2, and glucose 11, maintained at 37 °C and bubbled with a mixture of 95% O2–5% CO2

(pH�7.4). Contractions were recorded isometrically by aforce–displacement transducer (Grass, FT03) connected to aGrass polygraph (model 7B). Tissues were allowed to equili-brate for 60 min during which KRb was changed every 20min and maintained under optimal tension of 1 g prior initia-tion of the experimental protocol. After equilibration, uterinerings were bathed in a depolarizing solution (60 mM KCl)prepared by equimolar substitution of NaCl for KCl to ex-pose tissues to a single sub-maximal concentration of KCl(60 mM), and to elicit contractile response. Control contrac-tile responses were considered as two successive similar re-sponses. To study the effect of the compounds on a depolar-ization-induced contraction, a third maximal contraction toKCl 60 mM was recorded during 20 min to obtain steady re-sponse. After reaching plateau response, each compound (A,B, or C) was added in progressively increasing cumulativeconcentrations (3, 10, 30, 100 mM) to the bath solution and

112 Notes Biol. Pharm. Bull. 26(1) 112—115 (2003) Vol. 26, No. 1

∗ To whom correspondence should be addressed. e-mail: mariagcampos@yahoo.com © 2003 Pharmaceutical Society of Japan

Relaxation of Uterine and Aortic Smooth Muscle by Glaucolides D and Efrom Vernonia liatroides

María CAMPOS,*,a Martha OROPEZA,a Héctor PONCE,a Jaquelina FERNÁNDEZ,a

Manuel JIMENEZ-ESTRADA,b Héctor TORRES,b and Ricardo REYES-CHILPAb

a Unit of Medical Research in Pharmacology, National Medical Center S. XXI, Mexican Institute of Social Security; SanFrancisco 350–502, Col. Del Valle, Mexico City, 03100, Mexico: and b Institute of Chemistry, National University ofMexico; Mexico City, 04510, Mexico. Received July 29, 2002; accepted September 30, 2002

Vernonia spp. (Asteraceae) are used in herbolaria in Latin America in menstrual and stomach disorders,suggesting smooth muscle relaxing properties of some of their chemical constituents. For pharmacological sup-port for this belief, sesquiterpene lactones glaucolides D and E were assayed on isolated rat smooth muscle.Glaucolide E proved more potent than glaucolide D to relax high KCl- or noradrenaline-induced contractions inaorta and to relax the high KCl-contraction in uterus. Hirsutinolide-type sesquiterpene lactone also was testedbut displayed no effect. Relaxation of smooth muscle by structurally related sesquiterpene lactone parthenolidehas been attributed mainly to the aa-methylene gg-lactone moiety; because glaucolides D and E lack this func-tional group, their relaxant properties may rely on other alkylating sites such as C10 of the germacra-1(10),4-diene-4-epoxide skeleton.

Key words Vernonia; sesquiterpene lactone; glaucolide; hirsutinolide; smooth muscle; parthenolide

the effect of each concentration on the tonic contraction wasobserved for 5 min. In every set of experiments, a control tis-sue was run concurrently; this control was treated in thesame manner as test tissues but only vehicle was added. Tostudy the effects of compounds on agonist-induced contrac-tile response, contraction was induced by 10 mU/ml oxy-tocin. After 10 min, the compound was added in progres-sively increasing cumulative concentrations (3, 10, 30, 100mM) on contractile response.

Aorta Experiments Aortic rings (approximately 5 mmin length) were mounted in organ baths containing 5 ml KRbsolution, maintained at 37 °C and bubbled with a mixture of95% O2–5% CO2 (pH�7.4). The equilibration period wassimilar to that of uterine rings, but optimal tension was 2 g.Subsequently, a protocol similar to that of uterine rings wasperformed to elicit KCl (60 mM)-induced contractile re-sponses and to assay glaucolide effect. To study the effects ofglaucolides on agonist-induced contractile response, thetonic contraction was induced by 10�6

M noradrenaline con-centration, producing sub-maximal contraction. To avoid no-radrenaline oxidation in the bath solution, 50 mM ascorbicacid was added to noradrenaline solution freshly prepared foreach experiment. When contractile response to noradrenalinewas stable, the compound was added in progressively in-creasing cumulative concentrations (3, 10, 30, 100 mM) at 5min-intervals.

Drugs L-Noradrenaline bitartrate, 17b-estradiol-3-ben-zoate, and ascorbic acid were obtained from Sigma ChemicalCo. (St. Louis MO, U.S.A.). Dimethyl sulfoxide and ethanolwere purchased from Mallinckrodt Baker (Mexico), whileoxytocin was obtained from Sandoz (Mexico). Glaucolideswere dissolved in dimethyl sulfoxide/ethanol (2 : 5, v/v).Final concentration of DMSO/ethanol in the bathing solutiondid not exceed 0.2%, innocuous to tissue contractile activity.Estradiol benzoate was dissolved in corn oil (100 mg/ml).

Data Analysis Maximal tissue contractile response wasdetermined as the response elicited by 60 mM KCl or by ago-nist (10�6

M noradrenaline or 10 mU/ml oxytocin). Relax-ation induced by compounds A, B, and C was assessed aspercentage of relaxation of maximal contractile response.EC50 is the effective concentration 50 (concentration of com-pounds causing 50% relaxation of smooth muscle contractileresponse). Results are expressed as mean�S.D. of 4—10 ormore preparations each obtained from different animals.One-way analysis of variance and Bonferroni test for multi-ple comparisons were used to evaluate statistical differences,

and p�0.05 was considered significant.

RESULTS

Effect of Glaucolides on High KCl-Induced Contrac-tions Smooth muscle contractions induced by 60 mM potas-sium in both aorta and uterus were relaxed by glaucolides Dand E in a dose-dependent manner (Fig. 2). No effect was ob-served when hirsutinolide was assayed. EC50 and Emax valuesindicate that glaucolide E was more potent, and more effec-tive than glaucolide D to relax high KCl-induced contractionin both tissues (Table 1). When the effect of glaucolide Dwas compared between high KCl- and agonist-induced con-tractions, glaucolide D showed to be more potent acting onhigh KCl-induced contractions in aorta (p�0.001) and onoxytocin-induced contractions in uterus (p�0.05). Regardingglaucolide E, this showed to be more potent and effective act-ing on high KCl- than on noradrenaline-induced contractionsin aorta (p�0.01). The effect of glaucolide E on high KCl-and oxytocin-induced contractions in uterus was not statisti-cally different (Tables 1, 2).

Effect of Glaucolides on Agonist-Induced ContractionsGlaucolides D and E relaxed in a dose-dependent mannercontractions induced by noradrenaline (10�6

M) and oxytocin(10 mU/ml) in aorta and uterus, respectively (Fig. 3). Thehirsutinolide lacked effect on both tissues. Glaucolide Eshowed to be more potent and effective than glaucolide D onnoradrenaline-induced contractions in aorta (p�0.05). Nosignificant difference was observed between glaucolides Dand E on oxytocin-induced contractions in uterus (Table 2).

DISCUSSION

To our knowledge, this is the first report concerning relax-ant effects of glaucolides on smooth muscle. We have testedthe action of these compounds on smooth muscle contractioninduced by either high KCl or agonist (noradrenaline foraorta, oxytocin for uterus).

Glaucolides D and E relaxed contractions induced by highKCl or agonist in aorta and uterus; the hirsutinolide was de-void of activity. These results suggest that sesquiterpenoidslacking a-methylene g-lactone moiety could also relaxsmooth muscle contraction as long as they bear alternativealkylating sites such as C-10 provided by germacra-1(10),4-diene-4-epoxide skeleton. The importance of alternativealkylating- reactive sites in addition to a-methylene g-lac-

January 2003 113

Fig. 1. Structures of Glaucolides-D (A), E (B), [1R,4S,5R,6S,8S,10R]-8,10,13-Triacetoxy-1(4)-epoxy-1,5-dihydroxygermacr-7(11)-en-6(12)-olide (C), andParthenolide (D)

tone was first addressed by Fischer et al.,13) who examinedthe antimicobacterial activity of a variety of sesquiterpenelactones including parthenolide. In fact, it was proposed thatfacile transannular cyclization of parthenolide may explainthe high activity of this compound. A similar mechanismmay be operating on smooth muscle contractility in the caseof glaucolides D and E. It is noteworthy that hirsutinolide,transformed by a 1—4 transannular cyclization, lacked activ-ity, whereas glaucolide E, which bears a reactive a-methyl-ene in the acyl substituent, was more potent than glaucolide

D, exhibiting an epoxide.Spasmogenic responses of smooth muscle to high potas-

sium solutions can be explained in terms of calcium influxfrom extracellular milieu via voltage-operated calcium chan-nels;14,15) this mechanical response to KCl can be completelyinhibited by calcium entry blockers.16,17) The relaxing effectof several products isolated from medicinal plants has beenrelated to blockade of calcium influx to smooth musclecell.18,19) Additionally, agonists such as noradrenaline or oxy-tocin induce calcium influx by activating receptor-operated

114 Vol. 26, No. 1

Table 1. Relaxant Effect of Glaucolides D and E on High KCl-InducedContraction in Rat Smooth Muscle

Glaucolide D Glaucolide En

EC50 (mM) Emax (%) EC50 (mM) Emax (%)

Aorta 4 10.42�2.18a,b 59.36�8.00a 7.00�1.74 94.25�11.66Uterus 8 25.57�3.06c,d 83.28�12.96d 13.20�1.45 98.06�3.71

Values represent mean�S.D. a) p�0.05 vs. glaucolide E/high KCl/aorta.b) p�0.001 vs. glaucolide D/agonist/aorta. c) p�0.05 vs. glaucolide D/agonist/uterus. d) p�0.01 vs. glaucolide E/high KCl/uterus.

Table 2. Relaxant Effect of Glaucolides D and E on Agonist-Induced Con-traction in Rat Smooth Muscle

Glaucolide D Glaucolide En

EC50 (mM) Emax (%) EC50 (mM) Emax (%)

Aorta 6 26.11�4.69 50.12�6.02 20.87�3.54a 73.11�9.71a

Uterus 10 18.50�3.33 93.25�6.23 16.46�3.12 94.28�6.26

Values represent mean�S.D. a) p�0.05 vs. glaucolide D/agonist/aorta and glau-colide E/high KCl/aorta.

Fig. 2. Glaucolides D and E Relaxed in a Dose-Dependent Manner the High KCl-Induced Contraction in Rat Smooth Muscle

Bars represent the mean of 4—8 observations, vertical lines represent the standard deviation.

Fig. 3. Glaucolides D and E Relaxed in a Dose-Dependent Manner the Agonist-Induced Contraction in Rat Smooth Muscle

Bars represent the mean of 6—10 observations, vertical lines represent the standard deviation.

channels.20)

In the present study, glaucolides D and E relaxed the con-traction induced by either high potassium or agonist (nora-drenaline on vascular smooth muscle, oxytocin on uterus).Based on previous reports,14—20) we suggest that this findingmight be explained by the inhibition or blockade of calciuminflux via voltage- and receptor-operated channels in vascu-lar and uterine smooth muscles. However, at the tested doses,glaucolide E was more potent and effective than glaucolide Dto relax KCl-induced contraction in both tissues and in nora-drenaline-induced contraction in aorta, whereas in the oxy-tocin-induced contraction of uterine smooth muscle bothglaucolides displayed a similar effect. A likely explanationfor this finding might be that glaucolide E has higher affinitythan glaucolide D for calcium channels. However, this wouldnot explain the similar effect on oxytocin-induced contrac-tion in uterus. In this case, oxytocin-induced contraction in-volves calcium influx through receptor operated-channelsand calcium release from internal stores.21) Uterine intracel-lular calcium stores may be very efficient, i.e., oxytocin-in-duced contractile response in myometrium is extremely resis-tant to extracellular calcium removal.22) This might be thereason we observed no difference between the relaxing effectof glaucolides E and D on oxytocin-induced contractile re-sponse in uterine smooth muscle. On the other hand, thesecompounds might have other sites of action, such as proteinkinase C, phospholipase C, etc. Therefore, more studies areneeded to investigate the precise mechanism of action ofglaucolides on cellular pathways of calcium entry and/or onthe mechanism of calcium release from internal stores.

REFERENCES

1) Aguilar A., Camacho J. R., Chino S., Jacquez P., Lopez M. E.,“Herbario Medicinal del Instituto Mexicano del Seguro Social. Infor-

macion Etnobotanica,” IMSS, Mexico D. F., 1994, p. 64.2) Frutuoso V. S., Gurjao M. R., Cordeiro R. S., Martins M. A., Planta

Med., 60, 21—25 (1994).3) Maldonado J. E., Martínez R., Martínez V. M., Rev. Latinoamer.

Quím., 11, 58—59 (1980).4) Mabry T. J., Bedel-Baset Z. H., Padolina W. G., Jones S. B., Biochem.

Sys. Ecol., 2, 185—192 (1975).5) Alarcon M. C., Callegari J. L., Herz W., Planta Med., 56, 271—273

(1990).6) Bazon J. N., Callegari Lopes L., Vichnewski W., Dias D. A., Nagamiti

K., Cunha W. R., Herz W., Phytochemistry, 44, 1535—1536 (1997).7) Jiménez M., Ortega A., Navarro A., Maldonado E., Van Calesteren M.

R., Jankowski C. K., J. Natural Products, 58, 424—427 (1995).8) Borkosky S., Bardón A., Catalán C. A. N., Díaz J. G., Herz W., Phyto-

chemistry, 44, 465—470 (1997).9) Kotowickz C., Bardón A., Catalán C. A. N., Cerda-García-Rojas C.

M., Joseph-Nathan P., Phytochemistry, 47, 425—428 (1998).10) Jisaka M., Ohigashi H., Takegawa K., Huffman M. A., Koshimizu K.,

Biosci. Biotechnol. Biochem., 57, 833—834 (1993).11) Oketch-Rabah H. A., Christensen S. B., Frydenvang K., Dossaji S. F.,

Theander T. G., Cornett C., Watkins W. M., Kharazmi A., LemmichE., Planta Med., 64, 559—562 (1998).

12) Hay A. J., Hamburger M., Hostettmann K., Hoult J. R., Br. J. Pharma-col., 112, 9—12 (1994).

13) Fischer H. H., Lu T., Cantrell C. L., Castañeda-Acosta J., Quijano L.,Franzblaus S. G., Phytochemistry, 49, 559—564 (1998).

14) Amedée T., Mironneau C., Mironneau J., Br. J. Pharmacol., 88, 873—880 (1986).

15) Edwards D., Good D., Granger S., Hollingsworth M., Robson A.,Small R. C., Weston A. H., Br. J. Pharmacol., 88, 899—908 (1986).

16) Calixto J. B., Loch S., Br. J. Pharmacol., 85, 189—195 (1985).17) Granger S. E., Hollingsworth M., Weston A. H., Br. J. Pharmacol., 85,

255—262 (1985).18) Rojas A., Cruz S., Rauch V., Bye R., Linares E., Mata R., Phytomedi-

cine, 2, 51—55 (1995).19) Campos M., Valencia A., Ponce-Monter H., Uribe C., Osuna L.,

Calderon J., Phytotherapy Res., 11, 11—16 (1997).20) Karaki H., Weiss G. B., Gastroenterol. 87, 960—970 (1984).21) Anwer K., Sanborn B. M., Endocrinology, 124, 17—23 (1989).22) Anselmi E., D’Ocon M. P., Villar A., Prostaglandins, 34, 351—358

(1987).

January 2003 115