A

TeRw(t(waCf©

K

1

ie

aHramPp

0d

Available online at www.sciencedirect.com

Journal of Ethnopharmacology 117 (2008) 58–68

Hypotensive and vasorelaxant effects of the procyanidinfraction from Guazuma ulmifolia bark in normotensive

and hypertensive rats

Gil Alfonso Magos a,∗, Juan Carlos Mateos b,Elna Paez a, Gabriela Fernandez a, Carlos Lobato b,

Carmen Marquez b, Raul Guillermo Enrıquez b

a Departamento de Farmacologıa, Facultad de Medicina, Universidad Nacional Autonoma de Mexico,Apartado Postal No. 70-297, 04510 Mexico, D.F., Mexico

b Instituto de Quımica, Universidad Nacional Autonoma de Mexico, Mexico

Received 30 October 2007; received in revised form 13 January 2008; accepted 17 January 2008Available online 20 January 2008

bstract

he aim of the study: was to investigate the in vivo and in vitro cardiovascular activity of a procyanidin fraction (PCF) obtained from acetonextract of Guazuma ulmifolia bark which has traditionally been used as an antihypertensive agent.esults: 10 mg/kg PCF doses orally administered to sugar-fed hypertensive rats decreased both the systolic arterial pressure and the heart rate,hereas the same doses intravenously administered induced arterial hypotension which was attenuated by NG-nitro-l-arginine methylester

L-NAME 31 mg/kg) pretreatment. In these experiments we employed carbachol as a positive control test. The PCF reduced the contrac-ion induced by norepinephrine (1 × 10−7 M) in isolated aortic rings of normotensive (IC50 = 35.3 ± 12.4 ng/mL) and sugar-fed hypertensiveIC50 = 101.3 ± 57.2 ng/mL) rats. This relaxant activity was inhibited by either vascular endothelium removal or L-NAME (30 �M) pretreatment,

+

hile indomethacin (10 �M) or atropine (10 �M) had no effect. Preliminary analysis of the PCF by HPLC–DAD–MS and FAB mass spectrometryllowed the detection of the main components such as the complex of procyanidin oligomers consisting mainly of tetramers and trimers.onclusions: Guazuma ulmifolia bark possesses long-lasting antihypertensive and vasorelaxing properties linked to the endothelium related

actors, where nitric oxide is involved.2008 Elsevier Ireland Ltd. All rights reserved.

Rat a

th

eywords: Guazuma ulmifolia; Sterculiaceae; Antihypertensive; Vasorelaxant;

. Introduction

Hypertension is a major risk factor for stroke, myocardialnfarction, heart and kidney failure. Worldwide hypertension isstimated to cause 7.1 million premature deaths and 4.5 % of

Abbreviations: AcOEt, ethyl acetate; Cl3Fe, ferric chloride; DAD, dioderray detector; ESI, electrospray ionisation; FAB, fast atom bombardment;OAc, acetic acid; HPLC, high performance liquid chromatography; HR, heart

ate; HV, high voltage; LC, liquid chromatography; L-NAME, NG-nitro-l-rginine methylester; MAP, mean arterial pressure; Me2CO, acetone; MeOH,ethanol; MS, mass spectrum; NO, nitric oxide; PCF, procyanidin fraction;DA, photodiode array detection; RDA, retro Diels–Alder; SAP, systolic arterialressure; TLC, thin layer cromatography.∗ Corresponding author. Tel.: +52 55 5623 2165; fax: +52 55 5616 1489.

E-mail address: gamagos@servidor.unam.mx (G.A. Magos).

imcs(mlphpfiG“

378-8741/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.jep.2008.01.015

orta; Procyanidins

he disease burden (World Health Organization, 2002). Treatingypertension has been associated with about a 40% reductionn the risk of stroke and about a 15% reduction in the risk of

yocardial infarction (Collins et al., 1990). In consequence,urrent clinical practice guidelines identify lowering blood pres-ure as a priority in the treatment of people with hypertensionLonn, 2004). However, the hypertension remains inadequatelyanaged everywhere (Mancia et al., 2002), and in spite of the

arge number of antihypertensive drugs and combinations, mosteople in developing countries have poor access to modernealth care. Therefore, pharmacological validation of medicinal

lants or ethnomedical treatment methods could greatly bene-t populations with poor economic resources. The bark fromuazuma ulmifolia Lam (Sterculiaceae) commonly known as

guacimo” or “mutamba”, is used as antihypertensive in the

thnop

PibHnifa

Al1(1tstaa1fmlava

2

2

Tcrd(rfisw

2

wa5wonwoe(aa

sUtwtwn1Tcwdd

Zma9eiH

51e7

2

(0io1ccaarGWsisgwst(ebc

G.A. Magos et al. / Journal of E

anama folk medicine (Caballero-George et al., 2001), whilen Peru the entire plant is used by traditional healers for dia-etes and hypertension treatments (Agapito and Sung, 2003).owever, its popular preparation as antihypertensive agent isot described in detail; it is only mentioned that the entire plantnfusion is orally ingested. In contrast, the doses recommendedor respiratory and gastrointestinal affections (Caceres, 1999)re fully detailed (2–3 g three times per day in infusion).

Guazuma ulmifolia is a native tree widely used in Latinmerica for the treatment of a variety of diseases, particu-

arly for gastrointestinal disorder treatments (Heinrich et al.,992; Caceres et al., 1993; Hor et al., 1995); stomach achesComerford, 1996); diabetes Mellitus (Alarcon-Aguilara et al.,998); malaria and syphilis (Morton, 1977) and as uterine con-raction stimulants (Garcıa-Barriga, 1975). The phytochemicaltudies carried out with Guazuma ulmifolia bark have led tohe isolation of procyanidins consisting mainly of epicatechinnd catechin units (Hor et al., 1996). These compounds havelso been related to antihypertensive activities (Cheng et al.,993). However, this plant has neither been specifically studiedor its cardiovascular effects nor for its antihypertensive actionechanism. Therefore, the aim of the present study was to ana-

yze, the effects of a procyanidin fraction (PCF) obtained fromcetone extract of Guazuma ulmifolia bark on the rat’s cardio-ascular system in order to determine its contributions as anntihypertensive agent.

. Materials and methods

.1. Plant material and extraction

The bark of Guazuma ulmifolia was collected at Amatlan,epoztlan in the Mexican state of Morelos, September 2002. Theorresponding voucher specimen of the plant (09B561-2830)emains at the medicinal herbarium at the Instituto Mexicanoel Seguro Social (IMSS) for future reference. The dry bark1 kg) was previously defatted with hexane and macerated atoom temperature with Me2CO (5× 1000 mL, 24 h each). Theltered extracts were combined and evaporated to reduced pres-ure obtaining a deep red solid amorphous extract. Extract yieldas 70.2 g.

.2. Fractionation and isolation

Crude extract (70 g) was partitioned with AcOEt. Two partsere obtained from this process: the AcOEt soluble part (14.3 g),

nd the insoluble part (55.4 g). This latter part was put within00 mL MeOH–Me2CO 1:1 (v/v). From this mixture, a partas settled down. The rest was subjected to evaporation. 33 gf a dark red glassy residue was obtained. This material wasamed PCF and it gave a positive procyanidin green colourith Cl3Fe and red colour when it was examined by TLCn silica gel (Merck) with vanillin–H2SO4 reagent. TLC was

luted using Me2CO–toluene–AcOEt–MeOH–HOAc 5:3:3:1:1v/v/v/v/v). The PCF was subjected to solid phase extraction onBond Elute C18 cartridge (1 cm × 3 cm, Varian) to FAB+-MSnd HPLC–DAD–ESI/MS analysis. The spectrometric analy-

bs1o

harmacology 117 (2008) 58–68 59

is was performed using a Waters HPLC system (Milford, MA,SA) consisting of a 1525 binary pump and a 2996 PDA detec-

or. All experiments were carried out in positive ion modeith a mass spectrometer Esquire 6000 Ion trap (Bruker Dal-

onics, Billerica, MA, USA). FAB-MS (positive ion mode)as obtained by a JEOL JMS-SX102A instrument using 3-itrobenzylic alcohol as a matrix. 20 mg PCF was dissolved intomL of H2O–MeOH 80:20 and injected into the C18 cartridge.hen 2 mL of H2O–MeOH 80:20 (v/v) was also applied into theartridge for rinsing. The achieved retained sample was elutedith a mixture of 2 mL H2O–MeOH 25:75 (v/v). This mixtureisplaced the PCF and it shows an intense narrow ring goingown, which was monitored visually.

Analysis of PCF by analytical HPLC was performed using aorbax 300 Extend C18 column (150 mm × 4.6 mm i.d.; 5 �m)aintained at 25 ◦C. Solvent system: A 0.1% formic acid, B

cetonitrile. A gradient was performed during 0–45 min from0 to 0% A. After each run, the chromatographic system wasquilibrated with 10% B for 30 min. Flow rate of 1 mL/min;njection volume: 20 �L; sample concentration: 10 mg/mL in

2O–MeOH 25:75 (v/v). DAD conditions: 280 nm.ESI-MS conditions: positive ion mode; scan range:

0–1300 m/z; Nebulizer gas (N2) 30 psi; dry gas (N2)0 mL/min; dry temperature 365 ◦C; HV capillary 4000 V; HVnd plate offset −500 V; capillary exit 134.4 V and trap drive0.8.

.3. Anesthetized rats’ preparation

Experimental protocols were supported by CONACyTMO292 and Q42096) and followed the national (NOM-62-ZOO-1999 y NOM-087-SEMARNAT-SSA1-2002) andnternational official guidelines (Guide for the care and Usef Laboratory Animals. INH Publication 85-23, revised in985). The animals were anesthetized with a mixture ofhloralose–urethane (50:800 mg/kg) i.p. administered. Threeannulas were inserted into a femoral artery, a femoral vein,nd trachea in order to get blood pressure measurements,dministrate drug and allow spontaneous breathing. Mean arte-ial pressure (MAP) was recorded with a pressure transducerould–Statham P231D (Grass Instrument Division, Astro Med,est Warwick, R1) connected to the arterial cannula. The

ignal from the transducer was electronically damped and reg-stered by a Model 79 Grass polygraph. Heart rate (HR) wasimultaneously recorded through another channel from the poly-raph with a Grass 7P4 tachograph triggered by the pulseaves from the unfiltered transducer signal. After a 15 min

tabilization period, only one administration, the whole ace-one extract (50 mg/kg), PCF (10 mg/kg) or the saline solution0.2 mL/100 g) was injected i.v. as a bolus in the first series ofxperiments. An additional positive control group received car-achol (1 �g/kg-min during 2 h). In view of the short-lastingardiovascular effects of carbachol, this drug was administered

y infusion in order to elicit more prolonged responses. In theseeries MAP and HR recordings were immediately taken during20 min and their values were registered every10 min. In a sec-nd series, the rats were pretreated with L-NAME administered

6 thnop

iwdtghsa1f

2

wIrw2pcrratorathPab2matfstwbcmSeettr

2

2vi

mOtwmssKaOpmsdirTa

tArew((ataoTr

2

ccLB

twapvw

2

da

0 G.A. Magos et al. / Journal of E

.v. and after 15 min, PCF, saline solution or a carbachol infusionas injected. MAP and HR recordings were immediately takenuring 120 min. The 10 mg/kg PCF dose was chosen accordingo the dose–response curve (not shown) 0.5 logarithmic pro-ression data. This dose is close to the maximum immediateypotensive effect. Some animals died with 31 mg i.v. doses. Toum up, every group was integrated by six rats and their MAPnd HR were recorded 13 times over 120 m (before and every0 min). We obtained six values for each of those 13 recordingsrom which we present an average.

.4. Sugar-fed hypertensive rats preparation

Studies were carried out in male Wister 25-day-old ratseighing 90 ± 6 g. They were separated into two groups: group

, control rats were given tap water for drinking; and group II,ats were given 30% of commercially refined sugar in drinkingater during 16–20 weeks. All the animals were maintained at1–23 ◦C, subjected to a 12-h light:12-h dark cycle, fed withellets food (5001 Rodent Laboratory Chow) ad libitum andonducted according to the Guide for the care and Use of Labo-atory Animals (National Institutes of Health Publication 85-23,evised 1985). The sugar treatment ended right after group IInimals’ mean blood pressure was significantly increased upo 70%. This model is characterized by hyperinsulinemia, lossf tissue sensitivity toward insulin, hypertriglyceridemia, arte-ial hypertension and an oxidative stress increase (Banos etl., 1997). Systolic arterial pressure (SAP) measurements wereaken monthly; the rats were maintained at 32 ◦C in a LE 5650/6eater and scanner heating unit (Letica, Scientific instruments).ulse transducer and pressure cuff (LE 5160/R) were placedround the rat tail and connected to an LE 5007 automaticlood pressure computer. SAP was registered at 0, 0.5, 4, 8,4, 48, 72, 96, 120, 168 and 216 h using the standard tail cuffethod three times for each. A single PCF dose (10 mg/kg) was

dministered orally immediately after 0 h. An average for everyrio was calculated for each time. These averages were usedor graphics. According to a neuropharmacologic profile (nothown in this paper), 31 and 100 mg/kg of PCF oral administra-ions, produced noticeable motors and sensitive effects in ratshich may interfere with the blood pressure registered whereasehaviour changes are not evident with 10 mg; therefore wehose 10 mg/kg. Two additional groups of hypertensive and nor-otensive rats were treated parallely with p.o. saline solution.ix rats integrated each experimental group. To sum up, sixxperiments were carried out in 4 groups of six awaken ratsach (total = 24 experiments). For each experiment, SAP was 11imes registered from 0 to 216 h as detailed above. An average ofhree values was obtained in each of those 11 recordings for eachat.

.5. Rat aortic rings preparation

Hypertensive and normotensive rats were anesthetized with0 mg i.p. sodium pentobarbital and later on sacrificed by cer-ical dislocation. The upper thoracic aorta was removed and cutnto 0.5 cm long segments. These pieces were hung into a 3-

r

8a

harmacology 117 (2008) 58–68

L organ chamber between two nickel/chromium wire hooks.ne of the hooks was fastened to the bottom of the baths and

he other was attached to a Grass FT03 force transducer, whichas connected to a Grass model 79 polygraph (Grass Instru-ent Division). The baths contained modified Krebs Henseleit

olution (pH 7.4) in all experiments with the following compo-ition mM: NaCl, 121.4; KCl, 4.7; CaCl2, 2.5; MgSO4, 1.3;H2PO4, 1.2; NaHCO3, 25; glucose, 11; EDTA, 0.02 mM;

scorbic acid, 0.6 mM. The solution was bubbled with 95%2 + 5% CO2 and the temperature was maintained at 37 ◦C. Thereparations were subjected to 2 g resting tension, which wasaintained constantly throughout the experiments. A series of 5

timulations with norepinephrine (1 × 10−7 M) was carried outuring the 60 min stabilization period. The vascular endotheliumntegrity was assessed by verifying the contracted rings 70%elaxation while being treated with acetylcholine (1 × 10−6 M).he endothelium absence was confirmed when the relaxation bycetylcholine was less than 10%.

The hypertensive and normotensive rats aorta ring con-ractions were induced with norepinephrine (1 × 10−7 M).fter contractions reached a plateau, the PCF cumulative

esponse curves were obtained using rings with and withoutndothelium. Aortic rings with endothelium were incubatedith indomethacin (10 �M) or atropine (10 �M) or L-NAME

30 �M) for 20 min prior to the norepinephrine administration10−7 M) in experiments carried out using inhibitor compounds,nd then, PCF concentrations were added cumulatively. An addi-ional positive control group received carbachol (0.01, 0.1, 1nd 10 �M). The responses were expressed as relaxed percentsn the norepinephrine (1 × 10−7 M) contraction in all cases.he experimental groups were integrated with a total of eight

ings.

.6. Chemicals and solutions

The chloride of acetylcholine, ±norepinephrine, carbacholhloride, atropine sulphate, indomethacin, L-NAME, as well ashloralose and urethane were obtained from Sigma–Aldrich (St.ouis, MO). MeOH, Me2CO, AcOEt were purchased from J. T.aker. All other chemicals were reagent grade.

PCF and L-NAME were dissolved in 0.9% NaCl solu-ion for in vivo experiments. The volume of injectionas 2 mL/kg. Acetylcholine (1 mM), indomethacin (10 mM),

tropine (10 mM), L-NAME (30 mM) stock solutions were pre-ared daily within modified Krebs Henseleit solution to work initro. Indomethacin was mixed with sodium bicarbonate (0.5%/v). The solutions of PCF were made with Krebs solution.

.7. Statistical analysis

The experimental data are expressed as mean ± standardeviation (S.D.). Blood pressure (SAP and MAP) and HRre depicted in mmHg and beats/min. Aorta ring relaxation is

eported as norepinephrine contraction inhibition percentage.

SAP and HR values were registered at the 0, 0.5, 1, 2, 4, 6,, 24, 48, 72, 96, 120, 168 and 216 h over both normotensivend hypertensive awaken rats. MAP and HR of anesthetized

G.A. Magos et al. / Journal of Ethnopharmacology 117 (2008) 58–68 61

F nsivew d tetr

rowtv

T

ig. 1. HPLC profile recorded at 280 nm. Guazuma ulmifolia bark antihyperteere assigned to monomeric (m/z 291) dimeric (m/z 579), trimeric (m/z 867) an

ats were registered during 120 min every 10 min. The values

f each experimental group, i.e. anesthetized and awakened ratsere compared with their correspondent controls (saline solu-

ion groups) through unpaired t tests or one-way analysis ofariance followed by Dunnett’s test for multiple comparisons.

dgaa

fraction fragments detected by positive ion mode ESI-MS. Major m/z signalsamerics (m/z 1155) procyanidins.

hese tests were also applied on aortic rings relaxation pro-

uced by each PCF concentration with the pretreated and controlroups. A probability level of less than 0.05 in all cases wasccepted as significant. The aortic ring results were addition-lly evaluated by subjecting individual concentration response

62 G.A. Magos et al. / Journal of Ethnopharmacology 117 (2008) 58–68

F ractiom 1155

ctB

3

3

Hta1McatCmaDopm8FmMAp8T(d

3

dFpBaoci

sTNtaHjom(tjvloa9Mc

ig. 2. FAB+-MS obtained from the Guazuma ulmifolia bark antihypertensive f/z 307), and dimeric (3; m/z 579), trimeric (4; m/z 867) and tetrameric (5; m/z

urves to nonlinear regression analyses in order to calculatehe EC50. Regression analyses were carried out with a Primeriostatistics 4.02 (McGraw Hill Companies, Inc.).

. Results

.1. Chemical analysis

Fig. 1 shows the PCF recorded at 280 nmPLC–DAD–ESI/MS profile. It is important to point out

hat the better chromatographic profile resolution was obtainedccording to Shaolian and Hamburger (Zhou and Hamburger,996) Further PCF structural data were obtained from theS ion chromatogram through a LC-ion trap MS. The major

omponents were represented by the tetramerics (peaks 1–8)nd trimeric procyanidins (peak 9) identified when comparingheir MS with data reported in the literature (Hor et al., 1996;aballero-George et al., 2001). The MS (molecular ion at/z 1155) and fragmentation patterns were substantiated astetrameric identity. The fragment resulting from the retroiels–Alder (RDA) procyanidins main characteristic wasbserved at m/z 715 (Gu et al., 2003). Other fragmentationath by heterocyclic ring fission produced the fragments at/z 741 and m/z 453. Peaks at m/z 291, m/z 579 and m/z67 associated with several retention HPLC times (shown inig. 1) may be assigned to monomeric, dimeric and trimericolecular fragments. The PCF [M + H]+ ions at ion productS/MS spectra (not shown) support this fragmentation.FAB+ mass spectrum was obtained in order to identify

rocyanidins oligomers (Fig. 2). Thus peaks at m/z 289, 579,

67 and 1155 reveal the presence of such oligomeric structures.he monomeric epicatechin (m/z 289) and epigallocatechin

m/z 307) structure presence may be assessed from the FAB+

ata.

pats

n. Peak identification assigned to epicatechin (1; m/z 289), epigallocatechin (2;) procyanidins.

.2. Anesthetized rats assay

PCF obtained from acetonic extract keeps up the total car-iovascular activity in anesthetized rats. This fact is shown inig. 3 when the cardiovascular effects of raw acetone extract (leftanel) are compared to the effects of their PCF (right panel).oth products diminished MAP during the whole experimentnd produced HR a transitory decrease for five minutes. On thether hand, purified PCF in solid phase also keeps the cardiovas-ular effects up (data not shown). In contrast, the solids solublen AcOEt failed to exert rat cardiovascular activity.

The initial values of MAP and HR before administeringaline, PCF or carbachol to anesthetized rats are shown inable 1. This table shows the increase of the anesthetized rats L-AME pretreatment MAP values (p > 0.05, unpaired Student’s

-test). PCF from Guazuma ulmifolia (10 mg/kg i.v.) producedsignificant MAP long-lasting decrease as well as a transitoryR decrease in these animals, compared with the PCF prein-

ection HR values. However, when comparing the PCF effectsver HR values with HR saline values, PCF produced an HReaningful decrease which recovers the baseline after 60 min

Fig. 4 left panel without L-NAME; p < 0.05, unpaired Student’s-test, Dunnett’s). This response pattern was changed after prein-ection 31 mg/kg i.v. L-NAME since PCF effects over MAPalues were transitory as MAP’s decrease recovered the base-ine after 40 min. However, when comparing the PCF effectsver MAP values with MAP saline values, the PCF producedMAP meaningful decrease which recovers the baseline after

0 min. This statistic difference is due to L-NAME effect onAP values which increase in the saline group individuals. Pro-

yanidins transitorily decrease HR in the first 10 min (Fig. 4 right

anels; p < 0.05, unpaired Student’s t-test, Dunnett’s) in thesenimals. While the administration of L-NAME increased MAPo 60 mmHg as expected in control animals treated with salineolution. This effect lasted 90 min without HR changes (Fig. 4

G.A. Magos et al. / Journal of Ethnopharmacology 117 (2008) 58–68 63

Fig. 3. Whole acetone extract obtained from Guazuma ulmifolia bark (�, left panel) and their procyanidins fraction effects (PCF �, right panel) over mean arterialpressure (MAP-upper panels) and heart rate (HR-lower panels) in anesthetized normotensive rats. Both compared to their respective saline solution group (�). Thes aftero /minc nett’s

rswcd(

3

o

TIg

G

WWWWWW

V

hsai2tHu

ymbols represent the mean ± S.D. of six animals. Abscissae correspond to timer saline solution (0.1 mL/100 g); ordinates show MAP or HR in mmHg or beatsompared with saline group values. (p < 0.05) by unpaired Student’s t-test, Dun

ight panels). Carbachol infusion induced a significant hypoten-ion associated with decrement in HR. This hypotensive effectas attenuated by pretreatment of the animals with L-NAME. In

ontrast, this no selective Nitric Oxide (NO) synthase inhibitor,id not significantly alter the bradycardia elicited by carbacholFig. 4 right panels).

.3. Sugar-fed hypertensive rats assay

Initial SAP and HR values in the control and treated groupsf conscious rats are shown on Table 2. As expected, the animals

able 1nitial values of mean arterial pressure (MAP) and heart rate (HR) in the diverseroups of anesthetized rats

roup MAP HR

ithout L-NAME + saline solution 110 ± 14 408 ± 32ithout L-NAME + PCF 100 ± 15 388 ± 18ithout L-NAME + carbachol 110 ± 8 380 ± 20ith L-NAME + saline solution 149 ± 9* 414 ± 27ith L-NAME + PCF 138 ± 10* 420 ± 25ith L-NAME + carbachol 140 ± 7* 370 ± 12

alues are means ± S.D. of 6 experiments.* p < 0.05 by unpaired Student’s t-test.

dHutsSweh

3

pcn

i.v. administration of whole acetone extract (50 mg/kg) or their PCF (10 mg/kg)changes. Asterisks indicate significant differences among MAP and HR valuestest).

aving drinking water with refined sugar during 16–20 weeks,howed a high SAP (Reaven and Ho, 1991). 10 mg/kg PCF oraldministration in conscious hypertensive rats produced a signif-cant decrease in SAP and HR. Maximum effects were observed4 h after PCF administration. A progressive and slow recoveryo the initial values of SAP at 50 h was observed, whereas theR did not go back to basal values (Fig. 5 left panel; p < 0.05,npaired Student’s t tests, Dunnett’s). The same PCF dose pro-uced two-phase effect on SAP without significant changes inR in awaken normotensive rats (Fig. 5 right panel; p < 0.05,npaired Student’s t-test, Dunnett’s). A blood pressure transi-ory increment which is not significant comparing the salineolution group occurred during the first phase. On balance, theAP values decreased significantly which were fully recoveredithin the first 16 h after treatment during the second phase. The

ffects in these animals lasted shorter than in those observed inypertensive rats.

.4. Rat aortic rings assay

Stimulation of aortic rings with norepinephrine (1 × 10−7 M)roduced a sustained contraction (2.5 ± 0.2 g, n = 64). Theumulative PCF additions from Guazuma. ulmifolia caused pro-ounced concentration-dependent relaxant response in rings

64 G.A. Magos et al. / Journal of Ethnopharmacology 117 (2008) 58–68

Fig. 4. Shows Guazuma ulmifolia procyanidins fraction (PCF �), saline solution (�) and carbachol (�) effects over mean arterial pressure (MAP-upper panel) andheart rate (HR-lower panel) in anesthetized normotensive rats 15 min previously pretreated with either 0.1 mL/100 g of saline (left graphs) or L-NAME 31 mg/kg( cissaem mmHa nt’s t

w(aocc1hbol1a

TIg

G

WWWW

V

hlAwp

4

right graphs). In all cases symbols represent the six animals mean ± S.D. Absin of carbachol infusion i.v. administration. Ordinates show to MAP or HR in

nd HR values compared with saline group values. (p < 0.05) by unpaired Stude

ith endothelium from normotensive and hypertensive ratsFig. 6; p < 0.05, unpaired Student’s t-test). This vasorelax-nt activity was remarkably lower with intermediate dosesf procyanidins in hypertensive rat aortic rings. The con-entrations that produced 50 % inhibition of the maximumontractile response (IC50 ng/mL) were 35.3 ± 12.4 ng/mL and01.3 ± 57.2 ng/mL in isolated aortas of normotensive andypertensive rats. This relaxant activity was partially inhibitedy vascular endothelium removal (Fig. 7; right graph, p < 0.05)

r 30 �M L-NAME pretreatment, NO synthase inhibitor (Fig. 8;eft graph, p < 0.05, Dunnett’s). However, pretreatment with0 �M indomethacin (cyclooxygenase inhibitor) or with 10 �Mtropine (muscarinic antagonist) had no effect. On the other

able 2nitial value of systolic arterial pressure (SAP) and heart rate (HR) in the diverseroups of conscious rats

roup SAP HR

ater + saline solution (normotensive) 106 ± 8 335 ± 18ater + sugar + saline solution (hypertensive) 139 ± 7* 362 ± 11ater + PCF of Guazuma ulmifolia (normetensive) 102 ± 11 353 ± 26ater + sugar + PCF of G.ulmifolia (hypertensive) 138 ± 7* 348 ± 32

alues are means ± S.D. of 6 experiments.* p < 0.05 by unpaired Student’s t-test.

nclsdam

hdtdHtm

correspond to 120 min after 10 mg/kg PCF or 0.1 mL/100 g saline or 1 �g/kgg or beats/min changes. Asterisks indicate significant differences among MAP-test, Dunnett’s test.

and, the contribution of the endothelium to the PCF vasore-axant effect was compared with carbachol (positive control).s expected, carbachol produced a significant relaxant effecthich was completely blocked by L-NAME (Fig. 8; right graph,< 0.05).

. Discussion

High blood pressure is a major risk factor for stroke, coro-ary heart disease and renal vascular disease. The current andommon method for controlling hypertension is the use of aong-term drug therapy. It is well known that drugs have manyide effects which may complicate the patient’s medical con-ition. This is why physicians as well as medical professionalsnd even patients prefer and tend to use antique traditional herbaledicine.The Guazuma ulmifolia has been used in Panama to treat

ypertension empirically. Unfortunately, there is not enough evi-ence to prove Guazuma ulmifolia’s utility. This paper showshat a fraction obtained from Guazuma ulmifolia bark pro-

uces significant antihypertensive effects. The FAB+-MS andPLC–DAD–ESI/MS spectrometric analysis from this frac-

ion allowed us to identify tetrameric and trimeric procyanidinsainly (PCF). These compounds produced a biological and a

G.A. Magos et al. / Journal of Ethnopharmacology 117 (2008) 58–68 65

Fig. 5. Shows Guazuma ulmifolia procyanidins fraction (PCF �) and saline solution (�) effects over systolic arterial pressure (SAP-upper panels) and heart rate (HR-lower panels) in awake hypertensive (left panel) and normotensive rats (right panel). In both cases, symbols represent six animals mean ± S.D. Abscissae correspondto 220 h after 10 mg/kg PCF or 0.1 mL/100 g saline solution p.o. administration. Ordinates represent SAP or HR changes in mmHg or beats/min, respectively.Asterisks indicate significant differences among SAP and HR values compared with saline group values. (p < 0.05) by unpaired Student’s t-test, Dunnett’s test.

Fig. 6. Show Guazuma ulmifolia procyanidins fraction (PCF) in aortic ringsobtained from normotensive (�) and hypertensive (�) rats effects. Relaxationis expressed as a percentage of the norepinephrine (1 × 10−7 M) generated con-traction. Each point represents the mean ± S.D. of n = 8 observations. Abscissadenotes PCF log concentration. Asterisks indicate significant differences amongnormotensive and hypertensive aortic rings rats. p < 0.05 by unpaired Student’st-test.

Fig. 7. Show Guazuma ulmifolia procyanidins fraction (PCF) effects in nor-motensive rat aortic rings with endothelium (�) and without it (�) effects. Eachpoint represents the n = 8 observations mean ± S.D. Abscissae denote PCF logconcentration. Asterisks indicate significant differences between aortic ringsvalues with endothelium and without it. p < 0.05 unpaired Student’s t-test.

66 G.A. Magos et al. / Journal of Ethnopharmacology 117 (2008) 58–68

Fig. 8. Shows Guazuma ulmifolia procyanidins fraction (PCF) effects (left graph) or of carbacohol (right graph) when aorta rings of normotensive rats are previouslyp ME (©g s. Abd

pper

flrstpipasaal(io(trd

fdrataaa(ibrl

ed

dhhua2dcwr

hm2iarvartretn

veeThe

retreated for 15 min with Krebs (�), indomethacin (�), atropine (�) and L-NAenerated contraction. Each point represents the mean ± S.D. of n = 8 observationifferences among every treatment p < 0.05 by Dunnett’s test.

harmacological activity broad spectrum including both antihy-ertensive and vasorelaxant effects (Cheng et al., 1993; Cortest al., 2002) which make us consider them remarkably as beingesponsible for the cardiovascular activity.

Our results indicate clearly that one PCF oral dose (10 mg/kg)rom Guazuma ulmifolia bark produce an intense and long-asting antihypertensive effect (50 h) in sugar-fed-hypertensiveats. It also produces a long-lasting bradycardia. We con-ider that both antihypertensive and bradycardia effects arehe major findings in this research work. Unfortunately, asioneers, there might not be a common mechanistic and sat-sfactory explanation about the long-lasting effect. However,harmacokinetic factors might be involved in this effect. Inddition, recent studies in vivo (Shoji et al., 2006) havehown that monomers and smaller oligomeric procyanidinsre absorbed when an oral administration takes place. Itgrees with in vitro studies which indicate that low molecu-ar weight procyanidins such as tetramers, trimers, dimers and+)-catechin-epicatechin monomers are well absorbed by smallntestine cells. However, the procyanidins polymerization (sevenr more monomeric units) makes intestinal absorption difficultScalbert and Williamson, 2000). According to this evidence,he Guazuma ulmifolia bark procyanidins are oligomers whichemain stable in gastrointestinal juice which are absorbed pro-ucing a SAP decrease.

It is very important to point out that in the same sugared hypertensive rat, a PCF dose produces a remarkable HRecrease, when it is administrated. According to the fructose-fedat model, the hyperinsulinemia and insulin resistance are bothssociated with left ventricular hypertrophy. Within this hyper-rophy higher �1 cardiac adrenergic receptor densities as wells an epinephrine and norepinephrine urinary excretion increasere observed (Kamide et al., 2002). It is known that the �1drenergic cardiac receptor stimulation increases the rat’s HRWilliamson et al., 1994). It may be blocked with PCF admin-

stration. This might explain, at least partially, the long-lastingradycardia effect observed in PCF administrated hypertensiveats. Meanwhile, in normotensive rats there is no left ventricu-ar hypertrophy or �1 adrenergic receptors increase. That might

rtLs

). Relaxation is expressed as a percentage of the norepinephrine (1 × 10−7 M)scissae denote PCF or carbachol log concentration. Asterisks indicate significant

xplain why a PCF (10 mg/kg) oral administration does notecrease the HR in these normotensive rats.

On the other hand, there is a remarkable interest on procyani-ins found in certain kinds of food because of their potentialealth benefits; particularly those reducing risk factors such asypertension, hyperlipidemia and hyperinsulinemia, which endp with mortal cardiovascular diseases. It is believed that theirntioxidant activity is beneficial (Cos et al., 2004; Erlejman et al.,004; Banos et al., 2005; Puiggros et al., 2005). There is an evi-ence showing that different polyphenolic compounds, mainlyatechin oligomers prevent hypertension, cardiac hypertrophy asell as reactive oxygen species overproduction in fructose-fed

ats (Al-Awwadi et al., 2005).There are several mechanisms which have explained the

ypertension development in sugar fed rats, such as insulin-ediated changes in vascular smooth muscle tone (Tay et al.,

002). This effect is known as endothelial dysfunction and its observed when an insulin resistance syndrome occurs. Inddition, there are various evidences showing that the insulinesistance syndrome with endothelial dysfunction can be pre-ented (Nyby et al., 2005), attenuated (Hsieh, 2004) as wells reverted (Miatello et al., 2003) in hypertensive fructose-fedats. Those evidences agree with our results since in spite ofhe endothelial dysfunctional state in hypertensive sugar-fedats, PCF produced conspicuous hypotensive and vasorelaxantffects. In contrast, in the absence of the endothelial dysfunc-ional, the PCF (10 mg/kg) hypotensive effect is not observed inormotensive rats (controls).

On the other hand, in another very different model, an intra-enous PCF administration produced a remarkable hypotensiveffect in anesthetized rats. This indicates that the hypotensiveffect is not blocked by the chloralose-urethane anesthesia.his effect was not compared directly with the PCF anti-ypertensive effect obtained from awakened rats because thexperimental conditions between anesthetized and awakened

ats are very different. However, the PAM recording over anes-hetized rats allowed us to learn that the NO synthase inhibitor-NAME pretreatment reduced the PCF-induced hypotensionignificantly without modifying the bradycardia. This result

thnop

lh

tlatoai

dacdcplbiivcrbipmc

reratoeocfopmvdlr

adhAcaaea

pevAui2trP

pHaptbwcao

A

AJtCQ

R

A

A

A

B

B

C

G.A. Magos et al. / Journal of E

eads us to believe that NO is involved with the PCF producingypotension.

Nowadays there are several lines of evidence which showhat endothelial NO synthase decreased in both aortic endothe-ium and cardiac tissue in the fructose-fed rat model (Miatello etl., 2001). This is deeply linked to an excessive vascular oxida-ive stress which impairs the NO action. However, the resultsbtained from our experiments using anesthetized rats cannot bepplied to the hypertensive fructose-fed rat model where theres an endothelial dysfunction.

Previous in vitro studies have shown that procyanidins pro-uce an endothelium-dependent relaxation in rat thoracic aortand it is mediated by releasing NO from vascular endothelialells (Cortes et al., 2002). The in vitro vasorelaxant effects pro-uced by PCF were similar to those described by Cortes; thus, itan be postulated that the hypotensive effect observed in theresent study is mediated at least in part, by an endothelial-arginine/NO pathway. This hypothesis is strongly supportedy the following results. First, the intravenous NO synthasenhibitor L-NAME pretreatment reduced significantly the PCF-nduced hypotension of anesthetized rats. Secondly, the PCFasorelaxant effect produced on the norepinephrine-inducedontraction is attenuated with either vascular endotheliumemoval or L-NAME pretreatment. According to all what haseen said above, our results suggest that PCF hypotensive effects due to an active vasodilatation, which is mediated, at leastartially, by NO release from the vascular endothelium. Theechanism by which PCF stimulates NO release is not yet

onclusive.Some evidence indicates that there is an abnormal vascular

eactivity in both sugar-fed and fructose-feed rat models (Banost al., 1997; Kashiwagi et al., 1999). The endothelial-dependentelaxation is impaired in the fructose-feed model while thecetylcholine vascular response decreases. This has been relatedo a relative excess of superoxide anion through NADH/NADPHxidase activation which surpasses endothelial cells NO gen-ration. A constitutive NO synthase activity decrease is alsobserved (Kashiwagi et al., 1999). However, an impaired vas-ular reactivity has not always been found in vessels fromructose-feed rats (Navarro-Cid et al., 1999). In our experiments,ver the hypertensive rat aortic ring, a PCF intermediate doseroduced a lower effect compared to the one observed in nor-otensive rats. This fact indicates that endothelium-dependent

ascular reactivity is observed as a hypertensive sugar-feed ratecrease. However, with a higher PCF dose there is no vascu-ar reactivity difference between hypertensive and normotensiveats.

The current research work does not attempt to assessnother hypotensive mechanism different from the endothelium-ependent hypotensive mechanism. Instead, we only want toighlight that there might be a hypotensive effect related toT1 receptors. It is known that the renin–angiotensin systemontributes to a long-term blood pressure homeostasis as well

s a hypertension. Besides, there is an AT1 receptor as wells an angiotensina II increment in fructose-feed rats (Kamidet al., 2002). It has also been demonstrated that AT1 receptorntagonistic as losartan, reduces the elevated blood pressure and

C

harmacology 117 (2008) 58–68 67

revents left ventricular hypertrophy progression through directffects on renin–angiotensin system and on the sympathetic ner-ous system in the fructose-fed rat model (Miatello et al., 2003).

previous study showed that procyanidins from Guazumalmifolia bark in vitro produced (3H)-angiotensin II bindingnhibition on the human AT1 receptor (Caballero-George et al.,002). Unfortunately, there are no studies in vivo to demonstratehis phenomenon. However, procyanidins might inhibit the AT1enin–angiotensin system as well as cause hypotension whenCF from Guazuma ulmifolia bark is administrated.

We conclude that the phenolic compounds such as oligomericrocyanidins were detected in the PCF by FAB+-MS andPLC–DAD–ESI/MS analysis. The detailed chemical linkages

ssignment goes beyond this study’s scope. However, the antihy-ertensive and vasorelaxant activities caused by tetramerics andrimeric procyanidins explain why the use of Guazuma ulmifoliaark is so popular in Panama and Peru. However, further studiesith pure compounds will be required in order to confirm or elu-

idate the action mode to fully explain the intense long-lastingntihypertensive activity shown by a complex of procyanidinligomers obtained from Guazuma ulmifolia bark.

cknowledgements

The authors wish to express their sincere thanks to Abigailguilar Contreras for the taxonomic identification and to Mrs.

osefina Bolado for English revision of this text. We also wanto thank Consejo Nacional de Ciencia y Tecnologıa, CONA-yT, Mexico for financial support through projects MO292 and42096.

eferences

gapito, T., Sung, I., 2003. Fitomedicina: 1100 Plantas Medicinales. IsabelI.R.L., Lima-Peru, pp. 331–332.

larcon-Aguilara, F.J., Roman-Ramos, R., Perez-Gutierrez, S., Aguilar-Contreras, A., Contreras-Weber, C.C., Flores-Saenz, J.L., 1998. Study ofthe anti-hyperglycemic effect of plants used as antidiabetics. Journal ofEthnopharmacology 61, 101–110.

l-Awwadi, N.A., Araiz, C., Bornet, A., Delbosc, S., Cristol, J.P., Linck, N.,Azay, J., Teissedre, P.L., Cros, G., 2005. Extracts enriched in differentpolyphenolic families normalize increased cardiac NADPH oxidase expres-sion while having differential effects on insulin resistance, hypertension,and cardiac hypertrophy in high-fructose-fed rats. Journal of Agriculturaland Food Chemistry 53, 151–157.

anos, G., Carvajal, K., Cardoso, G., Zamora, J., Franco, M., 1997. Vascularreactivity and effect on plasma in a rat model of hypertriglyceridemia andhypertension. American Journal of Hypertension 10, 379–388.

anos, G., Medina-Campos, O.N., Maldonado, P.D., Zamora, J., Perez, I.,Pavon, N., Pedraza-Chaverri, J., 2005. Antioxidant enzymes in hypertensiveand hypertriglyceridemic rats: effect of gender. Clinical and ExperimentalHypertension 27, 45–57.

aballero-George, C., Vanderheyden, P.M.L., De Bruyne, T., Shahat, A.A., Vanden Heuvel, H., Solis, P.N., Gupta, M.P., Claeys, M., Pieters, L., Vauquelin,G., Vlietinck, A.J., 2002. In vitro inhibition of [3H]-angiotensin II bindingon the human AT1 receptor by proanthocyanidins from Guazuma ulmifolia

bark. Planta Medica 68, 1066–1071.

aballero-George, C., Vanderheyden, P.M.L., Solis, P.N., Pieters, L., Shahat,A.A., Gupta, M.P., Vauquelin, G., Vlietinck, A.J., 2001. Biological screeningof selected medicinal Panamanian plants by radioligand-binding techniques.Phytomedicine 8, 59–70.

6 thnop

C

C

C

C

C

C

C

E

G

G

H

H

H

H

K

K

L

M

M

M

M

N

N

N

N

P

R

S

S

T

W

WWorld Health Organization, Geneva.

8 G.A. Magos et al. / Journal of E

aceres, A., 1999. Plantas de Uso Medicinal en Guatemala. Universitaria,Guatemala, p. 126.

aceres, A., Fletes, L., Aguilar, L., Ramırez, O., Figueroa, L., Taracena, A.M.,1993. Samayoa B. Plants used in Guatemala for the treatment gastrointestinaldisorders. 3. Conformation of activity against enterobacteria of 16 plants.Journal of Ethnopharmacology 38, 31–38.

ollins, R., Peto, R., MacMahon, S., Hebert, P., Ffiebach, N.H., Eberlein, K.A.,1990. Blood pressure, stroke, and coronary heart disease Part 2: short-termreductions in blood pressure: overview of randomized drug trials in theirepidemiological context. Lancet 335, 827–838.

omerford, S.C., 1996. Medicinal plants of two Mayan healers from San Andres,Peten, Guatemala. Economic Botany 50, 327–336.

ortes, S.F., Valadares, Y.M., de Oliveira, A.B., Lemos, V.S., Barbosa, M.P.,Braga, F.C., 2002. Mechanism of endothelium-dependent vasodilatationinduced by a proanthocyanidins-rich fraction from Ouratea semiserrata.Planta Medica 68, 412–415.

os, P., De Bruyne, T., Hermans, N., Apers, S., Berghe, D.V., Vlietinck, A.J.,2004. Proanthocyanidins in health care: current and new trends. CurrentMedicinal Chemistry 11, 1345–1359.

heng, J.T., Hsu, F.L., Chen, H.F., 1993. Antihypertensive principles from theleaves of Melastoma candidum. Planta Medica 59, 405–407.

rlejman, A.G., Verstraeten, S.V., Fraga, C.G., Oteiza, P.I., 2004. The inter-action of flavonoids with membranes: potential determinant of flavonoidantioxidant effects. Free Radical Research 38, 1311–1320.

arcıa-Barriga, H., 1975. Flora Medicinal de Colombia, vol. 2. Instituto deCiencias Naturales, Universidad Nacional Colombia, Bogota, p. 538.

u, L., Kelm, M.A., Hammerstone, J.F., Zhang, Z., Beecher, G., Holden, J.,Haytowitz, D., Prior, R.L., 2003. Liquid chromatographic/electrospray ion-ization mass spectrometric studies of proanthocyanidins in foods. Journal ofMass Spectrometry 38, 1272–1280.

einrich, M., Rimpler, H., Barrera, N.A., 1992. Indigenous phytotherapy ofgastrointestinal disorders in a lowland mixed community (Oaxaca, Mexico):ethnopharmacologic evaluation. Journal of Ethnopharmacology 36, 63–80.

or, M., Heinrich, M., Rimpler, H., 1996. Proanthocyanidin polymers with anti-secretory activity and proanthocyanidin oligomers from Guazuma ulmifoliabark. Phytochemistry 42, 109–119.

or, M., Rimpler, H., Heinrich, M., 1995. Inhibition of intestinal chloridesecretion by proanthocyanidins from Guazuma ulmifolia. Planta Medica 61,208–212.

sieh, P.S., 2004. Attenuation of insulin-mediated pressor effect and nitric oxiderelease in rats with fructose-induced insulin resistance. American Journal ofHypertension 17, 707–711.

amide, K., Rakugi, H., Higaki, J., Okamura, A., Nagai, M., Moriguchi, K.,Ohishi, M., Satoh, N., Tuck, M.L., Ogihara, T., 2002. The rennin–angiotensinand adrenergic nervous system in cardiac hypertrophy in fructose-fed rats.American Journal of Hypertension 15, 66–71.

ashiwagi, A., Shinozaki, K., Nishio, Y., Okamura, T., Toda, N., Kikkawa, R.,1999. Free radical production endothelial cell as a pathogenetic factor forvascular dysfunction in the insulin resistance status. Diabetes Research and

Clinical Practice 45, 199–203.

onn, E., 2004. The clinical relevance of pharmacological blood pressure low-ering mechanisms. The Canadian Journal of Cardiology 20, 83SB–88SB.

ancia, G., Bombelli, M., Lanzarotti, A., Grassi, G., Cesana, G., Zanchetti, A.,Sega, R., 2002. Systolic vs. distolic blood pressure control in the hypertensive

Z

harmacology 117 (2008) 58–68

patients of the PAMELA population. Archives of Internal Medicine 162,582–586.

iatello, R., Risler, N., Castro, C., Cruzado, M., Gonzalez, S., Zumino, A.P.,2003. Chronic administration of losartan reverses cardiovascular changesin hypertensive fructose-fed rats. Cellular and Molecular Biology 49,945–952.

iatello, R., Risler, N., Castro, C., Gonzalez, S., Ruttler, M., Cruzado, M.,2001. Aortic smooth muscle cell proliferation and endothelial nitric oxidesynthase activity in fructose-fed rats. American Journal of Hypertension 14,1135–1141.

orton, J.F., 1977. Some folk-medicine plants of Central American markets.Quarterly Journal Crude Drug Research 15, 165–192.

avarro-Cid, J., Maeso, R., Perez-Vizcaino, R., Cachofeiro, V., Ruilope, L.M.,Tamargo, J., 1999. Effects of losartan on blood pressure, metabolic alter-ations, and vascular reactivity in the fructose-induced hypertensive rat.Hypertension 26, 1074–1078.

orma Oficial Mexicana NOM-062-ZOO-1999. Especificaciones tecnicaspara la produccion, cuidado y uso de los animales de laboratorio,Mexico.

orma Oficial Mexicana NOM-087-ECOL-SSA1-2002. Proteccion ambiental –Salud ambiental – Residuos peligrosos biologico-infecciosos – Clasificaciony especificaciones de manejo, Mexico.

yby, M.D., Matsumoto, K., Yamamoto, K., Abedi, K., Eslami, P., Hernandez,G., Smutko, V., Berger, M.E., Tuck, M.L., 2005. Dietary fish oil preventsvascular dysfunction and oxidative stress in hyperinsulinemic rats. AmericanJournal of Hypertension 18, 213–219.

uiggros, F., Llopiz, N., Ardevol, A., Blade, C., Arola, L., Salvado, M.J.,2005. Grape seed procyanidins prevent oxidative injury by modulating theexpression of antioxidant enzyme systems. Journal of Agricultural and FoodChemistry 53, 6080–6086.

eaven, G.M., Ho, H., 1991. Sugar-induced hypertension in Sprague–Dawleyrats. American Journal of Hypertension 4, 610–614.

calbert, A., Williamson, G., 2000. Dietary intake and bioavailability of polyphe-nols. The Journal of Nutrition 130, 2073S–2085S.

hoji, T., Masumoto, S., Moriichi, N., Akiyama, H., Kanda, T., Ohtake,Y., Goda, Y., 2006. Apple procyanidin oligomers absorption in ratsafter oral administration: analysis of procyanidins in plasma using theporter method and high-performance liquid chromatography/tandem massspectrometry. Journal of Agricultural and Food Chemistry 54, 884–926.

ay, A., Ozcelicay, A.T., Altan, V.M., 2002. Effects of l-arginine on blood pres-sure and metabolic changes in fructose hypertensive rats. American Journalof Hypertension 15, 72–77.

illiamson, A.P., Seifen, E., Lindemann, J.P., Kennedy, R.H., 1994. �1A-Adrenergic receptor mediated positive chronotropic effect in right atriaisolated from rats. Canadian Journal of Physiology and Pharmacology 72,1574–1579.

orld Health Organization, 2002. The World Health Report: Risks to Health.

hou, S., Hamburger, M., 1996. Application of liquid chromatography-atmospheric pressure ionization mass spectrometry in natural productanalysis evaluation and optimization of electrospray and heated nebulizerinterfaces. Journal of Chromatography 755, 189–204.