Immunohistochemical Localization ofEndothelin-1 in Human Placenta fromNormal and Growth-RestrictedPregnancies

MEHMET ERDEM,1* AHMET ERDEM,1 OZLEM ERDEM,2 GIZEM YILDIRIM,1

LEYLA MEMIS,2 AND OZDEMIR HIMMETOGLU1

1Department of Obstetrics and Gynecology, Faculty of Medicine, Gazi University, Bilkent 3 Konutlari Ufuk Sitesi,F3 Blok Daire 12, Bilkent, Ankara, Turkey 065302Department of Pathology, Faculty of Medicine, Gazi University, Ankara, Turkey

Received June 10, 2002; accepted April 1, 2003; published online June 13, 2003.

ABSTRACTThe purpose of this study was to investigate whether thelocalization of endothelin-1 (ET-1) immunoreactivitydiffers in human placental tissues from third-trimesternormal and intrauterine growth restricted (IUGR) preg-nancies. Immunohistochemistry for ET-1 was per-formed on human placentas from 30 IUGR and 30 un-complicated pregnancies matched for gestational age.The distribution and intensity of ET-1 immunoreactivitywas assessed by a semiquantitative scoring system.Doppler flow velocity waveform analysis of the umbilicalartery was performed in each patient before delivery.ET-1 was localized diffusely in placental specimens fromnormal and IUGR pregnancies. The localization of ET-1immunoreactivity was significantly higher in the capil-lary endothelial cells of villi as well as in the endothelial,decidual, and trophoblastic cells of the basal plate inplacentas from normal pregnancies than from IUGRpregnancies. There was no significant difference in pla-cental ET-1 immunoreactivity between IUGR pregnan-cies with normal and abnormal umbilical artery Dopplerflow velocity waveforms. Placental ET-1 immunoreactiv-ity was significantly higher in the decidual and tropho-blastic cells of the basal plate and the capillary endothe-lial cells of villi in normal pregnancies than in IUGRpregnancies with normal umbilical artery Doppler flowvelocity waveforms. However, only the decidual and tro-

phoblastic cells of the basal plate demonstrated signifi-cantly higher abundant localization of ET-1 immunore-activity in normal pregnancies than in IUGRpregnancies with abnormal umbilical artery Dopplerflow velocity waveforms. In conclusion, our findings sug-gest that the lower expression of ET-1 in placental tis-sues from IUGR pregnancies might be secondary to anadaptive mechanism to reduce the vasoconstrictor effectof ET-1.

Key words: endothelin-1, immunohistochemistry, intra-uterine growth restriction, placenta, umbilical arteryDoppler

INTRODUCTIONIntrauterine growth restriction is associated withan increase in fetal and neonatal morbidity andmortality rates [1]. It is a condition with a hetero-geneous pathophysiology and may be associatedwith structural changes in the uteroplacental vas-culature including narrowing of the spiral arteriesby intimal thickening and fibrinoid degeneration,and placental microvascular lesions characterizedby the obliteration of small muscular arteries inthe tertiary stem villi [2,3,4,5].

A number of vasoactive substances includingendothelins might contribute to abnormal states*Corresponding author, e-mail: erdemom@yahoo.com

Pediatric and Developmental Pathology 6, 307–313, 2003

DOI: 10.1007/s10024-002-0060-7

© 2003 Society for Pediatric Pathology

associated with elevated vascular resistance or va-sospasms. Endothelin-1 (ET-1), a 21 amino acidpeptide mainly produced by endothelial cells, is athree-endothelin peptide, and has potent and long-lasting vasoconstrictor activity [6]. ET-like immu-noreactivity and ET-binding sites have been iden-tified in placental vascular tissue [7,8,9,10].Elevated ET-1 levels have been observed in growth-restricted infants, particularly in those with abnor-mal Doppler flow velocity waveforms (FVW), sug-gesting a role of ET-1 in the development ofabnormal fetoplacental vascular resistance [11,12].However, neither a difference in receptor bindingcharacteristics nor an aberrant localization of ETreceptors has been observed in intrauterine growthrestricted (IUGR) placentas [11,13].

Although the relative contribution of ET ac-tivity in the elevation of fetoplacental vascular re-sistance is strongly indicated, few clinical studiesrelate the immunohistochemical localization ofET-1 in human placentas from normal and com-plicated pregnancies [14,15]. The purpose of thepresent study was to investigate the localization ofET-1 by immunohistochemistry in placental sec-tions from IUGR pregnancies and to compare theET-1 immunoreactivity observed in the humanplacentas of IUGR pregnancies with those of nor-mal pregnancies. Our other aims were to compareimmunohistochemical staining differences be-tween IUGR pregnancies with regard to umbilicalartery Doppler indices of increased fetoplacentalresistance.

METHODSPatientsThirty women with IUGR pregnancies were in-cluded in this study. Six pregnancies in the IUGRgroup were complicated with pregnancy-inducedhypertension. The estimated fetal weights based onserial sonographic evaluation and the birthweights were below the 10th percentile for gesta-tional age in all women in the study group. In allcases, gestational age was confirmed by an ultra-sound examination performed in the first trimes-ter. Pregnancies with fetal chromosomal abnor-malities or congenital malformations wereexcluded. Thirty pregnant women without compli-cations, who delivered appropriate for gestationalage infants and matched the study group for ges-

tational age at delivery, comprised the controlgroup. The study group was further divided intotwo subgroups according to umbilical artery Dopp-ler FVW: IUGR pregnancies with an abnormalFVW, and IUGR pregnancies with a normal FVW.

Doppler studiesPulsed Doppler ultrasound using Logic 500 (Gen-eral Electric, USA) Please provide city and state forsupplier General Electric. equipment was per-formed for Doppler blood flow analysis of the um-bilical artery within 3 days of delivery. In the um-bilical artery systolic/diastolic (S/D), pulsatilityindex (PI), and resistance index (RI) were calcu-lated from two distinct segments (from a free-float-ing loop and near insertion to the abdomen) dur-ing a period of fetal apnea and inactivity. At leastthree discrete waveform series were measured foreach umbilical segment and the average of the sixmeasurements was used for the final ratios. Theresults of Doppler parameters were evaluated bythe use of normal values from the literature [16].An abnormal S/D, PI, or RI was defined as a calcu-lated mean value above the 95th percentile forgestational age, and any patient with at least oneabnormal mean value for the three indices wasincluded in the abnormal Doppler FVW group.

TissuesHuman placentas from 30 IUGR pregnancies and30 normal pregnancies were collected at delivery.No gross abnormalities were observed on routineexamination. Placental tissue samples were takenfrom the center and periphery of each placenta,and representative areas of placenta were dis-sected. Specimens were then fixed in 10% formalinand embedded in paraffin. Four-micrometer-thicksections were cut and mounted on poly-L-lysine-coated glass slides and dewaxed by standard tech-niques for the immunohistochemical studies. Im-munohistochemistry was performed usingstreptavidin-biotin-peroxidase complex. Tissuesfrom normal and IUGR placentas were tested onthe same batch by hand to ensure there were nodifferences in the quality of immunostaining. Theslides were immersed in hydrogen peroxide for 30min to suppress endogenous peroxidase activityand were then washed in phosphate buffered sa-line (PBS) with a pH of 7.4. Slides were immersed

308 M. ERDEM ET AL.

in nonimmune protein blocking solution for 20min at room temperature. Sections were then in-cubated for 2 h with ET-1 (ET-1 mouse monoclo-nal antibody IgG1, Oncogene [Durmstadt, Germa-ny] CP44) antibody. The slides were then rinsedwith PBS and incubated with biotinylated anti-body. After the final washing, sections were coun-terstained with hematoxylin. Controls, in whichthe primary antibody was replaced by PBS, did notreveal specific staining. The immunohistochemicaldetection of ET-1 was performed in four tissuespecimens for each placenta and in two sectionsfor each specimen. Sections were examined con-currently by one of the authors (O.E.), who had noknowledge of the clinical history of the pregnancy.ET-1 immunoreactivity was investigated in differ-ent placental regions including the capillary endo-thelial cells of placental villi, syncytiotrophoblastsof villi, vessel endothelial cells of stem villi, and thebasal plate.

Statistical analysisThe ET-1 immunoreactivity of tissues was evalu-ated using a semiquantitative grading system thatconsidered the proportion of examined cells show-ing a positive reaction. The percentage of cytoplas-mic staining was graded as zero (no cells positive),1 (positive staining in � 10% of cells), 2 (positivestaining in � 10 to � 25% of cells), 3 (positivestaining in � 25 to � 50% of cells), 4 (positivestaining in � 50 to � 75% of cells), or 5 (positivestaining in � 75% of cells). The density of cytoplas-mic ET-1 immunostaining was determined by us-ing a semiquantitative four-step grading system:zero (no cells positive), 1 (mild cytoplasmic stain-ing), 2 (moderate cytoplasmic staining), and 3 (in-tense cytoplasmic staining). The extent and inten-sity of ET-1 staining were well correlated in alinear fashion with each other in every anatomicsite studied. Final ET-1 immunoreactivity scoresfor each placenta were produced by the multipli-cation of these two values.

Data are expressed as mean � SD. Statisticalanalysis of the data was performed with the soft-ware package SPSS (Statistical Package for SocialSciences; SPSS, Inc., Chicago, IL) for Windows7.0. Means between groups were compared usingStudent’s t-test. Mann–Whitney U- and Kruskal–Wallis one-way ANOVA tests were used to compare

ET-1 immunoreactivity scores between groups. A Pvalue � 0.05 was considered statistically signifi-cant.

RESULTSThere were no differences in mean maternal age(27.7 � 5.8 vs. 28.3 � 3.8 years, respectively) ormean gestational week at delivery (36 � 2.4 vs. 37.4� 1.6 wk, respectively) between pregnancies com-plicated with IUGR and uncomplicated normalpregnancies. Six placentas in the IUGR group andfour placentas in the control group were frompregnancies before 35 gestational wk. The meanbirth weight (� SD) of the infants in the IUGRgroup was significantly lower than that of the con-trol group, which was matched for gestationalweek at delivery (3311 � 424 to 2074 � 446, P �

0.001). Nineteen women in the IUGR group(63.4%) and 11 women in the control group(36.6%) underwent cesarean section. There was nodifference in ET-1 expression between placentasfrom vaginal and cesarean section deliveries (datanot shown).

Comparisons of ET-1 immunoreactivity indifferent placental regions between normal andIUGR pregnancies are summarized in Table 1. Inall normal and IUGR placental tissues, ET-1 im-munoreactivity was present diffusely in the endo-thelial cells of blood vessels localized in the meso-derm of normal placental villi at different ratios.Localization of ET-1 immunoreactivity in capillaryendothelial cells of villi was higher in placentalspecimens of the controls (P � 0.05) (Fig. 1). ET-1immunoreactivity was detected within the syncy-tiotrophoblast of the villi in all normal and IUGRplacental tissues, however, there was no significantdifference in their distribution between IUGR andcontrol placentas. ET-1 immunoreactivity was lo-calized in the endothelial cells of stem villi in allIUGR placentas and in the majority of the controlplacentas (all but two cases), but there was nosignificant difference in their distribution betweenIUGR and control placentas. ET-1 was localized ina diffuse manner in the endothelial cells of bothanchoring villi and maternal vessels, and decidualcells and trophoblastic cells of the basal plate, andthe localization of ET-1 was higher in the controlplacentas than in IUGR placentas (P � 0.01).

ET-1 IN INTRAUTERINE GROWTH RESTRICTION 309

IUGR was complicated by abnormal umbili-cal artery FVW in 12 pregnancies (RI � 95th per-centile in 7 patients and S/D � 3 in 11 patients). Nopatients had absent or reverse end diastolic flow.Comparisons of ET-1 immunoreactivity in differ-ent placental regions between normal and IUGRpregnancies according to Doppler findings are

summarized in Table 2. There were no significantdifferences in the distribution or proportion ofET-1 immunoreactivity in the endothelial cells ofblood vessels and syncytiotrophoblast of the villibetween placentas of IUGR pregnancies with nor-mal and abnormal umbilical artery FVW. The lo-calization of ET-1 immunoreactivity in the endo-thelial cells of blood vessels of stem villi, and theendothelial cells of blood vessels, and decidual andtrophoblastic cells of the basal plate, were alsosimilar. The localization of ET-1 immunoreactivitywas significantly higher in decidual cells and tro-phoblastic cells of the basal plate as well as in thecapillary endothelial cells of villi in the placentas ofthe controls compared to the placentas of IUGRpregnancies with normal umbilical artery DopplerFVW (P � 0.05). When placental tissues from preg-nancies with abnormal umbilical artery DopplerFVW were compared with those from the controls,the localization of ET-1 immunoreactivity was sig-nificantly higher only in the decidual cells andtrophoblastic cells of the basal plate in the controls(P � 0.05).

Six pregnancies in the IUGR group were com-plicated with pregnancy-induced hypertension.There was no significant difference in the distribu-tion of ET-1 immunoreactivity between the placen-tas of pregnancies either complicated with preg-nancy-induced hypertension or not (data notshown). All statistical comparisons were per-formed excluding premature cases in the IUGRgroup and the control group, but the results of thecomparisons did not change (data not shown).

Figure 1. Immunohistochemical localization of endothe-lin (ET)-1 in intrauterine growth restricted human pla-centa. Endothelial cells of capillary blood vessels (arrows)in placental terminal villi of placental base demonstrateda mild ET-1 immunoreactivity. The extent of staining wasgraded as 2 (positive staining in � 10 to � 25% of cells)and the intensity of staining was graded as 1 (mild cyto-plasmic staining); final endothelin-1 immunoreactivityscore was 2. Extravillous trophoblasts (t) of basal platedemonstrated a diffuse ET-1 immunoreactivity. The ex-tent of staining was graded as 5 (positive staining in �75% of cells) and the intensity of staining was graded as3 (intense cytoplasmic staining); final endothelin-1 im-munoreactivity score was 15 (streptavidin-biotin-peroxi-dase, 3,3� diaminobenzidine tetrahydrochloride, originalmagnification �100).

Table 1. Comparison of endothelin (ET)-1 immunoreactivity (IR) scores in different placentalregions between normal and intrauterine growth restricted (IUGR) pregnanciesa

Normal (n � 30) IUGR (n � 30)

P

Intensityof ET-1(� SD)

Extent ofET-1(� SD)

ET-1 IRscore(� SD)

Intensityof ET-1(� SD)

Extent ofET-1(� SD)

ET-1 IRscore(� SD)

Capillary endothelial cells ofplacental villi

1.4 � 0.5 2.9 � 1.3 4.7 � 3.3 1.3 � 0.5 1.8 � 1.1 2.8 � 1.7 � 0.05

Syncytiotrophoblasts of villi 2.0 � 0.6 4.4 � 0.8 9.1 � 3.9 1.8 � 0.7 4.0 � 1.2 7.7 � 4.3 NSb

Vessel endothelial cells ofstem villi

1.4 � 0.5 1.9 � 1.0 3.1 � 2.4 1.3 � 0.5 1.7 � 1.0 2.4 � 2.1 NS

Basal plate 2.0 � 0.6 3.9 � 1.2 8.3 � 4.1 1.6 � 0.6 3.3 � 1.5 5.2 � 3.2 � 0.01

aMann–Whitney U-test.bNot significant.

310 M. ERDEM ET AL.

Tab

le2.

Co

mp

aris

on

of

ET

-1im

mu

no

reac

tivi

tysc

ore

sin

dif

fere

nt

pla

cen

tal

regi

on

sb

etw

een

no

rmal

and

IUG

Rp

regn

anci

esw

ith

rega

rdto

Do

pp

ler

fin

din

gsa

IUG

R

No

rmal

pla

cen

ta(n

�30

)N

orm

alD

op

ple

r(n

�18

)A

bn

orm

alD

op

ple

r(n

�12

)

Inte

nsi

tyof

ET

-1(�

SD

)

Ext

ent

ofE

T-1

(�S

D)

ET

-11R

scor

e(�

SD

)

Inte

nsi

tyof

ET

-1(�

SD

)

Ext

ent

ofE

T-1

(�S

D)

ET

-11R

scor

e(�

SD

)

Inte

nsi

tyof

ET

-1(�

SD

)

Ext

ent

ofE

T-1

(�S

D)

ET

-11R

scor

e(�

SD

)P

Cap

illa

ryen

dot

hel

iace

lls

ofp

lace

nta

lvi

lli

1.4

�0.

52.

9�

1.3

4.7

�3.

3c,d

1.3

�0.

61.

9�

1.2

2.8

�2.

2c1.

3�

0.5

1.8

�0.

82.

6�

1.9d

�0.

05

Syn

cyti

otro

ph

obla

sts

ofvi

lli

2.0

�0.

64.

4�

0.8

9.1

�3.

91.

9�

0.7

4.0

�1.

38.

1�

4.7

1.7

�0.

64.

1�

0.9

7.1

�3.

5N

Sb

Ves

sel

end

oth

elia

lce

lls

ofst

emvi

lli

1.4

�0.

51.

9�

1.0

3.1

�2.

41.

4�

0.5

1.8

�1.

12.

6�

2.1

1.2

�0.

41.

7�

1.1

2.2

�2.

1N

S

Bas

alp

late

2.0

�0.

63.

9�

1.2

8.3

�4.

1e1.

6�

0.6

3.2

�1.

65.

4�

3.6e

1.5

�0.

73.

4�

1.2

4.9

�2.

7�

0.05

aK

rusk

al–W

alli

s(o

ne-

way

anal

ysis

ofva

rian

ce).

bN

otsi

gnifi

can

t.cB

etw

een

nor

mal

and

IUG

Rw

ith

nor

mal

Dop

ple

r,M

ann

–Wh

itn

eyU

-tes

t.dB

etw

een

nor

mal

and

IUG

Rw

ith

abn

orm

alD

opp

ler,

Man

n–W

hit

ney

U-t

est.

eB

etw

een

nor

mal

and

IUG

Rw

ith

nor

mal

Dop

ple

r,M

ann

–Wh

itn

eyU

-tes

t.

ET-1 IN INTRAUTERINE GROWTH RESTRICTION 311

DISCUSSIONThe current study has demonstrated that ET-1 wasdetected diffusely by immunohistochemistry inplacental specimens from IUGR pregnancies aswell as from normal pregnancies. In normal preg-nancies, ET-1 like immunoreactivity was formerlydetected immunohistochemically in the endothe-lium of fetal placental vessels of all sizes and in thecytoplasm of decidual cells in the basal maternalplate but not in the trophoblasts [7,9]. In a studyinvestigating the origin of ET-1 in human placenta,it was further demonstrated that ET-1 immunore-activity was specifically detected in the endothe-lium of the vessels as well as in the syncytiotropho-blast of the villi, and, after the purification oftrophoblasts, ET-1 immunoreactivity was observedin the majority of the cultured trophoblastasticcells [17]. Recently, ET-1 immunostaining hasbeen detected, not only in the endothelial cells ofcapillaries, venules, and arterioles, but also in thetrophoblastic cells of the basal plate and villi offull-term placentas, which was in accordance withour findings [14,18]. Consequently, in all thesestudies (including ours), it is indicated that ET-1immunoreactivity is widely distributed in the vas-cular endothelial cells of the placenta. Althoughthere is controversy over the immunostaining pat-tern of ET-1 in trophoblastic cells, ET-1 immuno-reactivity could have been demonstrated in tropho-blasts in the majority of studies, including thecurrent study. These data strongly suggest that thevascular endothelium as well as trophoblasts of thehuman placenta represent a source of ET-1 pro-duction or at least a site for ET-1 binding.

Although the contribution of ET activity isstrongly indicated in IUGR, to the best of ourknowledge, neither the distribution of ET-1 immu-noreactivity in human placenta from IUGR preg-nancies nor the comparisons of differences in thedistribution of ET-1 immunoreactivity betweenplacentas from IUGR and normal pregnancieshave been reported before. We found that the dis-tribution and intensity of ET-1 immunoreactivity,as assessed by a semiquantitative scoring system,was almost always higher in normal control pla-centas than in placentas from IUGR pregnancies.Moreover, the localization of ET-1 immunoreactiv-ity in placentas from normal pregnancies was sig-

nificantly evident in the capillary endothelial cellsof the villi, and the endothelial, decidual, and tro-phoblastic cells of the basal plate compared toplacentas from IUGR pregnancies. These findingsmay reflect a compensatory decline in the expres-sion of the vasoconstrictor peptide ET-1 in re-sponse to mechanisms triggering IUGR. Differ-ences in receptor distribution may theoreticallycause diminished ET-1 expression in IUGR preg-nancies. In fact, it was observed that there was nodifference between placental endothelin bindingcharacteristics in IUGR and normal pregnancies[11,13]. Consequently, in light of data obtainedfrom receptor studies in combination with ourdata regarding the localization of ET-1 immunore-activity in normal and IUGR placentas, it may besuggested that mechanisms including differencesin the production, action, or affinity of ET-1 tobinding sites other than aberrations in receptordistributions in placental tissues are implicated inthe pathogenesis of IUGR.

ET binding to the vascular smooth muscleETA and ETB receptors mediates vasoconstriction,while binding to endothelial ETB receptors mayresult in a paradoxical vasodilatation prior to asustained vasoconstriction by releasing endotheli-um-derived relaxation factors like nitric oxide andprostacyclin [19,20,21]. In animal studies, it hasbeen demonstrated that ET infusion induced adose-dependent decrease in fetal and placentalgrowth [22] and ETA and ETB receptor antagonismproduced fetal and placental growth restriction,while in the setting of long-term nitric oxide syn-thase inhibition, the antagonism of neither ETA

nor ETB improved fetal and placental growth [23].Although elevated cord blood ET-1 levels

were reported in growth-restricted infants, partic-ularly in those with severe fetoplacental vasocon-striction as reflected by absent or reverse end-dia-stolic flow velocity waveforms [11,12], ET receptorcharacteristics did not differ [11,13], or the local-ization of ET-1 immunoreactivity was higher innormal pregnancies than in IUGR pregnancies, asindicated in our study. Most studies in which bloodET-1 levels were assessed involved intrauterine fe-tal growth-restricted pregnancies complicatedwith absent or reverse end-diastolic blood flow. Infact, pregnancies complicated with an absent orreverse end-diastolic umbilical artery flow velocity

312 M. ERDEM ET AL.

waveform represent a relatively small group (1.9–9.4%) of pregnancies at high risk for intrauterinegrowth restriction [24]. Although no patients inour study population had absent or reverse end-diastolic umbilical Doppler flow, abnormal veloc-ity waveforms other than absent or reverse end-diastolic umbilical Doppler flow were not relatedto differences in the placental expression of ET-1.

One of the potential explanations of the diver-sity between umbilical ET-1 levels and placentalfindings could be that in a proportion of IUGRpregnancies, a hypertensive pregnancy disordercoexists, and differences in the circulating concen-tration of endothelins might reflect the maternaldisorder. It was observed that strong ET-1 expres-sion was evident in the endothelium of fetal vesselsand in the syncytiotrophoblast from near-term pla-centas, and in the extravillous cytotrophoblast ofthe basal and chorionic plates, regardless of gesta-tional age in preeclamptic placentas [15].

In conclusion, the distinct localization ofET-1 immunoreactivity in normal and IUGR pla-centas suggests that ET-1 expression may be in-volved in the modulation of fetal–placental bloodflow. The lower expression of ET-1 in placentaltissues from IUGR pregnancies might be second-ary to an adaptive mechanism to reduce the vaso-constrictor effect of ET-1. Further studies compar-ing immunohistochemical expression of ET-1 withcord blood levels are necessary to demonstrate thepotential role of ET-1 in the pathogenesis of intra-uterine growth restriction.

R E F E R E N C E S1. Piper JM, Xenakis EM, McFarland M, Elliott BD, Berkus

MD, Langer O. Do growth-retarded premature infants havedifferent rates of perinatal morbidity and mortality thanappropriately grown premature infants?. Obstet Gynecol1996;87:169–174.

2. Giles WB, Trudinger BJ, Baird P. Fetal umbilical artery flowvelocity waveforms and placental resistance: pathologicalcorrelation. Br J Obstet Gynaecol 1985;92:31–38.

3. McCowan LM, Mullen BM, Ritchie K. Umbilical artery flowvelocity waveforms and the placental vascular bed. Am JObstet Gynecol 1987;157:900–902.

4. Fok RY, Pavlova Z, Benircshke K, Paul RH, Platt LD. Thecorrelation of arterial lesions with umbilical artery Dopplervelocimetry in the placentas of small-for-dates pregnancies.Obstet Gynecol 1990;75:578–583.

5. Macara L, Kingdom JCP, Kaufman P. Structural analysis ofplacental terminal villi from growth-restricted pregnancieswith abnormal umbilical artery Doppler waveforms. Pla-centa 1996;17:37–48.

6. Inou A, Yanagisawa M, Kimura S. The human endothelin

family: three structurally and pharmacologically distinctisopeptides predicted by three separate genes. Proc NatlAcad Sci USA 1989;86:2863–2867.

7. van Papendorp CL, Cameron IT, Davenport AP. Localiza-tion and endogenous concentration of endothelin-like im-munoreactivity in human placenta. J Endocrinol 1991;131:507–511.

8. Robaut C, Mondon F, Bandet J, Ferre F, Cavero I. Regionaldistribution and pharmacological characterization of [125I]endothelin-1 binding sites in human fetal placental vessels.Placenta 1991;12:55–67.

9. Wilkes BM, Susin M, Mento PF. Localization of endothelin-1-like immunoreactivity in human placenta. J HistochemCytochem 1993;41:535–541.

10. Rutherford RA, Wharton J, McCarthy A. Differential local-ization of endothelin ETA and ETB binding sites in humanplacenta. Br J Pharmacol 1993;109:544–552.

11. McQueen J, Kingdom JCP, Connell JMC, Whittle MJ. Fetalendothelin levels and placental vascular endothelin recep-tors in intrauterine growth retardation. Obstet Gynecol1993;82:992–998.

12. Harvey-Wilkis KB, Nielsen HC, D’Alton ME. Elevated en-dothelin levels are associated with increased placental re-sistance. Am J Obstet Gynecol 1996;174:1599–1604.

13. Kohnen G, Mackenzie F, Collett GP. Differential distribu-tion of endothelin receptor subtypes in placentae from nor-mal and growth-restricted pregnancies. Placenta 1997;18:173–180.

14. Graf AH, Hutter W, Hacker GW. Localization and distribu-tion of vasoactive neuropeptides in the human placenta.Placenta 1996;17:413–421.

15. Barros JS, Bairos VA, Baptista MG, Fagulha JO. Immuno-cytochemical localization of endothelin-1 in human pla-centa from normal and preeclamptic pregnancies. Hyper-tens Pregn 2001;20:125–137.

16. Campbell S, Vyas S, Nicolaides KHJ. Doppler investigationof the fetal circulation. J Perinat Med 1991;19:21–26.

17. Malassine A, Cronier L, Mondon F, Mignot TM, Ferre F.Localization and production of immunoreactive endothe-lin-1 in the trophoblast of human placenta. Cell Tissue Res1993;271:491–497.

18. Marinoni E, Picca A, Scucchi L, Cosmi EV, Di Iorio R.Immunohistochemical localization of endothelin-1 in pla-centa and fetal membranes in term and preterm humanpregnancy. Am J Reprod Immunol 1995;34:213–218.

19. de Nucci G, Thomas R, D’Orleans-Juste P. Pressor effects ofcirculating endothelin are limited by its removal in thepulmonary circulation and by the release of prostacyclinand endothelium-derived relaxin factor. Proc Natl Acad SciUSA 1988;85:9797–9800.

20. Hirata Y, Emori T, Eguchi S. Endothelin receptor subtypeB mediates synthesis of nitric oxide by cultured bovineendothelial cells. J Clin Invest 1993;91:1367–1373.

21. Davenport AP, Maguire JJ. Is endothelin-induced vasocon-triction mediated only by ETA receptors in humans. TrendsPharmacol Sci 1994;15:9–11.

22. Neerhof MG, Silver RK, Caplan MS, Thaete LG. Endothe-lin-1-induced placental and fetal growth restriction in therat. J Matern Fetal Med 1997;6:125–128.

23. Thaete LG, Neerhof MG, Silver RK. Differential effects ofendothelin A and B receptor antagonism on fetal growth innormal and nitric oxide-deficient rats. J Soc Gynecol Inves-tig 2001;8:18–23.

24. Alfirevic Z, Neilson JP. Doppler ultrasonography in highrisk pregnancies; systemic review with meta-analysis. Am JObstet Gynecol 1995;172:1379–1387.

ET-1 IN INTRAUTERINE GROWTH RESTRICTION 313