Localization of the angiotensin II and its
receptor subtype expression in human
endometrium and identification of a novel
high-affinity angiotensin II binding site.
A Ahmed, … , N M Barnes, J R Newton
J Clin Invest.
1995;
96(2)
:848-857.
https://doi.org/10.1172/JCI118131
.
Angiotensin (ANG) II is not only a potent vasoconstrictor but may also be involved in the
regeneration of new blood vessels. In proliferative endometrium, ANG II-like
immunoreactivity was detected in glandular epithelium and stroma with negligible staining
around the vascular endothelium. In contrast, in secretory endometrium intense
immunostaining was seen in the perivascular stromal cells around the endometrial spiral
arterioles with negligible staining of the other cell types. Quantitative receptor
autoradiography using the nonselective radioligand [125I]-ANG II and subtype selective
competing compounds showed that endometrium contained predominantly AT2 receptors,
with relatively low expression of AT1 receptors and a novel non-AT1/non-AT2 angiotensin II
recognition site that was insensitive to AT1 or AT2 selective ligands. Levels of specific
[125I]-ANG II receptor binding displayed cyclic changes during the menstrual cycle,
reaching a maximum in early secretory endometrium and then decreasing in mid to late
secretory endometrium to levels seen in early to mid proliferative endometrium. In situ
hybridization showed AT1 receptor mRNA expression in the glands and in the endometrial
blood vessels. The cyclic changes in ANG II-like immunoreactivity together with expression
of both the known and the novel AT receptor subtypes imply that this octopeptide may play
a dual role both in the control of the uterine vascular bed and also in the regeneration of the
endometrium after endometrial shedding, acting as […]
Research Article
Localization
of
the Angiotensin
11
and
Its
Receptor
Subtype Expression
in
Human Endometrium and Identification
of
a
Novel
High-affinity Angiotensin
11
Binding Site
AsifAhmed,** Xiao F. Li,** Munjiba
Shams,*"
JohnGregory,s
TerryRollason,*
Nicholas M. Barnes,11 and John R. Newtont *Reproductive Physiopathology Group, Center for ClinicalResearch inImmunology and Signalling, tDepartmentsof Obstetrics andGynaecology, §Pathology, I1Pharnacology, The MedicalSchool, University of Birmingham, Edgbaston, Birmingham B15 27T,
UnitedKingdom
Abstract
Angiotensin (ANG) II isnotonly apotentvasoconstrictor
butmayalso be involved in theregeneration ofnewblood
vessels. Inproliferative endometrium, ANG II-like
immu-noreactivity was detected in glandular epithelium and stroma with negligible staining around the vascular endo-thelium. In contrast, in secretory endometrium intense
im-munostaining was seen in the perivascular stromal cells around the endometrial spiral arterioles with negligible stainingofthe other cell types.Quantitativereceptor
autora-diography usingthe nonselectiveradioligand [125I]-ANGII
and subtype selective competing compounds showed that endometrium containedpredominantly AT2 receptors,with
relatively low expression of AT1receptors andanovel
non-AT,/non-AT2
angiotensinIIrecognitionsitethatwas insen-sitive to AT1 or AT2 selective ligands. Levels of specific['"I]-ANG
II receptor binding displayed cyclic changesduring the menstrual cycle, reaching amaximum in early secretoryendometrium and then decreasing inmid to late secretory endometrium to levelsseeninearlytomid
prolif-erative endometrium. In situhybridizationshowed AT1 re-ceptor mRNAexpression in theglands and in the endome-trialblood vessels. Thecyclic changesinANG TI-like
immu-noreactivity togetherwithexpressionofboth the known and the novel ATreceptorsubtypes implythat thisoctopeptide
may play a dual role both in the control of the uterine vascular bed and also in the regeneration of the
endome-trium after endometrial shedding, acting as an angiogenic
andmitogenicmediator.(J.Clin. Invest. 1995.96:848-857.) Keywords: uterus*autoradiography *in situhybridization *menstruation* menorrhagia
Introduction
Excessive menstrualbleeding(> 80ml)canleadto iron
defi-ciencyanemia andmaynecessitatehysterectomy. Menorrhagia
This workwaspresented inpartattheSocietyforGynecologic
Investi-gationAnnualMeeting inChicagoon22-26 March 1994.
Addresscorrespondence toDr. Asif S.Ahmed, Departmentof Ob-stetrics andGynaecology, UniversityofBirmingham, Birmingham
Ma-ternityHospital, Edgbaston, Birmingham, B15 2TG,UnitedKingdom.
Phone/FAX: 021-627-2705.
Receivedfor publication20June1994 andinrevised form 19 May
1995.
canbe the result of organic diseases such as polyps, fibroids, pelvicpathology, or carcinoma, but in >50%of cases there is neither evidence of underlying pathology nor an imbalance of thehypothalamic-pituitary-ovarian axis. There is increasing evi-dence to suggest that the mechanisms controlling menstruation and its disorders areregulated by local factors within the endo-metrium itself (1).
The commencement of menstruation is characterized by
in-tense vasoconstriction of the spiral arterioles (2). After the
onsetof menstrual bleeding, the upper functional layer of the endometrium is shed, and cessation of bleeding is achieved by vasoconstriction of the remaining basal arteriolar fragments and notby thedeposition of platelet-fibrin plugs (3). This vasocon-striction is essential until blood loss is stemmed by the regenera-tion of the vessels and surrounding tissues (4).
The precise nature ofthe endometrial vasoconstrictor has
notbeenestablished. Prostaglandin
F2,
and more recently endo-thelins have been suggestedaspossible candidate endometrial vasoconstrictors (5). However, prostaglandin F2< is usually a relatively weak pressor agent (6). O'Reilly and co-workers (7) reported that therewere differences in the ratio of endothelin receptors(ETA andETB) receptor mRNA levels in proliferative andsecretoryendometriumusingreversetranscriptionpolymer-asechainreaction, but the functional relevance of these changes remains unclear.Althoughthe localfactors controlling the uter-ine vascular bedare poorly understood, it is increasingly clear that they play an important role both in the mechanisms of menstruation and in thepreparationofareceptive endometrium fortheimplantation of thedevelopingembryo.
Historically, angiotensin II (ANG
II)'
has beenassociated with thehomeostaticregulation of fluid and electrolyte balance andperipheral vascular resistance. More recentstudies suggest that it mayact as an angiogenicgrowth factor involved in the regeneration ofnewblood vessels (8-10). Administration of ANG IItothe rabbitcornea(9)or tochick embryo chorioallan-tois membrane (10) induced pronounced neovascularization and ANG II canbeadirectstimulus for cell growth (11). Asnonpeptidergic ANG II receptor (AT) antagonists were not
effectiveinblockingtheANG IIangiogenesisofpre- and post-capillaryvessels(10) definitiveproofofthisangiogenic effect needs furtherinvestigation.
ANG II interacts withspecific,high-affinity receptorsonthe surface of target cellsto exertitsbiologicaleffects. Advancesin
understanding the structureand function ofAT receptorshave
been facilitatedby two recent findings. First,the development ofnonpeptideinhibitors of ANG II hasclearly establishedthat
1.Abbreviations usedinthispaper:ANG II,angiotensin II; AT,ANG
II receptor. J. Clin. Invest.
© The AmericanSocietyfor ClinicalInvestigation, Inc.
there are at leasttwomajor classes ofATreceptorsreferred to as
AT1 and AT2 (12). Experiments with the
AT,-specific
inhibitorlosartan(previously designated DuP753) indicate that this sub-typeis responsible for at least the vast majority of the hemody-namic andcardiovascular effects of ANG11(12).Second,
sev-eral groups have used molecular techniques to isolate cDNA encoding
AT,
(13) and AT2 receptors (14). TheATI
receptor is sensitive to nanomolar concentrations of losartan (1,000-foldselectivityoverAT2 receptor)and is a Gprotein-coupled receptor which activates a signal transduction mechanismin-volving phosphoinositide hydrolysis and Ca2+ mobilization (15). In contrast, the AT2 receptor is sensitive to nanomolar
concentrations ofPD123177 (1,000-fold selectivity overAT,
receptor) and does not appear tobecoupledto atransduction systemvia a Gprotein but may promote cell growth(16). ANG
IImay beacrucial local factor in the endometrium since intense vasoconstriction of the spiral arterioles is a prerequisite for normal menstruation and the microvascular system responds with rapid capillary growth associated with endometrial regen-eration after menstruation. Mapping of the tissue distribution ofATreceptors is therefore essential foraunderstanding ofthe
role of ANG Hin this tissue.
Methods
Tissuecollection
Endometrium was collectedfromwomenundergoingdilatation and
cu-rettage for laporascopic sterilization. Tissues were obtained from 28 womenforimmunocytochemical studies,from 27 women for
quantita-tive receptorautoradiographic studies, and from 8 women for in situ
hybridization studies. Patientsselected hadnoapparentendocrinological
problems and no local organic pathologies. All tissues were obtained
from20-40-yr-old women with ahistory ofregularmenstrual cycles. Noneofthepatientshastakenany hormonalmedication foratleast 3
mo before surgery. Specimens of endometrium were obtained from proliferativeandsecretoryphasesofthecycleand theday ofthecycle
wasconfirmed by independent histologicalassessment(17, 18).Ethical
committeeapproval wasobtained fromthe SouthBirminghamEthical Committee.
The tissue wasrinsedinsterile salineandimmediatelyimmersed in 10%formaldehydeandroutinely processedforparaffinwaxembedding.
Tissue was also rapidly frozen overdry-ice, wrapped in Parafilmto prevent dehydration, and stored at -80'C until they were sectioned
for receptor autoradiographic or in situ hybridization studies. Frozen
endometrium and myometrial tissues weresurrounded in embedding
medium (Tissue-TeckO.C.T. Compound; Agag Scientific, Stanstead, Essex, UnitedKingdom) before 20-pm sectionswere cutusinga
cryo-stat(-15to-19°C)and thaw mountedontogelatin-coated glassslides for receptorautoradiographic studies, and
10-,4m
sectionswere cutand thaw mounted onto poly-L-lysine (Sigma Chemical Co. Ltd., Poole,United Kingdom)coatedglass slidesfor in situhybridizationstudies. Sectionswere stored (< 2wk)anddesiccated at -80°Cuntil used.
Immunocytochemistry
Serial3-ttmsectionsofformalin-fixed, paraffin-embeddedtissuewere
used forimmunohistochemistry aspreviously reported (19). Sections
weredeparaffinized byincubation for 5 min with histoclear and hydrated
through methanol to water. Endogenous peroxidase activity was
quenched by 0.3% (vol/vol) hydrogen peroxide in methanol for 10 min. The primary antibody was a rabbit polyclonal antibody raised
againstthe human ANG II and was purchased from Peninsula Labora-tories (Merseyside, UnitedKingdom). It is highly specific but
cross-reactscompletelywith human ANGIIIand veryslightly with angioten-sin 1(0.5%) and renin (1%). Nonimmunegoat serum(10% in 0.01 mol/literPBS,pH 7.2)wasusedas adilution of theprimary antibody
to reduce nonspecificbinding. Amplification of the primary antibody reaction was achieved using a goat anti-rabbit secondary antibody (di-luted 1:200 in 0.01 mol/literPBS, pH 7.2) for 30 min followed by a complex of streptavidin and biotinylated peroxidase (both from Dako Ltd., Bucks, UnitedKingdom). Finally, thebindingwasvisualized by the addition of 0.5mg/ml diaminobenzidine (Sigma Chemical Co. Ltd.) and 0.01% hydrogen peroxide in 0.01 mmol/liter PBS to the antigen-antibody complex. Between each step the sections were washed in 3 x200 ml of 0.1% (vol/vol) polyoxylene-10-oleoyl-ether in 0.01 mmol/ liter PBS, pH 7.2, over aperiod of 15 min. All incubations of antisera werecarried out at room temperature in a wet chamber mountedona
rocking tray which ensures a movement of antiserum over the whole section. The sections were counterstained with Mayer's hematoxylin, dehydrated, and mounted.
To testthespecificityof theimmunocytochemical staining, the pri-mary antibody was omitted from the sections orreplaced with goat nonimmune serum in controlexperiments.Controls were also performed by preabsorbing theprimary antibody withANG II at aconcentration of 16.6
pl
antibody/mgANG II for 24 h at 40C.Finally,asANG II is known tobe present inblood products andendometrial biopsies may be contaminated by blood products as well as the procedure beingtraumatic, endometrium was also obtained fromhysterectomyspecimens ascontrol.
Evaluation ofimmunostaining. An arbitrary fourpoint scale was
usedtogradetheintensityand theproportion of positivelystained cells. A score of fourindicated very intense staining of > 95% for stromal
cells, glandularepithelium,and blood vesselscomparedwitha scoreof zero whichindicates an absence ofstaining. A scoreofone tothree
indicated occasional, very weak, and moderate intensity of staining,
respectively. Both theimmunohistochemistry and the scoring of
posi-tively stainedcells wereperformed byasingleindividual withoutprior
knowledge of the phase of the menstrual cycle.
[J25I]-ANG
IIautoradiographyThreeendometrial sections were mounted on the same slide from each patient sample. Slide-mounted human endometrial sections were re-moved from storage and allowed - 30 min to equilibrate to room
tem-perature. To reduceendogenous levels ofANG II in the tissue, the
sections werepreincubated for 60 min in incubation buffer (mM): sodium chloride, 150; sodium dihydrogenphosphate, 50; magnesium chloride, 10;ethyleneglycol-bis-(,l-amino-ethyl ether)NN'-tetra-acetic acid,5; and 0.4% wt/vol bovine serum albumin, pH 7.4) at 25°C. The
slideswerethen incubated inincubation bufferwhich contained0.1nM
Iodotyrosyl'5I-[5-L-isoleucine] -ANGH(I251-ANGII, 2000 Ci/mmol; AmershamInternational,Bucks, United Kingdom) in the absence(total
binding)orpresence ofcompetingcompound(either1.0
/iM
unlabeled ANG II, 1.0jtMDuP753, 1.0HIMPD123177, or 1.0HtMDuP753 plus1.0
ILM
PD123177) for60min at25°C. Immediately afterincubation, the tissue sections were washed in ice-cold incubation buffer for 2 min anddipped(1 s) in ice-cold distilled water to remove buffer salts. The sections wererapidlydried in a streamofcold dry air and exposed toHyperfilm-[3H] (AmershamInternational) in x-ray cassettes together with ['25I]standards (Amersham International) for 40 d.
Autoradio-graphicfilms were developedinKodak LX 24 developer (5 min) and KodakUnifix (5 min).
Evaluationofautoradiography. After the development of
autoradio-graphic film,thesilvergraindensity was assessed using image analysis
(MCID; ImagingResearch Inc., St.Catherines, Ontario, Canada) from the three sectionsoneach slide anda meanvaluewasgenerated which represented n = 1. The film was quantified by reference to ['25I]
-standards (femtomoles permilligram of tissue equivalent values for intact grey matter;AmershamInternational). UnpairedStudent's t test
or ANOVA test followedby Dunnett's test, where appropriate, was
appliedtocompare thechangesin themeanvalues of silvergraindensity
atdifferent stages ofcyclein the groups. In situ hybridization
A 45-bp-long synthetic oligonucleotide probe complementary to the
synthe-sized and used to locate the mRNA that encode the receptor protein (20). This region of theATIAgene waschosen because it was found tobedivergent fromother homologous genes. It was found to be mis-matched by three base pairs to the corresponding human ATIgene (21).
The incubation conditions and washing temperatures were adjusted
ac-cordingly.Theoligoprobewasend labeled with["5S-dATP]by terminal
transferase using a commercially available kit (both from Du Pont,
Stevenage, United Kingdom). The specificity of the oligonucleotides waschecked by ensuring that the probe hybridized to the correct tran-script inpoly(A)+ RNA isolated from the endometrium under study on aNorthern blot.
In situhybridization wasperformed as previously reported (22). Endometrial and myometrial sections were prepared on poly-L-lysine-coated slides andfixed in4%paraformaldehydein PBS (pH 7.4) for
5 min, acetylatedin0.1 Mtriethanolamine(TEA),pH 8.0, with acetic
anhydride(0.25%wt/vol) for10 min. Slidesweredehydrated through anascending graded series of alcohol (70, 80,95, and 100% vol/vol)
followedby incubationinchloroform for 5min. Thesections were then
partially rehydrated sequentiallyin100%(vol/vol) and 95% (vol/vol) ethanoland air dried.As acontrol fornonspecific binding, before fixa-tion, somesectionsweretreatedwith100jig/mlofRNAse A(Sigma Chemical Co.Ltd.) for 1 h at370C.
After acetylation, the air-dried slides were hybridized in a moist
chamber under coverslips in the hybridization buffer containing
for-mamide (50% vol/vol), 3 M NaCl, 0.3 M sodium citrate, 0 1 mM
EDTA,dextransulfate(10% wt/vol), 0.5mlDenhardt'sreagent, 100
jig/mlofsalmon spermDNA, 100jug/mlof yeast transferRNA, 100 jig/ml of polyadenylic acid A, and 1 X 106 cpm/ml of35S-labeled antisense oligoprobe. Hybridization wascarriedout at27°C for 16 h.
Controlhybridizationwasperformed usingradiolabeled antisense
oligo-probe inthepresence of100-foldexcessof unlabeledantisensestrand. After hybridization, thecoverslips were removed andtheslides were then washed fourtimes in 1 x SSCfor25 minat awash temperature of56.5°C. The slides were thenfurther washed twice in 1 x SSC at
room temperature and airdried. Autoradiography wascarried outby
coating the slides in Ilford KS emulsion (Ilford Ltd., Cheshire, United Kingdom)at4°C for 5wk.Thesectionswerethenstainedlightly with 0.1% cresyl violet and mounted.
Results
Immunocytochemical localization of
ANGIIANG II-like immunoreactivity was detected in endometrium through the menstrualcycle. Therewasamarkeddifference in the pattern ofstainingbetween differentphasesof themenstrual
cycle. The positive control ofratadrenal sections showed in-tensepositive staining. Stainingwaslocalized intheglandular epitheliumand thestromaintheproliferative phaseof the
men-strualcycle, withnegligible stainingaround the vascular endo-thelium (Fig. 1,A andB). In contrast, thestaining intensified around the blood vessels of the endometrium in the secretory phase (Fig. 1, C-F). The cyclic changes in the pattern of staining in the stroma, glands, and blood vessels through the
menstrualcyclearegraphically illustratedaccordingto an arbi-trary grading system (Fig. 2). No difference in the staining
pattern was seen in sectionsof endometrium obtainedbyD&
C orhysterectomy. Nospecificstaining was seenwhen
preim-mune serum was substituted and staining was attenuated in control sections in whichprimary antibody was preincubated with thepeptide.
During theearly proliferative phase,heterogeneousstaining
wasfound in the stromaand theglandular epithelium and the
degreeofstaining varied from very weak to intense. Some of thestromal and glandular cells showed the intense typical brown
granularcytoplasmic stainingandno
staining
was seen aroundthe endometrial blood vessels (Fig. 1, A and B). In the mid
proliferativephase, the immunoreactivity was noted in some of
the cells inthe stroma and in the glandular epithelium. These
celltypesshowed relatively weak to moderate staining during
the mid proliferative phaseand the intensity of staining appeared todecrease compared withthe earlier phase of the cycle. Again,
no positive staining was seen around the endometrial blood
vessels. During the late proliferative phasenegligible
immuno-reactivity to ANG II was observed in the stroma, the glands,
and the blood vessels.
In contrast,positivestaining was localized around the
endo-metrial spiral arteriolesduring earlysecretory phase with
negli-gible staining in glandular epithelium and stroma (Fig. 1, C
and D). There wasanincrease intheintensityof staining which
wasdiffused around endometrial bloodvessels during the mid
secretory phase with minimal or entirely absent staining in the stroma. Moderate staining in the glands was observed in some cases. During the late secretory phasetheintensity of staining
was visuallyenhanced and localizedin the perivascular stromal
cells around the endometrial blood vessels (Fig. 1, E and F). The glandularepithelium showed weakstaining and the stroma
appearedrelatively negative to the primary antibody in the late secretory phase of the cycle (Fig. 1, E and F).
[125sI]ANG
IIautoradiographyLevels of AT receptor subtypes in 27 endometrial specimen from women (mean age 35.3 yr, range 18-39 yr) with normal cyclic menstruation were quantified as described in Methods. Of the 27 endometrial specimens, histological assessment of the specimens resulted in the following classifications: 1 was
earlyproliferative, 4 were mid proliferative, 7 were late prolifer-ative, 3 were early secretory, 6 were mid secretory, and 6 were late secretory.
Adjacent sections incubated with [1251]-ANG II labeled a saturable binding site in human endometrial-myometrial sec-tions (Fig. 3). Specific binding (defined by the inclusion of unlabeled ANG II, 1.0 jiM) was mainly associated with the AT2 receptorsince it was sensitive to the inclusion of PD123 177 (Fig. 3 C). Within the endometrium, the AT2 receptor specific ligand competed for mean ( ±SEM) 65 ±4.3% of the total bind-ing. In contrast, the specific losartan-sensitive binding repre-sented only 19+4.7% of the total binding. In addition to losart-an- and PD123177-sensitive specific [1251]-ANG II, an addi-tionalspecific sitewasconsistently detected that was insensitive to both losartan (1
jIM)
and PD 123177 (1 jiM) (Fig. 3 D) butsensitiveto 11LM
unlabeled ANGII(Fig. 3 E). This novel high-affinity AT recognition site represented 16±3.6% of specific binding of the total binding.ATreceptor binding levels. The mean (±SEM) levels of specific [1251]AT receptorbinding for AT1 and AT2 receptors showed cyclic variation. Although they paralleled each other, the levels of AT2 receptor weresignificantly higher than
AT,
Figure1. ANG 11-likeimmunoreactivityinendometrium.Serial sections ofendometriumfrom earlyproliferative (A and B), earlysecretory(C and D), and latesecretory(Eand F) phases wereincubatedwith primary antibody (A, C, and E) or withnonimmuneserum(B,D,and F).
Stainingis seen in theglands(G) and stromal cells and no staining is observed around the blood vessels (BV) as indicated by the arrows. Bar, 40
J.m..
(0.048±0.022 fmol/mg tissue equivalent; P < 0.05) and late secretory endometrium (0.051±0.020 fmol/mg tissue equiva-lent; P < 0.05). Similarly, the mean levels of AT2 receptor specific binding in the early secretory endometrium was 0.421±0.053 fmol/mg tissue equivalent and this too was
(A)
0
@0 0
0 0 0 0 0
*0-0 ---- se *@@e@
(B)
0*
..
...
.
..
* 0* 0 0 00
0@ 0e 0 0 000 0
(C) ..
00
.
00.
0 00 0 0 00 0
0 0000 0@0 00 0
EP
day 5-7 dayMP8-lo dayLP11-14 day 15-18 day 19-23 day 24-28ES MS LS
Menstrual
cycle
Figure2. Graphical representationofthecyclic changesin thecellular distributionofimmunostaining duringthemenstrualcycles.
Immuno-histological stainingscoresin thestroma(A),theglands (B),and
around thevessels(C)inindividualsamples (solid circles)of the
endometrium duringthe menstrualcycles.Thedegreeofchangein
intensity andnumbers of cellsstainingweregraded accordingtoan
arbitraryfourpointscale. Ascoreoffour indicates intensestainingof
>95%of theglands, stromacells,and vesselscomparedwitha score
ofonewhichindicatesoccasionalorveryweakstaining,andscoreof
zero meansnegative. EP, early proliferative; MP,midproliferative; LP,lateproliferative; ES, earlysecretory;MS,mid secretory;LS,late secretory.
inthe levels ofATIandAT2receptorspecific bindingbetween the lateproliferativeandearlysecretory endometrium.The
lev-els of
non-ATI/non-AT2
receptorspecific binding inendome-trium were very low throughoutthe menstrual cycle anddid
notchange significantly duringthe menstrualcycle (Fig.4C). In situhybridizationforAT, mRNA
Insitu hybridization was performedtoidentifythe site of
ex-pressionofAT,receptormRNA inendometrium.In the
prolif-erativeendometrium,thelevel of thehybridization signal was distributedoverboth theglandular epithelium andstroma(Fig. 5,A and B). The intensity ofhybridization decreased in the stromal cells in thesecretory endometrium(Fig. 5, CandD),
however AT1 receptor mRNA signal was maintained in the endometrialglands (Fig. 5, C andD). Endometrial blood ves-selsshowed intenseexpressionfor AT1 receptormRNAand in situhybridization signalswerehigheraround theblood vessels
Figure3. Autoradiogramsofthedistribution of[125I]-ANGIIbinding
toadjacentsectionsof the human endometrium. Darkregionsindicate
highdensities of labeled receptors.A,totalbinding;B,bindinginthe
presenceoftheAT,receptor selectiveligandDuP753 (1.0
MM);
C, bindinginthepresenceof theAT2receptor selectiveligandPD123177(1.0 IsM); D, bindingin thepresenceof DuP753(1.0 jM) plus
PD123177(1.0 sM); E, bindinginthepresenceof unlabeled ANGII (1 sM).Notabene: the difference indensitybetween A and B represents AT1receptorspecific binding;difference indensitybetween A and C represents AT2receptorspecific binding;difference between A and D represents specificAT, plus AT2receptordensity,and the difference between D and E represents thenon-ATI/non-AT2novel ATrecognition
site.Bar,30mm.
comparedwiththesurroundingtissue.Thesignalwasexpressed in thearteriolar smoothmuscle,which isnotsurprisingas vas-cular smooth muscle isacriticaltargetorganforANGII.There
4
3
2
1
0
U)
0 4
3
2
,E
11
4
3
2
T
I
C . 0e 0.0 E -0Xm
* * en
Eg
I
EP MP LP ES MS LS
n=1 n=4 n=7 n=3 n=6 n=6
C EP MP n=1 n=4 0.5- 0.4- 0.3- 0.2-0.1 -0.0 -B
-F
T
77Km..'
EP MP LP ES MS
n=1 n=4 n=7 n=3 n=6
-I
LS n=6
Figure4. Thelevels ofspecific[125I]-ANGH bindingsites in
endome-triumthroughout the menstrualcycles.Data represent the mean(±SEM)
ofAT,(A), AT2 (B), andnon-AT1/non-AT2(C)specific binding in
endometrial specimensfromwomen with normal menstrualcycles.
Sig-nificantly higher
levelsofAT1
andAT2specific binding
in theearly
secretory phase endometriumweredetectedcomparedwith those in the
endometria ofthemidproliferative, midsecretory,and latesecretory
phase. There was nosignificantdifference in theAT1orAT2specific
binding
levels between theearly
secretory
and lateproliferative.
EP,early
proliferative;MP,midproliferative;LP,lateproliferative;ES, early se-_ _ _ ~~~~~cretory;MS,midsecretory;LS,latesecretory. *P<0.05;**P<0.01; LP ES MS LS ***P<0.001 (one-way ANOVA followed byDunnett'sttest,compared
n=7 n=3 n=6 nf=6 with the level in the earlysecretoryendometrium).
was negligible hybridization signal in RNAse pretreated sec-tions(Fig. 5,E andF).Thespecificityof thesignalwasfurther confirmedby incubatinglabeledprobe alongwith 100-fold ex-cessof unlabeledprobe,which demonstrated that the
hybridiza-tionsignalwassignificantly lowercomparedwithhybridization inducedbythe labeledprobe alone(Fig. 5, G andH). Intense
hybridization signalwas seenin the basalisregionof the endo-metrium (Fig. 6,A and C) in and around the blood vessels. Low levels of expression of AT1 receptor mRNA were also seen in the myometrium with strong hybridization signal in isolated cells(Fig. 6,B andD).
Discussion
There isincreasingevidence inmanytissues, includinghuman ovaries(23)and human endometrium(24),tosuggestthe
exis-tence oflocally active tissuerenin-angiotensin systems (25).
Arecentstudyshowed that the stromal cells of human endome-triumarecapableofsynthesizingarenin-likesubstance. Renin
productionwassignificantlyincreasedbyprogesteroneand the maximumpotential to producetheangiotensin coincided with
the triggering events of menstruation (24, 26), indicating a
potential function in the control of menstruation. The present study clearlydemonstratesthat ANG II-like immunoreactivity
in human endometrium undergoes cyclic changes throughout
themenstrualcycle. Although immunocytochemical technique
isnotabletodifferentiate between the sites of synthesis ofANG
II and ANGHboundtoitsreceptors, the intensecytoplasmic immunostaining suggests that the immunoreactivity is largely
due toendogenous levels of ANGH within the cells. Support
for localrenin-angiotensin systemis alsoprovided bythe
find-ingsofRajuand Lee(27) showingthepresenceof
immunoreac-tive renin in endometrialcells, theenzymeresponsible for the conversion ofANGto ANGI. However, immunostaining due
to ANGIIboundtoreceptorscannotbe ruledout.
In thisstudy, stainingwaslocalized in theglandular epithe-lium and thestromaintheproliferative phase of themenstrual
cycle,withnegligible stainingaround the vascular endothelium. Incontrast,thestaining intensified around the bloodvessels of the endometrium in the secretory phase. These findings are in contrast to the results ofRaju and Lee (27) who showed
immunocytochemical localization ofrenin in thestromaand the
glands ofsomeof the sections ofproliferative endometrium but nostaininginsecretoryendometrium;thenegative reactivityin these sectionsmayrelatetoloss of antigenicity duetoformalin fixation. This issupported byourpreliminary results using fro-zen sections whichshow that renin immunoreactive proteinis
present in bothproliferative and secretory endometrium (our unpublished data).
Thecyclic changesinANG 11-likeimmunoreactivity imply
that thisoctapeptidemayplayadual role in the endometrium:
(a) as a controller of the uterine vascularbed and (b) as an
angiogenic and mitogenic mediator in the regeneration of the endometrium afterendometrial shedding. During the early to
midproliferative phasewhenthe endometriumundergoes rapid
cell division, angiotensin-like immunoreactivity was confined
tothestromaand theglandular epithelium. Thestromaand the
glandular epithelium showed typical brown granular cyto-plasmic stainingwhile there was nostainingaround the blood vessels. Inaddition, in situhybridization studies revealed that AT1 receptormRNAwas highly expressed both in theglands and stromal cellsinproliferative endometrium,while
-Figure5.Insitu localization ofAT,receptor mRNA expression in human endometrium. Bright-field (A, C, E, and G) and dark-field (B, D, F, and H) photomicrographs of sections of endometrium after in situ hybridization with35S-labeledoligonucleotide directed againstAT,in proliferative endometrium (A and B) andsecretoryendometrium (C and D). The high level ofAT, mRNA expression is seen around the glandular epithelium (G) and the endometrial blood vessels (BV). E and F indicate RNAse-treated serial section showing no hybridization while G and H show significant reduction in hybridization signalafterincubationof labeled probe along with 100-fold excess of unlabeled probe. Bar, 150 am.
Asthere israpid capillary growth associated with endome-trial regeneration aftermenstruation,it ispossiblethat ANGII
may modulate cell growth and neovascularization directly or
through the expression of other growth factors or protoonco-genes.ANG II was reported tostimulatebasicfibroblastgrowth
Figure 6.Insitu localization ofAT1receptor mRNAexpression in myometrium. A shows a bright-field view of anendometrial-myometrial section
atthe endometrial-myometrial junction hybridized witholigonucleotide directed againstAT1;Cshows adark-field view ofthe samesection; B and D showbright-fieldviewofmyometrialsection.Arrows indicate blood vessels at theendometrial-myometrial junction (A) andthe intense hybridization of isolated cells (B and D) within the myometrium. Bar, 150
Am.
endothelial cell proliferation (29). Also, ANG H has been shown to stimulate the expression of platelet-derived growth factor (30) as well as growth-related oncogenes in cultured smooth muscle cells(31). ANG H-stimulated 42-44-kD tyro-sine-phosphorylatedproteins such as the mitogen-activated pro-tein kinase inrataortic smooth muscle cells (32) suggest that ANG H may be involved in cell proliferation as this enzyme family has been showntoplayapivotal role in the regulation of cell division (33, 34).
As ANG H is a potent vasoconstrictor, the very intense staining seen around the endometrial bloodYessels inthe late luteal phase suggests its possible role in the initiation of men-struationby the vasoconstriction of the spiral arteriole before menstruation.Although >60% ofspecific
["2I]
-ANGIIbind-ing
was associated with theAT2 receptor, 20% was due toAT,receptor. Usinginsitu hybridization studies, the expression ofAT1 receptormRNA insecretory endometrium was demon-strated inthe endometrial glands and in the endometrial blood vessels. Strong hybridization signal for AT1 receptor mRNA
was alsoseen in and around the blood vessels at the endome-trial-myometrial junction in late secretory endometrium, which suggeststhat ANGHmay contribute to the vasoconstriction of thespiral arterioles.Asimilar resultdemonstrating binding sites foriodinated endothelins was reported in humanuterus(35),
however, the effect of endothelins on uterine blood vessels is notknown. On the otherhand, ANGHwasreported to induce adose-dependent contraction of human uterine artery (36). The apparent lack ofautoradiographic sensitivity indetecting AT1 receptor may be due tolack of translation of theAT1 receptor mRNA to receptorprotein. More likely, it is due to the relatively lowdensity ofAT1receptorsin the endometrium. Some autora-diographic studies have failed to demonstrate ANGHbinding
topreglomerular vessels in rat, rabbit, and human kidney (37) while others have reported that there is a predominance ofAT1
receptors inpreglomerular vessels (38).
Thefunction of AT receptor subtypes in the uterus is not yetestablished and the demonstration of specific sites does not necessarily imply a physiological role. As mentioned earlier, ANGHinducedadose-dependent contractionof human uterine
artery (36), however, the study did not look at the receptor subtype involved. The expression of AT1 receptor mRNA in vascular smooth muscle wouldbe consistent with a
vasocon-strictor role which has been demonstrated inthefeto-placental circulation (39), and in the ovine fetus infusion of ANG II
menstrual bleeding and hemostasis isachieved by vasoconstric-tion of the remaining basal arteriolar fragments (3). Although prostaglandinF2 and endothelins have been suggestedas puta-tiveendometrial vasoconstrictor candidates (5), neither pressor agent displays the cyclic changes in peptide orreceptor tobe apossible candidate for a "dual role peptide" as proposed by this study for the actions of ANG II in human endometrium. Theexpression of AT1 receptor mRNA around thebloodvessels and the intense localization of ANG TI-like immunoreactivity in the perivascular stromal cells around the blood vessels in late secretory endometrium supports a role of ANGIIin provid-ing the pressor mechanism to initiate menstruation and in the control of bleeding from the basal arteries.
The presence of clusters of cells (possibly inflammatory cells) among the myometrial cells, with substantially higher signal for AT, receptor mRNA than the surrounding cellular elements, is of some interest.Arecentstudy reported a similar expression of insulin-like growth factor and its receptors in
endometrium(41 ) and suggest that these cellsareinflammatory cells,possiblymonocytes and macrophages. Although the role of macrophages in uterine tissue has notyet been determined,
it is known that cocultures of endometrial stromal cells and macrophages stimulate stromal cell growth under serum-free conditions (41 ).
In addition to the control of vascular tone, the degree of menstrual bleeding isdetermined by endometrial regeneration
whichisessentialinthepreparationofareceptive endometrium for theimplantationofadeveloping embryo.The factors
con-trolling thisregenerationof the endometriumarepoorly under-stood. The AT2 receptors may promotecellgrowth ( 16)andare
highly expressed in human endometrium. Increasing evidence indicates thatestrogen-stimulated endometrial growth is medi-atedby locallyproduced growth factors (42). Increase in the intensity ofspecific
[1251I]AT
receptor binding from early and mid proliferative to late proliferative and reaching maximal levels in the early secretory endometriumimplies that the AT receptors may beupregulatedby estrogen in the first half of the menstrualcycle. The decrease in thespecific[1251I]
ATreceptorbinding in mid and latesecretory endometriumfromthelevels inearlysecretory endometrium suggests that progesterone may beantagonizingtheeffects of estrogen. The functional
signifi-cance of the high levels ofspecific
[1251]AT
receptorbinding duringthe lateproliferativeandearlysecretory endometrium isnotknown, but may relate tothepreparation ofendometrium forimplantation.
In addition to losartan- and PD123177-sensitive specific
[1251]
-ATreceptors,anadditionalhigh-affinity bindingsitewasconsistently detected thatwas insensitive to both losartan and
PD 123177 but was displaced with low concentration of unla-beled ANGII.This novelhigh-affinityATrecognitionsite rep-resented - 16% ofspecific total binding but itspotential
func-tionandpharmacologyremaintobe elucidated. It wasrecently proposedthat theangiogenic action of ANGIImay be mediated
via an unknown ATreceptor subtype ( 10). This non-AT1 /non-AT2 binding site in the endometrium mayrelate tothe factthat the endometrium undergoes extensive cyclic angiogenesisand warrantsfurtherinvestigation.
Acknowledgments
Weare grateful toDr.GF.. Steinfels(DuPont Merck Pharmaceuticals,
Wilmington, DE)for the giftsof losartan and PD123177. We would
like to thank the Consultants, Theatre, and Pathology staff of the Bir-mingham Maternity Hospital for their help with thisstudy.
This work was supported by grants from the Well Being project grant A2/94 and Mary Crosse. M. Shams is a post-doctoral research fellow funded by Well Being project grant A2/93.
References
1.Cameron,I.T. 1992. Medical management of menorrhagia. Curr. Obstet. Gynecol.2:136-140.
2. Markee, J. E. 1940. Menstruation inintraocular endometrial transplantsin the rhesus monkey. Contrib. Embryol. Carnegie Inst. 28:219-308.
3.Christiaens,G. C. M. L., J. J. Sixma, and A. A. Haspels. 1980. Morphology ofhaemostasisin menstrualendometrium.Br. J. Obstet.Gynaecol. 87:425-439. 4.Ludwig,H., H. Metzger, and M. Frauli. 1990.Endometrium:tissue remod-elling andregeneration.InContraceptionandMechanismsofEndometrial Bleed-ing. C.d'Arcangues,I. S. Fraser, J. R. Newton, and V. Odlind, editors. Cambridge UniversityPress, Cambridge, United Kingdom. 441-466.
5.Maigaard, S., A. Forman, and K.-E. Andersson. 1985. Different responses to prostaglandinF2 and E2 in human extra- andintramyometrialarteries. Prosta-glandins.30:599-607.
6.Cameron, I.T., and A. P. Davenport. 1992. Endothelins in reproduction. Rep. Med. Rev. 1:99-113.
7.O'Reilly,G., D. S.Charnock-Jones, A. T.Davenport,I.T.Cameron,and S. K. Smith. 1992. Presence of messenger ribonucleic acid for endothelin-1, endothelin-2, andendothelin-3in human endometrium and a change in the ratio of ET-A and ET-B receptor subtype across the menstrual cycle. J. Clin. Endocri-nol. & Metab.75:1545-1549.
8. Le Noble, J. L. M. L, J. W. M. Hekking, H. W. M.Van Straaten, D. W. Slaaf,and H. A. J.Struyker-Boudier.1991. Angiotensin II stimulates angiogenesis in the chorio-allantoic membrane of the chock embryo. Eur. J. Pharmacol. 195:305-306.
9. Fernandez, L. A., J. Twickler, and A. Mead. 1985. Neovascularization produced by angiotensinII.J.Lab. Clin. Med. 105:141-145.
10. Le Noble, F. A. C., N. H. J. S. Schreurs, H. W. M. Van Straaten, D. W. Slaaf, J. F. M. Smits, H. Rogg, and A. J. Struijker-Boudier. 1993. Evidence for a novel angiotensin II receptor involved in angiogenesis in chick embryo chorioallantoicmembrane. Am. J. Physiol. 264:R460-R465.
11.Baker, K. M., and J. F. Aceto. 1990. Angiotensin H stimulation of protein synthesis and cell growth in chick heart cells. Am. J. Physiol.259:H610-H618.
12. Wong, P. C., S. D. Hart, A. M. Zaspel, A. T. Chiu, R. J. Ardecky, R. D. Smith,and P. B. M. W. M.Timmermans. 1990. Functional studies of nonpeptide angiotensinII receptorsubtype-specificligands: Dup753(AII-i)and PD123177 (AII-2). J. Pharmacol. Exp. Ther.255:584-592.
13. Murphy, T. J., R. W. Alexander, K. K. Griendling, M. S. Runge, and K. E. Bernstein. 1991. Isolation of a cDNA encoding the vascular type- I angioten-sin II receptor. Nature(Lond.). 351:233-236.
14. Mukoyama, M., M.Nakajima,M.Horiuchi, H. Sasamura, R. E. Pratt, and V. J.Dzau. 1993.Expressioncloning oftype-2 angiotensinIIreceptor reveals a unique class of 7-transmembrane receptors. J. Biol. Chem.268:2539-2542.
15. Wong, P. C., A. T. Chiu, J. V. Duncia, W. F. Herblin, R. D. Smith, and P. B.M. W. M. Timmermans. 1992. AngiotensinII receptor antagonists and receptor subtypes. Trends Endocrinol. Metab. 3:211-217.
16. Bottari, S. P.,I. N. King, S. Reichlin, I. Dahlstroem, N. Lydon, and M.de Gasparo. 1992. TheangiotensinAT2 receptor stimulates protein tyrosin phosphatase activity and mediates inhibition of particulate guanylate cyclase. Biochem.Biophys. Res. Commun. 183:206-211.
17.Hendrickson,M. R., and R. L. Kempson. 1980. The normalendometrium: the base line. InSurgical Pathology of Uterus Corpus. M. R. Hendrickson and R. L.Kempson, editors. W. B.Saunders, Philadelphia. 36-98.
18.Noyes, R. W., A. T. Hertig, and J. Rock. 1950. Dating the endometrial biopsy. Fertil. Steril. 1:3-25.
19. Li, X. F., J. Gregory, and A. Ahmed. 1994.Immunolocalizationof vascular endothelial growthfactor in human endometrium. Growth Factors. 11:277-282. 20.Iwai, N., S. Y. Chaki, F. Konishi, S. Bardhan, C. Tibbetts, K. Sasaki, M. Hasegawa, Y. Matsuda, and T.Inagami. 1991. RatangiotensinIIreceptor cDNA sequence andregulationof the geneexpression.Biochem.Biophys. Res. Commun. 177:299-304.
21.Bergsma, D. J., C. Ellis, C. Kumar, P. R. Nuthulaganti, H. Kersten, N. Elshourbagy, E. Griffin, J. M. Stadel, and N. Aiyar. 1992. Cloning and character-ization of a humanangiotensinII type I receptor. Biochem. Biophys. Res. Commun. 183:989-995.
22. Shams, M., and A. Ahmed. 1994. Localization of mRNA for basic fibro-blastgrowthfactor in humanplacenta.Growth Factors. 11:105-111.
23. Palumbo, A. C., A. Jones, M. L. Lightman, M. L. Carcangiu, A. H. DeCherney,and F. Naftolin. 1989.Immunohistochemical localization of renin and angiotensin II in human ovaries. Am. J. Obstet. Gynecol. 160:8-14.
concentrationsin humanplasma andendometriumduringthe normal menstrual cycles.Br.J.Obstet. GynecoL 87:875-882.
25. Campbell, D. J.1987.Circulating andtissueangiotensin systems.J.Clin. Invest.79:1-6.
26.Shah, S. M., K.Higuchi,T.Inagami,and K. G. Osteen.1991.Effectof progesteroneonrenin secretion inendometrial stromal,chorionic trophoblast, and mesenchymalmonolayercultures.Am.J.Obstet. GynecoL 164:1145-1150.
27.Raju, G. C., and Y. S. Lee. 1989.Immunohistochemicaldemonstrationof renin in theendometrium.Ann.AcaLMed.18:345-347.
28.Stirling,D., R. R.Magness,R. Stow, M. R. Waterman, and E. R. Simpson. 1990. Angiotensin II inhibits leutenising hormone-stimulated cholesterol size chaincleavageexpressionand stimulatesbasicfibroblastgrowthfactorexpression in bovine leuteal cells inprimaryculture. J.Biol.Chem.265:5-81.
29. Ahmed, A., R. Plevin,M.Shaobi,S. A.Fountain, R. A.Ferriani, and S. K.Smith. 1994.Basic FGF activatesphospholipaseD inendothelialcells in the absenceofinositol-lipid hydrolysis. Am.J.PhysioL 266:C206-C212.
30.Bobik,A., S.Grinpukel,P.J.Little,A.Grooms,and G. Jackman. 1990. Angiotensin II and noradrenaline increase PDGF-BB receptors andpotentiate PDGF-BB stimulatedDNAsynthesisin vascular smooth muscle.Biochem Bio-phys.Res. Commun. 166:580-588.
31. Taubman,M.B., B.C. Berk,S.Izumo,T.Tsuda,R. W.Alexander, and R.Nadal-Ginard. 1989. Angiotensin II induces c-fosmRNA in aortic smooth muscle.J.BioLChetnL 264:526-530.
32. Molloy,C.J.,D. S. Taylor, and H. Weber. 1993.AngiotensinII stimula-tionofrapid protein tyrosine phosphorylationandproteinkinase activation in rat aorticsmooth muscle cells. J.BioLChem.268:7338-7345.
33. Pulvener, B. J., J. M. Kymakis,J. Avruch, E. Nikolakaki, and J. R. Woodgeti. 1991. Phosphorylation of c-jun by MAP kinase.Nature (Lond.). 353:670-674.
34. Wood, K. W., C. Sarnecki, T. M. Roberts, andJ. Blenis. 1992. Ras mediates nerve growthfactor receptor modulation ofthreesignal-transducing protein kinases: MAP kinase,Raf-landRSK. CelL 68:1041-1050.
35. Davenport, P. A.,I.T.Cameron, S. K.Smith, andJ. M. Brown.1991. Binding sites for iodinated endothelin-1,endothelin-2andendothelin-3 demon-strated on human uterineglandular epithelialcellsbyquantitative high-resolution autoradiography. J.EndocrinoL 129:149-154.
36. Poulsen,H.,N.Sjoberg,M.Stjernquist,and E. Zia. 1994. Atrial natriuretic peptide antagonizesthe contractile effect of angiotensinII in the humanuterine artery. Hum Reprod.(O4).9:1939-1943.
37. Grone, H. J., M.Simon,and E.Fuchs. 1992. Autoradiographic character-izationof angiotensin receptor subtypes infetaland adult humankidney.Amn J.
PhysioL 262:F326-F331.
38. DeLeon,H., andR.Garcia.1993.AngiotensinII receptorsubtypesin rat renalpreglomerularvessels. Receptor.2:253-260.
39.Adamson, S. L., R. J. Morrow, S. B.Bull, andB. L. Lurigille. 1989. Vasomotorresponses of theumbilical circulationinfetal sheep. Arm J.PhysioL 256:R1056-R1062.
40.Clark,K.G.,G.I.Irion, and C.E. Mack. 1990.Differential responses of uterine and umbilical vasculatures toangiotensinII andnorepinephrine.AmLJ.
PhysioL259:H197-H203.
41.Tang, X.-M.,M.J.Rossi,B. J. Masterson, and N.Chegini. 1994. Insulin-likegrowth factorI(IGF-I),IGF-I receptorsand IGF binding proteins 1-4 in human uterine tissue: tissuelocalizationandIGF-I action inendometrial stromal andmyometrial smoothmusclecellsinvitro. BioLReprod.50:1113-1125.