R

Ian

CU

a

ARRA

KESPAGT

1

tAmpfiIbghs

i

0d

Annals of Anatomy 193 (2011) 43–55

Contents lists available at ScienceDirect

Annals of Anatomy

journa l homepage: www.e lsev ier .de /aanat

ESEARCH ARTICLE

mmunohistochemical analysis of steroid receptors, proliferation markers,poptosis related molecules, and gelatinases in non-neoplastic andeoplastic endometrium

ornelia Amalinei, Corina Cianga, Raluca Balan, Petru Cianga, Simona Giusca, Irina-Draga Caruntu ∗

niversity of Medicine and Pharmacy “Gr. T. Popa” Iasi, 16 Universitatii Street, Iasi 700115, Romania

r t i c l e i n f o

rticle history:eceived 1 August 2008eceived in revised form 25 August 2010ccepted 28 September 2010

eywords:ndometrial carcinogenesisteroid receptorsroliferationpoptosiselatinasesIMP

a b s t r a c t

Endometrioid endometrial carcinoma developed from endometrial hyperplasia is associated withanomalies of proliferation, apoptosis, and matrix metalloproteinase (MMP) expression. Our study wasdesigned to investigate steroid receptor (ER, PR) expression and its correlation with proliferative activity(PCNA), apoptosis (Fas, FasL, Bcl-2, Bax, and p53), gelatinases (MMP-2 and MMP-9) and their tis-sue specific inhibitor (TIMP-1 and TIMP-2) immunoexpression in endometrial carcinogenesis. A totalof 38 cases were investigated, 10 non-neoplastic, 11 hyperplastic, and 17 carcinomatous endome-tria. Immunolabeling showed a higher expression of steroid receptors in hyperplasia and carcinomathan in non-neoplastic endometria and an ER/PR imbalance in carcinoma. The epithelial component ofendometrial carcinomas had the highest proliferative index. Bcl-2 had a stronger expression in hyper-plasia and carcinoma compared to non-neoplastic endometria and stromal tissue. The Bcl-2/Bax ratiowas lower in endometrial carcinoma. Fas and FasL expression was stronger in hyperplasia and fur-thermore in carcinoma. p53 expression was progressively stronger along the sequence non-neoplasticendometrial to hyperplasia-carcinoma. Both types of investigated MMPs showed an increased expres-sion in neoplastic endometria reaching a maximum level in carcinomas. MMP-9 immunostaining could

be correlated to myometrial invasion. TIMP-1 decreased and TIMP-2 increased in expression fromnon-neoplastic endometria to hyperplastic and carcinomatous endometrial, respectively. Our studydemonstrates that coordinated anomalies of steroid receptors, apoptosis and invasiveness factors arealready present in hyperplasia as cumulative steps along the way to malignant transformation andthat a complex MMP-2, MMP-9, TIMP-2/TIMP-1 imbalance seems to be responsible for the endometrial proliferation.

. Introduction

Endometrial carcinomas have the highest incidence amonghe malignancies of the feminine genital tract (Bockman, 1983).ccording to knowledge of epidemiologic, histopathologic, andolecular events endometrial carcinogenesis follows two different

athways. In the dualistic model of endometrial carcinogenesis therst pathogenetic type (endometrioid endometrial carcinoma, type

), representing 65–80% of all endometrial cancers, is characterized

y the following features: obesity, hyperlipidemia, hyperestro-enism (revealed by endometrial and ovarian stromal hyperplasia),ighly and moderately differentiated tumors (G1 and G2 in 82.3%),uperficial invasion of the myometrium (69.4%), high sensitivity to

∗ Corresponding author.E-mail addresses: cornelia.amalinei@histo-m.umfiasi.ro,

rina caruntu@yahoo.com (I.-D. Caruntu).

940-9602/$ – see front matter © 2010 Elsevier GmbH. All rights reserved.oi:10.1016/j.aanat.2010.09.009

© 2010 Elsevier GmbH. All rights reserved.

progestogens (80.2%), and favorable prognosis (5-year survival rateof 85.6%) (Bockman, 1983). The pathogenic correlation betweenG1 endometrial endometrioid carcinoma and atypical hyperplasiais currently sustained by numerous immunohistochemical stud-ies that demonstrate analogous molecular markers expression(Mutter, 2000). The pathogenetic proposal of endometrial carcino-genesis is that an accumulation of gene mutation is associatedwith morphologic features (Ronnett et al., 2002). A multi-stependometrial carcinogenesis involving coordinated intervention ofhormonal regulation, gene mutation, adhesion molecules, apop-tosis, imbalance between metalloproteinases (MMPs) and tissueinhibitors of MMPs (TIMPs) is currently accepted (Graesslin et al.,2006). The current research objectives in endometrial carcinogen-esis are the identification of characteristic molecular markers and

the analysis of the correlation between their expression and pro-gression of precursors or aggressiveness of carcinomas.

Estrogens are direct promoters of endometrial carcinogenesisthrough stimulation of rapid proliferation of epithelial cells, estro-

4 of Ana

gc12cfc

e(

acae

i(ive(oMeOeietrsgmbeai2Rb2ef

tacgTw2iEidre

snocfiMm

cells ± standard deviation (SD) in each histological group.PCNA estimation of the proliferative activity was performed by

choosing ten fields with maximum staining intensity, 100 epithelialcells being selected in each field. The PCNA index was expressed aspercent of positive cells (Yokoyama et al., 1998).

Table 1Antibodies used for immunohistochemical staining.

Marker Clone

Mo-a-hu ER� 1D5, Dako, DenmarkMo-a-hu PR PgR636, Dako, DenmarkMo-a-hu PCNA PC10, Dako, DenmarkMo-a-hu Bcl-2 124, Dako, DenmarkRa-a-hu Bax A3533, Dako North America, USAMo-a-hu Fas DX2, Dako, DenmarkMo-a-hu FasL NOK1, BD Biosciences Pharmingen

4 C. Amalinei et al. / Annals

en receptors (ERs) being overexpressed both in hyperplasia andarcinoma, in epithelial and stromal cell populations (Jazaeri et al.,999; Isaka et al., 2003; Atasoy and Bozdogan, 2006; Giuffre et al.,006). Progesterone receptors (PRs) and ERs show a generallyoncordant expression as a result of their interrelations and theirunctional status may influence the development of endometrialarcinoma (Ito et al., 2006, 2007).

Proliferating cell nuclear antigen (PCNA) also has an increasedxpression both in endometrial hyperplasias and carcinomasMitselou et al., 2003).

Apoptosis is crucial in human endometrium turnover. Alter-tions of apoptosis play an important role in endometrialarcinogenesis being revealed by Bcl-2/Bax, Fas/Fas ligand (FasL)nd p53 protein systems functionality (Mitselou et al., 2003; Zhangt al., 2004).

Invasive capacities defining any malignancy result from thentervention of proteases that degrade the extracellular matrixPuente et al., 2003). The matrix metalloproteinases (MMPs) arenvolved both in the endometrial physiology (characterized byariable MMPs expression along the endometrial cycle) (Freitast al., 1999; Henriet et al., 2002; Cornet et al., 2005) and pathologyGoffin et al., 2003). The endometrium expresses a large spectrumf MMPs: MMP-1, MMP-2, MMP-3, MMP-7, MMP-9, MMP-23,MP-26, and MT1-MMP (membranar type 1-MMP) (Velasco

t al., 1999; Henriet et al., 2002; Chegini et al., 2003; Curry andsteen, 2003; Li et al., 2004), being produced both by stromal andpithelial carcinomatous cells (Turpeenniemi-Hujanen, 2005). Thenterpretation of MMPs expression is difficult as remodelling of thextracellular matrix (ECM) is an essential feature in the menstrualurnover of the human endometrium (Cornet et al., 2005). Variousesearch methods have detected MMP-9 in endometrium with aignificant increase in menstruation. The stromal fibroblasts, thelandular epithelium, the endothelium, the leukocytes, and theacrophages localised in areas of collagen fibers disruption had

een identified as the cells responsible of its production (Cornett al., 2005; Curry and Osteen, 2003). Both MMP-2 (A gelatinase)nd MMP-9 (B gelatinase) are correlated to tumor growth andnvasion in endometrial and cervical neoplasia (Egeblad and Werb,002; Hojilla et al., 2003; Isaka et al., 2003; Lambert et al., 2003;ubin, 2003). However, an opposite effect of gelatinases is obtainedy in vitro generation of antiangiogenic polypeptides (Pozzi et al.,000) by cleaving plasminogen to an angiostatin fragment (Lucast al., 1998) and collagen type XVIII to an endostatin-containingragment (Heljasvaara et al., 2005).

The tissue inhibitors of metalloproteinases (TIMPs) are fourypes of natural inhibitors of MMPs: TIMP-1, TIMP-2, TIMP-3,nd TIMP-4. TIMPs partially regulate MMP activity by formingomplexes with their active forms and with their precursors (pro-elatinases) (TIMP-1 with proMMP-9 and TIMP-2 with proMMP-2).IMP-1 and TIMP-2 can form complexes with active MMP-9 andith both forms of MMP-2 (active and latent) (Määtä et al.,

000). High levels of TIMPs 1–3 presumably maintain tissuentegrity and ECM homeostasis in non-malignant endometrium.strogen and progesterone modulate MMP and TIMP expressionn human cycling endometrium. The progesterone effect is theown-regulation of gelatinases expression. Previous research haseported a high MMP-2, MMP-9, and TIMP-2 co-expression inndometrial adenocarcinoma (Määtä et al., 2000).

Our research is designed as a correlative investigation ofeveral types of markers characteristic for the transition fromon-neoplastic to neoplastic endometrium. Our study was based

n the main molecular events along the pathway of endometrialarcinogenesis being focused on the immunohistochemical pro-les of ER-PR, PCNA, Bcl-2/Bax, Fas/FasL, p53, MMP-2/TIMP-2, andMP-9/TIMP-1, and has attempted to demonstrate that severalalignant anomalies are already detectable in hyperplasia.

tomy 193 (2011) 43–55

2. Materials and methods

A total of 38 cases selected from the files of the IIIrd Obstetricsand Gynecology University Clinic of Iasi, Romania were included inour study, with the approval of the Ethics Committee of the Uni-versity of Medicine and Pharmacy “Gr. T. Popa”: 10 non-neoplasticendometria, with patient’s ages ranging from 33 to 55 years (mean45), 11 endometrial hyperplasias, with patient’s ages ranging from29 to 47 years (mean 43), and 17 endometrial carcinomas, withpatient’s ages ranging from 34 to 64 years (mean 50).

The endometrial fragments were routinely processed, paraffin-embedded, and sectioned at 4 �m. After deparaffinization antigenretrieval was performed by heating the sections for 20 min, in Anti-gen Retrieval Solution (Dako, Denmark), pH 6.10, at 96 ◦C, in awater-bath. This was followed by incubation in 3% H2O2, for 5 min,at room temperature (RT) to quench endogenous peroxidase. Non-specific binding sites were blocked by incubating the slides for 1 h,in 50 mM Tris–HCl (pH 7.4), containing 10% normal goat serum and1% BSA. The histological sections were incubated overnight at 4 ◦Cwith the primary antibodies that were used to reveal in humanendometrial tissues several molecules (Table 1): estrogen receptoralpha (ER�), PR, PCNA, Bcl-2/Bax, Fas/FasL systems, p53, MMP-2, MMP-9, TIMP-1, and TIMP-2. Specific reactions were detectedby successive incubations with a second biotinylated antibody(LSAB kit, Dako, Denmark) 30 min, at RT with enzymatic complexstreptavidin-peroxidase (30 min, at RT) and 3,3′-diaminobenzidinetetrahydrochloride (DAB) extemporaneously activated with 3%H2O2, for 30 min, at RT. The following step was the counterstainingwith Meyer’s haematoxylin. Negative controls were represented bysections from the same tissues incubated with a mixture of mouseisotypes (Universal Negative Control for N-Series Mouse PrimaryAntibodies-N1698, Dako, Denmark). Positive reactions were visu-alized as brown precipitates with nuclear pattern for ER�, PR, PCNA,and p53, nuclear or cytoplasmic labeling for Bcl-2 and cytoplasmicor membrane labeling for Bax, Fas, FasL, MMP-2, MMP-9, TIMP-1,and TIMP-2.

The IHC score was determined by analysing two variables: thepercentage of immunostained cells and the average staining inten-sity, considering the amount of positive cells as the most importantcriterion.

2.1. Semi-quantitative analysis

ER� and PR expressions were quantified as the percentage ofpositive cells per 1000 cells in each section (Shiozawa et al., 1996)and results were recorded as the mean percentage of positive

Mo-a-hu P53 DO-7, Dako, USAMo-a-hu MMP-2 D2705, Santa Cruz, BiotechnologyMo-a-hu MMP-9 K1504 Santa Cruz, BiotechnologyMo-a-hu TIMP-1 63515, R&D Systems, Minneapolis, USAMo-a-hu TIMP-2 89025.11, R&D Systems, Minneapolis, USA

of Ana

ci

m4

2

pd

vuw

3

cani

lastic

ial (%)

13.505.6617.2410.749.4315.0414.5518.8711.0518.7414.6617.88

enppwAts(ccip

n(nw((h

vpd

C. Amalinei et al. / Annals

The expressions of Bcl-2, Bax, Fas, FasL, and p53 molecules wereonsidered as positive when >10% of 1000 examined cells weremmunostained (Yamashita et al., 1999).

The percentage of positive cells was registered for approxi-ately 1000 cells per slide, subdivided into 10 selected fields at

00× magnification for MMPs and TIMPs (Graesslin et al., 2006).

.2. Statistical analysis

ANOVA one way (multiple comparison Tukey) was used to com-are data obtained by immunohistochemistry in the three differentiagnostic groups.

Student’s t-test (TTEST) was used to determine whether twoalues were likely to originate from the same two underlying pop-lations that had the same mean. For each marker a p value ≤ 0.05as considered statistically significant.

. Results

The results obtained by quantification of immunohistochemi-al expression of investigated markers by mean values (±SD) incomparative manner for non-neoplastic, hyperplastic, and carci-omatous endometrium in epithelial and stromal components are

llustrated in the following table:

Marker HistopathologyNon-neoplastic endometrium Hyperp

Epithelial (%) Stromal (%) Epithel

ER� 38.00 ± 19.32 31.00 ± 17.28 59.54 ±PR 66.50 ± 9.44 58.50 ± 21.08 87.09 ±PCNA 44.00 ± 19.26 36.50 ± 17.64 65.45 ±Bcl-2 27.50 ± 7.54 15.00 ± 5.27 36.36 ±Bax 31.00 ± 8.43 16.00 ± 6.14 49.09 ±Fas 37.50 ± 16.37 20.00 ± 14.33 43.18 ±FasL 23.00 ± 14.37 17.20 ± 20.18 32.27 ±P53 13.00 ± 15.84 9.60 ± 8.14 33.18 ±MMP-2 20.50 ± 15.89 12.50 ± 10.06 40.45 ±MMP-9 25.00 ± 12.90 17.50 ± 12.30 41.81 ±TIMP-1 49.50 ± 12.34 37.00 ± 12.06 35.00 ±TIMP-2 46.00 ± 9.66 32.00 ± 8.88 60.00 ±

ER� (Fig. 1a–d) showed a constant expression in non-neoplasticndometrium (Fig. 1a) both in the epithelial and stromal compo-ents, registering a slightly stronger stromal reaction in disorderedroliferative phases. ER� had a stronger expression in both com-onents of hyperplasias (Fig. 1b), exhibiting a significant differencehen compared to non-neoplastic endometria (p < 0.05) (Fig. 1d).relatively reduced ER� expression was noticed mainly inside

he epithelial component of functional endometria diagnosed asecretory phase, in disordered proliferative phase, in hyperplasian = 4 cases; 36.36%), and in G2-3 endometrial endometrioid car-inoma (47.05% of carcinomas). ER� was variably expressed inarcinomatous endometrium (Fig. 1c), being slightly stronger thann hyperplastic endometrium in both epithelial and stromal com-onents.

Generally, PR expression (Fig. 2a–d) was strong in non-eoplastic endometria both in epithelial and stromal componentsFig. 2a). PR expression was markedly increased in both compo-ents in hyperplasia (Fig. 2b), registering significant differenceshen comparing hyperplastic with non-neoplastic endometria

p < 0.05) (Fig. 2d). PR was moderately increased in carcinomasFig. 2c) with a slightly decreased reaction when compared to

yperplasia.

Our results showed that PCNA expression (Fig. 3a–d) wasariable in non-neoplastic endometria (Fig. 3a), higher in both com-onents of hyperplasias (Fig. 3b) and carcinomas, with a significantifference between non-neoplastic and carcinomatous endometria

tomy 193 (2011) 43–55 45

endometrium Carcinomatous endometrium

Stromal (%) Epithelial (%) Stromal (%)

52.27 ± 18.48 62.05 ± 20.39 63.52 ± 19.5085.45 ± 6.50 70.17 ± 25.07 65.29 ± 27.4657.27 ± 17.65 67.64 ± 16.30 64.41 ± 14.8828.18 ± 15.21 37.94 ± 15.91 21.76 ± 11.3139.09 ± 12.61 53.52 ± 22.62 43.52 ± 20.5223.63 ± 12.66 54.41 ± 15.99 37.35 ± 15.6220.45 ± 17.24 42.64 ± 17.33 35.29 ± 13.7418.63 ± 12.86 39.41 ± 23.37 24.41 ± 14.4532.72 ± 14.38 49.70 ± 20.42 35.58 ± 17.2138.63 ± 20.74 47.94 ± 19.20 48.82 ± 21.0320.63 ± 10.55 29.70 ± 11.92 21.17 ± 12.1841.81 ± 18.61 69.11 ± 20.25 48.23 ± 12.11

(p < 0.05) (Fig. 3d). The highest proliferative index was registered inthe epithelial component of carcinomas (Fig. 3c).

Bcl-2 and Bax expressions (Figs. 4a–d and 5a–d) were generallyreduced, with a significant difference between the two compo-nents for both molecules in non-neoplastic endometria, in favourof epithelial immunostaining (p < 0.05) (Figs. 4a,d and 5a,d). Bcl-2and Bax showed a slightly increased immunolabeling in epithe-lial component of neoplastic endometria. Bcl-2 and Bax expressionshowed significant differences between non-neoplastic, hyperplas-tic (Figs. 4b and 5b), and carcinomatous endometria (Figs. 4c and 5c)(p < 0.05) (Figs. 4d and 5d). The stromal reactions of Bcl-2 and Baxwere correlated to that of epithelial reactions.

A significant difference between epithelial and stromal expres-sion of Bcl-2 was noted in carcinomatous endometria (p < 0.05). Aslightly decreased expression of Bcl-2 in carcinomas vs. hyperplasiawas observed.

A significant difference for Bax expression between the twocomponents of hyperplastic endometria (p < 0.05) (Figs. 4d and 5d)was also noted.

Fas and FasL molecules showed a generally low immunolabel-ing (Figs. 6a–d and 7a–d). However, Fas expression was statisticallyhigher in epithelial component of all types of studied endometria(p < 0.05). Fas was higher in the epithelial component of neoplasticendometria with significant differences (p < 0.05) (Fig. 6d) between

non-neoplasia (Fig. 6a) and hyperplasia (Fig. 6b) and betweenhyperplasia-carcinoma. Fas positivity was slightly stronger in car-cinomatous epithelium (Fig. 6c) than in its stromal counterpart.Fas stromal expression was stronger in neoplastic endometria. Fasexpression showed significant differences (p < 0.05) when com-paring the mean values of non-neoplastic to hyperplastic andhyperplastic to carcinomatous endometria (Fig. 6d).

FasL showed an increased immunoexpression from the epithe-lium of non-neoplastic endometrium (Fig. 7a) to hyperplasia(Fig. 7b) and carcinoma (Fig. 7c). Non-neoplastic FasL expres-sion was significantly different (p < 0.05) from carcinomatousendometria (Fig. 7d). Although reduced, the stromal FasL expres-sion showed significant differences both between non-neoplasticto hyperplastic and hyperplastic to carcinomatous endometria(p < 0.05) (Fig. 7d). FasL immunolabeling was higher in the epithelialcomponent of hyperplastic endometria (p < 0.05) (Fig. 7d).

p53 expression (Fig. 8a–d) was reduced in both compo-nents of non-neoplastic endometria (Fig. 8a) and showed anincrease along the sequence non-neoplastic–hyperplastic (Fig. 8b)

–carcinomatous endometria (Fig. 8c). Significant differences werenoted between non-neoplastic and hyperplastic epithelial p53expression and between non-neoplastic–hyperplastic and non-neoplastic–carcinomatous stromal p53 immunostaining (p < 0.05)(Fig. 8d). A significant difference between epithelial and stromal

46 C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55

Fig. 1. ER immunoexpression in endometrium. (a) ER expression in proliferative endometrium (×100). (b) ER strong stromal expression in simple endometrial hyperplasia( a (×1h l (S) enr vs. 3E)(

pc

ircbMcemcc

nw(trhncsie

eos

×100). (c) ER� strong epithelial expression in endometrial endometrioid carcinomyperplastic (2) and carcinomatous (3) groups, both inside epithelial (E) and stromaespectively; *I corresponds to p 0.0005 (1S vs. 3S), *F corresponds to p 0.0062 (1EANOVA one way, multiple comparison Tukey).

53 expression was noted when comparing hyperplastic and car-inomatous endometria immunostaining (p < 0.05) (Fig. 8d).

Both types of investigated MMPs showed a similar aspect ofmmunolocation (Figs. 9a–d and 10a–d). MMP-2 had a relativelyeduced expression in non-neoplastic endometria (Fig. 9a), espe-ially inside the stromal component, with significant differencesetween stromal and glandular expression (p < 0.05) (Fig. 9d).MP-2 increased in expression in hyperplastic (Fig. 9b) and car-

inomatous endometria (Fig. 9c). Significant differences in MMP-2xpression were observed between non-neoplastic and carcino-atous endometria (p < 0.05) (Fig. 9d). Although MMP-2 positive

arcinomatous stromal cells were constantly observed, a signifi-antly stronger epithelial reaction was noted (p < 0.05) (Fig. 9d).

MMP-9 had also a stronger expression in the epithelial compo-ent of non-neoplastic endometria showing significant differenceshen compared to its stromal expression (Fig. 10a) (p < 0.05)

Fig. 10d). MMP-9 increased in expression from non-neoplastico hyperplastic andometria (Fig. 10b) and carcinoma (Fig. 10c)espectively. The differences between the MMP-9 expressions inyperplastic and carcinomatous endometrium compared to theon-neoplastic endometrium were significant (p < 0.05) in bothomponents (Fig. 10d). MMP-9 had a tendency of similar expres-ion in both components of hyperplasia, with a slightly strongermmunostaining in carcinomatous stroma when comparing to

pithelial tumoral cells.

TIMPs (Figs. 11a–d and 12a–d) presented a significantly higherxpression inside the epithelial component in all three categoriesf investigated endometria (p < 0.05). TIMPs expression hadignificant differences between non-neoplastic–hyperplastic

00). (d) Immunohistochemical profile of ER� (mean %± SD) in non-neoplastic (1),dometrial components shows a progressive increase in hyperplasia and carcinoma

, *D corresponds to p 0.0097 (1E vs. 2E), and *G corresponds to p 0.0134 (1S vs. 2S)

or non-neoplastic–carcinomatous endometria (p < 0.05)(Figs. 11d and 12d).

TIMP-1 slightly decreased along the sequence non-neoplastic(Fig. 11a) –hyperplastic (Fig. 11b) –carcinomatous endometria(Fig. 11c) in both epithelial and stromal components.

Conversely, TIMP-2 increased in both components alongthe sequence non-neoplastic (Fig. 12a) –hyperplastic (Fig. 12b)–carcinomatous endometria (Fig. 12c). Significant differencesbetween TIMP-2 epithelial expression in non-neoplastic com-pared to hyperplastic or to carcinomatous endometria (p < 0.05)(Fig. 12d) were observed. Significant differences between TIMP-2stromal expression in non-neoplastic vs. carcinomatous endome-tria (p < 0.05) (Fig. 12d) were also noted.

4. Discussion

The pathogenesis of endometrial hyperplasia involves hyperoe-strogenism (Matias-Guiu et al., 2001) and occurs through severalalterations in cell physiology that includes self sufficiency to growthsignals, insensitivity to growth-inhibitory signals, defects in DNArepair, evasion of apoptosis, limitless replicative potential, sus-tained angiogenesis, tissue invasion, and metastasis (Atasoy andBozdogan, 2006). As a putative precursor lesion of endometrialendometrioid carcinoma, endometrial hyperplasia is expected to

show most if not all of these changes (Salvesen et al., 1999; Ioachin,2005; Giuffre et al., 2006).

The proliferation of both normal and carcinomatous endome-trial cells is stimulated by estrogens through their nuclear receptors(Nephew et al., 2000; Bircan et al., 2005). ER� isoform is considered

C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55 47

Fig. 2. PR immunoexpression in endometrium. (a) PR strong stromal expression in proliferative andometrium (×400). (b) PR strong expression in both epithelial and stromalcomponents in simple endometrial hyperplasia (×100). (c) PR strong epithelial expression in epithelial component of endometrial endometrioid carcinoma (×100). (d)I c (2) ac trium( s. 3E),T

tfisiH

(pdlE

islnwdcd2Te

veh

mmunohistochemical profile of PR (mean % ± SD) in non-neoplastic (1), hyperplastiomponents shows an increase in hyperplasia compared to non-neoplastic endome1S vs. 2S), *D corresponds to p 0.0450 (1E vs. 2E), *E corresponds to p 0.0151 (2E vukey).

o be more expressed in endometrial hyperplasia than ER� iso-orm (Pearce and Jordan, 2004; Hu et al., 2005), showing a gradualncreasing expression from normal proliferative endometrium toimple hyperplasia and furthermore to complex hyperplasia. Bothsoforms decrease in atypical hyperplasia (Bozdogan et al., 2002;u et al., 2005).

The relatively reduced ER� expression in the investigated casesespecially in the epithelial component) could reflect the initialroliferative or secretory phase, hormonal receptors anomalies inisordered proliferative phases, a heterogenous staining in irregu-

arly bleeding endometrium (Galant et al., 2004) or a dominance ofR� expression.

Both epithelial and stromal components showed a strong ER�n hyperplasias and in G1-2 endometrial endometrioid carcinomas,uggesting a similar mechanism of hormonal stimulation in bothesions. ER� had a variable, slightly increased expression in carci-omatous endometria when compared to hyperplastic endometria,ith a similar expression in the epithelial and stromal components,emonstrating stromal involvement in carcinogenesis. A possibleo-expression of both ER isoforms in our cases may partially explainifferences in our results compared to other reports (Neis et al.,000; Nunobiki et al., 2003; Uchikawa et al., 2003; Abe et al., 2006).he endothelial cells ER� positivity confirm the complex hormonal

ndometrial regulation.

Although a coordinated ER–PR expression is reported in pre-ious reports (Ito et al., 2007), the relatively discordant ER�–PRxpression registered in both components of non-neoplastic andyperplastic endometria suggests the relative excess of PR com-

nd carcinomatous (3) groups, both inside epithelial (E) and stromal (S) endometrialand a decrease in carcinoma compared to hyperplasia; *G corresponds to p 0.0027and *H corresponds to p 0.0092 (2S vs. 3S) (ANOVA one way, multiple comparison

plex, possible because of increased ER breakdown, decreased ERsynthesis, enhanced quantity of estradiol (E2) dehydrogenase,active estrogen sulphurylation or previous exogenous hormonalsupply.

As PR complex suppresses estrogen action by several mecha-nisms (Hale et al., 2002), the risk factors for endometrial cancerare recently discussed in terms of inadequate progesterone oppo-sition (Hale et al., 2002; Ito et al., 2007). The strong PR expression ofboth components in hyperplasia and a slightly decreased expres-sion in carcinoma when compared to hyperplasia correspond to thereported imbalance between progesterone–estrogen stimulationin carcinomatous endometria.

According to the literature data (Mohsin et al., 2004; Ito et al.,2007), the lowest PR expression was registered in G3 carcinomas,especially inside the epithelial component, corresponding to moreadvanced stages at diagnosis and supplementing data that considerPR an independent prognostic factor in endometrial carcinoma.

PCNA expression was observed both in glandular and stromalendometrial cells of investigated cases confirming previous lit-erature data (Cao et al., 2002). PCNA showed a heterogeneousstaining, exhibiting a variable positivity in neighbouring glands. Thesharply decreased PCNA expression in the secretory phase glandu-lar epithelium and strong expression inside stromal cells suggests

an involvement of positive cells in proliferation and indicates a dif-ferent growth-regulating mechanism. The parallel co-expressionof steroid receptors and PCNA noted both in disordered prolifera-tive phase and in simple endometrial hyperplasia may favour theincrimination of the previous as precursor of the latter. As expected,

48 C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55

Fig. 3. Proliferation activity in endometrium. (a) PCNA variable expression in normal proliferative endometrium (×100). (b) PCNA epithelial and stromal expression inendometrial hyperplasia (×400). (c) PCNA variable expression in G2 endometrioid endometrial carcinoma (×100). (d) Immunohistochemical profile of PCNA (mean % ± SD)i epithh *I corrc

Pcpcm

ceiObRdl2cs

em2csncctr

n non-neoplastic (1), hyperplastic (2) and carcinomatous (3) groups, both insideyperplasia and carcinoma respectively; *D corresponds to p 0.0152 (1E vs. 2E),orresponds to p 0.0144 (1S vs. 2S) (ANOVA one way, multiple comparison Tukey).

CNA showed an increased expression mainly inside the epithelialells corresponding to the advanced proliferative phase, in hyper-lasias, and even higher in carcinomas. PCNA overexpression in thearcinomatous invasion front was constantly observed, as a testi-ony of the increased aggressiveness of this type of malignancy.Apoptosis maintains cellular homeostasis by eliminating senes-

ent cells, during late secretory and menstrual phases of thendometrial cycle (Harada et al., 2004). Alterations of apoptosis,nvolving Bcl-2, Fas, and the caspase family (Yamashita et al., 1999;tsuki, 2001) play an important role in endometrial carcinogenesis,eing revealed by PTEN and p53 anomalies (Salvesen et al., 1999;ogers et al., 2000; Lax, 2004). Bcl-2 is an inhibitor of apoptosisuring the proliferative phase and its expression decreases in the

ate secretory and menstruating phases (Konno et al., 2000; Otsuki,001). Bax is a Bcl-2 family member that promotes cell death sus-eptibility. The levels of Bax protein increase dramatically in theecretory phase and induces endometrial cell turnover.

Both Bcl-2 and Bax showed a slightly increased expression inpithelial neoplastic endometrial cells, confirming their involve-ent in the progressive deregulation of apoptosis (Song et al.,

002). The slightly decreased expression of stromal Bcl-2 in car-inomas compared to that noted in hyperplasias demonstrates aupplementary repression of apoptosis in the investigated carci-

omatous endometria. The slightly increased Bax expression inarcinomatous compared to hyperplastic endometria may reflectommon mechanisms in tumoral progression by clonal selec-ion of the most aggressive cell populations. Although previousesearches had reported a lower Bcl-2/Bax ratio in endometrial car-

elial (E) and stromal (S) endometrial components shows a significant increase inesponds to p 0.00064 (1S vs. 3S), *F corresponds to p 0.0047 (1E vs. 3E), and *G

cinoma (Vaskivuo et al., 2002), only the stromal ratio was decreasedwhile the epithelial ratio was similar in the investigated cate-gories, supplementing data concerning strong stromal influencein endometrial carcinomas. Similar Bcl-2/Bax epithelial ratio inhyperplasia and carcinoma shows that apoptosis deregulation isa common anomaly both in endometrial carcinoma and precursorlesion.

Fas and FasL immunostaining showed their co-localization tothe apical membrane, with a higher staining of the epithelium.Fas expression in the non-neoplastic endometria correspondedto the previous reports of low expression in proliferative andearly secretory phases (Yamashita et al., 1999; Song et al., 2002;Harada et al., 2004). Reduced Fas and FasL molecules expres-sions may suggest the down-regulation of Fas-mediated apoptosisby possible generation of sFasL from different types of cells inthe investigated cases. Fas and FasL were more expressed in theepithelial component of neoplastic endometria both when com-paring hyperplastic to non-neoplastic endometrium and carcinomato hyperplasia, demonstrating the progressive intervention ofFas/FasL system in endometrial carcinogenesis. Stromal expressionwas constant, demonstrating continuous stromal cells inter-vention in carcinogenesis. FasL epithelial expression increasedfrom non-neoplastic endometrium to hyperplasia and carcinoma

and the stromal expression, although reduced, showed signifi-cant differences both between non-neoplastic–hyperplastic andhyperplastic–carcinomatous endometria. FasL expression couldsuggest that both positive epithelial cells and activated lympho-cytes are capable to induce apoptosis in Fas positive cells. Although

C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55 49

Fig. 4. Bcl-2 immunohistochemical expression in endometrium. (a) Bcl-2 expression in proliferative endometrium (×200). (b) Bcl-2 expression in a curettage specimendiagnosed as simple endometrial hyperplasia (×100). (c) Bcl-2 expression in G2 endometrioid endometrial carcinoma (×400). (d) Immunohistochemical profile of Bcl-2(mean % ± SD) between non-neoplastic (1), hyperplastic (2) and carcinomatous (3) groups, both inside epithelial (E) and stromal (S) endometrial components shows a slightlyprogressive increase in hyperplasia and carcinoma respectively; *C corresponds to p 0.0031 (3E vs. 3S), *D corresponds to p 0.0049 (1E vs. 2E), *I corresponds to p 0.0456 (1Svs. 3S), *F corresponds to p 0.0302 (1E vs. 3E), and *G corresponds to p 0.0185 (1S vs. 2S) (ANOVA one way, multiple comparison Tukey).

Fig. 5. Bax immunohistochemical expression in endometrium. (a) Bax expression in proliferative endometrium (×200). (b) Bax immunohistochemical expression in complexendometrial hyperplasia (×100). (c) Bax heterogenous expression in G1 endometrioid endometrial carcinoma (×100). (d) Immunohistochemical profile of Bax (mean % ± SD)between non-neoplastic (1), hyperplastic (2) and carcinomatous (3) groups, both inside epithelial (E) and stromal (S) endometrial components shows a slight progressiveincrease in hyperplasia and carcinoma respectively; *D corresponds to p 0.0001 (1E vs. 2E), *I corresponds to p 0.0009 (1S vs. 3S), *F corresponds to p 0.0105 (1E vs. 3E), and*G corresponds to p 0.0205 (1S vs. 2S) (ANOVA one way, multiple comparison Tukey).

50 C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55

Fig. 6. Fas immunohistochemical expression in endometrium. (a) Fas expression in proliferative endometrium (×100). (b) Fas expression in endometrial hyperplasia (×100).(c) Fas expression in G1 endometrial carcinoma (×200). (d) Immunohistochemical profile of Fas (mean % ± SD) in non-neoplastic (1), hyperplastic (2) and carcinomatous(3) groups, both inside epithelial (E) and stromal (S) endometrial components shows a progressive increase in hyperplasia and carcinoma respectively; *C corresponds to p0.0196 (3E vs. 3S), *B corresponds to p 0.0396 (2E vs. 2S), *F corresponds to p 0.0172 (1E vs. 3E), *H corresponds to p 0.0174 (2S vs. 3S), and *I corresponds to p 0.0080 (1S vs.3S) (ANOVA one way, multiple comparison Tukey).

Fig. 7. FasL immunohistochemical expression in endometrium. (a) FasL expression in proliferative endometrium (×200). (b) FasL expression in endometrial hyperplasia(×100). (c) FasL expression in G1 endometrial carcinoma (×200). (d) Immunohistochemical profile of FasL (mean % ± SD) in non-neoplastic (1), hyperplastic (2) and carci-nomatous (3) groups, both inside epithelial (E) and stromal (S) endometrial components shows a slightly progressive increase in hyperplasia and carcinoma respectively; *Fcorresponds to p 0.0044 (1E vs. 3E), *H corresponds to p 0.0272 (2S vs. 3S), and *I corresponds to p 0.0248 (1S vs. 3S) (ANOVA one way, multiple comparison Tukey).

C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55 51

Fig. 8. p53 immunohistochemical expression in endometrium. (a) p53 expression in proliferative endometrium (×100). (b) p53 expression in endometrial hyperplasia (×100).(c) 53 strong expression in G3 endometrioid endometrial carcinoma (×400). (d) Immunohistochemical profile of p53 (mean % ± SD) in non-neoplastic (1), hyperplastic (2)and carcinomatous (3) groups, both inside epithelial (E) and stromal (S) endometrial components shows a progressive increase to hyperplasia and carcinoma respectively;*D corresponds to p 0.0154 (1E vs. 2E), *F corresponds to p 0.0018 (1E vs. 3E), and *I corresponds to p 0.0022 (1S vs. 3S) (ANOVA one way, multiple comparison Tukey).

Fig. 9. MMP-2 immunohistochemical expression in endometrium. (a) MMP-2 expression in proliferative endometrium (×200). (b) MMP-2 expression in simple endometrialhyperplasia (×100). (c) MMP-2 epithelial expression in G1 endometrioid endometrial carcinoma (×200). (d) Immunohistochemical profile of MMP-2 (mean % ± SD) in non-neoplastic (1), hyperplastic (2) and carcinomatous (3) groups, both inside epithelial (E) and stromal (S) endometrial components shows a progressive increase in hyperplasiaand carcinoma respectively; *D corresponds to p 0.0044 (1E vs. 2E), *F corresponds to p 0.0004 (1E vs. 3E), *G corresponds to p 0.0014 (1S vs. 2 S), *I corresponds to p 0.0001(1S vs. 3S) (ANOVA one way, multiple comparison Tukey).

52 C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55

Fig. 10. MMP-9 immunohistochemical expression in endometrium. (a) MMP-9 expression in proliferative endometrium (×400). (b) MMP-9 expression in complexendometrial hyperplasia (×400). (c) MMP-9 expression in G1 endometrioid endometrial carcinoma (×100). (d) Immunohistochemical profile of MMP-9 (mean % ± SD)in non-neoplastic (1), hyperplastic (2) and carcinomatous (3) groups, both inside epithelial (E) and stromal (S) endometrial components shows a marked increase in hyper-plasia and carcinoma respectively; *D corresponds to p 0.0263 (1E vs. 2E), *F corresponds to p 0.001 (1E vs. 3E), *I corresponds to p 0.007 (1S vs. 3S), and *G corresponds to p0.0101 (1S vs. 2S) (ANOVA one way, multiple comparison Tukey).

Fig. 11. TIMP-1 immunohistochemical expression in endometrium. (a) TIMP-1 expression in proliferative endometrium (×200). (b) TIMP-1 expression in simple endometrialhyperplasia (×100). (c) TIMP-1 expression in G1 endometrial carcinoma (×200). (d) Immunohistochemical profile of TIMP-1 (mean % ± SD) in non-neoplastic (1), hyperplastic(2) and carcinomatous (3) groups, both inside epithelial (E) and stromal (S) components shows a progressive decrease in endometrial hyperplasia and carcinoma respectivelyexcepting S in carcinomas; *I corresponds to p 0.0039 (1S vs. 3S), *F corresponds to p 0.0018 (1E vs. 3E), *G corresponds to p 0.0371 (1S vs. 2S), and *D corresponds to p 0.0237(1E vs. 2E) (ANOVA one way, multiple comparison Tukey).

C. Amalinei et al. / Annals of Anatomy 193 (2011) 43–55 53

Fig. 12. TIMP-2 immunohistochemical expression in endometrium. (a) TIMP-2 expression in proliferative endometrium (×x200). (b) TIMP-2 expression in complex endome-t 00). (h ) comr vs. 3E)(

hismtoedips

cTmbIost2aew

mt9aaG

rial hyperplasia (×100). (c) TIMP-2 expression in G1 endometrial carcinoma (×2yperplastic (2) and stromal (3) groups, both inside epithelial (E) and stromal (Sespectively; *C corresponds to p 0.003 (3E vs. 3S), *F corresponds to p 0.0053 (1EANOVA one way, multiple comparison Tukey).

igher pre-apoptotic vs. lower anti-apoptotic proteins are reportedn endometrial carcinoma (Song et al., 2002), both molecules pre-ented a slightly increased value in our study. This discrepancyay be attributed to a different distribution in tumor cells, to

he histological grade of the tumor, or to the different method-logies performed for the identification of apoptotic factors (Songt al., 2002). Based on Bcl-2, Bax, Fas, FasL expressions, our resultsemonstrate a partial involvement of both apoptotic pathways

n endometrial carcinogenesis with epithelial Bcl-2/Bax systemrevalence both in hyperplasia and in carcinoma and of Bcl-2/Baxystem dominance in carcinoma.

Activation of p53 can promote apoptosis of damaged or agedells serving as an effective tumor suppressor (Oren et al., 2002).he tumor suppressor protein p53 seems to be implicated in thealignant transformation of endometrium, the mutant protein

eing highly expressed mainly in type II endometrial carcinoma.n our study p53 expression was reduced in both componentsf non-neoplastic endometria and showed an increase along theequence non-neoplastic–hyperplastic–carcinomatous endome-ria, corresponding to the previous literature data (Boruban et al.,008). There was also a significant difference between epithelialnd stromal expression both in hyperplastic and carcinomatousndometria demonstrating p53 dominant role in the epithelial cellshen compared to stromal cells.

The enhanced expression of MMPs has been correlated with thealignant phenotype of human cancer cells. Our study confirms

he results of several studies that have localized MMP-2 and MMP-mRNA to stromal cells, including endothelial cells, macrophages,

nd fibroblasts, particularly in areas within or surrounding tumorggregates and also to epithelial tumor cells (Di Nezza et al., 2002;raesslin et al., 2006).

d) Immunohistochemical profile of TIMP-2 (mean % ± SD) in non-neoplastic (1),ponents shows a progressive increase in endometrial hyperplasia and carcinoma, *I corresponds to p 0.0327 (1S vs. 3S), and *D corresponds to p 0.0385 (1E vs. 2E)

Both gelatinases showed a similar aspect of immunoloca-tion. MMP-2 exhibited a low positivity in the investigatednon-neoplastic endometria, both as percentage of positive cellsand of staining intensity. An increased epithelial expressionof MMP-2 was noticed in neoplastic endometria. Moreover,MMP-2 expression was similar to other reports (Graesslinet al., 2006), being stronger in hyperplastic endometria withatypia and showing a progressive increase from hyperplas-tic to malignant endometrium. The relatively higher MMP-2expression noted inside the glandular component of endometrialhyperplasias compared to non-neoplastic endometrial expres-sion suggests a complex MMP-2 role in precursor lesions ofendometrial carcinomas. The differences in MMP-2 expressionbetween the two components may be explained by epithelialMMP-2 production in certain types of endometrial hyper-plasias.

According to the literature reports, MMP-9 expression showedan increasing expression from non-neoplastic to hyperplastic andto carcinomas respectively, with significant differences betweennon-neoplastic and carcinomatous endometria (p < 0.05). MMP-9expression was higher inside the non-neoplastic epithelium butboth components showed an increased expression in hyperplasiaand carcinoma respectively. Similarly to the literature data (Freitaset al., 1999; Curry and Osteen, 2003; Galant et al., 2004), MMP-9strong expression in stromal, immune and vascular cells suggeststheir active involvement in endometrial proliferation. Our results

suggest that MMP-9 is supplementarily produced by other typesof cells than MMP-2, having a stronger expression in carcinoma-tous stromal cells. Supplementary, MMP-9 expression in smoothmyocytes was apparently correlated to myometrial invasion mech-anisms.

5 of Ana

1ngAel(Mio

teBstnthacd

acsce

5

ento

pftt

eaOadic

nhipSo

bmcnsmea

4 C. Amalinei et al. / Annals

TIMPs presented a significant epithelial expression. TIMP-slightly decreased epithelial expression along the sequence

on-neoplastic–hyperplastic–carcinomatous endometrium sug-ests a MMP-9/TIMP-1 imbalance in endometrial carcinogenesis.lthough previous reports considered MMP-2, MMP-1, and TIMP-1xpression in endometrial carcinoma being correlated to histo-ogic stage, depth of invasion, histologic type and age of patientsGraesslin et al., 2006), carcinomas were characterised by increased

MP-2 and decreased TIMP-1 expression in our study. TIMP-1 sim-lar stromal values in hyperplasia and carcinoma support the ideaf a common stromal environment in both lesions.

TIMP-2 showed an intense staining of tumor cells of all his-ologic grades associated with variable degrees of stromal andndothelial staining, as previously reported (Määtä et al., 2000;oruban et al., 2008). TIMP-2 increased in expression along theequence non-neoplastic–hyperplastic–carcinomatous endome-ria, in both components, with significant differences betweenon-neoplastic vs. hyperplastic or carcinomatous endometria inhe epithelial component. Although the literature data supports theypothesis that an increased MMP-2 expression concomitant withreduced TIMP-2 expression selects a subgroup of endometrial car-inoma with high risk of aggressiveness (Graesslin et al., 2006), ourata suggest a stimulating TIMP-2 effect on MMP-2.

Our results support previous data that MMPs, along with TIMPs,re likely to play a complex role in the progression of endometrialarcinoma, by facilitating tumor invasion and subsequent metasta-is. An already existing imbalance between MMP-2, MMP-9, TIMP-2omplex/TIMP-1 in hyperplasia and enhanced in carcinomatousndometria was evident in the investigated cases.

. Conclusions

Endometrium represents the most remarkable example ofxcessive patterns of physiologic changes as a pathologic mecha-ism, as endometrial remodelling process is intrinsically associatedo proliferation, considering neoplastic development as a repetitionf histogenesis.

The coordinated intervention of steroids and their receptors,roliferating factors, apoptosis-related factors and invasivenessactors imbalances are involved in hormonal disorders and fur-hermore in hyperplasia, as steps along endometrial malignantransformation.

Our study adds new data to recent reports suggesting variablestrogens effects, as promoters of proliferation, some hormonalctions on endometrial cells being mediated by stromal cells.ur results confirm the influence of steroid receptor expressionnd their imbalance on the proliferation index, in the disor-ered proliferative phase, hyperplasia and carcinoma, with an

ncreased ER expression in hyperplasia and an ER/PR imbalance inarcinoma.

The imbalances between apoptosis molecules show a domi-ant intervention of decreased Bcl-2/Bax epithelial ratio similar inyperplasia and carcinoma and a decreased Bcl-2/Bax stromal ratio

n carcinomas. The Bcl-2/Bax system anomalies are associated to53 mutations and to a weaker intervention of Fas/FasL system.upplementary, PCNA high expressions are characteristic featuresf both endometrial hyperplasia and carcinoma.

Among MMPs, gelatinases seem to cooperate in endometrium,eing co-localised in epithelial cells. MMP-9 seems to be supple-entarily expressed in stromal and muscular cells, in endometrial

arcinomas, as a correlation with its enhanced role in invasive-

ess. TIMPs expression suggests that TIMP-1 decrease, with aimilar stromal value both in hyperplasia and carcinoma, as a com-on feature. MMP-9/TIMP-1 imbalance is characteristic both for

ndometrial hyperplasia and carcinoma. TIMP-2 increases, as itcts more than a simple MMP-2 inhibitor, having a complex role

tomy 193 (2011) 43–55

in endometrial carcinogenesis. We may conclude that proliferationand invasiveness may be the result of a complex MMP-2, MMP-9,TIMP-2/TIMP-1 imbalance.

Future research is necessary to elucidate the complex interac-tions between molecules involved in proliferation and apoptosis,opening new perspectives in the early diagnosis and treatment ofendometrial neoplasias.

Acknowledgements

The research was supported by the CNCSIS (National Council forScientific Research in Higher Education) grant 1218/2007.

References

Abe, H., Shibata, M.-A., Otsuki, Y., 2006. Caspase cascade of Fas-mediated apoptosisin human normal endometrium and endometrial carcinoma cells. Mol. Hum.Reprod. 12, 535–541.

Atasoy, P., Bozdogan, O., 2006. Molecular markers in endometrial hyperplasia.Gynecol. Oncol. 11, 61–67.

Bircan, S., Ensari, A., Ozturk, S., 2005. Immunohistochemical analysis of c-myc, c-junand estrogen receptor in normal, hyperplastic and carcinomatous endometrium.Pathol. Oncol. Res. 11, 32–39.

Bockman, J.V., 1983. Two pathogenetic types of endometrial carcinoma. Gynecol.Oncol. 15, 10–17.

Boruban, M.C., Altundag, K., Kilic, G.S., Blankstein, J., 2008. From endometrial hyper-plasia to endometrial cancer: insight into the biology and possible medicalpreventive measures. Eur. J. Cancer Prev. 17, 133–138.

Bozdogan, O., Atasoy, P., Erekul, S., Bozdogan, N., Bayram, M., 2002. Apoptosis relatedproteins and steroid hormone receptors in normal, hyperplastic and neoplasticendometrium. Int. J. Gynecol. Pathol. 21, 375–382.

Cao, Q.J., Einstein, M.H., Anderson, P.S., Runowicz, C.D., Balan, R., Jones, J.G., 2002.Expression of COX-2, Ki-67, Cyclin D1, and P21 in endometrial endometrioidcarcinomas. Int. J. Gynecol. Pathol. 21, 147–154.

Chegini, N., Rhoton-Vlasak, A., Williams, R.S., 2003. Expression of matrixmetalloproteinase-26 and tissue inhibitor of matrix metalloproteinase-3 and -4in endometrium throughout the normal menstrual cycle and alteration in usersof levonorgestrel implants who experience irregular uterine bleeding. Fertil.Steril. 80, 564–570.

Cornet, P.B., Galant, C., Eeckhout, Y., Courtoy, P.J., Marbaix, E., Henriet, P., 2005. Reg-ulation of matrix metalloproteinase-9/gelatinase B expression and activationby ovarian steroids and LEFTY-A/endometrial bleeding-associated factor in thehuman endometrium. J. Clin. Endocrinol. Metab. 90, 1001–1011.

Curry, T.E., Osteen, K.G., 2003. The matrix metalloproteinase system: changes, reg-ulation, and impact throughout the ovarian and uterine reproductive cycle.Endocr. Rev. 24, 428–465.

Di Nezza, L.A., Misajon, A., Zhang, J., Jobling, T., Quinn, M.A., Ostör, A.G., Nie, G., Lopata,A., Salamonsen, L.A., 2002. Presence of active gelatinases in endometrial carci-noma and correlation of matrix metalloproteinase expression with increasingtumor grade and invasion. Cancer 94, 1466–1475.

Egeblad, M., Werb, Z., 2002. New functions for the matrix metalloproteinases incancer progression. Nat. Rev. Cancer 2, 161–174.

Freitas, S., Meduri, G., Nestur, E., Bausero, P., Perrot-Applanat, M., 1999. Expression ofmetalloproteinases and their inhibitors in blood vessels in human endometrium.Biol. Reprod. 61, 1070–1082.

Galant, C., Berliere, M., Dubois, D., Verougstraele, J.C., Charles, A., Lemoine, P.,Kokorine, I., Eeckout, Y., Courtoy, P.J., Marbaix, E., 2004. Focal expression andfinal activity of matrix metalloproteinases may explain irregular dysfunctionalendometrial bleeding. Am. J. Pathol. 165, 83–94.

Giuffre, G., Arena, F., Scarfi, R., Simone, A., Todaro, P., Tuccari, G., 2006. Lactoferrinimmunoexpression in endometrial carcinomas: relationships with sex steroidhormone receptors (ER and PR), proliferation indices (Ki-67 and AgNOR) andsurvival. Oncol. Rep. 16, 257–263.

Goffin, F., Munaut, C., Frankenne, F., Perrier d’Hauterive, S., Beliard, A., Fridman,V., Nervo, P., Colige, A., Foidart, J.M., 2003. Expression pattern of metallopro-teinases and tissue inhibitors of matrix metalloproteinases in cycling humanendometrium. Biol. Reprod. 69, 976–984.

Graesslin, O., Cortez, A., Fauvet, R., Lorenzoto, M., Birembaut, P., Darai, E., 2006.Metalloproteinase-2, -7 and -9 and tissue inhibitor of metalloproteinase-1 and-2expression in normal, hyperplastic and neoplastic endometrium: a clinical-pathological correlation study. Ann. Oncol. 17, 637–645.

Hale, G.E., Hughes, C.L., Cline, J.M., 2002. Endometrial cancer: hormonal factors, theperimenopausal “window of risk”, and isoflavones. J. Clin. Endocrinol. Metab.87, 3–15.

Harada, T., Kaponis, A., Iwabe, T., Taniguchi, F., Makrydimas, G., Sofikitis, N.,

Paschopoulos, M., Paraskevaidis, E., Terakawa, N., 2004. Apoptosis in humanendometrium and endometriosis. Hum. Reprod. Update 10, 29–38.

Heljasvaara, R., Nyberg, P., Luostarinen, J., Parikka, M., Heikkila, P., Rehn, M., Sorsa,T., Salo, T., Pihlajaniemi, T., 2005. Generation of biologically active endostatinfragments from human collagen XVIII by distinct matrix metalloproteinases.Exp. Cell. Res. 307, 292–304.

of Ana

H

H

H

I

I

I

I

J

K

L

L

L

L

M

M

M

M

M

C. Amalinei et al. / Annals

enriet, P., Cornet, P.B., Lemoine, P., Galant, C., Singer, C.F., Courtoy, P.J., Eeckhout,Y., Marbaix, E., 2002. Circulating ovarian steroids and endometrial matrix met-alloproteinases (MMPs). Ann. N. Y. Acad. Sci. 955, 119–138.

ojilla, C.V., Mohammed, F.F., Khokha, R., 2003. Matrix metalloproteinases and theirtissue inhibitors direct cell fate during cancer development. Br. J. Cancer 10,1817–1821.

u, K., Zhong, G., He, F., 2005. Expression of estrogen receptors ER� and ER�in endometrial hyperplasia and adenocarcinoma. Int. J. Gynecol. Cancer 15,537–541.

oachin, E., 2005. Immunohistochemical tumor markers in endometrial carcinoma.Eur. J. Gynaecol. Oncol. 26, 363–371.

saka, K., Nishi, H., Nakada, T., Feng Li, Y., Ebihara, Y., Takayama, M., 2003. Matrixmetalloproteinase-26 is expressed in human endometrium but not in endome-trial carcinoma. Cancer 97, 79–89.

to, K., Utsunomiya, H., Suzuki, T., Saitou, S., Akahira, J., Okamura, K., Yae-gashi, N., Sasano, H., 2006. 17beta-Hydroxysteroid dehydrogenases in humanendometrium and its disorders. Mol. Cell. Endocrinol. 248, 136–140.

to, K., Utsunomiya, H., Yaegashi, N., Sasano, H., 2007. Biological roles of estrogen andprogesterone in human endometrial carcinoma—new developments in potentialendocrine therapy for endometrial cancer. Endocr. J. 54, 667–679.

azaeri, O., Shupnik, M.A., Jazaeri, A.A., Rice, L.W., 1999. Expression of estrogen recep-tor alpha mRNA and protein variants in human endometrial carcinoma. Gynecol.Oncol. 74, 38–47.

onno, R., Yamakawa, H., Utsunomiya, H., Ito, K., Sato, S., Yajima, A., 2000. Expressionof survivin and Bcl-2 in the normal human endometrium. Mol. Hum. Reprod. 6,529–534.

ambert, V., Wielockx, B., Munaut, C., Galopin, C., Jost, M., Itoh, T., Werb, Z., Baker, A.,Libert, C., Krell, H., Foidart, J.M., Noel, A., Rakic, J.M., 2003. MMP-2 and MMP-9synergize in promoting choroidal neovascularization. FASEB 17, 2290–2292.

ax, S.F., 2004. Molecular genetic pathways in various types of endometrial carci-noma: from a phenotypical to a molecular-based classification. Virchows. Arch.444, 213–223.

i, W., Savinov, A.Y., Rozanov, D.V., Golubkov, V.S., Hedayat, H., Postnova,T.I., Golubkova, N.V., Linli, Y., Krajewski, S., Strongin, A.Y., 2004. Matrixmetalloproteinase-26 is associated with estrogen-dependent malignancies andtargets �1-antitrypsin Serpin. Cancer Res. 64, 8657–8665.

ucas, R., Holmgren, L., Garcia, I., Jimenez, B., Mandriota, S.J., Borlat, F., Sim, B.K., Wu,Z., Grau, G.E., Shing, Y., Soff, G.A., Bouck, N., Pepper, M.S., 1998. Multiple formsof angiostatin induce apoptosis in endothelial cells. Blood 92, 4730–4741.

äätä, M., Soini, Y., Liakka, A., Autio-Harmainen, H., 2000. Localisation of MT1-MMP, TIMP-1, TIMP-2, and TIMP-3 messenger RNA in normal, hyperplastic,and neoplastic endometrium. Enhanced expression by endometrial adeno-carcinomas is associated with low differentiation. Am. J. Clin. Pathol. 114,402–411.

atias-Guiu, X., Catasus, L., Bussaglia, E., Lagarda, H., Garcia, A., Pons, C., Munoz,J., Arguelles, R., Machin, P., Prat, J., 2001. Molecular pathology of endometrialhyperplasia and carcinoma. Hum. Pathol. 32, 569–577.

itselou, A., Ioachim, E., Kitsou, E., Vougiouklakis, T., Zagorianakou, N., Makrydimas,G., Stefanaki, S., Agnantis, N.J., 2003. Immunohistochemical study of apoptosis-related Bcl-2 protein and its correlation with proliferation indices (Ki67, PCNA),tumor suppressor genes (p53, pRb), the oncogene c-erbB-2, sex steroid hormonereceptors and other clinicopathological features, in normal, hyperplastic andneoplastic endometrium. In Vivo 17, 469–477.

ohsin, S.K., Weiss, H., Havighurst, T., Clark, G.M., Berardo, M., Roanhle, D., To, T.V.,Qian, Z., Love, R.R., Allred, D.C., 2004. Progesterone receptor by immunohisto-chemistry and clinical outcome in breast cancer: a validation study. Mod. Pathol.17, 1545–1554.

utter, G.L., 2000. Histopathology of genetically defined endometrial precancers.Int. J. Gynecol. Pathol. 19, 301–309.

tomy 193 (2011) 43–55 55

Neis, K.J., Brandner, P., Ulrich-Winkelspecht, K., 2000. The oestrogen receptor (ER)in normal and abnormal uterine tissue. Eur. J. Cancer 36, S27–36.

Nephew, K.P., Choi, C.M., Polek, T.C., 2000. Expression of fos and jun proto-oncogenesin benign versus malignant human uterine tissue. Gynecol. Oncol. 76, 388–396.

Nunobiki, O., Taniguchi, E., Ishii, A., 2003. Significance of hormone receptor statusand tumor vessels in normal, hyperplastic and neoplastic endometrium. Pathol.Int. 53, 846–852.

Oren, M., Damalas, A., Gottlieb, T., Michael, D., Taplick, J., Leal, J.F., Maya, R., Moas,M., Seger, R., Taya, Y., Ben-Ze’Ev, A., 2002. Regulation of p53: intricate loops anddelicate balances. Ann. N. Y. Acad. Sci. 973, 374–383.

Otsuki, Y., 2001. Apoptosis in human endometrium: apoptotic detection methodsand signaling. Med. Electron. Microsc. 34, 166–173.

Pearce, S.T., Jordan, V.C., 2004. The biological role of estrogen receptors alpha andbeta in cancer. Crit. Rev. Oncol. Hematol. 50, 3–22.

Pozzi, A., Moberg, P.E., Miles, L.A., Wagner, S., Soloway, P., Gardner, H.A., 2000.Elevated matrix metalloproteinase and angiostation levels in integrin alpha 1knockout mice cause reduced tumor vascularisation. Proc. Natl. Acad. Sci. U.S.A.97, 2202–2207.

Puente, X.S., Sanchez, L.M., Overall, C.M., Lopez-Otin, C., 2003. Human and mouseproteases: a comparative genomic approach. Nat. Rev. Genet. 4, 544–558.

Rogers, P.A., Lederman, F., Plunkett, D., Affandi, B., 2000. Bcl-2, Fas and caspase 3expression in endometrium from levonorgestrel implant users with and withoutbreakthrough bleeding. Hum. Reprod. 15 (Suppl. 3), 152–161.

Ronnett, B.M., Zaino, R.J., Ellenson, L.H., Kurman, R.J., 2002. Endometrial carcinoma.In: Kurman, R.J. (Ed.), Blaustein’s Pathology of the Female Genital Tract, 5th edn.Springer-Verlag, New York, pp. 501–559.

Rubin, H., 2003. Complementary approaches to understanding the role of proteasesand their natural inhibitors in neoplastic development: retrospect and prospect.Carcinogenesis 24, 803–816.

Salvesen, H.B., Iversen, O.E., Akslen, L.A., 1999. Prognostic significance of angio-genesis and Ki-67, p53, and p21 expression: a population-based endometrialcarcinoma study. J. Clin. Oncol. 17, 1382–1390.

Shiozawa, T., Li, S., Nakayama, K., Nikaido, T., Fujii, S., 1996. Relationship between theexpression of cyclins/cyclin-dependent kinases and sex-steroid receptors/Ki67in normal human endometrial glands and stroma during the menstrual cycle.Mol. Hum. Reprod. 2, 745–752.

Song, J., Rutherford, T., Naftolin, F., Brown, S., Mor, G., 2002. Hormonal regulation ofapoptosis and the Fas and Fas ligand system in human endometrial cells. Mol.Hum. Reprod. 8, 447–455.

Turpeenniemi-Hujanen, T., 2005. Gelatinases (MMP-2 and -9) and their naturalinhibitors as prognostic indicators in solid cancers. Biochimie 87, 287–297.

Uchikawa, J., Shiozawa, T., Shih, H., 2003. Expression of steroid receptor coactiva-tors and corepressors in human endometrial hyperplasia and carcinoma withrelevance to steroid receptors and Ki-67 expression. Cancer 98, 2207–2213.

Vaskivuo, T.E., Stenbäck, F., Tapanainen, J.S., 2002. Apoptosis and apoptosis-relatedfactors Bcl-2, Bax, tumor necrosis factor-alpha, and NF-kappaB in humanendometrial hyperplasia and carcinoma. Cancer. 95, 1463–1471.

Velasco, G., Pendas, A.M., Fueyo, A., Knauper, V., Murphy, G., Lopez-Otin, C., 1999.Cloning and characterization of human MMP-23, a new matrix metallopro-teinase predominantly expressed in reproductive tissues and lacking conserveddomains in other family members. J. Biol. Chem. 274, 4570–4576.

Yamashita, H., Otsuki, Y., Matsumoto, K., Ueki, K., Ueki, M., 1999. Fas ligand, Fas anti-gen and Bcl-2 expression in human endometrium during the menstrual cycle.

Mol. Hum. Reprod. 5, 358–364.

Yokoyama, Y., Takahashi, Y., Morishita, S., Hashimoto, M., Niwa, K., Tamaya, T., 1998.Telomerase activity in the human endometrium throughout the menstrual cycle.Mol. Hum. Reprod. 4, 173–177.

Zhang, H.G., Wang, J., Yang, X., Hsu, H.C., Mountz, J.D., 2004. Regulation of apoptosisproteins in cancer cells by ubiquitin. Oncogene 23, 2009–2015.