Expression of Ezrin and Estrogen Receptors During Cervical Carcinogenesis

Abstract

Rationale:

Development of cervical squamous carcinoma (CXCA) is accompanied by changes in estrogen receptors (ERs, ERα and ERβ) and ezrin expression; however, reports have been conflicting. Using histologically documented staging of cervical biopsies, we determined ezrin and ER relationships during CXCA development.

Methods:

Immunoreactive (ir) ezrin, ir-ERα, and ir-ERβ were studied in normal epithelium, carcinoma in situ/cervical intraepithelial neoplasia (CIN) 1 to 3, and local invasion or metastatic CXCA. Results were compared using H scoring. Cultures of Caski metastatic CXCA cells were treated with estradiol and/or tamoxifen and studied for ER-driven ir-ezrin and the morphologic response.

Results:

Koilocytosis was present and indicated viral presence. The ezrin H score increased from CIN1 to CIN3, reaching significant differences from normal by CIN3 (P = .004) and 2× normal in metastatic CXCA. Estrogen receptor α and ERβ H scores fell, reaching significance by CIN3 (ERα, P = .0001; ERβ, P = .024). During estradiol treatment, ezrin in Caski cells increased and localized to the periphery, in ruffles and processes. The selective ER modulator tamoxifen blocked the estradiol-induced changes.

Conclusions:

During cervical carcinogenesis, the usual relationship between estrogen and ezrin induction is abridged. This is consistent with the effects of human papilloma virus viral proteins such as E6 and E7 that upregulate SIX1, a protein that induces ezrin. Cervical carcinogenesis is progressive but arrests at the preinvasive stage for varying lengths of time. These studies suggest that changes in ezrin may be associated with the development of the invasive phenotype and penetration of the basement membrane. They also raise the possibility that inhibiting ezrin expression could be a target for the prevention of invasive CXCA.

Keywords

carcinogenesis cervical cancer estradiol ezrin

Introduction

An estimated 13 000 cases of invasive epidermoid carcinoma of the cervix, cervical squamous carcinoma (CXCA), were diagnosed, and an estimated 41 000 deaths from cervical cancer were expected in 2015.^1,2 Human papilloma virus (HPV), especially HPV16, is the main cause of this lesion.³ Despite the availability of an anti-HPV vaccine, the likelihood of eradication of this disease remains a distant hope; therefore, study of the mechanisms underlying the pathogenesis of CXCA remains a high priority.

Ezrin is a member of the ezrin–radixin–moesin protein family. It drives cellular specialization and functions as a signal transducer. Ezrin is involved in the regulation of phenotypic changes in cellular architecture. Moreover, it facilitated cellular movement and motility, recognizes internal and external signals, and controls cellular growth.^4
-6 These functions are essential to maintain homeostasis in normal cells and are highly involved in the process of tumorigenesis, invasion, and distant metastasis.^6
-8 Abnormal ezrin forms have been found in cancers and nonmalignant conditions^9,10 (A. Fadiel and F. Naftolin, personal communication). We and others have shown high expression of ezrin in CXCA and other cancers, especially metastatic lesions.¹¹

Although estrogen receptors (ERs) have been shown to be present in the normal cervical epithelium and CXCA, reports have indicated low expression of ERs in developing and invasive CXCA. However, preclinical trials using the selective estrogen receptor modulator (SERM) raloxifene showed inhibition of mouse CXCA supporting retention of ER activity.^12
-14

Since estrogen is a well-known regulator of maturation (keratinization) of the cervical epithelium and ezrin is important in normal and malignant cell polarization, adhesion, and motility, we sought to determine the changes in these key proteins during the progress of CXCA. The present report correlates the expression of ezrin and ERs with the progression of cervical carcinogenesis and supports a causative role for ezrin and ER expression in the development of CXCA.

Materials and Methods

Under an institutional review board–approved protocol, archival cervical tissue and sections were obtained from Soonchunhyang Medical Center Hospital. A total of 57 specimens containing the squamocolumnar junction (SCJ; with the exception of the metastatic lesions were studied. Of these, 10 specimens were classified into each of the following 5 groups: normal, cervix intraepithelial neoplasia (CIN) I to III, and metastatic CXCA. Another 7 specimens were classified as invasive CXCA. The stage of cancer was determined by a single pathologist according to the International Federation of Gynecology and Obstetrics system. Specimens with CIN were labeled according to the disposition of normal versus abnormal cells (CIN1 = cells in lower one-third of epithelial height abnormal, CIN2 = cells in lower two-thirds of epithelial height abnormal, CIN3 = cells in more than two-thirds of epithelial height abnormal). All specimens were obtained by cold conization or hysterectomy, fixed in 10% formalin and cut as 5 microsections. Caski cells were purchased from American Type Culture Collection (Rockville, Maryland).

Tissue Processing and Immunohistochemistry for Ezrin

Immunohistochemical staining was conducted on paraffin sections (10 specimens in each of the normal, CIN I-III, and metastatic CXCA groups and 7 invasive CXCA specimens). Each section contained a well-conserved SCJ and a full thickness layer of epithelial plus stromal tissue. Sections were collected on poly-l-lysine-coated slides (Sigma-Aldrich Corp, St Louis, Missouri), deparaffinized in dimethylbenzene, and then rehydrated through a graded ethanol series. After rinsing in phosphate-buffered saline (PBS) 3 times for 5 minutes each, antigen retrieval from the tissue was carried out by autoclaving in 0.01 M sodium citrate buffer (pH 6.0) for 20 minutes. After cooling for 1 hour at room temperature, the tissue was washed in PBS 3 times for 5 minutes each. Endogenous peroxidases were exhausted by treating the sections with 3% hydrogen peroxide (Fisher ChemAlert Guide, New Jersey) for 30 minutes, and then the tissue was washed 3 times with PBS. The slides were then incubated in a humidified chamber with blocking buffer (5% bovine serum + 0.5% Tween-20 in PBS) for 1 hour at room temperature and then incubated overnight at 4°C with rabbit antihuman ezrin polyclonal antibody (1:200 dilution; Millipore Corporation, Minneapolis, MN). Next, the tissue was rinsed with PBS 3 times for 10 minutes each, the sections were incubated with biotinylated universal antibody (antirabbit/mouse immunoglobulin G [IgG]) for 1 hour, and the sections were treated with an avidin–biotin complex (Vectastain kit for 30 minutes. Then the tissue was washed with PBS 3 times for 10 minutes each and the sections were developed using diaminobenzidine for 12 minutes at room temperature (DAB kit; Vector Laboratories, Burlingame, California). The tissue was subsequently lightly counterstained with Mayer hematoxylin (Merck, Darmstadt, Germany), dehydrated, and mounted using Permount (Fisher Chemicals, New Jersey).

Immunofluorescent Staining of Ezrin in the Caski Cell Line

A modification was used in our estrogen-free method for tissue culture.¹⁵ Prior to study, the cells were maintained in minimal essential medium (MEM) supplemented with penicillin, streptomycin, and 10% fetal calf serum (GIBCO BRL, Gaithersburg, Maryland). Cells were cultured in 5% CO₂ in humidified air, and the medium was changed twice a week. After flask incubation to subconfluence (60%-70%) in MEM containing 10% fetal bovine serum, the medium was changed to phenol red-free, serum-free Opti-MEM 1× for more than 72 hours during which time the cells grew to 100% confluence without evidence of cell death (apoptotic bodies, floating cells, etc). Following 24 hours of preincubation in Opti-MEM, 1×, the cells were exposed to 0.05% trypsin (supplemented with 5.3 nM EDTA–4Na, mycoplasma-free; GIBCO BRL, Gaithersburg, Maryland) for 5 minutes and lightly shaken to loosen them from the flask bottom. After cells were detached, they were washed with fresh Opti-MEM 1× to stop the action of trypsin. The cells were transferred to plastic Falcon tubes (Fisher Scientific, Chino, California) in Opti-MEM 1× and gently centrifuged. The cell pellet was resuspended in culture medium and transferred to 4-chambered plates for the estrogen and tamoxifen treatment to establish an estrogen-controlled environment, and the cells were then incubated for an additional 24 hours at 37°C with serum-free Opti-MEM 1× with added 0.001% ethanol (timed-vehicle control),⁷ estrogen (10⁻⁸ M), or estrogen plus tamoxifen (10⁻⁶ M) or tamoxifen (10⁻⁶ M). The cells were fixed with Cytofix solution, washed with 1× PBS 3 times, and then blocked with 1% bovine serum albumin. Subsequently, the cells were incubated with anti-ezrin mouse antibodies (Neomarkers, lot 661 × 503) and diluted (1:200) overnight at 4°C. Next, the cells were incubated with secondary antibodies (Alexa Fluor antimouse IgG; 1:600) for 30 minutes at room temperature and observed by fluorescence microscopy.

Scoring and Statistical Analysis

Modified “H testing” was used for semiquantitative evaluation of proteins in the stained sections. Two blinded observers using ×40 objectives (final magnification ×400) evaluated the intensity of staining—0 = no staining, 1+ = weak but detectable staining, 2+ = moderate or distinct staining, and 3+ = strong staining. H scores were calculated using the following equation: H = ΣPi (i + 1), where i is the intensity of staining with a value from 0 to 3 and Pi is the percentage of stained cells, varying from 0% to 100%. In each slide, 5 different random areas and 100 cells in each area were evaluated and the average score was calculated. Comparisons of the means were performed by 1-way analysis of variance. Statistical significance was defined as P < .05.

Results

Microscopical examination confirmed that all tissues were staged correctly and that the sections were satisfactory for immunostaining and H scoring (see Figure 1 and Discussion section). Granular immunostaining in normal endocervical glands at the SCJ was used as a positive control, and ezrin, ERα, and ERβ were expressed in all stages. Koilocytosis was present in all sections of the SCJ, that is, all tissues had viral signatures (data not shown).

Figure 1.

Immunoreactive-ezrin staining (brown) in the cervix (×20): (A) normal cervix; (B) cervix intraepithelial neoplasia (CIN) I; (C) CIN II; (D) CIN III; (E) invasive cancer; (F) metastatic cancer. The density of ezrin staining increased with disease progression.

Ezrin Expression in Cervical Tissue and Caski Cells

Ezrin immunoreactivity was positive and observed to increase in the following order of stages of transformation: normal < cervix < squamous epithelium < CIN 1 < 2 < 3 < invasive < metastatic squamous cell carcinoma (Figure 1).

Normal epithelial cells showed a light staining pattern at the basal or parabasal cell borders. In CIN1, the ezrin scores were slightly higher and staining appeared more intense around the luminal boarder of the cells. Compared to normal tissues, ezrin expression showed statistically significant increases by the attainment of CIN3. The difference in H score between ezrin staining in CIN2 and CIN3 nearly doubled between the stages before CIN2 or after CIN3 but this difference did not reach statistical significance (P = .223; Figure 1). The mean H scores from CIN3 invasion was 2 or greater with an up-going trend. High levels of ezrin expression were noted in cancer tissue compared with normal and low-grade CIN samples (P < .05).

Estrogen Receptors α and β Expression in Cervical Tissue

In normal cervical epithelium, ER was mainly observed in basal cells of the squamous epithelium and endocervical glands (data not shown). The strongest staining for ERα and β was present in the normal epithelium; this continuously decreased as cancer progressed. Overall, immunostaining for both ERs showed a sustained and almost complete decline during the transformation to invasion and metastases (P < .05; Figure 2B and C).

Figure 2.

Ezrin expression in Caski cells with immunohistochemistry (×20). A, Control (cultured Caski cells): ezrin was observed in the perinuclear and cytoplasmic areas, and cells were round or ovoid without spindles. B, Following 24-hour treatment with estrogen (E2, 10⁻⁸ M), cells formed pseudopodia and ruffled cell membranes, and ezrin translocated to the cell membrane. C, Following 24-hour treatment with tamoxifen (10⁻⁶ M), cells stained more heavily than controls but did not form processes. D, Following 24-hour treatment with tamoxifen (10⁻⁶ M) and E2 (10⁻⁸ M), cells were round without spindles, and morphological changes appeared with pseudopodia.

Estrogen Treatment of Caski Cells

In estrogen-free cultures, immunoreactive (ir) ezrin was weakly detected in the cytoplasmic and perinuclear areas of the flat, ovoid-shaped Caski CXCA cells. Following 24 hours of estradiol treatment, ir-ezrin appeared to be more dense and peripheral. Significant phenotypic changes were observed such as spindles, protruding processes, nuclear pleomorphism, and irregular cellular shapes. Tamoxifen treatment had little apparent effect, with some translocation to the cell periphery. The cells appeared smaller and more rounded than those that were untreated. Estradiol plus tamoxifen-treated Caski cells were not different from untreated cells (Figure 3).

Figure 3.

A, Ezrin mean H score values in cervical tissue. *P < .05 versus control. B, Estrogen receptor α (ERα) mean H score values in cervical tissue. *P < .05 versus control. C, Estrogen receptor β (ERβ) mean H score values in cervical tissue. *P < .05 versus control.

Discussion

In this project, we delineated the expression of ERs and ezrin during the development of cervical cancer. To categorize the study sections, we followed the pathological diagnosis of normal, CIN 1 to 3, and invasive CXCA. Although this classification has been replaced for the purpose of interpretation of Papanicolaou smears,¹⁶ it served quite well for the present project in which we directly visualized the morphohistology being tested. The milestones chosen for this analysis included CIN1 (malignant cells limited to the basal one-third of the section), CIN2 (malignant cells limited to the basal two-thirds of the section), and CIN3 (malignant cells occupying entire thickness of the section). There was no evidence of penetration of the basement membrane in any of the “CIN” sections. We carefully reviewed each tissue section to ascertain which stage of preinvasive cancer was being examined.

Expressions of ezrin and ER were studied in this project because of our previous demonstration of increased ezrin in many cancers, including estrogen-sensitive female genital cancers and because estrogen/ER induces ezrin.^15,17 Animal studies have shown the requirement for ER in HPV E6/7-induced CXCA.¹⁸ In support of the role of ER in CXCA, Spurgeon and others have shown that the SERM raloxifene retards the growth of viral E6/E7-induced CXCA in mice.^13,14 The present project documented the decline in ER expression during the development of CXCA. To confirm the functional activity of ER in CXCA, we showed that treating Caski cells with estrogen-induced ezrin expression and development of the invasive phenotype, that is, ER in CXCA, are functional. However, the decline in ER may explain the disappearance of ectocervical glycogen, that is the hallmark of cervical epithelial transformation.

Despite the decrease in ER expression, the expression of ezrin increased during carcinogenesis; in CXCA, the relationship between ezrin and estrogen is dichotomous. SIX1 induction by viral E6/7 may explain this apparent contradiction of the relationship of ER with ezrin.¹⁹

Both ERα and β showed a sustained decrease in ir-ERα and β during the progression from normal epithelium to invasive CXCA. There have been conflicting reports regarding the expression of both ERα and β in CXCA. These may have been related to antibody-targeted epitopes.²⁰ However, ER-null animal studies have confirmed the necessity for ER in the development and growth of CXCA.¹⁸ There are also several reports showing that SERM cause decreased growth and apoptosis of CXCA.²⁰ Moreover, estradiol resulted in clear estrogen actions in CXCA cells in culture, and this was reversed by the SERM tamoxifen.^{13,14,21

-24}

Sex steroids²⁵ and growth factors²² regulate the expression of ERs. Since ERα and β are products of separate genes on separate chromosomes, a single mutational event that affects both genes is not likely. As well, there is no evidence of a systemic loss of ER or ER function in CXCA. Epigenetic regulation of ERs by cancers has been most widely studied in breast cancer. The 2 prominent regulators are methylation of the promoter site and dysregulation of microRNAs (miRNAs).^26
-28 It has been shown that in prostate cancer, methylation of 2 sites on the ERβ promoter caused the progressive loss of ERβ observed during carcinogenesis.²⁹ Although there are dysregulations of miRNA that affect ezrin, see below, thus far, none that affect ERs in CXCA have been reported. Although G protein-related ER (GPER) has been linked to carcinogenesis,²² no direct effect of HPV E proteins on GPER has been reported. Taken together, these considerations indicate that the change in ER expression must reflect local conditions, probably related to the effects of HPV E6/7.^{18,19,30

-33}

Conclusions

We have elucidated the progressive fall of ir-ER α and β during the development of CXCA. This was accompanied by a progressive rise in ir-ezrin expression. This dichotomy is contradictory to previous findings that show estrogen to induce ezrin expression. This result implies that another agent is in play. SIX1 induces ezrin and has, itself, been shown to be induced by HPV E6/7.

Ezrin regulates the cell phenotype, including normal characteristics such as cell polarization, junctional specializations, and cell shape; but, excess ezrin could contribute to the development of the metastatic phenotype including lack of cell polarization, tenuous cell adhesion specializations, and the metastatic phenotype. These considerations raise the possibility that the rise and/or mutations of ezrin play a role in the transition from preinvasive to invasive cancer. We have recently submitted evidence of the latter possibility.

Human papilloma virus has been confirmed to be the causative agent in numerous cancers not generally associated with sex steroid action. The present findings is reasonable to test for the presence and action of sex steroid receptors and of ezrin in these cancers.

SERMs have been shown to inhibit the growth and progression of CXCA in animal models. While this is a transient response, the same is true in the treatment of other estrogen-sensitive lesions such as breast cancer. It is reasonable to suggest epidemiologic studies into the incidence of cervical cancer in women undergoing long-term antiestrogen treatment and the initiation of clinical antiestrogen trials in cases of cervical neoplasia.

Footnotes

Acknowledgments

Seung Do Choi and Ahmed Fadiel contributed equally to this project.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Supported by Varadi Ovarian Initiative for Cancer Education

References

American Cancer Society. Cancer Facts & Figures 2015. Atlanta, GA: American Cancer Society; 2015.

Dunleavey

. Incidence, pathophysiology and treatment of cervical cancer. Nurs Times. 2004;100(44):38–41.

Mirkovic

Howitt

Roncarati

. Carcinogenic HPV infection in the cervical squamo-columnar junction. J Pathol. 2015;236(3):265–271.

Fadiel

Lee

Demir

. Ezrin is a key element in the human vagina. Maturitas. 2008;60(1):31–41.

Fehon

McClatchey

Bretscher

. Organizing the cell cortex: the role of ERM proteins. Nat Rev Mol Cell Biol. 2010;11(4):276–287.

Ohtani

Sakamoto

Rutherford

Chen

Satoh

Naftolin

. Ezrin, a membrane-cytoskeletal linking protein, is involved in the process of invasion of endometrial cancer cells. Cancer Lett. 1999;147(1-2):318.

Fadiel

Chen

Ulukus

Ohtani

Hatami

Naftolin

. Ezrin overexpression by transformed human ovarian surface epithelial cells, ovarian cleft cells, and serous ovarian adenocarcinoma cells. Reprod Sci. 2012;19(8):797–805.

Martin

Harrison

Mansel

Jiang

. The role of the CD44/ezrin complex in cancer metastasis. Crit Rev Oncol Hematol. 2003;46(2):165–186.

Majores

Schick

Engels

. Mutational and immunohistochemical analysis of ezrin-, radixin-, moesin (ERM) molecules in epilepsy-associated glioneuronal lesions. Acta Neuropathol. 2005;110(6):537–546.

10.

Clucas

Valderrama

. ERM proteins in cancer progression. J Cell Sci. 2015;128(6):1253.

11.

Chen

Fadiel

Feng

Ohtani

Rutherford

Naftolin

. Ovarian epithelial carcinoma tyrosine phosphorylation, cell proliferation, and ezrin translocation are stimulated by interleukin 1alpha and epidermal growth factor. Cancer. 2001;92(12):3068–3075.

12.

Chung

. Targeting female hormone receptors as cervical cancer therapy. Trends Endocrinol Metab. 2015;26(8):399–401.

13.

Munger

. Are selective estrogen receptor modulators (SERMs) a therapeutic option for HPV-associated cervical lesions and cancers? Am J Pathol. 2014;184(2):358–361.

14.

Spurgeon

Chung

Lambert

. Recurrence of cervical cancer in mice after selective estrogen receptor modulator therapy. Am J Pathol. 2014;184(2):530–540.

15.

Song

Fadiel

Edusa

. Estradiol-induced ezrin overexpression in ovarian cancer: a new signaling domain for estrogen. Cancer Lett. 2005;220(1):57–65.

16.

Nayar

Wilbur

. The pap test and Bethesda 2014. “The reports of my demise have been greatly exaggerated.” (After a quotation from Mark Twain). Acta Cytol. 2015;59(2):121–132.

17.

Ohtani

Sakamoto

Rutherford

. Ezrin, a membrane-cytoskeletal linking protein, is highly expressed in atypical endometrial hyperplasia and uterine endometrioid adenocarcinoma. Cancer Lett. 2002;179(1):79–86.

18.

Chung

Wiedmeyer

Shai

Korach

Lambert

. Requirement for estrogen receptor alpha in a mouse model for human papillomavirus-associated cervical cancer. Cancer Res. 2008;68(23):9928–9934.

19.

Davicioni

Triche

Merlino

. The homeoprotein SIX1 transcriptionally activates multiple protumorigenic genes but requires ezrin to promote metastasis. Cancer Res. 2006;66(4):1982–1989.

20.

López-Romero

Garrido-Guerrero

Rangel-López

. The cervical malignant cells display a down regulation of ER-α but retain the ER-β expression. Int J Clin Exp Pathol. 2013;6(8):1594–602.

21.

Caiazza

Ryan

Doherty

Winter

Sheahan

. Estrogen receptors and their implications in colorectal carcinogenesis. Front Oncol. 2015;5:19.

22.

Dey

Barros

Warner

Ström

Gustafsson

JÅ

. Insight into the mechanisms of action of estrogen receptor β in the breast, prostate, colon, and CNS. J Mol Endocrinol. 2013;51(3):T61–T74.

23.

Fogarty

Matulis

Kraus

. Activation of estrogen receptor α by raloxifene through an activating protein-1-dependent tethering mechanism in human cervical epithelial cancer cells: a role for c-Jun N-terminal kinase. Mol Cell Endocrinol. 2012;348(1):331–338.

24.

Chung

Lambert

. Prevention and treatment of cervical cancer in mice using estrogen receptor antagonists. Proc Natl Acad Sci U S A. 2009;106(46):19467–19472.

25.

Saito

Sakamoto

MacLusky

Naftolin

. Gap junctions and myometrial steroid hormone receptors in pregnant and postpartum rats: a possible cellular basis for the progesterone withdrawal hypothesis. Am J Obstet Gynecol. 1985;151(6):805–812.

26.

Lowery

Miller

Dwyer

Kerin

. Dysregulated miR-183 inhibits migration in breast cancer cells. BMC Cancer. 2010;10:502.

27.

Mathe

Scott

Avery-Kiejda

. miRNAs and other epigenetic changes as biomarkers in triple negative breast cancer. Int J Mol Sci. 2015;16(12):28347–28376.

28.

Murad

Kokkinaki

Gunawardena

. miR-184 regulates ezrin, LAMP-1 expression, affects phagocytosis in human retinal pigment epithelium and is downregulated in age-related macular degeneration. FEBS J. 2014;281(23):5251–5264.

29.

Zhu

Leav

Leung

. Dynamic regulation of estrogen receptor-beta expression by DNA methylation during prostate cancer development and metastasis. Am J Pathol. 2004;164(6):2003–2012.

30.

Auvinen

Carpen

Korpela

Ronty

Vaheri

Tarkkanen

. Altered expression of ezrin, E-cadherin and β-catenin in cervical neoplasia. Neoplasma. 2013;60(1):56–61.

31.

Kong

Liu

. High expression of ezrin predicts poor prognosis in uterine cervical cancer. BMC Cancer. 2013;13:520.

32.

Shen

Chen

. Upregulated expression of Ezrin and invasive phenotype in malignantly transformed esophageal epithelial cells. World J Gastroenterol. 2003;9(6):1182–1186.

33.

Díaz

Ceja-Ochoa

Restrepo-Angulo

. Estrogens and human papilloma virus oncogenes regulate human ether-a-go-go-1 potassium channel expression. Cancer Res. 2009;69(8):3300–3307.