Coexpression index of estrogen receptor alpha mRNA isoforms in simple, complex hyperplasia without atypia, complex atypical hyperplasia and adenocarcinoma

Abstract

Objective.

Estrogen receptor isoforms are postulated to play an important role in modulating the estrogen response. To clarify the molecular mechanisms through which malignant changes are activated in endometrium, this study aims to examine the expression profiles of wild-type ER-alpha and their splice variants and to assess the number of coexisting mRNA isoforms of ER-alpha in normal endometrium as well as in endometrial hyperplasia and endometrial endometrioid adenocarcinoma.

Methods.

Human endometrium and specimens including endometrial hyperplasia and endometrial cancer were obtained during surgery. Endometrial data were classified into four groups: simple hyperplasia (n = 24), complex hyperplasia (n = 15), atypical hyperplasia (n = 11), endometrial endometrioid adenocarcinoma (n = 19) (grade 1, grade 2 morphological degree) and proliferative endometrium (n = 24) as a control group. Total cellular RNA was extracted from endometrial tissues using Total RNA Prep Plus. A real-time quantitative RT-PCR assay was developed to quantify the wild-type ER-alpha and ER-alpha mRNA isoforms copy numbers. We have evaluated the variation in ERs mRNA level between normal endometrium and endometrial hyperplasia and adenocarcinoma. We also evaluated the “sharing indicator”. It is a factor of mRNA ER-alpha holding shares in whole mRNA it assume quotient of ER-alpha slicing variant to all variants of mRNA ER-alpha.

Results.

It was found that the number of coexisting mRNA isoforms was significantly higher in adenocarcinoma endometrium than that evaluated for various degrees of hyperplasia endometrium and normal proliferative endometrium (p < 0.05, the Kruskal–Wallis test).

Conclusion.

The risk for progression of endometrial hyperplasia to atypical hyperplasia and eventually endometrioid adenocarcinoma may be accompanied by an increase in the number of alternative splicing variants of mRNA ER-alpha.

Introduction

Estrogens play an integral role in the cyclical changes, pathogenesis and growth of most endometrial hyperplasia and carcinoma [1]. The physiological effects of these steroids are mediated by a ligand-inducible nuclear transcription factor, the estrogen receptor (ER).

Until recently, it was thought that a single ER was responsible for all of the biological actions of estrogens. However, the identification of ER-beta has indicated that the cellular responses to ER ligands are far more complex. Additional ER isoforms, generated by alternative mRNA splicing have been defined in several tissues and are postulated to play an important role in modulating the estrogen response. Although their biological function has not been fully explained, some studies suggest a role of specific isoforms in the pathogenesis of estrogen-dependent breast cancer [2]. ER-alpha splicing variants coding for proteins lacking specific functional domains have been found in estrogen target tissues in human subjects [3], [4], [5], [6], [7]. Several ER-alpha splicing variants might bind wild-type ER-alpha (wtER-alpha) generating heterodimers and may thus be involved in estrogen-dependent pathologies.

It has been shown, using immunohistochemical techniques, that ER-alpha protein lacking exon 5 is present in breast cancer [8]. Rey et al. [9] showed that wtER-alpha was expressed in the endometrium of patients with unexplained infertility. However, deletion of a specific domain due to alternative mRNA splicing may generate variant receptors with altered activity [9]. These variants might upset the ability of wtER-alpha to bind estrogens, thus affecting the formation of stable ER-alpha homodimers or altering its transactivation functions [10]. ERs seem to be the most important factors in the mechanism of estrogen action.

This study showed the changes in ER-alpha and its isoforms expression level in the hyperplasia–atypia–adenocarcinoma sequences in human endometrium.

Although most ER-alpha splicing variants are not translated, alternative splicing could control the level of their functional mRNA present in cells and thus also the expression of functional receptor proteins [9]. Any alterations of composition, structure or quantity of the estrogen receptor have influence on cellular signalization and tissue function.

Several authors suggest that the presence of ER mRNA isoforms might lead to reduced levels of estrogen binding or to the inhibition of wtER binding to its cognate response element. Expression of ERs, as well as of the mRNA splice variants, may influence the biologic properties of some tumors and affect their ability to respond to estrogen and antiestrogen therapies.

It has been hypothesized that this also may play a role in the progression of breast cancer [11], [12], [13]. Most identification of ER-specific mRNA exon deletion variants has been performed on breast tumor samples or on breast cancer cell lines.

The findings of multiple isoforms of ERs and multiple sites on DNA where these receptors can act introduce levels of complexity to the mechanisms of estrogens action that were previously unimagined. Although the tissue distribution of these ER isoforms has not yet been completely explained, the idea that the expression of specific isoforms could be involved in tissue-specific responses to estrogen has not escaped the scrutiny of endocrinologists.

Thus, the accurate determination of ERs and their isoforms expression is very important from the clinical point of view. It is clear that investigation into the role of ERs and their isoforms in endometrium are essential to fully understand the pathomechanisms of endometrium hyperplasia and carcinoma. Only limited information is available about ER mRNA exon deletion variants in benign and malignant endometrium [14]. Both estrogen receptors have been shown to be expressed in the endometrium, and it has been suggested that changes in the relative levels of their isoforms might be associated with various types of hyperplastic changes. ER-alpha splicing variants have been studied in estrogen target tissues, but their expression in various types of endometrial hyperplasia is unknown.

Recently, the tendency has been to classify endometrial carcinoma into two different types [15]. Type I is endometrioid carcinoma (EEC), developing often from complex and atypical endometrial hyperplasia, being more often more well differentiated and noninvasive to superficially myoinvasive and rare production of metastases and is associated with estrogen stimulation. Type II is non-endometrioid carcinoma (NEEC), largely papillary serous carcinomas, arises in atrophic endometria. Type II includes estrogen-independent carcinomas being more aggressive and more prone to metastasis. The molecular alterations involved in the development of type I are different from those of the type II carcinomas and determine their biological features [16], [17]. Type I usually express estrogen receptors, microsatellite instability and mutations in the PTEN, K-ras, beta-catenin genes [18], [19], [20]. Type II has very frequently mutations in the P53 gene, abnormal concentration of P53 protein and loss of heterozygosity [21]. Obviously, the two types may share pathologic and molecular features.

An important aim in understanding of carcinogenesis of endometrium is to define the molecular forms and profile of ER and their isoforms expressed in various types of endometrial hyperplasia and adenocarcinoma.

In this study, we have evaluated the occurrence of mRNA ER-alpha splicing variants in normal endometrium, endometrial hyperplasia (simple, complex and atypica) and adenocarcinoma to enrich our knowledge and to clarify the molecular mechanisms through which malignant changes is activated in the endometrium.