Functional analysis of an alternatively spliced estrogen receptor lacking exon 4 isolated from MCF-7 breast cancer cells and meningioma tissue

Progesterone Receptor Synthesis in Human Meningiomas: Relation to the Estrogen-Induced Proteins pS2 and Cathepsin-D and Influence of Epidermal Growth Factor, Forskolin and Phorbol Ester in vitro

Autonomous expression of progesterone receptors (PR) in human meningiomas is well established. To evaluate whether, similar to progesterone receptors, other estrogen-inducible proteins are also autonomously expressed in meningiomas, concentrations of pS2 and cathepsin-D (Cath-D) were measured in 52 meningiomas. No pS2 protein was detectable in 52/52 tested meningiomas. The Cath-D protein was measurable in all 52 meningiomas, but the mean concentration of Cath-D in meningioma cytosols was 2.4-fold lower than that of a group of 54 breast tumors (p < 0.001). These results indicate that autonomous expression is a PR-related rather than an estrogen receptor-related phenomenon and, consequently, that estradiol is probably not responsible for PR synthesis in human meningiomas. To evaluate the role of other, non-estradiol-dependent signalling pathways in PR synthesis, the effects of EGF, Forskolin and phorbol ester on PR synthesis were tested in vitro. No PR was detectable after the addition of EGF to six different primary cultures. Forskolin and TPA addition caused a morphological change in meningioma cells, but did not induce PR or pS2 synthesis in two different primary meningioma cultures. We conclude that PR synthesis in human meningiomas cannot be triggered by switching on the signalling pathways activated by these growth factors.

Reduced alternative splicing of estrogen receptor alpha in the endometrium of women with endometriosis

Endometriosis is a condition which involves the presence of uterine stroma and glands outside of the uterine cavity and represents one of the most prevalent disorders of the female reproductive tract. The key symptom of endometriosis is pain, including dysmenorrhea, deep dyspareunia, and chronic pelvic pain. As such, endometriosis has significant economic consequences within the healthcare system and can influence the daily quality of life in affected patients. However, the pathophysiology of this disease and the mechanisms in which this condition generates pain are very unclear. This study, involving 30 women with endometriosis and 28 controls without endometriosis, aimed to investigate relative levels of estrogen receptor alpha (ERα) splice variants in the endometrium of women with and without endometriosis and investigate potential links to the severity of pain. Wild type (wt)-ERα was dominantly expressed in human endometrium while the expression of ERα-del.4, ERα-del.7, and ERα-del.3,4 was significantly reduced in endometriosis patients compared with healthy patients (p < 0.05). Furthermore, the relative ratios of wtERα:ERα-del.4, and wtERα:ERα-del.3,4 were associated with the severity of pain in endometriosis patients (p < 0.05). Consequently, analyzing differences in the relative levels of four types of ERα splice variant in the endometrium of patients with endometriosis may help in the development of endometriosis-targeted treatment and the development of appropriate therapies.

Estradiol Membrane-Initiated Signaling in the Brain Mediates Reproduction

Over the past few years our understanding of estrogen signaling in the brain has expanded rapidly. Estrogens are synthesized in the periphery and in the brain, acting on multiple receptors to regulate gene transcription, neural function, and behavior. Various estrogen-sensitive signaling pathways often operate in concert within the same cell, increasing the complexity of the system. In females, estrogen concentrations fluctuate over the estrous/menstrual cycle, dynamically modulating estrogen receptor (ER) expression, activity, and trafficking. These dynamic changes influence multiple behaviors but are particularly important for reproduction. Using the female rodent model, we review our current understanding of estradiol signaling in the regulation of sexual receptivity.

Molecular mechanism of estrogen-estrogen receptor signaling

17β‐Estradiol (E2), as the main circulating estrogen hormone, regulates many tissue and organ functions in physiology. The effects of E2 on cells are mediated by the transcription factors and estrogen receptor (ER)α and ERβ that are encoded by distinct genes. Localized at the peri‐membrane, mitochondria, and the nucleus of cells that are dependent on estrogen target tissues, the ERs share similar, as well as distinct, regulatory potentials. Different intracellular localizations of the ERs result in dynamically integrated and finely tuned E2 signaling cascades that orchestrate cellular growth, differentiation, and death. The deregulation of E2–ER signaling plays a critical role in the initiation and progression of target tissue malignancies. A better understanding of the complex regulatory mechanisms that underlie ER actions in response to E2 therefore holds a critical trajectory for the development of novel prognostic and therapeutic approaches with substantial impacts on the systemic management of target tissue diseases.

β-arrestin regulates estradiol membrane-initiated signaling in hypothalamic neurons

Estradiol (E2) action in the nervous system is the result of both direct nuclear and membrane-initiated signaling (EMS). E2 regulates membrane estrogen receptor-α (ERα) levels through opposing mechanisms of EMS-mediated trafficking and internalization. While ß-arrestin-mediated mERα internalization has been described in the cortex, a role of ß-arrestin in EMS, which underlies multiple physiological processes, remains undefined. In the arcuate nucleus of the hypothalamus (ARH), membrane-initiated E2 signaling modulates lordosis behavior, a measure of female sexually receptivity. To better understand EMS and regulation of ERα membrane levels, we examined the role of ß-arrestin, a molecule associated with internalization following agonist stimulation. In the present study, we used an immortalized neuronal cell line derived from embryonic hypothalamic neurons, the N-38 line, to examine whether ß-arrestins mediate internalization of mERα. β-arrestin-1 (Arrb1) was found in the ARH and in N-38 neurons. In vitro, E2 increased trafficking and internalization of full-length ERα and ERαΔ4, an alternatively spliced isoform of ERα, which predominates in the membrane. Treatment with E2 also increased phosphorylation of extracellular-signal regulated kinases 1/2 (ERK1/2) in N-38 neurons. Arrb1 siRNA knockdown prevented E2-induced ERαΔ4 internalization and ERK1/2 phosphorylation. In vivo, microinfusions of Arrb1 antisense oligodeoxynucleotides (ODN) into female rat ARH knocked down Arrb1 and prevented estradiol benzoate-induced lordosis behavior compared with nonsense scrambled ODN (lordosis quotient: 3 ± 2.1 vs. 85.0 ± 6.0; p < 0.0001). These results indicate a role for Arrb1 in both EMS and internalization of mERα, which are required for the E2-induction of female sexual receptivity.

Nuclear receptor coactivators: regulators of steroid action in brain and behaviour

Steroid hormones act in specific regions of the brain to alter behaviour and physiology. Although it has been well established that the bioavailability of the steroid and the expression of its receptor is critical for understanding steroid action in the brain, the importance of nuclear receptor coactivators in the brain is becoming more apparent. The present review focuses on the function of the p160 family of coactivators, which includes steroid receptor coactivator-1 (SRC-1), SRC-2 and SRC-3, in steroid receptor action in the brain. The expression, regulation and function of these coactivators in steroid-dependent gene expression in both brain and behaviour are discussed.

Transcriptional activity of human brain estrogen receptor-α splice variants: evidence for cell type-specific regulation

Estrogen receptor α (ERα) isoforms with complex types of alternative splicing are naturally present in the human brain and may affect canonical receptor signaling. In the present study we investigated transcriptional activity of common ERα splice variants from this group with different molecular defects: MB1 (intron retention), TADDI (small deletion between exons 3 and 4 with an insert), the Δ (deletion) 3(⁎)-7(*)/819 (complete skipping of exons 4, 5 and 6 and partial deletion of exons 3 and 7) and the Δ3-6 (lacking exons 3, 4, 5 and 6) in HeLa and M17 cells upon stimulation with (17β)estradiol or insulin-like growth factor 1 (IGF-1). In HeLa cells, all these splice variants showed the dominant negative function that was more pronounced for the TADDI. In M17 cells the dominant negative variants appeared to be the MB1 and the Δ3-6, whereas TADDI turned out to be a clearly dominant positive variant. In M17 cells mRNA levels of Δ3-6 and Δ3(*)-7(*)/819 variants increased following (17β)estradiol administration. In Hela cells (17β)estradiol up-regulated the IGF-1 receptor mRNA levels in cultures transfected with MB1, TADDI and Δ3(*)-7(*)/819. Our data demonstrate that ERα splice variants show differential levels of the transcriptional activity in a cell type-specific way and that IGF-1 signaling pathways are differentially employed in a cell-type specific manner depending on the level of the discrete ERα splice variants expressed. Functional properties of various ERα splice variants and their cell type-specificity should, thus, be considered as potential confounders of estrogen therapy effects on the brain.

Identification of novel transcript variants of estrogen receptor α, β and progesterone receptor gene in human endometrium

The human progesterone receptor (PR) and estrogen receptor genes (ESR1 and ESR2) are known to code for a multitude of transcript variants resulting from alternative splicing. Many of them are translated into nuclear receptor proteins with altered structure and function. Expression of these alternative estrogen and progesterone receptors modulates the cellular response to sexual steroid hormones. Recent studies also suggested their significance in development of hormone-dependent diseases like gynecological cancers. We report identification of 12 new transcript variations of the PR, ESR1, and ESR2 gene in human endometrium which result from differential exon-skipping. We succeeded in cloning of four new double or triple exon-deletion transcript variants of ERα, four single, double or triple exon-skipped mRNA isoforms of ERβ, and four new transcript variations of PR gene. Sequence analysis suggested that at least four of them, ERαΔ5/6, ERαΔ5/6/7, PRΔ7, and PRΔ6/7 are translated into receptor proteins which might exert ligand-independent effects on steroid hormone signalling. Comparison of pre- and post-menopausal endometrium revealed differential expression of PRΔ6/7, ERαΔ5/6/7, ERαΔ3/4/5, and ERβΔ1-0N. We also report differential expression of the exon-skipped isoforms in a panel of human cancer cell lines derived from the breast, ovary, and endometrium. Our identification of additional transcript variations further increases the complexity of steroid hormone receptor gene expression and signalling.

Oestrogen receptor splice variants in the pathogenesis of disease

The full-length oestrogen receptor (ER) exists in most vertebrates as two separately encoded isoforms. ER splice variants represent truncated or otherwise modified versions of the full-length alpha or beta isoforms of the parent receptor. ERalpha is found on chromosome 6q and encodes a 595 amino acid protein, while ERbeta is found on chromosome 14q and encodes a 530 amino acid protein. These receptors possess differing ligand affinities, are differentially expressed in a tissue-specific fashion and may act antagonistically. Their altered expression has been implicated in the pathophysiology of a diverse range of conditions from cancer progression in hormone-responsive tissues to neurodegenerative disease. Variously co-expressed with full-length ERs, ER splice variants may have a positive or negative influence on transcription either by modifying the effect of the parent receptor or through their own intrinsic activity. To date, the vast majority of studies have used generic primers or antibodies against the full-length receptors and would not distinguish ER-mediated effects associated with various splice variants. Thus the evidence base of the influence of ER splice variants in normal developmental physiology and in the pathogenesis of disease is weak and greater understanding of their role will undoubtedly lead to new therapeutic strategies for disease intervention and treatment. This review aims to compile the current evidence for the presence of ER splice variants in humans, their physiological roles and clinical sequelae.

Estrogen receptor-alpha splice variants in the human brain

In the present review we discuss recent findings showing that, in addition to the canonical estrogen receptor-alpha (ERalpha), the level of various ERalpha splice variants is changed in the human brain in aging and Alzheimer's disease (AD) at both the mRNA and protein level and that they should be considered for the understanding of estrogen effects on the brain and estrogen therapy pitfalls. Indeed, the expression pattern of certain splice forms is brain area-specific. Thus, the major isoform found in the mamillary body (MB) appeared to be del.7 (deletion of exon 7), while in the hippocampus del.4 (exon 4 omitted) was expressed at the highest level. Furthermore, while transcripts missing exons 7 and 2 declined with aging in the MB of patients >60 years old, no age-related alterations were determined for a number of splice variants in the hippocampus. A novel MB1 isoform with a 168-bp deletion within the transactivation function 1 of ERalpha turned out to accumulate in the histaminergic tuberomamillary nucleus of postmenopausal women. Finally, the level of alternatively spliced ERalpha may also change in AD in a brain area-specific manner and so affect the sensitivity to estrogen therapy.

Oestrogen synthesis, oestrogen metabolism and functional oestrogen receptors in bovine aortic endothelial cells

In order to investigate the mechanisms by which oestrogenic hormones influence the vascular system, we have studied their metabolism and the functioning of oestrogen receptors in bovine aortic endothelial cells from primo-secondary cultures, a widely studied model of vascular pathophysiology. We have demonstrated the enzymic activity of oestradiol-17 beta-hydroxysteroid dehydrogenase, 17-ketoreductase and aromatase in these cells. Immunocytochemical analyses, using two different monoclonal antibodies that recognize epitopes in the A/B domain of the oestrogen receptor, showed that this molecule has a predominantly cytoplasmic localization even after the addition of oestrogen to the culture medium. We showed that the hormone-receptor complexes were functional by demonstrating their transactivating ability in transfection experiments using the luciferase gene reporter and an oestrogen-responsive element transcriptional enhancer, although the amplitude of the response was in the range of only 140-150%: this was not a consequence of the presence of a specific limiting factor, but instead might be related to the peculiar subcellular localization of the oestrogen receptor.

Oestrogens and the structural and functional plasticity of neurons: implications for memory, ageing and neurodegenerative processes

Oestrogens have numerous effects on the brain, beginning during gestation and continuing on into adulthood. Many of these actions involve areas of the brain that are not primarily involved in reproduction, such as the basal forebrain, hippocampus, caudate putamen, midbrain raphe and brainstem locus coeruleus. This paper describes three actions of oestrogens that are especially relevant to brain mechanisms involved in memory processes and their alterations during ageing and neurodegenerative diseases: (1) the regulation of cholinergic neurons by oestradiol in the rat basal forebrain, involving induction of choline acetyltransferase and acetylcholinesterase according to a sexually dimorphic pattern; (2) the regulation of synaptogenesis in the CA1 region of the hippocampus by oestrogens and progestins during the four- to five-day oestrus cycle of the female rat. Formation of new excitatory synapses is induced by oestradiol and involves N-methyl-D-aspartate receptors; removal of these synapses involves intracellular progestin receptors; (3) sex differences in hippocampal structure, which may help to explain differences in the strategies that male and female rats use to solve spatial navigation problems. During the period of development when testosterone is elevated in the male, aromatase and oestrogen receptors are also elevated, making it likely that this pathway is involved in the masculinization of hippocampal structure.

Estrogen receptor alpha and its splice variants in the hippocampus in aging and Alzheimer's disease

Clinical and experimental studies show that estrogens can have beneficial effects on hippocampus-dependent cognitive functions that may be mediated by estrogen receptor (ER)alpha. We investigated whether menopause and Alzheimer's disease (AD) cause changes in this ER subtype. Immunocytochemical staining of ERalpha, aromatase and Golgi complex (GC) was performed on paraffin embedded hippocampal tissue from women of the pre-, peri- and postmenopausal age. Canonical ERalpha mRNA amplicons, ERalpha splice variants (del.2, del.4, del.7, MB1) and aromatase transcripts were measured by Q-PCR in frozen hippocampal samples of AD and matched control cases. Nuclear ERalpha, aromatase and the GC enhanced during aging in women indicating availability of locally synthesized estrogens that may up-regulate ERalpha by which neuronal metabolism can be augmented in the hippocampus after the menopause. In AD cases canonical and alternatively spliced ERalpha mRNA, and aromatase gene expression were down-regulated suggesting a deficit in local estrogen levels and diminished signalling through ERalpha. The major ERalpha splice variants in the hippocampus were found to be MB1 and del.4. A novel ERalpha isoform TADDI was isolated and sequenced from two female patients. It lacks 31 bp at the junction between exons 3 and 4 with an insertion of 13 nucleotides from the middle of the exon 2.

Observations on the presence of E domain variants of estrogen receptor-α in the breast tumors

BACKGROUND AND OBJECTIVES

Estrogen receptor-alpha (ER-alpha) that exists as multiple splice variants, has been widely used as a prognostic marker in the management of breast cancer. Here we have analyzed the hormone binding E domain splice variants of ER-alpha in the breast tumors with reference to the immunoreactive receptor.

METHODS

Thirty breast cancer patients undergoing surgery at the All India Institute of Medical Sciences, New Delhi, were analyzed for the splice variants of E domain by RT-PCR. The ER level was determined by ELISA and the samples were considered positive if the receptor levels were >or= 15 fmol/mg protein.

RESULTS

Our results show that exon 4 and 5 deletions were prevalent in both ER-positive and ER-negative categories. While most ER-positive cases expressed wild-type (wt) exon 6 + 7, nearly 40% of ER-negative cases showed deletion of exon 6 + 7. Therefore, deletion of exon 6 + 7 or masking of epitopes could lead to underestimation of ER by ELISA. All the metastasis and recurrence cases had undetectable levels of ER. A significant number of node-positive cases expressed immunoreactive ER and wt exon 6 + 7 (r = 0.509, P < 0.37).

CONCLUSIONS

Estimation of ER levels combined with composite analysis of ER variants may be a better prognostic marker for breast cancer.

Estrogen receptor-alpha splice variants in the medial mamillary nucleus of Alzheimer's disease patients: identification of a novel MB1 isoform

Previously we have reported an increased nuclear estrogen receptor-alpha (ERalpha) in the medial mamillary nucleus (MMN) in Alzheimer's disease (AD). In the present study, we addressed the presence of specific ERalpha mRNA splice variants in this brain area of five AD cases compared with five controls using the RT-PCR and quantitative RT-PCR approach. Indeed, the occurrence of isoforms with the deletion of exons 7 (del.7), 4 (del.4), or 2 (del.2) was determined in all patients. However, there were no significant differences in the relative transcription levels of each of the mentioned splice variants between AD and control cases, although the ratio of the del.7 isoform to the canonical ERalpha mRNA was higher in controls. Given that exons 7 and 4 encode the ligand-binding domain of the ERalpha, whereas exon 2 encodes the DNA-binding domain, abundant expression of these splice variants suggests that much of the available ERalpha in the MMN of AD and elderly control patients is nonfunctional because they will be unable to bind either the ligand (del.7 and del.4 variants) or the estrogen-responsive elements on appropriate DNA (del.2 variant). Yet, the wild-type ERalpha mRNA appeared to be 2- to 3-fold up-regulated in AD, confirming the rise in the nuclear immunocytochemical staining and pointing to the potential for a beneficial effect of estrogen replacement therapy on the MMN-associated cognitive functions in AD because it represents the availability of potentially functional ERalpha in the MMN. Noteworthy, the expression of the wild-type, del.7, and del.2 mRNAs declined with advanced age in both AD and control patients. Interestingly, we have identified in two AD and two control patients a novel ERalpha splice variant that we called MB1 (mamillary body, exon 1) with a 168-nucleotide deletion corresponding to a U2-type intron inside exon 1 encoding the major portion of the transactivation function 1 domain of the receptor.

Estrogen receptor mutations in human disease

As early as the 1800s, the actions of estrogen have been implicated in the development and progression of breast cancer. The estrogen receptor (ER) was identified in the late 1950s and purified a few years later. However, it was not until the 1980s that the first ER was molecularly cloned, and in the mid 1990s, a second ER was cloned. These two related receptors are now called ERalpha and ERbeta, respectively. Since their discovery, much research has focused on identifying alterations within the coding sequence of these receptors in clinical samples. As a result, a large number of naturally occurring splice variants of both ERalpha and ERbeta have been identified in normal epithelium and diseased or cancerous tissues. In contrast, only a few point mutations have been identified in human patient samples from a variety of disease states, including breast cancer, endometrial cancer, and psychiatric diseases. To elucidate the mechanism of action for these variant isoforms or mutant receptors, experimental mutagenesis has been used to analyze the function of distinct amino acid residues in the ERs. This review will focus on ERalpha and ERbeta alterations in breast cancer.

Expression of estrogen receptor beta exon-deleted variant mRNAs in ovary and uterine endometrium

Various estrogen receptor beta exon-deleted variant (ER-beta EDV) mRNAs were expressed in human ovary and uterine endometrium. Estrogen receptor beta (ER-beta) completely or partially deleted exon n is expressed as ER-beta EnDV or En'DV, respectively. The mRNAs for ER-beta single exon-deleted variant (EDV), ER-beta E2DV, E4DV, E5DV and E6DV; for ER-beta double exon-deleted variants, ER-beta E1'+2DV, E4+5DV and E5+6DV; and for ER-beta triple exon-deleted variants, ER-beta E2'+3+4DV and E4+5+6DV were detected. In ER-beta E2DV, E4+5DV, E5DV and E6DV mRNAs, the new stop codon is made in the exon following the deleted exon(s), and the new proteins may lack the corresponding domains. In ER-beta E1'+2DV, E2'+3+4DV, E4DV, E4+5+6DV and E5+6DV mRNAs, the original stop codon is still present, and the new proteins may conserve the new short amino acid sequences surrounding the deleted exons. ER-beta E1'+2DV, E2DV, E2'+3+4DV, E4DV, E4+5DV and E4+5+6 are unlikely to work as a transcription factors. On the other hand, ER-beta E5DV, E6DV and E5+6DV lack only the ligand-binding domain, and might work as dominant positive or negative factors. Therefore, ER-beta E5DV, E6DV and E5+6DV, constitutively expressed in human ovary and uterine endometrium might, in part regulate estrogen responsiveness.

Steroid receptors and metastatic potential in endometrial cancers

The relative overexpression of estrogen receptor (ER)-alpha exon 5 splicing variant, the disrupted synchronization of ER-beta and ER-alpha expressions, and the suppression of progesterone receptor (PR) form A expression as a transcriptional repressor might be related to metastatic potential of uterine endometrial cancers, leading to poor patient prognosis related to estrogen refractoriness.

The truncated estrogen receptor alpha variant lacking exon 5 is not involved in progesterone receptor expression in meningiomas

Human meningioma tissues are mostly estrogen receptor (ER) negative and progesterone receptor (PR) positive in ligand binding and enzyme immuno assays. To explain this apparently ER independent PR expression, we investigated the existence of a 'hidden' ER variant, which would be capable of activating transcription of the PR gene. Total RNA of seven meningiomas, two breast cancer tissues and of MCF7 cells was analyzed by RT-PCR using primers situated in exon 4 and exon 6. Differential hybridization of the PCR transcripts with probes in exon 4 and 5 respectively, revealed a wild type ER (wtER) fragment and an exon 5 deleted ER variant (ERDelta5). PCR products of two meningiomas were cloned for sequence analysis. The result confirmed the existence of a wtER and ERDelta5.RT-PCR followed by Southern analysis was performed on mRNA of 23 meningiomas to determine the amount of ERDelta5 relative to wtER, which was compared to the PR content of the tissues. In contrast to our initial hypothesis and literature data on breast cancer, there was no relationship between the ERDelta5/wtER ratio and PR protein concentration. It is therefore concluded that ERDelta5 mRNA does not play the dominant role in PR synthesis in meningioma tissue.

Expression of estrogen receptor alpha mRNA and protein variants in human endometrial carcinoma

Breast cancer tissue has been shown to contain alternatively spliced estrogen receptor alpha (ER-alpha) mRNA variants, which have altered biological activities compared to the full-length ER-alpha. The development of endometrial cancer, as well as drug resistance in breast cancer patients undergoing tamoxifen therapy, may represent altered ER-alpha function secondary to specific exon deletions. While the literature is replete with ER mRNA variant data, little information is available regarding the presence and function of endometrial ER variant proteins. We evaluated the presence of human ER-alpha mRNA and protein variants in six premenopausal, six postmenopausal, and six endometrial carcinoma samples. Reverse transcription-polymerase chain reaction, DNA hybridization, and sequencing techniques identified exon 4, exon 5, and exon 7 mRNA splice variants in all patients as well as MCF-7 and Ishikawa cell lines. Presence of translated proteins for full-length ER-alpha, as well as splice variants, was investigated by Western blot analysis using antibodies directed against the N-terminus, hinge region, and C-terminus portions of the ER. These experiments confirmed the presence of immunopositive protein bands of approximately 64-66 kDa in all patients corresponding to wild-type ER-alpha. A protein band migrating at 41 kDa, consistent with an exon 5 splice variant, was only seen in endometrial adenocarcinoma samples. Premenopausal and postmenopausal endometrial samples did not contain detectable amounts of ER splice variant protein. Human ER-alpha mRNA variants are present in all human endometrial samples, but detectable levels of variant proteins are only observed in patients with endometrial adenocarcinoma.

Stage- and region-specific expression of estrogen receptor alpha isoforms during ontogeny of the pituitary gland

The expression time course of estrogen receptor alpha (ER alpha) was analyzed by RT-PCR in fetal and newborn rat pituitaries. In addition to the classical ER alpha messenger RNA (mRNA), three shorter transcripts were detected and subsequently cloned. Sequence analysis showed that they corresponded to ER alpha mRNAs lacking exon 3 (which encodes a zinc finger in the DNA-binding domain), exon 4 (which encodes the nuclear localization signal and part of the steroid-binding domain), or both exons 3 and 4. As analyzed by RT-PCR and ribonuclease protection assay, the respective expression levels of the different transcripts varied dramatically during pituitary development; short forms appeared 4 days before full-length ER alpha mRNA. On Western blots from rat pituitaries of different ages, an ER alpha-specific antiserum labeled four protein bands of the expected molecular weights, revealing that all four ER alpha mRNAs are translated in vivo. Immunocytochemistry, using the same antiserum, showed the ER alpha to be present first in the cytosol of intermediate lobe cells (around embryonic day 16). Only 5 days later, nuclear staining became detectable in the anterior lobe. We argue that the observed cytosolic staining will be essentially due to short ER alpha isoforms, which are indeed more abundantly expressed in the intermediate lobe. These data suggest that during pituitary development, the activity of the ER alpha might be specifically regulated by differential splicing of its primary transcript, resulting in a differential subcellular localization of the isoforms.

Variant estrogen and progesterone receptor messages in human vascular smooth muscle

BACKGROUND

Estrogens stimulate growth of breast or uterine cells but have the opposite effect on vascular smooth muscle cells, in which they protect against coronary artery disease with or without concomitant administration of progesterone. A possible cause of differences in hormone action is variable tissue-specific expression of hormone receptor. Therefore, we analyzed the structure of estrogen receptors (ERs) and progesterone receptors (PRs) in human vascular smooth muscle.

METHODS AND RESULTS

RNA was isolated from human vascular smooth muscle, and the functional domains of ER-alpha and PR were characterized by reverse transcriptase and polymerase chain reaction. Interestingly, in addition to wild-type ER-alpha and PR, 5 variant ER-alpha and 2 variant PR transcripts were found. These variants contained precise deletions of exons encoding regions of the hormone-binding domain. The PR transcripts lacked exon 4 (PRDelta4) and exon 6 (PRDelta6). The ER-alpha transcripts were missing exon 4 (ERDelta4), exon 5 (ERDelta5), exon 6 (ERDelta6), exon 7 (ERDelta7), and exons 6 and 7, (ERDelta6,7). ER-beta variants were also detected. The PR variants were functionally characterized, and PRDelta6 was found to be a dominant-negative transcription inhibitor of wild-type receptors. Variant PR was present in premenopausal women but absent in postmenopausal women.

CONCLUSIONS

Variant PR and ER transcripts are extensively expressed in human vascular smooth muscle. The complex tissue-specific effects of sex hormones may be mediated by the expression of heterogeneous forms of their cognate receptors. The presence of variant ERs and PRs may be of importance in altering the physiological effects of estrogens or progestins in vascular smooth muscle.

Transcriptional regulation by a naturally occurring truncated rat estrogen receptor (ER), truncated ER product-1 (TERP-1)

Truncated estrogen receptor product-1 (TERP-1) is a naturally occurring rat estrogen receptor (ER) variant transcribed from a unique start site and containing a unique 5'-untranslated region fused to exons 5-8 of ERalpha. TERP-1 is detected only in the pituitary, and TERP-1 mRNA levels are highly regulated during the estrous cycle, exceeding those of the full-length ERalpha on proestrus. These data suggest that TERP-1 may play a role in estrogen- regulated feedback in the pituitary. We examined the ability of TERP-1 to modulate gene transcription in transiently transfected ER-negative (Cos-1) and ER-positive pituitary (alphaT3 and GH3) cell lines. In Cos-1 cells transiently cotransfected with TERP-1 and either ERalpha or ERbeta, low levels of TERP-1 (ratios of < 1:1 with ER) enhanced transcription of model promoters containing estrogen response elements by an average of 3- to 4-fold above that seen with ER alone. At higher concentrations of TERP-1 (> 1:1 with ER) transcription was inhibited. TERP-1 also had a biphasic action on transcription in the alphaT3 and GH3 pituitary cell lines, although the stimulatory action was less pronounced. TERP-1 actions were dependent on ligand-activated ER as TERP-1 did not bind estradiol in transfected Cos-1 cells or in vitro, and estrogen antagonists prevented the stimulatory effects of TERP-1. Coimmunoprecipitation studies suggest that TERP-1 does not bind with high affinity to the full-length ERalpha. However, TERP-1 may compete with ER for binding sites of receptor cofactors because steroid receptor coactivator-1 (SRC-1) rescued the inhibitory actions of TERP-1. The ability of TERP-1 to both enhance and inhibit ER-dependent promoter activity suggests that TERP-1 may play a physiological role in estrogen feedback in the rat pituitary.

Estrogen dependency in uterine endometrial cancers

Estrogen dependency in uterine endometrial cancers involves complicated tumor biology. A transformed phenotype of uterine endometrial cancers is supported by estrogen-dependent oncogene (c-Ha-ras, c-fos and c-jun) expressions. The relative overexpression of estrogen receptor exon 5 splicing variant and the damaged expression of progesterone receptor A, which belong to a lack of estrogen dependency, are related to metastatic potential. The estrogen-related metastatic processes, detachment, invasion and angiogenesis in some uterine endometrial cancers can be inhibited by progestins.

Expression of exon 5 deleted estrogen receptor variant messenger RNA in human uterine myometrium and leiomyoma

To examine the relationship between uterine leiomyoma, an estrogen-dependent tumor and its estrogen receptor, the relative amounts of wild type estrogen receptor (WT) mRNA and exon 5 deleted estrogen receptor variant (D5-ER) mRNA to G3PDH mRNA were examined in human uterine myometrium and leiomyoma specimens obtained from 46 patients in 3 age groups (group A: 41-45 years old, group B: 46-50 years old, group C: 51-54 years old) using a quantitative reverse transcription polymerase chain reaction method (RT-PCR). D5-ER mRNA was co-expressed with WT mRNA in all myometrium and leiomyoma specimens. In myometrium, the relative amount of WT decreased with aging, but in leiomyoma, it was high in group B. The relative amount of D5-ER mRNA and the ratio of D5-ER mRNA to WT mRNA (D5/WT ratio) were significantly higher in group C in both myometrium and leiomyoma. The percentage of the patients whose D5/WT ratio was higher in leiomyoma than in myometrium (L/M ratio>1.0) increased with age. These findings suggest that D5-ER increases to supplement the decreasine in WT in uterine tissues toward menopause and that D5-ER plays a more active role in leiomyoma than in myometrium during the perimenopausal period.

Quantitation of estrogen receptor mRNA in breast carcinoma by branched DNA assay

A quantitative nucleic acid hybridization assay for determination of estrogen receptor (ER) mRNA in breast carcinoma is described. The assay, which is based on the branched DNA (bDNA) technology, requires 20 mg of tissue, is simple, highly specific, and reproducible, and correlates reasonably well with an established methodology (r = 0.87). The assay has a dynamic range of 3 x 10(3)-6 x 10(7) copies of ER mRNA per well. ER message as high as 2.5 x 10(6) copies per well could be detected in normal breast tissues. Thus a sensitivity of 3 x 10(3) ER copies per well was sufficient to analyze clinical specimens. In the present studies, accurate measurement of tissue weight enabled direct reporting of the ER mRNA values as the end point results. The bDNA assay provides a useful tool for the detection and quantitation of ER mRNA in research and routine clinical laboratories.

Estrogen receptor variants

Estrogen receptor (ER)3 gene expression in breast epithelium is an intricately regulated event. The human ER gene is transcribed from at least three different promoters which are expressed in a cell- and tissue-specific manner, and result in mRNA isoforms with unique 5'-untranslated exons. The ER is overexpressed in about two thirds of breast tumors, and even in early premalignant breast lesions compared with adjacent normal breast epithelium. Furthermore, normal breast epithelium as well as breast cancer tissue contains alternatively spliced ER mRNA variants where single or multiple exons are skipped. It is still unclear if any or all of the ER mRNA splicing variants are translated in vivo, and if a change in the balance of ER variants could effect tumor development and progression to hormone-independent growth. Although infrequent in primary breast cancer, single amino acid changes within the ER in metastatic disease which might influence cell proliferation may also contribute to neoplastic progression of the mammary epithelium.

Distribution and hormone regulation of estrogen receptor immunoreactive cells in the hippocampus of male and female rats

Estrogen regulates the synaptic plasticity and physiology of the hippocampus as well as learning behaviors that are mediated by the hippocampus. The density of dendritic spines and synapses, the number of N-methyl-D-aspartate (NMDA) binding sites, the levels of NMDA receptor subunit NR1 protein, muscimol binding to the gamma-amino butyric acid (GABA)A receptor, and levels of glutamic acid decarboxylase message in the CA1 region of the hippocampus are altered with estrogen treatment. In addition, some of these parameters exhibit sex differences in their response to estrogen treatment. To establish that estrogen can have a direct effect on the hippocampus and to determine whether or not sex differences in estrogen responsiveness are due to sex differences in estrogen receptor (ER) levels, we used immunocytochemistry with the AS409 antibody to map the location of ER-immunoreactive (ER-ir) cells in the hippocampus of male and female rats. We found that (1) the ERs appear to be in interneurons rather than pyramidal or granule cell neurons, (2) ER-ir cells are located in greatest concentration in the hilus of the dentate gyrus and the stratum radiatum of the CA1 region, (3) the density of ER-ir cells exhibits a rostral to caudal gradient in the hilus and the CA1 regions, (4) there are no sex differences in either the number or immunostaining intensity of ER-ir cells in the hippocampus, (5) the ER levels are down-regulated by estrogen in both male and female rats, and (6) the mean intensity of staining for the ER-ir cells in the hippocampus is about 25% of that in the ER-ir cells of the hypothalamus. From this, we can conclude that estrogen can have a direct effect on hippocampal neurons and that any sex differences in estrogen responsiveness is due to something other than sex differences in ER levels or function in the hippocampus.

Constitutive co-expression of estrogen and progesterone receptor mRNA in human meningiomas by RT-PCR and response of in vitro cell cultures to steroid hormones

Although it is well recognised that human meningiomas are rich in progesterone receptor (PgR), controversy has existed about the presence of the estrogen receptor (ER) in these tumours. We have investigated the presence of both ER and PgR in a series of 20 human meningiomas, spanning the main histological groups, using reverse transcription linked PCR (RT-PCR). Total RNA was extracted from whole tissues and reverse transcribed to yield cDNA. This was amplified using primers specifically designed to detect ER and PgR. All samples co-expressed ER and PgR mRNA, irrespective of tumour classification, patient age or sex. In general, transcripts for PgR appeared considerably stronger than those for ER, and although this was a purely qualitative study, it suggests increased expression of PgR. Addition of exogenous 17beta-estradiol or progesterone to meningioma cell cultures showed that 2/4 cultures responded to these steroids. Our results confirm that human meningiomas do express gene transcripts for ER, and that previous failures to detect ER in these tumours may be due to the lack of sensitivity of the techniques employed. However, these receptors may not be functional in all tumours.

Specific modulation of estrogen receptor mRNA isoforms in rat pituitary throughout the estrous cycle and in response to steroid hormones

We have identified several estrogen receptor (ER) mRNA isoforms in rat pituitary and characterized their regulation by gonadal steroids. The ER mRNAs correspond to splice variants in which either exon 4, exons 3 and 4, or exons 5 and 6 are deleted. A previously isolated pituitary-specific truncated mRNA, TERP-1, containing a unique 5'-end and exons 5 through 8 of the full-length ER, was also studied. The exon deletion variants were expressed in males and females, in pituitary, uterus, testes, heart, hypothalamus, and liver. An antibody to the ER C-terminus bound to full-length (64 kDa) and smaller (50 55 kDa and 40-45 kDa) ER proteins in uterus and pituitary and a pituitary-specific ER of 20-24 kDa corresponding to TERP-1. Estrogen (E) treatment in vivo stimulated full-length ER 2-3-fold, and TERP-1 7-10-fold, but had no effect on any exon deletion variant. Progesterone treatment, alone or with E, had no consistent effect on any ER mRNA form. TERP-1 mRNA was also dramatically and specifically modulated during the estrous cycle, increasing approximately 500-fold between the morning of diestrous and the afternoon of proestrus. Thus, ER mRNA variants exist in estrogen-responsive tissues; the pituitary contains at least one tissue-specific ER which is regulated by steroids and which may contribute to changes in regulated biological activity.

A novel estrogen receptor variant mRNA lacking exons 4 to 6 in breast carcinoma

A novel estrogen receptor mRNA splice variant which lacks the entire exon 5 and part of exon 4 and 6 was identified using reverse transcription PCR in human breast carcinomas. The variant was translated in vitro and produced a protein of approximately Mr 31000 which lacked the ligand binding domain. The binding of the variant estrogen receptor (ER) to a synthetic estrogen-responsive element (ERE) was compared with that of the wild-type ER (wtER). The variant ER bound weakly to the synthetic ERE, both in the presence and absence of estradiol, whilst the wtER bound strongly in the absence and the presence of estradiol. When wtER and variant ER were simultaneously translated in vitro, no heterodimerization was observed using band shift assay. Addition of increasing amounts of variant ER protein to the wtER in the ERE binding reaction showed that the variant protein competed with the binding of the wtER to the synthetic ERE. Furthermore, variant ER are not transcriptionally active. The variant was also expressed in 96% of the 102 breast tumours analysed, of which 62 were tamoxifen-resistant tumours. The expression of this variant was significantly higher (relative to ER) in untreated ER-positive breast tumours which were both progesterone receptor (PgR) negative and pS2 negative phenotype.

Estrogen receptor variants and mutations

There is a large and increasing body of experimental and clinical data supporting the existence or variant estrogen receptor (ER) proteins in both normal and neoplastic estrogen target tissues including human breast. Therefore, future examination of ER signal transduction and/or measurement of ER protein must take into account variant ER expression. The functions of variant ER proteins, either physiological or pathological, remain unclear, although a role(s) for some ER variants in breast tumorigenesis and breast cancer progression would be consistent with the accumulated data. Possible tissue specific expression leads to the speculation that ER variants may have a role in tissue specific estrogen action. The following review focuses on the current knowledge available in the scientific literature with respect to the type and characteristics of estrogen receptor variants and mutations that have been identified to occur naturally in tissues and cell lines.

Oestrogen receptor variants and mutations in human breast cancer

Several oestrogen receptor variant and mutated mRNA species have been identified in human breast samples and cell lines. Over-expression and altered expression of some of these mRNAs have been correlated with breast tumourigenesis and progression. The following review focuses on the current knowledge available in the scientific literature with respect to the type and characteristics of oestrogen receptor variants and mutations that have been identified as occurring naturally in human breast tissues and cell lines.

Estrogen receptor mRNA splice variants in pre- and postmenopausal human endometrium and endometrial carcinoma

Breast cancer tissue has been shown to contain alternatively spliced estrogen receptor (ER) mRNA variants which may result in alternate ER proteins. These ER variants lack specific functional domains and may alter breast cancer cells responses to both estrogen and antiestrogens. Specifically, ER variants might play a role in Tamoxifen resistance in breast cancer patients, as well as the development of endometrial carcinoma, an estrogen-dependent tumor, in patients taking this medication. We investigated the presence of ER variants in normal human endometrium and endometrial carcinoma. Ribonuclease protection assays (RPA) demonstrated ER mRNA variants in the DNA and hormone-binding domains. The reverse transcription-polymerase chain reaction (RT-PCR) assay was used to examine the ER complementary DNA (cDNA) from 25 patients, and generated two major products in both the exon 2 to 5 and 4 to 8 regions. Southern blot analysis of PCR products revealed exon 4 and 7 deletions in all 25 endometria without any qualitative differences in variant expression among premenopausal, postmenopausal, and adenocarcinoma samples. Cloning and sequencing of cDNA variants definitively identified exact deletions of either exon 4 or exon 7. These results demonstrate significant levels of ER mRNA splice variants as well as full-length ER mRNA in normal and neoplastic endometria.

Expression of estrogen receptor wild type and exon 5 splicing variant mRNAs in normal and endometriotic endometria during the menstrual cycle

Expression of estrogen receptor exon 5 splicing variant (ER-E5SV) mRNA contributes to dominant positive properties and metastatic potential of gynecological cancers. Therefore, in endometriotic endometria, which conserve implantation and tumor-like spreading potentials, the expression of ER-E5SV and estrogen receptor wild type (ER-WT) mRNA was analyzed, to investigate the biological implications. Estrogen receptor wild type mRNA in normal endometria was downregulated after ovulation, but not in endometriotic endometria. Therefore the wild type cascade in endometriotic endometria might be partly disorganized. The level of ER-E5SV mRNA in endometriotic endometria was lower than that in normal endometria, regardless of the stage in the menstrual cycle. However, there was no significant difference in the ratio of ER-E5SV to ER-WT mRNA in both normal and endometriotic endometria. On the other hand, the level of ER-E5SV mRNA in both normal and endometriotic endometria did not vary during the menstrual cycle. The dominant positive behavior of ER-E5SV might be masked by the functional cascade of ER-WT in normal endometria, but not in endometriotic endometria. This status might result in an incomplete response to endogenous steroids, and contribute to implantation and spreading potentials of endometriosis.

Expression of estrogen receptor exon 5 splicing variant (ER E5SV) mRNA in gynaecological cancers

Estrogen receptor exon 5 splicing variant (ER E5SV) mRNA has been found in tumours and the corresponding normal tissues, being transcriptionally active without ligand binding. Therefore, the expression of ER E5SV mRNA in gynaecological cancers was studied. The presence of ER E5SV mRNA was demonstrated in the normal ovary, uterine endometrium and cervix and their corresponding cancers. The ratio of ER E5SV/ER WT mRNA expression increased in some cases of metastatic tumour, but did not decrease in any case. Relative overexpression of ER E5SV mRNA might contribute to dominant positive properties and metastatic potential. Therefore, detection of ER E5SV mRNA abundance might be a useful indicator of metastatic potency in gynaecological cancers.

Multiple splicing variants of the estrogen receptor are present in individual human breast tumors

Transcript variants of the estrogen receptor (ER) were investigated in 109 primary breast tumors using reverse transcription-polymerase chain reaction (RT-PCR) and primers allowing analysis of each internal exon. A high incidence of different ER variants was observed, each individual tumor often manifesting multiple variants coexisting with corresponding wildtype (wt) ER. These variants, by sequence analysis confirmed to represent exon splicing deletions, included ER deltaE2 (found in 41% of the tumors examined), ER deltaE3 (74%), ER deltaE4 (72%), ER deltaE5 (66%), and ER deltaE7 (88%). No evidence of transcripts lacking exon 6 was found, although a ER deltaE5,7 variant manifesting simultaneous deletion of exons 5 and 7 was observed. The presence of specific ER variants was not significantly correlated to the status of ER and progesterone receptor (PgR) protein expression, as assessed by routine analysis, although a trend towards a higher incidence of ER deltaE3 and increased expression of ER deltaE7 in ER+/PgR- tumors was observed, suggesting a dominant inhibitory effect on normal ER function to be involved. Moreover, ER deltaE4 was more common in ER+ tumors, possibly due to a cytoplasmic sequestring of this variant lacking a nuclear localization sequence. The presence of ER variants was not associated to clinicopathological variables, and equally frequent in tumors from patients having recurred or remained recurrence-free during adjuvant tamoxifen therapy. In conclusion, although ER splicing variants are abundant in breast cancer, the present study provides no evidence for a direct role of these ER variants in tumor development and tamoxifen resistance. It remains possible, however, that minor cell clones within the tumor, undetected by analysis of tumor homogenates, displaying extreme difference in content of ER variants, could be selected for during therapy or metastasis.

Alternative splicing of the estrogen receptor primary transcript normally occurs in estrogen receptor positive tissues and cell lines

Several laboratories have described estrogen receptor mRNA variants created by skipping internal exons. Some of the putative proteins encoded for by these variants have been functionally characterized by transfection analyses. The variant lacking exon 5 would lead, if translated, to a truncated receptor which shows dominant positive transactivation activity in the absence of hormone. It has been postulated that the variant could account for anti-estrogen resistant tumor growth and for expression of the progesterone receptor in estrogen negative tumors. In order to understand the possible role this and other variants may have in the tumorigenesis of mammary tissue we have carried out a thorough analysis of variants expressed in a tumor cell line (MCF-7), in a tumor sample and in a sample of normal breast tissue derived from mammary reduction surgery. We performed rt-PCR analyses followed by hybridization with exon specific oligonucleotide probes. By these means we have detected nine different variants co-expressed in MCF-7 cells and at least the major variants were equally expressed in normal and neoplastic breast tissue. The same is true for the variant lacking exon 5 which, however, resulted to be a variant of low expression in the three samples analyzed. Variant formation appeared to be restricted to the estrogen receptor messenger since several other members of the superfamily of nuclear receptors did not show variant formation. We also have analyzed the effect of the most abundantly expressed variant, the exon 4 lacking variant, on normal estrogen receptor function, on the growth and on the response to estradiol and to tamoxifen of MCF-7 cells. Although over-expressed at high levels this variant has, if any, only marginal effects on the expression of endogenous estrogen regulated genes and on growth and response to the hormone and its antagonist. Although the lack of function of this variant cannot be extrapolated to other variants, their involvement in tumor formation appears rather unlikely since they are also expressed in normal tissue and the single variant is expressed in addition to many others, some of which might have opposing effects. Variant formation is, however, specific for the estrogen receptor and apparently regulated with tissue specificity as our expression analysis in normal mouse tissues shows. Therefore the variants probably have a physiological significance yet to be discovered.

Human vascular smooth muscle cells express an estrogen receptor isoform

In women, estrogen (E2) exerts a clinically relevant anti-atherogenic effect. The atheroprotective effects of E2 are mediated both by E2-induced changes in systemic factors and by direct effects of E2 on the blood vessel wall. In studies to characterize E2 signaling pathways in vascular smooth muscle cells (VSMC), we recently demonstrated that human VSMC express a functional estrogen receptor [1]. In the present study, we applied a reverse transcription/PCR-based strategy to identify isoforms of the E2 receptor in human VSMC. We now report that in addition to the classical E2 receptor, human VSMC derived from both mammary artery and saphenous vein express an estrogen receptor isoform containing an in-frame deletion of Exon 4 (ER delta 4). RNase protection assays confirm the presence of ER delta 4 message in VSMC and demonstrate it is nearly as abundant as the classical E2 receptor. Transient transfection experiments in VSMC and HeLa cells demonstrate that, in contrast to the classical 67 kDa nuclear-localized E2 receptor, ER delta 4: (a) is a 55 kDa protein that is widely distributed throughout the cell; (b) does not transactivate an E2 response element-driven reporter plasmid in response to E2; and (c) does not modulate transactivation of the ERE-reporter by the classical (wild type) estrogen receptor. Thus, human VSMC express an E2 receptor isoform that does not appear to alter gene transcription. The presence of a novel isoform of the E2 receptor may have important implications for studies of E2-mediated signaling in VSMC.

Oestrogen receptor independent expression of progestin receptors in human meningioma--a review

Human meningiomas are rich in progestin receptors (PR), which are expressed in this tissue in an oestrogen independent fashion. In the search for an explanation of this observation, the existence of a protein in human meningioma cytosol which is capable of binding to a synthetic oestrogen responsive element (ERE) has been demonstrated. Using reverse transcriptase, PCR mRNA encoding for the wild-type oestrogen receptor (ER) was found. In addition, several splice variants of ER mRNA have been identified in human meningioma tissue, including variants lacking exons 4, 5 and 7. We found the ER delta 4 protein to have no transcriptional activity and the ER delta 7 protein reportedly is dominant negative. These mutants therefore probably are not responsible for the autonomous PR synthesis in human meningioma. The ER delta 5 protein, by contrast, has been reported to have oestrogen independent transcriptional activity and it is tempting to speculate that this protein is similar or identical to the ERE binding protein we have found in human meningioma. The role of wild type ER mRNA is presently unclear. Activation of other signal transduction pathways in meningioma does not lead to an increased PR concentration. The promoter area of the meningioma PR gene should be investigated for the possible sensitivity to other transcription factors.

Differential expression of estrogen receptor mRNA splice variants in the tamoxifen resistant human breast cancer cell line, MCF-7/TAMR-1 compared to the parental MCF-7 cell line

Breast cancer patients with an estrogen receptor (ER) positive tumor can be treated with the anti-estrogen tamoxifen, but development of anti-estrogen resistance is a serious problem. We have analyzed a tamoxifen resistant human breast cancer cell line MCF-7/TAMR-1 for alterations in ER which might explain the tamoxifen resistance. The MCF-7/TAMR-1 cells expressed both wild-type ER mRNA and protein, and by RT-PCR we were able to clone ER cDNAs corresponding to the following mRNA splice variants: ER delta E2, ER delta E4, ER delta E5, ER delta E7 and a new double splice variant lacking both exon 4 and 7 (ER delta E4,7) The existence of the ER delta E4,7 variant was confirmed by RNase protection assay. Semi-quantitative RT-PCR revealed that ER delta E2 mRNA was expressed at a higher level in MCF-7/TAMR-1 cells, whereas the ER delta E5 mRNA was expressed at a significantly lower level in MCF-7/TAMR-1 cells compared with MCF-7 cells. The differential expression of the two ER mRNA splice variants indicates that they may be involved in anti-estrogen resistance, although the present knowledge of their biological function does not provide us with an explanation.