Progesterone Receptor Action in Leiomyoma and Endometrial Cancer

Abstract

Progesterone is a key hormone in the regulation of uterine function. In the normal physiological context, progesterone is primarily involved in remodeling of the endometrium and maintaining a quiescent myometrium. When pathologies of the uterus develop, specifically, endometrial cancer and uterine leiomyoma, response to progesterone is usually altered. Progesterone acts through mainly two isoforms of the progesterone receptor (PR), PRA and PRB which have been reported to exhibit different transcriptional activities. Studies examining the expression and function of the PRs in the normal endometrium and myometrium as well as in endometrial cancer and uterine leiomyoma are summarized here. The clinical use of progestins and the transcriptional activity of the PR on genes specific to endometrial cancer and leiomyoma are described. An increased understanding of the differential expression of PRs and response to progesterone in these two diseases is critical in order to develop more efficient and targeted therapies.

I. Introduction

The progesterone receptor (PR) has been the focus of extensive analysis over the past few decades given its significance in reproductive tissues. The uterus is one of the most highly responsive organs to progesterone. Based on PR function, certain modalities of treatment for uterine pathologies have involved synthetic progestins or selective progesterone receptor modulators (SPRM). These compounds have proven to be effective in certain cases of endometrial cancer or uterine leiomyoma. Studies investigating the expression of PRs, and action of progesterone through its receptor in endometrial cancer and leiomyoma are summarized here. A brief description of PR expression and progesterone action in the normal endometrium and myometrium followed by a description of the clinical studies using progestins and SPRMs and the transcriptional activity of the PR on genes specific to endometrial cancer and leiomyoma will be presented.

II. The Uterus

The uterus is the major female reproductive organ where the fetus develops during pregnancy. During development, the uterus develops from the middle to upper portion of the paramesonephric duct, also known as the Mullerian ducts.¹ The uterus further organizes into distinct layers: the outer-most layer which consists of smooth muscle is the myometrium and the innermost layer which lines the uterine cavity is the endometrium (Fig. 1A). The endometrium consists of a layer of columnar luminal epithelium supported by cellular stroma containing tubular glands (Fig. 1B). The luminal and glandular cells of the endometrium originate from the paramesonephric duct epithelium while the stroma originates from the mesenchyme surrounding the urogenital ridge. It is also from this mesenchyme that the myometrium forms. The myometrium consists of an organized network of smooth muscle cells with supporting stromal and vascular tissue (Fig. 1B). During pregnancy, the myometrium stretches by expanding the size and number of the smooth muscle cells and contracts in a coordinated fashion during labor. After pregnancy the uterus returns to its nonpregnant size. Both the endometrium and myometrium are highly responsive to the steroid hormones, estrogen and progesterone, and represent one of the most dynamic sites of hormone action during the menstrual cycle and pregnancy.

Fig. 1.

(A) The human uterus is comprised of an outer smooth muscle layer termed the myometrium and the innermost layer which lines the uterine cavity termed the endometrium. (B) Cross section of human uterine tissue shows distinct morphology of the myometrium and endometrium. The myometrium consists of smooth muscle cells with supporting stroma and vasculature. The endometrium consist mainly of epithelial glands and stroma.

III. Progesterone Action on the Endometrium and Myometrium

A. Physiological Response to Progesterone

The ovary is the major source of estrogen and progesterone in the human, synthesizing and secreting these hormones in a cyclical fashion.² Granulosa cells from developing primary follicles biosynthesize and secrete estrogen and after ovulation these granulosa cells mature to form the corpus luteum which actively secretes progesterone and estrogen during the secretory phase of the menstrual cycle. If there is no pregnancy, the corpus luteum regresses resulting in the decline of estrogen and progesterone levels. If there is a pregnancy, the corpus luteum continues to grow and function for several months, after which time, it will regress as the placenta begins to synthesize estrogen and progesterone.

The endometrium undergoes extensive remodeling in response to ovarian steroid hormones. Estrogen promotes proliferation and growth of the endometrial lining while progesterone antagonizes estrogen driven growth as well as promotes differentiation in preparation of an impending implantation.³ Specifically, when progesterone levels are high during the luteal phase of the menstrual cycle, the glandular epithelium transforms from relatively inactive cells full of free ribosomes to very active polarized cells, containing giant mitochondrial profiles, intracellular deposits of glycogen/glycoprotein-rich material, and a complex intranuclear channel system.⁴ Morphologically, the glands become tortuous and have large lumens due to increased secretory activity. In parallel, the underlying stroma becomes very edematous as a result of increased capillary permeability and the cells begin to appear large and polyhedral, a transformation process termed decidualization. Decidualization begins in the stroma around the spiral arteries when progesterone levels are high during the midluteal phase, and spreads to the upper two-thirds of the endometrium.⁵ If embryo implantation occurs, the reaction is intensified and becomes the decidua of pregnancy. The decidualized cell biochemically expresses new proteins and two of the most abundantly secreted proteins are insulin-like growth factor binding protein-1 (IGFBP1) and prolactin (PRL) and as a result are considered markers of decidualization.⁶ If there is no pregnancy, levels of estrogen and progesterone decline due to the regression of the corpus luteum causing the upper two thirds of the endometrium to be shed.

The overall effect of progesterone on the myometrium is to maintain quiescence in both the pregnant and nonpregnant uterus. The nonpregnant uterus contracts throughout the menstrual cycle.⁷ During the late follicular phase, a progressive increase in uterine contractility parallels the rise in estradiol levels. After ovulation, during the luteal phase, the uterus undergoes a characteristic period of quiescence, strongly implicating progesterone as a mediator of this quiescence. At menses, after circulating levels of progesterone decrease, contractility increases and participates in the emptying of the uterine contents. In the pregnant uterus, it is well accepted that progesterone keeps the myometrium quiescent in order to promote and sustain pregnancy. Progesterone promotes myometrial relaxation and thought to actively block the transformation of the myometrium to a contractile phenotype. Although circulating levels of progesterone do not decrease before labor onset,^8-10 the withdrawal remains a principal mechanism for the control of human parturition. Synthetic progesterone antagonists, such as RU486 initiate myometrial contractions at all stages of pregnancy.^11,12

B. Progesterone Receptors

The physiologic actions of progesterone are mediated by interaction with the PR, a member of the nuclear hormone receptor superfamily of ligand-activated transcription factors.^13,14 There are two predominant PR isoforms, designated PR-A and PR-B, transcribed from the same gene by two distinct promoters, with the only difference being that human PR-B is larger by an additional 164 amino acids at the amino terminus^15-17 (Fig. 2). As a result, PR-A and PR-B have distinct transcriptional activities.^18-25 Three activation function (AF) domains have been identified in PR as AF1, AF2, and AF3. In many contexts, PRB functions as an activator of progesterone-responsive genes, while PRA is transcriptionally inactive. In addition, PRA also functions as a strong transdominant repressor of PRB as well as the human estrogen receptor (ER) transcriptional activity.²³ The precise mechanism underlying the differential activities of the two PR isoforms is not fully understood. Studies have suggested that PRA and PRB adopt different conformations within the cell which may contribute to its different transactivation functions. Tetel et al.¹⁹ demonstrated that the interaction of the amino terminus to the carboxyl terminus in PRB and PRA is different. Giangrande et al.²⁴ demonstrated that PRA is unable to efficiently recruit the coactivators SRC-1 and GRIP1 upon agonist binding despite the fact that both PRA and PRB contain sequences within the LBD that binds coactivators. In addition, PRA interacts efficiently with the corepressor SMRT permitting it to function as a transdominant repressor.²⁴

Fig. 2.

Functional domains of PRB and PRA. PRA lacks the 164 amino acids in the N-terminus and thus the activation function (AF)3 domain. The DNA binding domain (DBD), ligand binding domain (LBD) AF1 and AF2 regions are present in both PRA and PRB.

At the promoter level, PR binds to a palindromic consensus sequence called the progesterone response element (PRE). It is thought that multisite PREs promote a more stable complex.^23,26 Surprisingly, a survey of proximal promoter regions of endogenous genes regulated by PR reveal a lack of tandem palindromic PREs and an abundance of PRE half-sites. This suggests that additional mechanisms are in place for PR recruitment to specific sites and activation of genes. Reports have shown that for the glucocorticoid receptor (GR), “composite” response elements which consist of a single hormone response element in tandem with heterologous binding sites leads to synergy between the GR and a variety of factors, including Sp1, NF1, CACCC-box, and AP1.^27-29

C. Coregulators of PR

Nuclear receptors recruit the coregulators that perform all of the subsequent reactions needed to induce or repress expression of genes. Coactivators, such as the SRC family (steroid receptor coactivator) and CBP/p300, enhance transcription by liganded receptors while corepressors (such as SMRT and NCOR) repress transcription.³⁰ These coregulators exist and function in large multiprotein complexes which are recruited to the target gene in rapid sequence by nuclear receptors. This complex contains many enzymes that are required for gene expression. Such enzymatic reactions for transcription include chromatin modification and remodeling, initiation of transcription, elongation of RNA chains, RNA splicing, and termination of the transcriptional response. Consequently, it is suggested that genes encoding for coregulators of hormones receptors are the true master genes of eukaryotes.³⁰

It has been shown that upon ligand treatment, PR interacted preferentially with SRC-1, which recruited CBP and significantly enhanced acetylation at K5 of histone H4.³¹ In contrast, activated GR preferentially associated with SRC-2 (TIF-2/GRIP-1), which subsequently recruited pCAF and led to specific modification of histone H3, suggesting that specific coactivators are differentially recruited by different steroid hormone receptors which then recruit distinct histone acetyltransferases to modulate the transcription of steroid-responsive genes.

Investigation of the SRC family of coactivators, which consist of SRC-1, SRC-2, and SRC-3 in mice has demonstrated its relevance in steroid responsive tissues. SRC-1 knockout mice, although viable exhibited decreased responses to steroid hormone treatment.³² Ablation of SRC-2 resulted in a partial lethal phenotype due to death in the first month of life.³³ Those that survived to adulthood showed slowed growth and hypofertility due to placental defects. Female SRC-3 null mice have reduced fertility. However, SRC-3 is not expressed in the mouse endometrium and the uteri of SRC-3^−/− mice are able to undergo the artificially induced decidual reaction.³⁴ Thus, it was concluded that SRC1 and SRC2 are the critical members of the p160 coactivator family for the regulation of uterine function.³⁴ Recent studies have demonstrated the role of SRC-2 in the adult uterus using a floxed allele of SRC-2 crossed to the PR-Cre mouse which abrogated SRC-2 function only in cell lineages that express the PR.³⁵ Absence of SRC-2 in PR-positive uterine cells was shown to result in infertility due to an early block in embryo implantation. The uterus of these mice was unable to undergo the necessary cellular and molecular changes that precede complete progesterone-induced decidual progression. The expression of a number of decidualization markers, Bmp2, Cox2, and follistatin was significantly reduced in the partially decidualized PR(Cre/+) SRC-2(flox/flox) mice. While Bmp2 induction was negligible, Cox2 and follistatin were partially induced, which was suggested to be due to SRC-2 being essential for the induction of pathways that lead to Bmp2 expression but that additional coregulators may be required for elaborating the Cox2 and follistatin expression pathways. The incomplete decidual response shown in both the PRCre/+SRC-2flox/flox mouse and the SRC-1KO³² suggests that both SRC coregulators may be required together in PR-mediated signaling cascades that result in a fully decidualized uterus. To support this hypothesis, removal of SRC-1 in these PR(Cre/+) SRC-2(flox/flox) mouse uterus resulted in the complete absence of a decidual response, confirming that both uterine SRC-2 and -1 cooperate in progesterone-initiated transcriptional programs.

It is also widely observed that other transcription factors are able to interact with and modulate function of nuclear receptors. Reports have shown that for GR, PR, and ER-alpha, other transcription factors, including members of Sp, NF, CACCC-box, and AP1 families can bind to response elements that occur in tandem to the nuclear receptor response element allowing synergy/antagonism between the steroid receptors and the other transcription factors.^27-29 In the absence of canonical PRE sequences, PR can tether to some transcription factors such as Sp1, Ap-1, Stat5, p65 subunit of NF-κB,^36-38 and FOXO1.^39,40 Through mouse models, it has been demonstrated that FKBPs, which are immunophilins interact with PR and influence its localization.⁴¹ Specifically, FKBP4 and FKBP5 interact with PR and are expressed in the uterine stroma during implantation. Furthermore, the FKBP4 knockout mice are infertile due to the inability to support implantation or undergo adequate decidualization.

In one pioneering study, a genome-wide scan of chromosomes 21 and 22 was performed in order to identify ER binding regions.⁴² In doing so, the authors discovered that FOXA1 was necessary for mediating the estrogen response in breast cancer cells. Furthermore, the FOXA1 binding site was the most conserved motif proximal to the regions that had an ER element. Another forkhead protein, from a different subfamily, FOXO1 can interact with steroid hormone receptors such as ER-alpha, retinoic acid receptor, thyroid hormone receptor, and PR and elicits either repressive or activating effects on nuclear receptor mediated gene expression.^39,43,44 In endometrial fibroblasts, FOXO1 and PR interacted with each other and bound to tandem DNA sequences in the IGFBP1 promoter.^39,40 In addition, FOXO1 modulated PR transactivation of a PRE responsive reporter.³⁹ Gene array studies revealed that many of the genes significantly regulated by FOXO1 during decidualization of endometrial stromal cells are also dependent on PR.⁴⁰ Given that both transcription factors are involved in important cellular processes in the endometrium associated with growth inhibition and differentiation, the cross talk between these two molecules may be an important mode of endometrial remodeling.

D. Progesterone Receptors in the Endometrium and Myometrium

Progesterone is central to the remodeling that occurs in the endometrium for uterine receptivity and acts on both the epithelial and stromal compartments. Studies in mice with selective ablation of PR isoforms revealed that PR-A is necessary for ovulation and modulates the antiproliferative effects of progesterone in the uterus while PR-B is required for normal mammary gland development and function.^45,46 Recent evidence has confirmed the existence of a functional third isoform, designated PR-C, which appears to play a critical role in the onset of parturition.⁴⁷ The presence of multiple PR isoforms potentially increases the specificity and versatility of hormone action in a target tissue. PRs A and B are expressed in cells of the endometrium and its expression is dependent on the hormonal status and cell type. In the glandular epithelium, PR expression is stimulated by estrogen during the proliferative phase but is downregulated by its own ligand in the secretory phase. Prior to ovulation, PR-A and PR-B levels are approximately equivalent in glandular epithelium but only PR-B persists in these cells in the mid-secretory phase, suggesting that PR-B is most important for the progesterone driven phenotypic changes in the glands at this time. In the stroma, PR-A is the dominant isoform throughout the cycle.^48-50 Conditions associated with endometrial pathology such as endometrial cancer is due to inadequate progesterone response as subsequently described.

Myometrial expression of PR is less dynamically regulated than the endometrium during the menstrual cycle. Studies have demonstrated that both PRA and PRB are expressed and levels are consistent throughout the menstrual cycle.^51,52

IV. Endometrial Cancer

Endometrial cancer is the most common cancer of the female reproductive tract, with estimated 39,080 new cases diagnosed in 2007. Despite the frequent detection of early-stage cancers and the evolving use of adjuvant chemotherapy for advanced disease, the death rate from this malignancy has increased and currently claims 7400 lives among US women per year.⁵³ The incidence of endometrial cancer is rising as life expectancy increases and as key risk factors, including obesity, become more prevalent. A better understanding of the pathophysiology underlying endometrial cancer is the first step to identifying key biomarkers that can improve diagnostic efforts and prevent development of this disease.

Endometrial cancer is diagnosed by pathological examination. Approximately 80% of all endometrial carcinomas are endometrioid type, which arise from endometrial glands (Fig. 3A). The malignant phenotype and the varying degrees of differentiation are easily recognizable by microscopy.⁵⁴ Endometrioid carcinoma is graded histologically from grade 1 to grade 3 depending on the percentage of solid nonsquamous areas and cytological atypia.⁵⁵ Most endometrioid carcinomas are well to moderately differentiated and present alongside hyperplastic endometrium. These tumors, also known as type 1 endometrial carcinomas, are associated with chronic exposure to estrogen and a lack of opposing progesterone. One source of estrogen is fat tissue, in which peripheral androgens act as aromatase substrates to produce estrone. Estrone is then converted to estradiol by 17-hydroxysteroid dehydrogenase.⁵⁶ Through the androgen-to-estrone pathway, postmenopausal women produce approximately 100 mg of estrone per day, or more if they are obese.⁵⁷ Chronic estrogen exposure is also exacerbated by comorbid conditions in obese women, namely hypertension and diabetes. Polycystic ovarian syndrome (PCOS) is also associated with higher levels of estrogen and androgen, as well as low levels of opposing progesterone, leading to a higher risk of endometrial cancer in these patients. Finally, the long-term use of tamoxifen, with its paradoxical estrogen agonist effect in the uterus, can also lead to type I endometrial cancer.⁵⁸

Fig. 3.

Endometrial cancer and uterine leiomyoma. (A) Endometrial adenocarcinoma arises from the endometrial glands exhibiting malignant behavior. (B) Uterine leiomyomas arise from benign overgrowth of smooth muscle cells and can range dramatically in size.

Type II endometrial cancer occurs primarily in elderly postmenopausal women, and is neither related to estrogen nor preceded by endometrial hyperplasia. Type II tumors are high grade tumors with serous or clear-cell morphology and carry a poor prognosis. Other morphological variants of endometrial cancer are placed into this category but occur at much lower rates.^58,59 At the molecular level, type 1 tumors are commonly associated with abnormalities of DNA-mismatch repair genes, including k-ras, PTEN, and beta-catenin. Type 2 tumors are associated with abnormalities of p53 and HER2/neu, although they are not present in all cases.⁵⁸

V. Progesterone Receptor Action in Endometrial Cancer

A. Progestin Therapy in Women

Studies have proven the efficacy of progesterone in protecting the endometrium against the hyperplastic effects of estradiol by inducing glandular and stromal differentiation.⁶⁰ Accordingly, progestins play an essential and effective role in the management of endometrial hyperplasia.^61,62 In a study of 52 postmenopausal women diagnosed with atypical hyperplasia or hyperplasia without atypia, 90% of patients had complete remission after treatment with 40 mg megestrol acetate per day for 42 months.⁶³ However, close follow-up is usually recommended in women who are treated with progestins especially for atypical hyperplasia since there is a significant increased risk of progression to carcinoma.⁶⁴ Progestins have been used as adjuvant for endometrial cancer in hopes to prevent recurrence. However, several studies have shown that progestin treatment is not beneficial to the overall survival of women postsurgery.^65,66 Progestins are also used as primary therapy, especially for premenopausal women as a fertility-sparing treatment. Approximately 25% of endometrial cancer cases affect premenopausal women, particularly in the setting of obesity, chronic anovulation, and polycystic ovary syndrome.^62,67-70 Progestin therapy would only be given when the tumor is well differentiated with positive receptors. There are few studies looking at the efficacy of progestin treatment in these women with the majority of published studies being case reports. In a review of articles published between Jan 1966 and Jan 2007 describing patients with endometrial cancer treated with hormonal therapy, 133 patients were identified, who were treated for an average duration of 6 months, and who demonstrated an average response time of 12 weeks.⁷¹ Of these 133 patients, 51% demonstrated a lasting complete response, 25% showed a temporary response, and 24% never responded to treatment. It is evident that a larger study is required to demonstrate the true benefit of progestin therapy in endometrial cancer. In regards to sparing fertility, there is no doubt that progestin therapy can be used;⁷² however, close follow-up is required because of the substantial rate of recurrence.

B. Progesterone Receptors in Endometrial Cancer

Morphological and biochemical evaluations demonstrated that in endometrial cancer, PRA is localized to the nucleus, even in the absence of progesterone.⁷³ In contrast, a large proportion of PRB is cytoplasmic in the absence of ligand, but is rapidly translocated to the nucleus in the presence of progesterone. All endometrial cancer specimens demonstrated cytoplasmic PRB in 50% or more of the cells, and five of the seven tumors that were moderately to poorly differentiated demonstrated no PRB staining in the nuclei. Nuclear PRB was thus significantly associated with increasing tumor differentiation. PRA and PRB exhibit different activating properties and mediate the transcription of different sets of genes in endometrial cancer cells. Smid-Koppman et al.⁷⁴ demonstrated that in the presence of progesterone, PRB expressing Ishikawa cells displayed almost complete inhibition of cell growth, while PRA expressing Ishikawa cells only displayed 50% inhibition of cell growth. In an additional study by Hanekamp et al.,⁷⁵ it was demonstrated that while PRB expressing Ishikawa cells caused more tumor growth in vivo than PRA expressing Ishikawa cells, tumor growth was inhibited after administration of MPA only in the tumors expressing PRB. There is an ongoing debate as to the PR status in endometrial tumors with one study suggesting that PRB is predominant in advanced endometrial tumors,⁷⁶ another study pointing to the loss of both isoforms in advanced endometrial cancer,⁷⁷ and a third study that indicates only PRA is expressed in poorly differentiated endometrial carcinoma cell lines.⁷⁸

C. Genes Regulated by Progestins in Endometrial Cancer

Nevertheless, in vitro studies have clearly demonstrated the efficacy of progestins to influence endometrial cancer cell behavior. When endometrial cancer cells are transfected with specifically PR-A or PR-B, progesterone can promote cell cycle inhibition, endometrial cancer cell invasion, differentiation to a secretory phenotype, induction of replicative senescence, and can down-regulate the expression of cellular adhesion molecules.^79,80 Regulation of genes such as cyclin D1, matrix metalloproteinase-1 (MMP-1), -2, -7, and -9, and Ets-1 in response to progestins have been implicated to mediate the inhibition of cell growth and invasiveness.⁸¹ Primary cells from endometrial tumors also respond to progestins by significantly reducing proMMP-9, proMMP-2, and MMP-2 release.⁸² Progestins have been shown to induce glycodelin expression in Ishikawa cells^83-85 which causes inhibition of G1/S progression and upregulation of CDKIs thereby reducing cell proliferation.⁸⁶ Progestins can increase FOXO1 protein levels in Ishikawa cells, specifically through PRB⁸⁷ and promote cell cycle arrest and apoptosis in these cells. Interestingly, levels of FOXO1 protein are dramatically lower in 77%⁸⁸ or 95.9%⁸⁷ of endometrial tumor tissues studied compared to normal tissues from cycling endometrium. Shiozawa et al.⁸⁹ reported that p27 expression in hyperplastic epithelia was negligible before MPA treatment, whereas it was greatly increased after treatment. Watanabe et al. demonstrated that it is through PRB that p21 and p27 expression increases.⁹⁰ Microarray studies revealed that short term (4 h) and high dose (30 μg/ml)) exposure of Ishikawa cells to progesterone result in 247 differentially expressed genes of which 126 were upregulated and 121 were downregulated. Of these, 135 genes were involved in biological processes like cell cycle, cell proliferation and differentiation, developmental processes, immune responses, intracellular protein traffic, and transport.⁹¹ Hanekamp et al.⁹² reported that MPA inhibits expression of several metastasis-related genes in a set of endometrial cancer cell lines. Treatment of Hec50co cells transfected with PR with progesterone for 12 h significantly regulated genes associated with cell signaling, DNA remodeling, apoptosis, tumor-suppressor, and transcription factors. Interestingly, there was a consistent modulation of cytokines consistent to an antiinflammatory environment. Specifically, proinflammatory genes such as TNFalpha, IL-1beta, and MCP-1/MCAF-1 were downregulated and antiinflammatory genes such as TRAP1 and SMAD4 were induced⁹³. Progestins have been shown to modulate proteins in the apoptotic cascade in human endometrial precancers. Women with hyperplasia treated with either system MPA or a levonorgestrel intrauterine device exhibited increased apoptosis in the glandular cells with decreased expression of the antiapoptotic genes, Bcl-2 and BAX.⁹⁴ Overexpression of PRA and PRB in endometrial cancer cells resulted in a significant progesterone-dependent inhibition of expression of a cadre of cellular adhesion molecules, including fibronectin, integrin alpha3, integrin beta1, integrin beta3, and cadherin 6.^79,95 Thus, it is apparent that progesterone, through PRA and PRB modulate genes that are involved in processes associated with cell cycle, apoptosis, cell adhesion, differentiation, and inflammation in order to regulate endometrial cancer cell behavior.

D. Transcriptional Activity of Progesterone Receptors in Endometrial Cancer

It has been demonstrated that transcriptional activation by progesterone can involve other transcription factors. For example, liganded PR decreases the transcriptional activity of the activating protein-1 (AP-1) transcription factor family, and in particular, c-Jun. In addition, progesterone strongly inhibited total AP-1 as well as c-Jun recruitment to the cyclin D1 promoter, whereas it enhanced AP-1 occupancy on the p53 and p21 promoters, as shown by chromatin immunoprecipitation assays. This study concluded that in endometrial cancer cells, modulation of AP-1 activity is a potential pathway of progesterone-induced growth inhibition in endometrial cancer cells.⁹⁶

Another mechanism of progesterone action has been proposed to involve inhibition of NFkappaB transcriptional activity. Specifically, expression of A20 and ABIN-2 were induced through PRB and these factors bind in a complex and inhibit NFkappaB transcriptional activity.⁹⁷ EMSAs revealed the complete inhibition of NFkappaB dimer binding to DNA by both PRA and PRB. The inhibition of NFkappaB and its tumorigenic inflammatory and antiapoptotic effects by PR may be one pathway by which progesterone treatment is effective against endometrial hyperplasia and cancer.

Glycodelin (GdA) is a progesterone induced gene in normal endometrial epithelial cells and endometrial cancer cells. Studies have shown that ligand-activated PR stimulates GdA promoter activity through functional Sp1 sites.⁹⁸ As on numerous other genes, PR can tether to Sp1 to regulate promoters that do not have PRE sequences.

In another study, it was shown that progesterone upregulated COMT protein expression in Ishikawa cells primarily through PRA. COMT converts genotoxic catecholestrogens to anticarcinogenic methoxyestrogens (2-ME2) in the endometrium. COMT promoter activity was differentially regulated by the three half-site PREs. Accordingly, high doses of 2-ME2 inhibited Ishikawa cell proliferation.⁹⁹

A novel mechanism for PR-A and PR-B mediated gene transcription in the uterus has been proposed to involve selected KLF members. Specifically, Kruppel-like factor 9 (also known as BTEB1) interacts with ligand-activated PRB to increase PRB transactivity. This facilitates the recruitment of the transcriptional integrator CREB-binding protein within the PR dimer, and is dependent on the structure of the ligand bound by PRB. By contrast, BTEB1 does not influence agonist bound PRA transactivity, but augments PRA inhibition of PRB-mediated transactivation. Also, BTEB1 potentiates ligand-independent PRA transcriptional activity in the presence of CREB-binding protein. Similar observations were made with the BTEB1-related family members Krüppel-like family (KLF) 13/FKLF2/BTEB3 and Sp1 on PRB transactivity.¹⁰⁰

VI. Conclusions and Perspectives of Progesterone Action in Endometrial Cancer

Endometrial adenocarcinoma is highly associated with unopposed estrogen action. The significance of progesterone in preventing estrogen-driven proliferation is underlined by its efficacy in eradicating endometrial hyperplasia and some endometrial cancers. While its role in preventing endometrial cancer may involve independent mechanisms to those that promote tumor cell death and regression, it is obvious that progesterone action is complex and involves numerous pathways and players. Despite the limitations of in vitro systems, which utilize endometrial cancer cell lines that have been propagated over many years, endometrial cancer cell behavior in response to progestins and the specific genes that are regulated have proven to be remarkably similar in these cell lines as those grown in xenograft models as well as in the tumor behavior from women. Thus far, in vivo and in vitro studies have shown that progesterone, through its receptor can regulate genes associated with cell cycle, apoptosis, cell adhesion, differentiation, and inflammation. Progesterone binds to either receptor A or B, which can then bind to specific sequences on promoters, in the presence of numerous other coregulators. Depending on the PR isoform and the predominant coregulators it associates with, progesterone can enhance or repress transcription of genes. The complexity of PR action is demonstrated by its differential action depending on the promoter region, length of progesterone exposure, and cell type. Further investigation is required to elucidate mechanisms of PR action in endometrial cancer.

It is also noteworthy that progesterone is key in differentiating the endometrial stroma. While many studies have focused on progesterone action in the stroma in the normal cycling uterus as it pertains to pregnancy and fertility, very little is known on its role in the stroma as it pertains to endometrial cancer. Evidence is strong that progestins are effective in treating endometrial hyperplasia and some endometrial cancers and studies have demonstrated that progestins can regulate endometrial cancer cell behavior. Given the significant role that progesterone plays in the stroma, it would be worth investigating how progesterone acts through the stroma to influence endometrial cancer cells. Thus, progesterone responsiveness may be dictated not only by the hyperplastic or malignant epithelium but also by the stroma.

VII. Uterine Leiomyoma

Uterine fibroids, also known as leiomyomas, are benign tumors originating from the myometrium. They are composed of smooth muscle cells and large amounts of extracellular matrix (ECM) (Fig. 3B). These tumors can range from a few millimeters to over 20 cm in size. Leiomyomas are common and can occur in up to 77% of women.¹⁰¹ The incidence in African-American women is 60% at age 35 and over 80% by age 50 whereas Caucasian women have an incidence of 40% by age 35 and almost 70% by age 50.¹⁰² Although the tumors are considered benign, they cause significant morbidity, pain and discomfort, and excessive menstrual bleeding. Risk factors for leiomyomas include early menarche, family history, ethnicity, increased body mass index, and tissue injury. Leiomyomas are the primary indication for over 200,000 hysterectomies in the USA.¹⁰³ Studies have identified possible factors responsible for the development of leiomyomas, including chromosome rearrangements, congenitally elevated ERs, hormonal changes, and injury.¹⁰⁴ Once the disease has set in, hormones and growth factors play a prominent role in the growth and expansion of leiomyomas.

To date, medical treatments for leiomyomas are limited and this is due to the fact that the mechanisms regulating the development and growth of these tumors remain unclear. There exists only one FDA approved drug for the treatment of uterine leiomyoma and thus there is a desperate need for new treatments for one of the most prevalent chronic public health problems in US women. It is hoped that a better understanding of leiomyomas at the molecular level would lead to a more effective treatment of this disease.

VIII. Progesterone Receptor Action in Leiomyoma

A. Relevance of Progesterone in Uterine Leiomyomas

Although the initial steps in the pathogenesis of uterine fibroids are most likely due to chromosomal aberrations and/or the effects of specific genes,¹⁰⁵ their development is highly dependent on ovarian steroid hormones. Traditionally, estrogen has been considered the major mitogenic factor in the uterus. However, a growing body of evidence from biochemical, histological, clinical, and pharmacological studies indicates that progesterone and PR play a key role in uterine fibroid growth and development.¹⁰⁶ Several investigators have shown an increased concentration of both PR-A and PR-B in leiomyoma tissue compared with adjacent myometrium.^51,107,108 Furthermore, there was an increase in mitotic activity in fibroid tissue relative to the adjacent myometrial tissue during the luteal phase¹⁰⁹ and after treatment with medroxyprogesterone acetate.¹¹⁰ Increased expression of the proliferation marker Ki67 in leiomyoma compared with the normal myometrium has also been described, and its upregulation was linked to progesterone.¹⁰⁷ Epidermal growth factor (EGF) mRNA was increased in leiomyomata only during the secretory phase of the cycle, suggesting that progesterone, not estrogen, controls the expression of this important growth factor.¹¹¹ In addition, in vitro studies showed that progesterone suppresses apoptosis and stimulates proliferation of leiomyoma cells.^112-116 Progesterone markedly increased BCL2 protein expression in primary leiomyoma cell cultures.^112-116

Clinical studies with both progestins and RU486 indicate that progesterone may be at least as important as estrogen for regulating fibroid growth. When used as add-back therapy in combination with GnRH agonists, the synthetic progestins medroxyprogesterone acetate and norethindrone attenuate or reverse the inhibitory effects of GnRH agonists on leiomyoma size.^117,118 The effects of pregnancy on leiomyoma size have been studied as a possible model for in vivo exposure to high levels of progesterone.^119-122 The greatest increase in volume of uterine leiomyomata occurred before the 10th week of gestation.¹¹⁹ Those investigators who followed the leiomyoma size longitudinally after the first trimester, however, did not observe a further difference between the second and third trimester.^120-122

The strongest current evidence for possible in vivo mitogenic effect of progesterone on leiomyoma growth comes from clinical trials indicating that four different antiprogestins, RU486, asoprisnil, proellex, and CDB2914 consistently reduced tumor size.^123-131 The original studies of Murphy and coworkers in the 1990s suggested that RU486 might be used in the medical management of uterine leiomyomata. Pilot studies indicated that the size of leiomyomata decreased significantly after treatment with RU486.^123-125 Early studies indicated that different doses of RU486 decreased leiomyoma size as well as associated excessive uterine bleeding.¹²⁶ A similar endometrial histology, characterized by hyperplastic glands and stroma, was observed in patients treated with the antiprogestins RU486 and asoprisnil.¹²⁷ It was subsequently shown that asoprisnil also acts primarily as a progesterone antagonist in the endometrium.¹³² A number of investigators have attempted to avoid the side effect of endometrial hyperplasia by decreasing the dose of RU486 to 5 mg/day; this dose has been shown to successfully decrease leiomyoma size and uterine bleeding associated with these tumors.^125,130,131 Importantly, treatment with RU486 given at a dose of 5 mg/day did not cause endometrial hyperplasia.¹³¹ Despite the number of mechanisms proposed for these effects,^133-143 a full understanding of the pathophysiology responsible for progesterone-dependent growth and the mechanisms underlying the observed therapeutic effects of antiprogestins remain unclear.

B. Progesterone Action on Genes Associated with Proliferation, Apoptosis, and ECM Deposition

Although data focusing on the genes regulated in leiomyoma by progesterone are limited, studies investigating differential gene expression in leiomyomas during the menstrual cycle have provided groundwork for identifying those that are influenced by steroid hormones. In one study, the temporal and spatial expression of proliferative and proapoptotic molecules that could participate in leiomyoma pathogenesis was determined.¹⁴⁴ For example, levels of Fas ligand (FasL) protein, which is associated with apoptosis, was higher during the secretory phase compared with the proliferative phase in the leiomyoma. Furthermore, higher expression of FasL was found in the leiomyoma compared to myometrium. Levels of proliferating cell nuclear antigen (PCNA), which is associated with cell proliferation, was higher during the proliferative phase in leiomyoma and levels were higher than that of paired myometrium. Lower PTEN expression, which is the phosphatase that is associated with the PI3K/AKT pathway, was detected in the leiomyoma compared to the myometrium. In this study, it was speculated that the higher FasL level in the leiomyoma may correspond to suppression of local immunity by inducing apoptosis of immune cells, while a higher level of PCNA and a lower level of PTEN may be related to increased mitogenesis and decreased apoptosis in leiomyoma. Other studies have demonstrated that leiomyoma tissues have higher PCNA levels than myometrium throughout the menstrual cycle.¹⁴⁵ Furthermore, treatment with estradiol or progesterone increases PCNA expression in leiomyoma cells compared to controls.¹¹² Asoprisnil, a SPRM decreased the PCNA positive rate in cultured leiomyoma cells with no difference in myometrial cells.¹⁴⁶

The antiapoptotic bcl-2 gene in leiomyoma has been investigated by several groups. It has been demonstrated that bcl-2 is more highly expressed in leiomyoma than myometrium.^115,147,148 Progesterone and estrogen regulate bcl-2 expression differently. Progesterone upregulates bcl-2 mRNA while estrogen downregulates bcl-2 protein.¹¹⁵ Furthermore, Yin et al.¹⁴⁹ found that liganded PR binds to the bcl-2 promoter and enhances bcl-2 transcription in primary cultured leiomyoma cells. Overexpression of the dominant negative ER in cultured leiomyoma cells decreased bcl-2.¹⁵⁰ It has been speculated that this reduction in ER activity results in decreased PR expression and hence a decrease in bcl-2 expression. Asoprisnil decreased antiapoptotic bcl-2 with a corresponding increase in TUNEL staining, cleaved caspase 3, and cleaved PARP supporting a role for PR in preventing apoptosis in these cells.^133,146

ECM components are of high interest in leiomyoma pathology due to large quantities of matrix proteins found in leiomyoma. It has been demonstrated that certain ECM components and proteins are regulated by steroid hormones. Collagen type I and III mRNAs were upregulated in leiomyoma compared to myometrium during the proliferative phase of the menstrual cycle.¹⁵¹ In concert with this is the higher expression of MMPs in leiomyoma compared to myometrium during the secretory phase, while tissue inhibitors of MMPs (TIMPs) are more highly expressed in leiomyoma during the proliferative phase.¹⁵² Collagen binding protein fibrododulin (FMOD), which is involved in collagen fibril network formation, is more highly expressed in leiomyoma and myometrium during the proliferative phase of the menstrual cycle.¹⁵³ Levens et al.¹⁵³ also demonstrated that GnRHa treatments decreased FMOD. Asoprisnil treatment of primary leiomyoma cultures decreased TIMP1, TIMP2, collagen I, and collagen III while increasing MMP-1, MT1-MMP, EMMPRIN supporting that progesterone may be in involved in ECM deposition and turnover.¹⁵⁴

Recently, investigators have uncovered that miRNA’s may play a role in leiomyoma pathogenesis. MicroRNAs are small noncoding RNA’s that inhibit translation mostly through binding to target mRNA 3′ UTR. Marsh et al.¹⁵⁵ and Wang et al.¹⁵⁶ suggested that certain miRNA’s are differentially expressed in leiomyoma compared to myometrium. Wang et al.¹⁵⁶ focused on the let-7 family of miRNA’s and found that certain let-7 family miRNA’s may be correlated with tumor size. Pan et al.¹⁵⁷ also found that miRNA’s are differentially expressed in leiomyoma compared to myometrium and are regulated by sex steroids. More research is needed to address target genes of differentially regulated miRNA’s.

C. Growth Factor Regulation in Leiomyoma by Progesterone

Given the growth properties of leiomyomas, the expression and regulation of growth factors have been studied. Here, the regulation of these growth factors by both progesterone and estrogen is highlighted. Steroid hormones can regulate EGF and its receptor (EGFR) which have been implicated in leiomyoma growth. Both the local growth factor and its receptor are expressed in leiomyoma and myometrial tissues.¹⁵⁸ During the secretory phase of the menstrual cycle, EGF mRNA is higher in leiomyoma than myometrium.¹¹¹ Interestingly, progesterone does not increase EGF-R while increasing EGF and estrogen increases EGFR but does not increase EGF.¹⁴⁵ This supports previous results¹⁵⁹ that progesterone treatment but not estradiol increased EGF mRNA. In addition, asoprisnil decreased EGF mRNA.^134,146 A dominant negative ER decreased immunoreactive EGF in cultured primary and immortalized leiomyoma cells.¹⁵⁰ These data suggest that in the case of EGF and EGFR, estrogen and progesterone alter the response to the same growth factor pathway in different ways.

Insulin like growth factors IGF-I, IGF-II, and IGF-II receptor but not receptor type I have been detected in leiomyoma tissues at levels higher than the myometrium.^160,161 Studies show that IGF-I treatment can increase leiomyoma proliferation.^162-164 IGF-I gene expression was most abundant in leiomyomata obtained during the late proliferative phase of the cycle and was undetectable in leiomyomata from hypoestrogenic patients. SPRM, asoprisnil, decreased IGF-I mRNA in leiomyoma while having no effect in myometrial cells.^134,146 IGF-II mRNA expression did not vary with phase of the menstrual cycle.¹⁶⁵ Expression levels of IGF-II receptor were not altered with progesterone and estrogen treatments in cultured leiomyoma cells.¹⁶⁶

Both platelet derived growth factor (PDGF) and receptor are expressed in leiomyoma and myometrium and have been implicated in leiomyoma growth.¹⁶⁷ PDGF is a potent mitogen for smooth muscle cells and leiomyoma cells.^168-170 While it has been shown that leiomyoma tissues express higher levels of PDGF-A and B chain mRNA levels compared to matched myometrial tissue,^161,171 other studies show conflicting observations.^144,172-174 Women treated with GnRHa exhibited a reduction in uterine volume which was statistically related to the decrease in PDGF expression.¹⁶⁵ While estrogen can upregulate PDGF in cultured leiomyoma cells,¹⁶⁴ progesterone has also been implicated in regulating PDGF expression as shown by the increased expression of PDGF-BB expression during the secretory phase compared to the proliferative phase of the menstrual cycle in leiomyoma tissue.¹⁷¹

The transforming growth factor beta (TGF-beta) family increases the expression of ECM components and are involved in reproductive tissue development and growth.^175,176 Since fibroid tumors are composed mostly of ECM, examining connections between leiomyoma growth and TGF-beta cytokines and receptors have been of interest. Consistent expression of TGF-beta receptors type I–II and TGF-beta 1, 2, 3 have been found in myometrium^177,178 although expression in leiomyoma remains discrepant. For example, two studies showed increased expression of TGF-beta 1 mRNA in leiomyoma compared to myometrium, while another group showed the contrary.^168,177,179 In addition, TGF-beta 2, 3, and their receptors have been found to be more highly expressed in leiomyoma than in myometrium.¹⁷⁷ The expression of TGF-beta 3 has been more consistent demonstrating higher expression in leiomyoma compared to myometrium.¹⁸⁰ Furthermore, highest levels of TGF-beta 3 were found during the secretory phase of the menstrual cycle suggesting progesterone involvement.^168,181 Accordingly, the SPRM, asoprisnil, decreased TGF-beta 3 mRNA in leiomyoma cells.^134,146 Treatment with estrogen and progesterone differentially altered TGF-beta levels in myometrium and leiomyoma.^139,180

D. Activation of Signaling Pathways in Leiomyoma by Progesterone and Estrogen

The role of nuclear hormone receptors in activating signaling pathways as a mechanism for leiomyoma tumor growth have been proposed in recent years. In concert with the hormonally regulated growth factors described in the above sections EGF, FGF, IGF-I, HGF, and PDGF receptors were found to be highly expressed in leiomyoma tissues compared to myometrium in a receptor tyrosine kinase array.¹⁶¹ Specifically, expression of IGF-IR beta was more abundant in leiomyoma as well as phosphorylated IGF-IR and downstream effectors were more highly activated in leiomyoma. The mitogen activated protein kinase (MAPK) pathway regulates a variety of proteins involved in apoptosis and cell growth.¹⁸² Estradiol has been shown to rapidly activate the MAPK pathway in primary leiomyoma cells, including the rapid protein tyrosine phosphorylation of a subset of intracellular proteins, such as GAP, PI3K, and PLCgamma.¹⁸³ Interestingly, activation of this pathway was related to E2-induced PDGF secretion. In this study, it was proposed that PDGF, alone or in association with other growth factors, is the main growth factor involved in the proliferation response of leiomyoma cells to E2 stimulation. In accordance to this, ER-alpha phosphorylation was higher in leiomyoma tissues derived from patients in the proliferative phase of the menstrual cycle and this correlated with an increased phosphorylation of p44/p42 MAPK proteins in leiomyoma.¹⁸⁴ In addition, phosphorylated p44/42 colocalized with ER-alpha phosphorylated on serine 118, suggesting that MAPK can phosphorylate ER-alpha in leiomyoma. ER-alpha can also bind to the p85-alpha regulatory subunit of PI3K, allowing for PI3K activation in MCF-7 cells.¹⁸⁵

There is increasing evidence that progesterone also has rapid, membrane initiated effects independent of gene transcription to alter production of second messenger and cell signal transduction pathway. Some of these rapid nongenomic effects of progesterone have been shown to be mediated through the same nuclear PR that regulates gene transcription.^186,187 Pioneering work by Edwards’ group¹⁸⁷ demonstrated that PRB can directly bind to the SH3 domain of Src kinase and thereby activate the kinase. Similarly, progesterone-mediated regulation of the PI3K/AKT pathway has been demonstrated in breast cancer cells as well as in rat endometrial stromal cells.^186-190 Boonyaratanakornkit et al.¹⁸⁷ showed that p85 can interact with PR in a GST-pull-down system. Recently, it was shown that progesterone can rapidly phosphorylate AKT in leiomyoma cells.¹⁹¹ AKT phosphorylation was abrogated by PR antagonist RU 486 and PI3K inhibitor LY290004 in primary leiomyoma cells. Furthermore, the downstream targets of AKT, FOXO1, and GSK3 beta were phosphorylated upon progestin treatment indicating activation of down-stream signaling components. In leiomyoma, protein levels of AKT as well as the phosphorylated form were higher than myometrium and phosphor-AKT levels dropped in leiomyoma samples taken from menopausal women.¹⁴⁷ In concert with this is deactivated PTEN, a negative regulator of PI3K which was also increased in leiomyoma compared to myometrium, although these differences were less dramatic in menopause.¹⁹² GnRHa therapy decreased PI3K activity and AKT phosphorylation supporting that AKT activation is hormone dependent.¹⁹³ Given the involvement of the AKT pathway in cell proliferation and survival, its activation by progesterone may be another mechanism by which this hormone promotes leiomyoma growth.

Although estrogen and progesterone have distinct functions, the two hormones and their receptors interact with one another. Hodges et al. found that R5020 and MPA inhibited estradiol induced proliferation of ELT3 cells and inhibited ER activated gene transcription. These results suggest that liganded PR transdominantly suppresses ER signaling in leiomyoma. Of note is that estradiol increased expression of both PR isoforms.¹⁹⁴ Accordingly, overexpression of dominant-negative ER decreased PR expression in human leiomyoma cells.¹⁵⁰

IX. Conclusions and Perspectives on Progesterone Action in Uterine Leiomyoma

The importance of progesterone in promoting leiomyoma growth was initially substantiated by clinical studies as described above. The effectiveness of antiprogestins and SPRMs in reducing leiomyoma size provides strong support of progesterone being mitogenic in leiomyoma. Since the field of progesterone action in leiomyoma is one that has been understudied, it remains unclear as to how progesterone promotes growth of these tumors. Thus far, many studies have demonstrated that expression of genes and proteins are similar in leiomyoma compared to matched myometrium, however, it is the levels of expression and regulation during the menstrual cycle that differ. Whether it is demonstrated by the differential expression of genes during the menstrual phase or in response to progestin or antiprogestin treatment in vivo and in vitro, it is apparent that progesterone promotes expression of genes associated with growth and survival of leiomyoma. Regulation of growth factors and their receptors by progesterone have been proposed as another mode of progesterone action. At the transcriptional level, there are very few studies investigating the role of PR on promoters of genes in leiomyoma. Studying the recruitment of coregulators to various PR binding regions has given insight to how PR functions at the transcriptional level in leiomyomas. Finally, the rapid effects of progesterone on signaling molecules in leiomyoma provide yet another mode of progesterone action on leiomyoma growth. The involvement of classical PR in mediating these rapid effects are physiologically significant and whether progesterone membrane receptors are involved in progesterone mediated activation of kinases are unknown. Given the high incidence of leiomyoma in women, the morbidity that is associated with this disease and the financial burden of over 200,000 hysterectomies per year in the US alone, it is imperative that alternate therapies are developed. This can only be done with a better understanding of the molecular mechanisms associated with this disease.

X. Future Directions

The significance of progesterone in the uterus is indisputable. While its action remains complex and context-specific, it is crucial to decipher how PR works in uterine pathologies in order to be able to treat these diseases effectively. The stark contrast in the physiological response to progesterone in endometrial cancer compared to leiomyoma has important clinical implications when using progestins or antiprogestins as a mode of therapy. A clear example is the use of RU486 for decreasing leiomyoma size, which although effective, can promote endometrial hyperplasia. More information is needed on the differential action of PR in endometrial cancer cells and leiomyoma cells. Comparative studies on the mechanisms of action of PR in epithelial cells and the mesenchymal-derived fibroblasts would provide further insight to the differences in PR action in these two diseases. Alternatively, given the differing response to progesterone in the breast and the endometrium, it would be worthwhile to conduct studies comparing PR action in breast and endometrial cancers. At the molecular level, since PR action is dictated by coregulators, an extensive analysis of proteins that complex with PR on different gene promoters after specific times of progesterone treatment and then combined with gene expression studies would be informational. Global analysis of PR binding regions using chromatin immunoprecipitation techniques, identification of coregulators using mass spectrometry, and analysis of gene expression using microarray combined together with bioinformatics analysis would be one effective approach to decipher differential PR action in uterine cells. Although unraveling the complexity of PR may seem daunting and insurmountable, the information gathered thus far provides solid groundwork to tackle this challenge. Use of innovative and state-of-the-art technology will be key in moving this field forward.