The Divergent Effects of Ovarian Steroid Hormones in the MCF-7 Model for Luminal A Breast Cancer: Mechanistic Leads for Therapy

The growth modulating effects of the ovarian steroid hormones 17β-estradiol (E2) and progesterone (PRG) on endocrine-responsive target tissues are well established. In hormone-receptor-positive breast cancer, E2 functions as a potent growth promoter, while the function of PRG is less defined. In the hormone-receptor-positive Luminal A and Luminal B molecular subtypes of clinical breast cancer, conventional endocrine therapy predominantly targets estrogen receptor function and estrogen biosynthesis and/or growth factor receptors. These therapeutic options are associated with systemic toxicity, acquired tumor resistance, and the emergence of drug-resistant cancer stem cells, facilitating the progression of therapy-resistant disease. The limitations of targeted endocrine therapy emphasize the identification of nontoxic testable alternatives. In the human breast, carcinoma-derived hormone-receptor-positive MCF-7 model treatment with E2 within the physiological concentration range of 1 nM to 20 nM induces progressive growth, upregulated cell cycle progression, and downregulated cellular apoptosis. In contrast, treatment with PRG at the equimolar concentration range exhibits dose-dependent growth inhibition, downregulated cell-cycle progression, and upregulated cellular apoptosis. Nontoxic nutritional herbs at their respective maximum cytostatic concentrations (IC90) effectively increase the E2 metabolite ratio in favor of the anti-proliferative metabolite. The long-term exposure to the selective estrogen-receptor modulator tamoxifen selects a drug-resistant phenotype, exhibiting increased expressions of stem cell markers. The present review discusses the published evidence relevant to hormone metabolism, growth modulation by hormone metabolites, drug-resistant stem cells, and growth-inhibitory efficacy of nutritional herbs. Collectively, this evidence provides proof of the concept for future research directions that are focused on novel therapeutic options for endocrine therapy-resistant breast cancer that may operate via E2- and/or PRG-mediated growth regulation.


Introduction
In the development of breast cancer, the ovarian steroid hormones 17β-estradiol (E 2 ) and progesterone (PRG) represent important growth-regulatory hormones [1]. Although the growth-promoting effects of E 2 are well-established, the cellular effects of PRG are divergent and mostly context dependent. The receptors for both these hormones function as ligand-regulated nuclear transcription factors that bind to specific DNA response elements and operate via co-activators and co-repressors to affect the transcriptional modulation of down-stream target genes. The critical molecular processes responsible for the cellular effects of E 2 and PRG include complex autocrine and paracrine pathways involving specific cell-cycle regulators and NFkB-mediated growth modulation. Thus, two steroid hormones effectively interact via well-defined molecular cross-talk to regulate cell proliferation, differentiation, and breast carcinogenesis [2][3][4]. The molecular interaction of estrogen receptor (ER) and progesterone receptor (PR) signaling in hormone-receptor-positive breast cancer results in a modulatory effect of PR on the action of ER to attenuate tumor growth via multiple pathways. In addition to ER-and PR-mediated signaling, the receptor-independent cellular metabolism of these hormones generates metabolites with distinct cellular effects relevant to the process of carcinogenesis [5][6][7][8].
The hormone-receptor-positive Luminal A and Luminal B molecular subtypes of clinical breast cancer respond to selective estrogen-receptor modulators, aromatase inhibitors, and to HER-2 targeted therapeutics. However, long-term anti-estrogen therapy and human epidermal growth factor-2 (HER-2)-based targeted therapy is associated with intrinsic and/or acquired drug resistance that compromises the therapeutic efficacy and favors progression of therapy-resistant disease. The progression of therapy-resistant disease is frequently associated with the emergence of drug-resistant stem cell populations [9]. These limitations of current therapy for breast cancer emphasize an unmet need to identify stem cell targeting testable alternatives. Drug-resistant stem cell models have been developed for HER-2-enriched and triple-negative breast cancer subtypes [10,11].
The expression status of select hormone and growth-factor receptors has provided a basis for the clinically relevant classification of the cellular models for breast cancer subtypes [12,13]. Furthermore, stable expressions of the clinically relevant HER-2 oncogene [14][15][16] or the gene for aromatase enzyme [17,18] have provided valuable models to examine the role of oncogenes and estrogens in breast cancer.
The human mammary carcinoma-derived MCF-7 cell line that is ER/PR positive and expresses non-amplified HER-2, represents a model for the Luminal A molecular subtype of clinical breast cancer [12,13]. Several mechanistically distinct Chinese nutritional herbs have documented preferential growth-inhibitory efficacy in the isogenic MCF-7 phenotypes that exhibit a modulated ER-α function [19]. However, little evidence is available on the effects of nutritional herbs, either on PRG activity or function. Based on the evidence of negative growth regulation by PRG, the potential therapeutic utility of progestogens and progesterone receptor (PR) agonists may represent beneficial treatment options. It is conceivable that the non-toxic nutritional herbs may also be effective in the hormonereceptor-positive breast cancer via PR function.
The goal of the present review is to provide (i) a systematic discussion of the published literature relevant to the development and characterization of the MCF-7 model; (ii) mechanistic leads for the significance of anti-proliferative metabolites of E 2 and PRG; and (iii) growth-inhibitory efficacy of dietary phytochemicals and nutritional herbs as stem-cell-targeting testable alternatives for the chemo-endocrine therapy-resistant Luminal A molecular subtype of breast cancer.

Cellular Models
In the mouse mammary gland organ culture model, the ovarian steroid hormones E 2 and PRG are critical for the induction of normal ductal morphogenesis. Lactogenic hormones, prolactin and hydrocortisone, induce lobulo-alveolar growth and mammary-specific differentiation [20]. In response to treatment with chemical carcinogens, the mammary gland organ cultures develop lactogenic hormone-independent pre-neoplastic alveolar lesions [21], and the transplantation of cells from 7-12 dimethyl benzanthracene (DMBA)induced alveolar lesions produce rapidly growing tumors [22]. However, the impact of similar interactions between steroid and polypeptide hormones on the initiation or progression of human mammary carcinogenesis remains to be fully elucidated.
Global gene-expression profiling of clinical breast cancer has facilitated the molecular classification of breast cancer subtypes [23]. Cellular models developed from human breast carcinoma-derived cell lines continue to represent valuable resources to identify clinically relevant mechanistic pathways and molecular targets for therapeutic efficacy [12,13]. MCF-7 cells expressing mutant HER-2 represent an additional model for the Luminal B breast cancer subtype [14]. The CYP19 A1 aromatase enzyme is critical for peripheral and intra-tumoral estrogen biosynthesis. This enzyme converts adrenal androstenedione to estrone (E 1 ) and testosterone to E 2 . E 1 is converted to E 2 by 17β-hydroxysteroid de-hydrogenase [5,6]. MCF-7 cells expressing the aromatase gene represent a model for aromatase-positive post-menopausal breast cancer [17,18]. Furthermore, the stable expression of the HER-2 oncogene in human mammary epithelial 184-B5 cells induces tumorigenic transformation [15,16]. The 184-B5/HER cell line represents a valuable model to examine the role of the HER-2 oncogene in the initiation and progression of human breast cancer.

The Growth Characteristics of the MCF-7 Model
Carcinoma-derived cell lines commonly exhibit hyper-proliferation and persistence anchorage-independent growth in vitro, and tumor formation in vivo. Table 2 compares the growth pattern of human breast epithelium-derived non-tumorigenic 184-B5 cells and human breast carcinoma-derived tumorigenic MCF-7 cells. The MCF-7 cells exhibit hyper-proliferation, as evidenced by a decrease in the population doubling time, increase in the saturation density, and accelerated cell cycle progression. Additionally, MCF-7 cells exhibit downregulated cellular apoptosis, decreased estrogen metabolite ratio, and a robust increase in anchorage-independent colony formation, the latter being a specific in vitro surrogate end point for in vivo tumor formation. These data suggest that MCF-7 cells exhibit a loss of homeostatic growth control and persistent cancer risk.

Growth Modulation by Estradiol and Progesterone
The distinct growth modulatory effects of E 2 and PRG at the physiologically relevant concentrations were examined on MCF-7 cells maintained in a culture medium supplemented by 0.7% serum (serum concentration of E 2 and PRG < 0.01 nM in the culture medium). Treatment with E 2 at the physiological concentration range of 1 nM to 20 nM, resulted in a concentration-dependent increase in the viable cell number ( Figure 1A). In contrast, treatment with PRG at the equimolar concentration range displayed a concentrationdependent decrease in the viable cell number ( Figure 1B).

Growth Modulation by Estradiol and Progesterone
The distinct growth modulatory effects of E2 and PRG at the physiologically relevant concentrations were examined on MCF-7 cells maintained in a culture medium supplemented by 0.7% serum (serum concentration of E2 and PRG < 0.01 nM in the culture medium). Treatment with E2 at the physiological concentration range of 1 nM to 20 nM, resulted in a concentration-dependent increase in the viable cell number ( Figure 1A). In contrast, treatment with PRG at the equimolar concentration range displayed a concentrationdependent decrease in the viable cell number ( Figure 1B).

Cell Cycle Progression and Cellular Apoptosis
The data provided in Figure 2 provide evidence for the potential mechanistic leads that are responsible for the effects of E2 and PRG. As illustrated in Figure 2A

Cell Cycle Progression and Cellular Apoptosis
The data provided in Figure 2 provide evidence for the potential mechanistic leads that are responsible for the effects of E 2 and PRG. As illustrated in Figure 2A, treatment with E 2 decreases the G 1 :S + G 2 /M ratio due to an increase in the S phase of the cell cycle. In contrast, treatment with PRG increases the G 1 :S + G 2 /M ratio due to G 1 arrest and a decrease in the S and G 2 /M phases of the cell cycle. As illustrated in Figure 2B, treatment with E 2 inhibits cellular apoptosis, while treatment with PRG increases cellular apoptosis.
with E2 decreases the G1:S + G2/M ratio due to an increase in the S phase of the cell cycle. In contrast, treatment with PRG increases the G1: S + G2/M ratio due to G1 arrest and a decrease in the S and G2/M phases of the cell cycle. As illustrated in Figure 2B, treatment with E2 inhibits cellular apoptosis, while treatment with PRG increases cellular apoptosis. ER and PR belong to a superfamily of ligand-regulated nuclear transcription factors that are responsible for the expression of cognate downstream target genes. These genes include pS2, GRB2, and cyclin D1 for E2 [3,24], and a receptor activator of nuclear factor kB (RANK) and its ligand RANK-L for PRG [2][3][4]. In addition to the genomic mechanisms, the cellular metabolism of E2 and PRG plays a significant modulatory role in breast carcinogenesis. For example, the metabolites generated from E2 and PRG have documented divergent growth-modulatory effects on breast cancer cells. ER and PR belong to a superfamily of ligand-regulated nuclear transcription factors that are responsible for the expression of cognate downstream target genes. These genes include pS2, GRB2, and cyclin D1 for E 2 [3,24], and a receptor activator of nuclear factor kB (RANK) and its ligand RANK-L for PRG [2][3][4]. In addition to the genomic mechanisms, the cellular metabolism of E 2 and PRG plays a significant modulatory role in breast carcinogenesis. For example, the metabolites generated from E 2 and PRG have documented divergent growth-modulatory effects on breast cancer cells.

Cellular Metabolism of Estradiol
The CYP450-mediated enzymatic metabolism of E 2 generates mechanistically distinct metabolites that exert specific biological effects on the non-tumorigenic or tumorigenic mammary epithelial cells. During E 2 metabolism, 4-hydroxylated E 2 and 16α-hydroxylated E 1 function as proliferative agents, while 2-hydroxylated E 2 and 2-hydroxyalted E 1 function as anti-proliferative agents [5,6]. These metabolites, because of their distinct biological effects, alter the ratio of anti-proliferative to proliferative metabolites. Similar metabolic alterations have also been documented in mammary epithelial cells that exhibit a tumorigenic transformation induced by stable transfection with Ras, Myc, and HER-2 oncogenes [25][26][27]. Thus, the altered 2-OHE 1 :16α-OHE 1 ratio may represent a novel experimentally modifiable endocrine biomarker for the efficacy of testable alternatives for breast cancer prevention/therapy.
Because of the distinct effects of hydroxylated metabolites of E 1 , a ratio of these metabolites may represent a valuable end-point marker. Table 3 compares the status of the 2-OHE 1 :16α-OHE 1 ratio in cellular models for breast cancer that differ in their relative risk for cancer development. These data demonstrate that, depending on the relative risk of developing cancer, the estrogen metabolite ratios are substantially decreased.

The Cellular Metabolism of Progesterone
During PRG metabolism, 5α-hydroxylated metabolite functions as a proliferative agent, while 3α-hydroxylated metabolite functions as an anti-proliferative agent [7,8]. Thus, in the tumorigenic cells, the ratio of anti-proliferative:proliferative metabolites of PRG is altered in favor of the proliferative metabolites.
Similar to the E 2 metabolites, PRG metabolites 5α-dihydro progesterone (5α-PRG) and 3α-dihydro progesterone (3α-PRG) have documented growth-modulatory divergent effects in the MCF-7, MDA-MB-231, T47D, and MCF-10A models. At the mechanistic levels, growth promotion by the 5α metabolite is associated with increased DNA synthesis and the mitotic index and the downregulation of apoptosis, while growth inhibition by the 3α metabolite is associated with a decrease in these end points [7,31,32]. The growth modulating effects of PRG metabolites have also been documented in vivo in the tumors produced by transplanted MDA-MB-231 cells. [8].
The divergent effects of E 2 and PRG have been documented in T47D, ZR75-1, and MCF-7 cells. For example, the E 2 -mediated inhibition of cellular apoptosis is associated with the increased expression of anti-apoptotic BCL-2. In contrast, the PRG-mediated induction of cellular apoptosis is associated with the decreased BCL-2: BAX ratio, predominantly due to the increased expression of pro-apoptotic BAX [7]. In tissue explants of ER-positive tumors, E 2 increases, while PRG decreases, cell proliferative activity [32,33].
The autocrine growth-modulatory effects of PRG involve the expression of cyclindependent kinase inhibitors p18 and p27 [34], and the modulation of ER-α and RNA polymerase III [35]. In addition to these autocrine effects, the paracrine cellular effects of PRG involve the NFkB pathway via modulation in the expression of RANK and its ligand RANKL [36], and the downregulated select interferon-stimulated genes [37]. Since E 2 metabolites 4-OHE 2 and 2-OHE 2 , E 1 metabolites 16α-OHE 1 and 2-OHE 1 , and PRG metabolites 5α-PRG and 3α-PRG exhibit opposing growth-regulatory effects, molecular/metabolic pathways for E 2 and the PRG function may provide potential mechanistic leads responsible for E 2 -and PRG-mediated growth modulation.
Collectively, the evidence for the enzymatic conversion of E 2 and PRG that generate metabolites with divergent growth-modulatory effects supports the significance of the ratio of proliferative and anti-proliferative metabolites. Thus, the experimental upregulation of anti-proliferative metabolites may provide mechanistic leads for novel therapeutic interventions.
The published data on inhibitory efficacy Chinese nutritional herbs Epimedium grandiflorum (EG), Lycium barbarum (LB), and Cornus officinalis (CO) on MCF-7 cells, a model for Luminal A breast cancer, provide mechanistic leads to identify nontoxic testable alternatives. The anti-proliferative effects of these herbs are associated with the altered cellular metabolism of E 2 in favor of the formation of the non-proliferative metabolite 2-OHE 1 [46][47][48].
The data provided in Table 4 illustrates that treatment with CO, EG, and LB induces the upregulation of the 2-hydroxylation pathway, leading to an increase in 2-OHE 1 formation and the downregulation of the 16α-hydroxylation pathway leading to a decrease in the 16α-OHE 1 formation. The CYP19 A1 Aromatase enzyme represents a critical enzyme for peripheral and intra-tumoral estrogen biosynthesis in post-menopausal breast cancer, and thereby provides growth-promoting estrogens via the conversion of E 2 and E 1 from testosterone and androstenedione, respectively [49,50]. In this context, it is notable that the progesterone metabolite 20α-DHP, predominantly detected in normal breasts, functions as a potent inhibitor of aromatase [51]. An extract from the inner bark of the South American Tabebuia avellanedae (TA) tree displays pro-apoptotic and anti-aromatase activity in the MCF-7 AROM model for aromatase expressing post-menopausal breast cancer. In this model, the TAmediated induction of cellular apoptosis is associated with the upregulated expression of the pro-apoptotic BAX gene and downregulated expression of the anti-apoptotic BCL-2 gene. Treatment with TA is also associated with downregulated expressions of several E2 responsive genes, such as ESR-1, AORM, PR, PS2, GRB2, and cyclin D1. The potency of aromatase inhibition by TA, based on the content of the active agent naphthofuran dione, is substantially higher than the pharmacological inhibitors of aromatase, letrozole, and exemestane [52]. Collectively, these data on aromatase inhibition provide evidence for aromatase as a potential target for experimental modulation.

Drug-Resistant Stem Cells
Stem cells are responsible for preserving the regulated program of epithelial proliferation, differentiation, and apoptosis, which are critical for cellular homeostasis in normal breasts via Wnt/β-catenin, Notch, and NFkB signaling pathways, wherein the interactive influence of E 2 and PRG plays an important role [53]. In drug-resistant cancer stem cells, these signaling pathways are dysregulated, and RAS-, PI3K-, AKT-, and mTOR-mediated survival pathways are activated [9,54,55].
Acquired resistance to conventional chemo-endocrine therapy and molecularly targeted therapy results in the emergence of drug-resistant cancer stem cells. Reliable stem cell models provide valuable experimental approaches to identify stem cell targeting testable alternatives. The tamoxifen-resistant (TAM-R) stem cell model for Luminal A breast cancer, the lapatinib-resistant (LAP-R) stem cell model for HER-2-enriched breast cancer, and the doxorubicin-resistant (DOX-R) stem cell model for triple-negative breast cancer exhibit upregulated expressions of select stem cell markers [10]. These models may provide clinically relevant experimental approaches to evaluate the therapeutic efficacy of stem-cell-targeting natural herbal products, including dietary phytochemicals and Chinese nutritional herbs.
In the Luminal A molecular subtype of breast cancer selective estrogen-receptor modulator tamoxifen (TAM) has wide clinical applications. However, long-term treatment with TAM is associated with acquired tumor resistance. The data in Figure 3 illustrate that the tamoxifen-resistant stem cell model derived from MCF-7 cells exhibits an increased expression of select stem cell markers, such as TS, CD44, NANOG, and OCT-4. This model may provide clinically relevant experimental approaches to evaluate the therapeutic efficacy of natural herbal products targeted towards breast cancer stem cells.

Conclusions
The present review discussed the published evidence relevant to the roles of E2 and PRG in breast cancer biology, the significance of reliable stem cell models for therapyresistant breast cancer, and the mechanistic evidence for the growth-inhibitory efficacy of natural products, including dietary phytochemicals and nutritional herbs. Collectively, this review provides a proof concept that natural products may represent testable alternatives for therapy-resistant breast cancer.
A comprehensive overview of the conceptual background of the cellular models for Luminal A, Luminal B, and aromatase-expressing post-menopausal breast cancer subtypes, current targeted therapeutic options, therapy-resistant stem cells, naturally occurring dietary phytochemicals, and nutritional herbs as therapeutic alternatives for therapyresistant breast cancer and future research directions are summarized in Table 5.  This model may provide clinically relevant experimental approaches to evaluate the therapeutic efficacy of natural herbal products targeted towards breast cancer stem cells.

Conclusions
The present review discussed the published evidence relevant to the roles of E 2 and PRG in breast cancer biology, the significance of reliable stem cell models for therapyresistant breast cancer, and the mechanistic evidence for the growth-inhibitory efficacy of natural products, including dietary phytochemicals and nutritional herbs. Collectively, this review provides a proof concept that natural products may represent testable alternatives for therapy-resistant breast cancer.

Future Prospects
By extending the evidence provided in the present review, future research directions will involve patient-derived tumor xenograft (PDTX) and organoid (PDTO) models [62][63][64][65][66]. Additionally, future investigations will develop reliable stem cell models from therapyresistant breast cancer subtypes. The outcome of these investigations on patient-derived samples is expected to provide strong evidence for clinically relevant data and their potential for clinical translation. Data Availability Statement: The data sets used and/or analyzed in the current study are available from the corresponding author on reasonable request.