New small‐molecule compound Hu‐17 inhibits estrogen biosynthesis by aromatase in human ovarian granulosa cancer cells

Abstract Estrogen‐dependent cancers (breast, endometrial, and ovarian) are among the leading causes of morbidity and mortality in women worldwide. Aromatase is the main enzyme that catalyzes the biosynthesis of estrogen, which drives proliferation, and antiestrogens can inhibit the growth of these estrogen‐dependent cancers. Hu‐17, an aromatase inhibitor, is a novel small‐molecule compound that suppresses viability of and promotes apoptosis in ovarian cancer cells. Therefore, this study aimed to predict targets of Hu‐17 and assess its intracellular signaling in ovarian cancer cells. Using the Similarity Ensemble Approach software to predict the potential mechanism of Hu‐17 and combining phospho‐proteome arrays with western blot analysis, we observed that Hu‐17 could inhibit the ERK pathway, resulting in reduced estrogen synthesis in KGN cells, a cell line derived from a patient with invasive ovarian granulosa cell carcinoma. Hu‐17 reduced the expression of CYP19A1 mRNA, responsible for producing aromatase, by suppressing the phosphorylation of cAMP response element binding‐1. Hu‐17 also accelerated aromatase protein degradation but had no effect on aromatase activity. Therefore, Hu‐17 could serve as a potential treatment for estrogen‐dependent cancers albeit further investigation is warranted.


| BACKGROUND
Gynecological malignancies such as ovarian, breast, and endometrial cancers are the leading cause of cancer-related deaths in women. 1 Ovarian cancer is the second most commonly diagnosed cancer in women worldwide, 2 and the mortality rate is worse than other reproductive cancers because of difficulty in early diagnosis. 3,4 Surgery followed by chemotherapy is currently the standard therapy for ovarian cancers. 5 Platinum-based drugs (e.g., cisplatin and carboplatin) and paclitaxel are traditional first-line therapies. 6,7 Although ovarian cancer diagnosed in the early stage may be treated successfully, as the disease progresses, most advanced-stage patients develop resistance to first-line drugs. 8 Therefore, there is an urgent need to explore more therapeutic agents to overcome drug resistance.
Estrogens play a pivotal role in the development of gynecological cancers. 9,10 Therefore, a new strategy developed to treat estrogen-mediated carcinogenesis involves inhibiting cell proliferation by reducing the estrogen levels. [11][12][13] There are two major therapeutic approaches to disrupt estrogen function. The first is to block estrogen binding to the estrogen receptor (ER) via ER antagonists. The second is to reduce estrogen biosynthesis with aromatase inhibitors. [13][14][15] Aromatase, encoded by the CYP19A1 gene, is a rate-limiting enzyme that catalyzes estrogen biosynthesis and is highly expressed in the placenta, breast, and granulosa cells of ovarian follicles. 9,16,17 Multiple studies indicate that aromatase inhibitors are better tolerated than ER antagonists, in addition to being less toxic and highly effective in curing estrogen-dependent cancer. 18,19 There are two types of aromatase inhibitors, steroidal (e.g., exemestane) and nonsteroidal (e.g., letrozole) that can be used to treat estrogen-dependent cancers in postmenopausal women. [20][21][22] However, the inhibition of aromatase results in an increased risk of osteoporosis and cardiovascular disease. [23][24][25] Therefore, novel aromatase inhibitors with greater clinical efficacy and fewer side effects are needed.
Phytolaccaesculenta (known as shanglu in China) is an important traditional Chinese medicine, and a decoction of its root is used to treat inflammation-related conditions. 26 Hu-17, a novel synthetic compound, was derived from the root of phytolaccaesculenta. We found that Hu-17 can strongly inhibit the proliferation of cells and promote apoptosis in ovarian epithelial carcinoma cell lines and animal models. It was authorized in 2018 (China patent number: ZL201510256415.X). However, the effect of Hu-17 on ovarian granulosa cell carcinoma is not clear.
Similarity Ensemble Approach (SEA) is a computational strategy that use chemical similarity among ligands organized by their targets to calculate similarities and predict drug off-target or targeted activities. [27][28][29] In this study, we used SEA to predict drug targets of Hu-17 and assess its intracellular signaling in a steroidogenic human ovarian granulosa-like tumor KGN cell line treated with Hu-17.

| Materials
Forskolin, exemestane, formestane, cisplatin, and PD98059 were purchased from Sigma Chemical Co. MG132, Z-VAD-FMK and cycloheximide were purchased from Selleckchem. Paclitaxel was kindly provided by the Shanghai Key Laboratory of Gynecologic Oncology, Ren Ji Hospital. Hu-17 was synthesized by the laboratory of Professor Yanghua Yi, Second Military Medical University. Hu-17 (molecular weight 1084 Da; structure in Figure 1) was stored at −20°C as a stock solution (20 mmol/L) in dimethyl sulfoxide.

| Cell culture
Human granulosa cells (hGCs) were collected from patients with ovarian hyperstimulation syndrome (OHSS) and without OHSS undergoing their first in vitro fertilization/intracytoplasmic sperm injection cycle at the Center for Reproductive Medicine, Ren Ji Hospital. All participants provided written informed consent to participate in this study. The isolation protocol for hGCs was performed as described previously. 30 KGN, a steroidogenic human ovarian granulosa-like tumor cell line, was kindly provided by the Shandong University, China. MCF-7 and SUM-159 cells were purchased from Cell Bank, Chinese Academy of Sciences. The hGCs and KGN cells were maintained in phenol red-free DMEM/F12 supplemented with 10% charcoal-stripped fetal bovine serum (FBS). MCF-7 and SUM-159 cells were cultured separately in DMEM and F12 supplemented with 10% FBS at 37°C and 5% CO 2 . All media and FBS were purchased from Gibco.

| Measurement of intracellular cAMP concentration
The intracellular cAMP level in KGN cells was measured using an EIA kit (Cayman, Ann Arbor, MI, USA) after treatment with Hu-17 (1.5 µmol/L) or forskolin (50 mmol/L) in serum-free DMEM in the presence of the phosphodiesterase inhibitor 3-isobutyl-1-methyl xanthine (500 mmol/L, Sigma). The assay was performed as described previously. 31

| In vitro aromatase activity assay
The CYP19/MFC High Throughput Inhibitor Screening kit and Baculovirus-infected insect cell-recombinant human CYP19 (with oxidoreductase) were purchased from BD Biosciences (Gentest). The in vitro activity of aromatase was determined by measuring the conversion rate of a fluorometric substrate to its fluorescent metabolite. Experimental procedures were consistent with the manufacturer's protocols.

| Cell viability and morphological changes
Cell viability was measured using the MTT assay (Sangon Biotech). The assay was performed as described previously. 32 The morphological changes in KGN cells treated with Hu-17, CHX, or MG132 were visualized using an inverted microscope connected to a digital camera (Carl Zeiss).

| Extraction of protein and western blotting
Protein extraction and western blotting were performed as described previously. 34 The following primary antibodies were used: anti-AROM was purchased from Abcam; anti-CREB-1, anti-P90RSK, anti-pP90RSK, anti-ERK, anti-pERK, anti-MEK, anti-pMEK, anti-EGFR, and anti-c-Raf were purchased from Cell Signaling Technology; and anti-AKT, anti-pAKT, anti-P38, anti-pP38, and β-actin were purchased from Santa Cruz Biotechnology. Secondary antibodies were also purchased from Cell Signaling Technology. A chemiluminescent detection system (Thermo Fisher Scientific, Carlsbad, CA, USA) was used to detect bands with peroxidase activity.

Phospho Explorer antibody microarray
The Phospho Explorer Antibody Microarray (PEX100) was designed and manufactured by Full Moon BioSystems Inc. Whole-cell lysates isolated from KGN cells cultured with 1.5 µmol/L Hu-17 for 0, 3, 6 h were collected in a Protein Extraction Buffer (Full Moon BioSystems Inc.). The array consisted of 1318 phospho-specific antibodies; detailed information about the array can be found at https://www.fullm oonbio.com/produ ct/phosp ho-explo rer-antib ody-array. The antibody array experiment and analysis were performed according to an established protocol. 35

| Prediction of Hu-17-target interactions using SEA
The SEA software was used to predict molecular targets for each phytochemical and was standardized as previously described. 27 Briefly, SEA uses the chemical similarity of a bait molecule against a set of ligands from the ChEMBL database (https://www.ebi.ac.uk/chemb l/) annotated to a target to predict whether the bait molecule will modulate that target. An E-value is derived from a statistical model that represents the probability of observing this raw score by random chance alone. A smaller E-value indicates stronger overall chemical structural similarity between two sets of compounds. The Tanimoto coefficient, which represents the similarity between any two molecular fingerprints, on a 0.0 (completely dissimilar) to 1.0 (completely similar) scale, was calculated.

| Steroid assays
Testosterone (Sigma Chemical Co.) at a concentration of 10 −7 ng/mL was added as a substrate for estrogen synthesis in KGN cells. After treatment with Hu-17 for 24 h, the levels of estrogen were measured using ECLIA (Roche), according to the manufacturer's instructions.

| Statistical analysis
Data were compared by either one-way ANOVA or the t test by Duncan's multiple-range tests. p values < 0.05 were considered to be statistically significant. Statistics were performed using GraphPad Prism 5 (GraphPad, San Diego, CA, USA) and the levels of significance are indicated in the figure legends.

| Hu-17 inhibits cell proliferation and induces cell apoptosis in KGN cells
Previously, we found that Hu-17 strongly inhibited ovarian epithelial carcinoma cell proliferation and promoted apoptosis in several ovarian epithelial carcinoma cell lines (including A2780, SKOV3, and HO8910) and mouse models (China patent number: ZL201510256415.X). However, whether Hu-17 can affect ovarian granulosa cell carcinoma is not clear. Therefore, in order to investigate the role of Hu-17 on ovarian granulosa cell carcinoma, we choose KGN, a human ovarian granulosa-like tumor cell line, as a study mode in this study. The survival of KGN cells was assessed using the MTT assay after exposure to a series of doses of Hu-17 for 24, 48, and 72 h ( Figure 2A). Hu-17 decreased the survival of KGN cells in a both dose-and time-dependent manners. After treatment with Hu-17 (0, 1, 1.5, and 2 μmol/L) for 24 h, the proportion of apoptotic cells was 6.2%, 16.5%, 27.6%, and 36.8%, respectively ( Figure 2B), indicating that Hu-17 induced apoptosis in KGN cells in a dose-dependent manner. PARP-1 cleavage is an early molecular marker of apoptosis. Treatment with 1.5 and 2 μmol/L Hu-17 for 24 h resulted in considerable PARP-1 proteolytic cleavage ( Figure 2C). We also treated cells with cisplatin and paclitaxel, two chemotherapy drugs currently used in the clinical setting, for comparison with Hu-17. The proportion of apoptotic cells following treatment with 1.5 μmol/L Hu-17, 100 μmol/L paclitaxel, and 15 μmol/L cisplatin were similar (27.5%, 25.2%, and 29.1%, respectively) ( Figure 2D). Hu-17, cisplatin, and paclitaxel also increased the abundance of cleaved-PARP-1 ( Figure  2E). These data suggested that Hu-17 could inhibit proliferation and induce apoptosis in KGN cells.

| Predicting potential targets of Hu-17 using SEA
To gain insights into the mechanism of Hu-17, we used SEA to identify potential targets. SEA predicted that Hu-17 might affect hormone signaling, especially estrogen synthesis, by targeting AR, ER, PR, and CYP19A1 in human cells (Table  1). SEA also predicted that Hu-17 had the potential to affect protein tyrosine phosphatases, including PTPN2, PTPN1, and PTPN6 (Table 1).

| Hu-17 inhibits estrogen synthesis and aromatase expression in KGN cells
KGN cells, not other ovarian cell lines (A2790, SKOV3, and HO8910), are known to be steroidogenic ovarian granulosa-like tumor cells that expresses abundant aromatase. 36 Therefore, based on the SEA predictions, we used KGN cells to study the role of Hu-17 in aromatase expression and estrogen synthesis. We found that estradiol levels were 52% lower in the culture medium of hGCs treated with 1.5 μmol/L Hu-17 for 24 h compared with control (no-treated) cells ( Figure 3A). Additionally, estradiol levels were 79% lower in KGN cells treated with 1.5 μmol/L Hu-17 than in control cells, although estradiol levels did not change when treated with 15 μmol/L cisplatin or 100 μmol/L paclitaxel treatment ( Figure 3B). These experiments demonstrated that Hu-17, but not first-line chemotherapy drugs such as cisplatin and paclitaxel, could reduce estrogen biosynthesis, despite them all having similar effects on cell viability. Therefore, the decrease in estrogen production by Hu-17 is not due to the decrease in the numbers of KGN cells.
by decreasing aromatase protein expression rather than by directly blocking its activity.
In consistent with estrogen production, 1.5 μmol/L Hu-17 also significantly decreased aromatase expression compared with control in KGN cells, but 15 μmol/L cisplatin or 100 μmol/L paclitaxel had no obvious effects ( Figure 3G). To identify the mechanism by which Hu-17 induces apoptosis, we simultaneously exposed KGN cells to 1.5 μmol/L Hu-17 and 50 μmol/L Z-VAD-FMK, a universal caspase inhibitor, for 24 h. Z-VAD-FMK significantly reduced Hu-17-mediated apoptosis but did not block the Hu-17-induced reduction in aromatase expression ( Figure 3H). These results suggested that Hu-17 could inhibit estrogen production via the reduction of aromatase expression in an apoptosis-independent manner in KGN cells.

| Hu-17 inhibits CREB-1 phosphorylation in KGN cells
The cAMP/PKA/CREB pathway is involved in the regulation of aromatase expression in ovarian granulosa cells. 10 Therefore, we investigated the effect of Hu-17 on cAMP/ PKA/CREB pathway activation in KGN cells. Hu-17 significantly inhibited CREB-1 phosphorylation at 3 and 6 h ( Figure 4B) but had no effect on CREB-1 mRNA and total protein in these cells ( Figure 4A,B). Furthermore, 1.5 μmol/L Hu-17 reduced intracellular cAMP levels at 1, 3, and 6 h in KGN cells ( Figure 4C). Forskolin, an activator of adenylate cyclase, not only increased the basal cAMP levels but also reversed the Hu-17-induced decrease in intracellular cAMP in KGN cells ( Figure 4D). Treatment with

| Hu-17 blocks CREB-1 phosphorylation via inhibition of ERK signaling
In addition to PKA, the ERK and p38 kinase pathways are also involved in the regulation of CREB-1 phosphorylation and aromatase expression. [37][38][39][40] Analysis of the phospho-antibody microarray revealed that treatment with Hu-17 inhibited the phosphorylation of several key proteins in the ERK pathway ( Figure 6A P90RSK (phospho-Thr573). We evaluated the phosphorylation of CREB-1, ERK1/2, and upstream effectors P90RSK and MEK1 in KGN cells through selective confirmatory western blotting analysis to investigate whether Hu-17 treatment affected CREB-1 phosphorylation through the ERK pathway. As shown in Figure 6B, the phosphorylation of ERK, P90RSK, and MEK1 was decreased in KGN cells after Hu-17 treatment for 3 h ( Figure 6B). Moreover CREB-1 phosphorylation and aromatase protein expression were decreased in KGN cells treated with the ERK inhibitor PD98059 for 24 h ( Figure 6C). These data suggested that Hu-17 acted through the ERK pathway to regulate aromatase expression in KGN cells.

| Hu-17 causes proteasome-mediated degradation of aromatase in KGN cells
Interestingly, we found that Hu-17 decreased CYP19A1 mRNA at the earlier time points including 1, 2, 3, and 6 h but not at long time point such as 24 h ( Figure 3E and Figure 7A), while it decreased aromatase protein level at both short time and long time points including 2, 3, 6, and 24 h ( Figure 3F and Figure 7B) in KGN cells, suggesting that in addition to affect CYP19A1 mRNA transcription, Hu-17 may also could affect aromatase protein level on a post-transcriptional level. Therefore, to better understand the mechanism of Hu-17induced aromatase degradation, we treated KGN cells with Hu-17 in the absence or presence with CHX (20 ng/mL), a de novo transcription inhibitor, as well as MG132 (10 μmol/L), a proteasome inhibitor, and found that CHX did not block Hu-17-induced reduction of aromatase protein while MG132 significantly attenuated the Hu-17-induced decreased in aromatase protein levels ( Figure 7C), indicating that Hu-17 didn't affect protein synthesis but induced aromatase protein degradation via proteasome pathway in KGN cells.

| Aromatase protects KGN cells from apoptosis
Treating KGN cells with MG132 not only attenuated Hu-17-induced aromatase protein degradation, but also reversed apoptosis ( Figure 7D). Therefore, to determine the relationship between aromatase stability and apoptosis, we explored the functional consequences of CYP19A1 silencing on apoptosis in KGN cells, siRNA-mediated knockdown of CYP19A1 ( Figure 8A,B) increased the proportion of apoptotic cells from 5.1% to 13.2% ( Figure 8C). Consistently, Based on the results of the SEA prediction, we used phosphokinase antibody arrays to identify signaling pathways affected by Hu-17 treatment. Since the estrogen signaling pathway is significant in estrogen-dependent gynecological tumors, we focused on the estrogen pathway and performed detailed mechanistic studies to explore new functions of Hu-17. We also used cisplatin and paclitaxel to demonstrate that only Hu-17, but not first-line chemotherapy drugs, reduced aromatase expression and estrogen levels, despite the ability of all three drugs to induce apoptosis in KGN cells was similar. These results provided evidence that the reduction in aromatase expression and estrogen production by Hu-17 is not due to the decreased numbers of KGN cells. Chemotherapy drugs, including cisplatin and paclitaxel, are the current first-line treatments for tumor of the ovary and breast. 41 Cisplatin creates intrastrand and interstrand DNA crosslinks to block DNA replication and exerts its anticancer effects. 42 Paclitaxel, a microtubule-targeted drug, stimulates microtubule polymerization to block mitosis and induce apoptosis. 43 Side effects of these drugs, such as neurotoxicity, myelosuppression, and drug resistance, are major obstacles in successful treatment outcomes. Therefore, endocrine therapy is an important adjuvant approach in preventing recurrences of estrogen-dependent tumors. Further, endocrine therapy has more tissue specificity and less toxicity than chemotherapeutic regimens. 8,44,45 Aromatase inhibitors are superior to tamoxifen and are the most widely used hormonal agents in the treatment of estrogen-dependent tumors. [46][47][48] Therefore, that Hu-17 inhibits aromatase expression and reduces estrogen levels make it a promising candidate to treat estrogen-dependent cancers.
Aromatase is a rate-limiting enzyme of estrogen synthesis, and CREB-1 is the key transcriptional factor in the regulation of aromatase in the ovaries. Aromatase expression is controlled by a promoter proximal to the start of translation in the ovary, which binds the transcription factors CREB-1 and SF-1. 49,50 CREB-1 is phosphorylated and activated by PKA and other protein kinases, such as those in the ERK, p38, and PI3K pathways. 10,37,39,40 We found that Hu-17 inhibited CREB-1 phosphorylation without affecting total CREB-1 expression. Additionally, Hu-17 reduced intracellular cAMP levels. As intracellular cAMP homeostasis plays an important role in the activation of PKA in response to follicle stimulating hormone in ovarian granulosa cells, it appears that Hu-17 regulates the cAMP pathway to decrease CYP19A1 transcription in KGN cells. The expression of CYP19A1 was suppressed by PD98059, a MEK/ERK pathway inhibitor, which is consistent with the results of previous studies in human primary granulosa luteal cells. 51 However, some data show that only inhibitors of AKT/PI3K and p38 MAPK, not inhibitors of JNK (SP600125) and ERK (PD98059), significantly decreased aromatase expression in KGN cells. 52 In our study, there was a significant difference in expression of CYP19A1 mRNA before 24 h, while after 24 h, we could not observe these differences. The MEK/ERK pathway could regulate CYP19A1 to an extent, although the PKA pathway played a vital role in CYP19A1 expression. These results indicated that Hu-17 reduced CYP19A1 transcription by inhibiting CREB-1 phosphorylation mainly through suppression of the PKA and ERK pathways, which is consistent with the results of the Phospho Explorer antibody microarray.
SEA also predicted that Hu-17 had the potential to affect protein tyrosine phosphatases. Tyrosine-specific protein phosphatases catalyze the removal of the phosphate group attached to a tyrosine residue. These enzymes are crucial regulatory components in signal transduction pathways, such as the MAPK pathway and cell cycle-regulation. As such, they are important in the regulation of cell growth, proliferation, differentiation, and transformation. 53 In addition to decreasing CYP19A1 mRNA expression, we also found that Hu-17 could increase degradation of aromatase protein in KGN cells, which is similar to what we observed with exemestane, which promotes aromatase degradation through the ubiquitin/26S proteasome pathway without affecting CYP19A1 mRNA in MCF-7 cells. 55 It will be interesting to investigate whether Hu-17 and exemestane promote aromatase degradation by the same mechanism; however, no direct evidence was obtained. Therefore, the mechanism by which Hu-17 regulates aromatase ubiquitination and proteasomal degradation needs to be further studied.
Although we demonstrated that the decrease in aromatase expression by Hu-17 was not dependent on apoptosis, we found that knockdown of CYP19A1 increased apoptosis in KGN cells, which suggested that the decrease in aromatase expression caused by Hu-17 could result in apoptosis. TODA et al. found that mice lacking aromatase activity because of targeted disruption of CYP19A1 exhibited anovulation and increased expression of pro-apoptotic genes such as bax in the ArKO ovaries compared with their wild-type siblings. 56 Therefore, CYP19A1 might have a role in regulating apoptosis. Furthermore, Hu-17 had the same effects on the breast cancer cells lines MCF-7 and SUM-159. Moreover aromatase inhibitors are highly effective and less toxic than chemotherapy and are often offered to ER-positive breast cancer patients to sustain a better quality of life. 57

| CONCLUSION
We demonstrate that the newly synthesized compound Hu-17, as a potent inhibitor of estrogen biosynthesis, reduced CYP19A1 expression and accelerated aromatase protein degradation in the human ovarian granulosa cancer cell line KGN. This research would draw attention to the study and development of new small-molecule compounds as aromatase inhibitors. Therefore, Hu-17 could serve as a potential treatment for estrogen-dependent cancers as well as be helpful in designing new pharmaceutical tools for the prevention and treatment of estrogen-dependent cancers.