The Flavonoid Baicalein Negatively Regulates Progesterone Target Genes in the Uterus in Vivo.

We previously reported that baicalein acts as a PR antagonist in vitro and that it increased expression of the GR target gene, glucocorticoid-induced leucine zipper (Gilz) in the uterus in vivo, which could be blocked by the PR and GR antagonist, RU486. 31 To investigate the effects of baicalein on uterine cell proliferation, CD-1 mice (n=5/group) were injected IP with 10% DMSO, 1 mg/kg progesterone (P4), 25 mg/kg baicalein, 1 mg/kg P4 combined with 25 mg/kg baicalein, or 1 mg/kg P4 combined with 10 mg/kg RU486 for 7 days. In the control group, immunohistochemistry for PCNA staining found that 49% of the luminal and glandular epithelial cells were proliferating ( Figure. 1A ), but in the P4 treated group, a significant number of the epithelial cells (93%) were quiescent ( Figure 1B ). When treated with P4 and baicalein, there were more proliferating epithelial cells (32%) compared to the P4 group (7%), which were not different than control ( Figure. 1C ). Baicalein alone resulted in 47% of the luminal epithelial cells stained for proliferation marker PCNA. This result indicated that baicalein alone does not reduce the percentage of PCNA positively stained luminal epithelium, but baicalein attenuated the anti-proliferative effects of P4 on epithelial proliferation. Mice were also treated with P4 in combination with the PR antagonist, RU486, and the uteri had 23% proliferating epithelial cells, which was not statistically different compared to the control group ( Figure 1D ). In the uterus, it is well understood that steroid hormones, such as estradiol (E2) and P4, have opposing actions in regulating cell proliferation and differentiation in a paracrine manner. 32 – 35 Estradiol, acting through stromal ER alpha, drives epithelial proliferation through the secretion of factors such as Fgf10, Hox10a and Bmp8a. 36 On the other hand, P4 inhibits expression of Fibroblast growth factors (FGFs) and blocks estrogen-induced epithelial proliferation through PR. 37 Treating with baicalein combined with P4 reduced the anti-proliferative effect of P4. In addition, studies showed that baicalein blocked proliferation and induced apoptosis in the cervical carcinoma Hela cell line. 38 , 39 Women with certain gynecological diseases such as endometriosis and uterine fibroids are treated with progestin antagonists. 40 , 41 These diseases often involve inflammation that contributes to abdominal pain. 42 , 43 Heart and neural crest derivatives-expressed protein 2 (HAND2) is a transcription factor that is exclusively expressed in the sub-epithelial stromal regions in the uterus and has a critical role in regulating uterine epithelial functions. 44 HAND2 is induced by P4 in ovariectomized mice and blocked by RU486. 37 We evaluated HAND2 protein expression by immunohistochemistry (IHC) after mice were treated with baicalein and our results showed that HAND2 protein was limited to a thin layer of sub-epithelial stromal cells in the control group ( Figure. 2A ), but its expression was upregulated and extended to deep regions of the stromal in the mouse uteri treated with P4 ( Figure. 2B ). In mice treated with the combination of P4 and baicalein, HAND2 staining was more diffuse than in the P4 group ( Figure. 2C ). HAND2 protein levels were comparable between the control group and the P4 combined with RU486 group ( Figure. 2D ). HAND2 expression was limited to the small region close to epithelium when treated with baicalein ( Figure. 2E ), indicating that baicalein alone did not increase HAND2 abundance compared the control group. Studies have reported that P4 upregulates HAND2, and HAND2 inhibits the expression of stromal FGFs that induce luminal epithelial proliferation through the ERK1/2 pathway in a paracrine manner. 37 , 45 Together with the proliferation data, the data indicate that P4 induces HAND2 and inhibits luminal epithelial proliferation, while baicalein blocks the P4-induced HAND2 expression and the inhibition of proliferation similar to RU486, the established PR antagonist. Zinc finger and BTB domain-containing 16 (ZBTB16) is a transcriptional factor that belongs to the family of Krüppel‐like zinc finger proteins and is involved in cell cycle control differentiation of myeloid cells, and spermatogenesis. 46 ZBTB16 is induced by P4 in the female reproductive tract and is essential for stromal cell decidualization. 47 Kommagani and colleagues reported in a study of human endometrial stromal cells that ChIP-Seq identified more than 10 progesterone response elements within the Zbtb16 gene, indicating that it may be a direct target of PR signaling. 48 We evaluated the effects of baicalein on ZBTB16 abundance, and our results showed that there was very little ZBTB16 in the control group ( Figure. 3A ), but its abundance was elevated in the P4 treated group ( Figure. 3B ). When treated with P4 and baicalein, the abundance level of ZBTB16 was reduced compared to P4 treatment alone, similar to the P4 and RU486 group ( Figure. 3D ). When treated with baicalein alone, ZBTB16 was at basal level, similar to the control ( Figure. 3E ). These results suggest that baicalein acts as a PR antagonist that decreases P4-induced ZBTB16 expression in the uterus. Furthermore, Qiu and colleagues reported that ZBTB16 was one of the downstream targets promoted by FOXA1 and higher FOXA1 expression was correlated with a higher incidence of endometrial cancer. 49 ZBTB16 was also found to be regulated by both GR and PR and it functioned as a tumor suppressor that inhibits proliferation and metastasis in breast cancer cells. 50 , 51 To evaluate the effects of baicalein on PR protein levels, uterine sections were immunostained for PR. The results showed that PR was expressed uniformly in the luminal and glandular epithelial cells in the control group ( Figure. 4A ). Treatment of P4, P4 with baicalein, or P4 with RU486 did not alter PR expression ( Figure. 4B - D ). Similarly, treatment of baicalein alone did not cause a difference in PR expression in the uterus compared to the control or P4 group ( Figure. 4E ). In our previous study, we also showed that baicalein did not change PR expression in breast cancer cells. 31 The results are consistent with the in vitro data. GR is a constitutively expressed transcription factor. 52 GR is expressed in the uterus of both mouse and human. 53 To evaluate the effect of baicalein on GR protein, uterine sections were stained for GR. The results showed that GR was abundantly expressed in both the stromal and epithelial cells in the control ( Figure. 5A ). In the P4 treated group, GR was expressed in the uterine stromal cells, but very little was in the luminal epithelial cells ( Figure. 5B ). When mice were treated with P4 and baicalein, GR was expressed intensively in the stroma as well as luminal epithelium, similar to the P4 and RU486 ( Figure. 5 C and D ). When mice were treated with baicalein alone, GR was expressed in both stroma and epithelium, resembles the control group ( Figure. 5E ). These data suggest that baicalein alone does not alter the expression of GR, but it could reverse the P4 induced stromal-only expression of GR to both stromal and epithelial expression. The exact effects of P4 on GR localization is not clear and the biological significance of epithelial and stromal expression remains unknown, however, GR expression in the uterus is associated with poor prognosis in ER-expressing endometrial tumors. 54 Future studies are required to understand the role of GR in the uterus. To evaluate and compare the transcriptomic profiles of the mouse uteri, we extracted mRNA from the control, P4, and baicalein groups and subjected it to RNA sequencing (n=4). Results for significantly altered transcripts are summarized in Figure 6 . Baicalein treatment significantly upregulated 46 genes and downregulated 131 genes. We found a large portion of genes that were upregulated by baicalein were downregulated by P4. Of the 131 mRNA significantly downregulated by baicalein, 44 were upregulated by P4; while 18 of 46 transcripts significantly upregulated by baicalein were downregulated by P4. The full list of commonly altered transcripts is presented in Table 1 . Full lists of significantly altered mRNA by each group are available as Supporting Information tables . DAVID functional annotation analysis showed that 46% (24 of 52) GO terms downregulated by baicalein were upregulated by P4, and 48% (13 of 27) KEGG pathways downregulated by baicalein were upregulated by P4. The top three oppositely regulated common KEGG pathways are focal adhesion, proteoglycan in cancer and cGMP-PKG signaling pathway, with nine, five and five genes in the pathway respectively. Full lists of altered GO/KEGG terms by each group are available as Supporting Information tables . Among these common genes changed in the opposite direction by baicalein and P4, studies have reported that many of them are regulated by P4 in uterine biology. 55 – 57 For instance, Cdkl2 transcript was upregulated by baicalein treatment, but downregulated in the P4 treatment group. A previous study also found that Cdk12 was downregulated by P4 in mouse uterus. 58 In human endometrial stromal cells (HESC), knockdown of PR leads to downregulation of Ccdc69. 59 Our data showed baicalein treatment downregulated Ccdc69 similar to knockdown of PR. In another study with microarray data, Cdr2 expression was enhanced by P4, but it was downregulated by baicalein in our study. 60 Progesterone is critical for successful pregnancy, and PR signals regulate many genes that important for implantation and decidualization. 4 , 61 We found that baicalein downregulated integrin Itga7 in the mouse uterus, but it was upregulated in human myometrium and decidualized stromal cells during pregnancy, which again is consistent with working to oppose P4 action. 62 , 63 Similarly, Tcf23, another baicalein-downregulated gene, was shown to be critical and upregulated in decidualization in HESC by P4. 64 Baicalein also regulated genes involved in uterine pathologies such as fibroids and endometrial cancer. Fibroids are benign smooth muscle tumors originated from the myometrium. Their development is highly dependent on ovarian hormones, and PR actions play a key role in fibroid growth. 9 A genomic and transcriptomic study revealed Hspb7 was associated with cell proliferation in fibroids, but it was downregulated by baicalein in our study. 65 Baicalein upregulated Mal2 in the mouse uterus, and Mal2 was previously found to be linked with proliferation, migration, and invasion in endometrial cancer, which is known to be related to reduced P4 action. 66 Interestingly, a few lncRNAs involved in PR signaling or tumorigenesis were also oppositely regulated by baicalein and P4. For example, Fam107a was downregulated by baicalein, but it was reported to be upregulated by MPA in myometrial explants in pregnant women. 67 Fam212b has been shown to be one of the core regulators of endometrial carcinogenesis, and was downregulated by baicalein treatment in our study. 68 Significantly downregulated transcripts by baicalein were subjected to DAVID KEGG pathway analysis and our result identified 27 pathways. In addition to the common pathways oppositely regulated by baicalein and P4, baicalein downregulated 14 other pathways including cAMP signaling pathway, oxytocin signaling pathway GnRH signaling pathway and ECM-receptor interaction ( Figure. 6E ). The role of baicalein in these pathways needs further investigation. Overall, these data suggest that baicalein regulates a subset of genes that are important in uterine physiology and pathology in the opposite direction of P4 and has antagonistic effect on PR in the uterus in vivo . RNAseq data was further analyzed by GSEA to evaluate significantly altered gene sets ( Table 3 and Figure. 7 ). Our data showed that 2 of 8 hallmark gene sets were negatively enriched in baicalein group but positively enriched in P4 group. These gene sets were UV response and myogenesis, and the normalized enrichment scores were 1.76 and 1.47 for P4, −1.94 and −2.29 for baicalein, respectively. Three of 12 hallmark genes were positively enriched in the baicalein group but negatively enriched in the P4 group. These gene sets were E2F targets, G2M checkpoint and DNA repair, and the normalized enrichment scores were −2.59, −2.26 and −1.61 for P4 and 1.95, 1.88 and 1.41 for the baicalein group respectively. The full list of significantly enriched gene sets is shown in Table 3 . These overlapping gene sets indicate a role of baicalein in regulating cell cycle and proliferation. Furthermore, some of the significantly enriched gene sets by baicalein alone are also involved in gynecological diseases. For instance, it is well understood that Epithelial To Mesenchymal Transition (EMT) plays an important role in the progression of many cancers including endometrial cancer. 69 , 70 Baicalein suppresses metastasis of breast cancer cells by inhibiting EMT 71 . Wnt/beta-catenin signaling is involved in several aspects of the genesis of fibroids, and inactivation of Wnt/beta-catenin signaling suppresses endometrial cancer cell growth in vitro. 72 , 73 Interestingly, our GSEA data showed both EMT and Wnt/beta-catenin signaling gene sets were negatively enriched by baicalein, suggesting a potential benefit on gene regulated in fibroids and endometrial cancer. These data suggest that baicalein has additional functions other than mediating PR signaling. Future studies are needed to elucidate the effects of baicalein on other gene sets and pathways. In order to evaluate whether baicalein could exert glucocorticoid side effects such as weight gain and diabetes, the body weight, and organ weights were measured at the end of the animal study. The weights of the body, uteri and liver for each group are summarized in Table 4 . There was no significant difference in body weight or organ weights in any treated group. The blood level of glucose was also measured and there was no difference in glucose level in any treated group. Although common glucocorticoid side effects include obesity, antagonism of insulin action and osteoporosis, our results showed no difference in body weight or serum glucose level of the mice. More research is warranted to investigate the role of baicalein on GR. Chinese Skullcap has been widely used as a medicinal plant in Asian countries for centuries, and the main bioactive compound is baicalein. 22 Baicalein is orally consumed and well absorbed from the stomach and small intestine, and the plasma concentration of baicalein reaches maximum 0.75–3 h after administration. It is predominantly metabolized in the liver and small intestine by glucuronidation via uridine 5’-diphospho-glucuronosyl-transferase systems. 74 In two studies where healthy adults were given baicalein chewable tablets at either a single dose of 100–2800 mg or multiple doses of 200–800 mg daily, researchers found no signs of liver or kidney toxicity and minimal mild side effects, indicating oral intake of baicalein was safe and well tolerated. 75 , 76 Our data show that baicalein opposes P4 action on the luminal epithelium and it blocks the expression of PR target genes HAND2 and ZBTB16. RNA sequencing analysis indicates that baicalein regulates a subset of PR target genes in the opposite direction of P4. This study is the first to show baicalein can repress some actions of PR in the murine uterus.

Ovariectomized CD1 mice (age 6–8 weeks from Envigo) were used for the animal studies. Baicalein (purity ≥ 95%) was purchased from Cayman (70610, Cayman Chemical, Ann Arbor, MI). In the first animal study, five mice were randomly assigned into each treatment group and received 10% DMSO, 1 mg/kg progesterone, or 25 mg/kg baicalein for 7 days through IP injection. RNA was extracted from the uteri of these mice for RNA sequencing analysis. In the second animal study, five mice were randomly assigned into each treatment group and received 10% DMSO, 1 mg/kg progesterone, 25 mg/kg baicalein, 1 mg/kg progesterone with 25 mg/kg baicalein, or 1 mg/kg progesterone with 10 mg/kg RU486 for 7 days through IP injection. After treatment, the animals were weighted and euthanized. The uterine tissue was collected and weighted. One uterine horn was snap-frozen in liquid nitrogen and stored at −80 °C for RNA extraction and the other uterine horn was fixed in 10 mL of 10% buffered formalin for 24 h, transferred into 70% EtOH and processed for histology using a Shandon 1000 Processor (Thermo) as described before. 77 The processed tissue was embedded with paraffin into 5 mm thick blocks and then sectioned into 5 μm thick sections using a microtome. The slides were dried for at least 24 h before being processed for further analysis. All animal studies were approved by the UIC Institutional Animal Care and Use Committee (Protocol number 18–205). Serum samples were collected for blood glucose measurement at the end of the animal study. Briefly, blood was drawn immediately after euthanasia from the posterior vena cava and cooled on ice for 30 min, then 15 min at room temperature (rt) to clot. The samples were centrifuged at 1000 x g for 10 min to remove the clot, and the supernatant liquid component (serum) was submitted to the Diagnostic Laboratory of the Biologic Resources Laboratory at UIC for measurement. Five samples per group were measured. Immunohistochemistry (IHC) was performed for PCNA, HAND2, FKBP5 on uterine samples as previous described. 78 Briefly, slides of uterine horns were deparaffinized using three xylenes washes and rehydrated though a series of decreasing concentrations of EtOH, then subjected to heat-induced antigen retrieval with sodium citrate buffer at 100 °C for 30 min and allowed to cool to rt. This was followed by inactivation of endogenous peroxidase activity with 0.3% H 2 O 2 /MtOH for 15 min in dark. The samples were then rinsed with phosphate buffered saline with Tween-20 (PBST) and incubated in blocking solution consisting of 5% horse serum (Vectastain ABC kit, Vector Laboratories, Inc.) diluted in 1%BSA/PBST at RT for 60 min. The tissue sections were incubated with following primary antibodies overnight at 4 °C: PCNA (1:200, 13110 Cell Signaling), FKBP5 (1:200, 14155–1 Protein Tech), PR (1:200, AB101688 Abcam), GR (1:100, 12041 Cell Signaling), HAND2 (1:200, ab200040 Abcam) and ZBTB16 (1:200, PA5–112862 Invitrogen). Next day, slides were rinsed with PBST prior to incubation with anti-goat biotinylated secondary antibody (Vectastain ABC kit, Vector Laboratories, Inc.) at 1:200 dilution in PBST for 60 min at rt. Slides were then rinsed and incubated in ABC solution (PBS: A: B=50:1:1) (Vectastain ABC kit, Vector Laboratories, Inc.) for 30 min at rt. For visualization of the immunoreactivity, all slides were subjected to chromogen 3’3-diaminobenzidine (DAB) (Vector Laboratories, Inc. Burlingame, CA) for 30 seconds. Slides were rinsed in tap water for 10 min to stop the DAB reaction. Thereafter, the slides were counterstained with hematoxylin for 1 min followed by dehydration and cover-slipping. After drying for 24 h, the slides were cleaned and imaged using Nikon E600 Eclipse microscope with CMOS C-Mount microscope camera. Uteri of mice treated with 10% DMSO, 1 mg/kg progesterone, or 25 mg/kg baicalein for 7 days in the first animal study were subjected to RNA isolation and RNA sequencing. RNA sequencing of uterine tissue was profiled using n=4 per treatment group. Total RNA was extracted from uterine tissues of mice using the Qiagen RNeasy Mini kit (Qiagen #74104) according to the manufacturer’s instructions. The concentration of mRNA was determined by a nanodrop. RNA libraries (three technical replicates/treatment) were created. RNA quality determination, mRNA enrichment, library construction, sequencing, and transcriptome statistical analysis were performed at the Genomics Core Facility at Northwestern University. Samples with RINs of 7 or greater were prepared with TruSeq mRNA-Seq Library Prep (Illumina) with 1 ug of RNA and 12 cycles of PCR amplification. The libraries were barcoded, pooled and sequenced on the HiSeq Sequencing 50 followed by statistical analysis. For RNA-Seq data, gene set enrichment of differentially expressed genes was performed using DAVID Webservice and GSEA. Gene sets with an FDR adjusted p-value <0.01 were considered significant. All data were analyzed utilizing GraphPad Prism software 8 (GraphPad Prism). Data are presented as means ± standard error of the means (SEM). Unpaired student’s t test and one-way ANOVA were performed. Tuckey’s test was used for multiple comparison. A statistical significance was assigned at p≤0.05.