Estrogen Promotes Pituitary Prolactinoma by Upregulating TLR4 / NF-κB/p38MAPK Pathway

Background: Prolactinomas have harmful effects on human health, and the pathogenesis is still unknown. Furthermore, the morbidity of women is much more than man, maybe related with estradiol level. Thus, it is important to reveal the pathogenesis and develop new therapeutic methods for prolactinomas. Methods: Immunouorescence analysis or Immunohistochemistry analysis were performed on the ERβ, TLR4 and prolactin (PRL) expressions of pituitary gland in C57BL/6 mice and human prolactinoma specimen. In the present study, the role of TLR4 in prolactinoma was determined using estradiol-induced mice models in C57BL/6 wild-type (WT) and TLR4−/− mice. MMQ cells were treated with estradiol, fulvestrant, LPS or transfected with different TLR4 small interfering RNA, which to study ERβ, TLR4 and PRL expression in MMQ cells. Co ‐ immunoprecipitates analysis was used to investigate the interaction between ERβ and TLR4. Results: Immunouorescence analysis or Immunohistochemistry analysis showed that PRL and TLR4 expression were co-located and increased in the pituitary gland of mice and human prolactinoma specimen compared with the control specimen. It was shown that PRL and TLR4 expression was co-located and increased signicantly in the pituitary gland of estradiol-injected prolactinoma mice compared with the control mice. Knockout of TLR4 signicantly inhibited tumor overgrowth, and PRL expression was decreased in estradiol-induced mice through regulating TLR4/NF ‐ κB/p38MAPK pathway. Estradiol promoted PRL expression through regulating TLR4/NF ‐ κB/p38MAPK pathway in vitro study, and pre-Inhibiting ERβ or TLR4 reverse the effect, while simultaneously activating ERβ and TLR4 enhanced PRL expression than activating single ERβ or TLR4. Furthermore, ERβ co-immunoprecipitates with endogenous TLR4 was assessed by co-immunoprecipitation analysis. Conclusions: These results suggest that estradiol promoted prolactinoma development by activating the TLR4/NF ‐ κB/ p38MAPK pathway through Erβ and TLR4 knockout inhibited the proliferation and secretion of prolactin in prolactinoma. E2+siTLR4 blotting. that TLR4 knockdown downregulated the TLR4/NF ‐ κB/p38MAPK estradiol, inhibited PRL protein expression in MMQ cells blotting that the cells PRL(P<0.05), ERβ(P<0.05), TLR4(P<0.05), Myd88(P<0.05), NF ‐ κBp65 and p38MAPK(P<0.05) protein compared with the only E2 or LPS treated our ndings


Prolactinoma animal models
The Estradiol-treated rat is a well-known model of pituitary lactotroph hyperplasia and hyperprolactinaemia [10,29]. The prolactinoma model was established by intraperitoneal injection of estradiol oil in mice (concentration of 1 mg/1 mL, 0.1 mL for each mouse once), once every 5 days and lasted for 50 days. The mice were divided into four groups (n=10): control group, estradiol-induced prolactinomas group, TLR4-/-group, estradiol-induced TLR4-/-group. All mice were prepared into prolactinomas model by intraperitoneally injecting estradiol oil except control group and TLR4-/-group mice. The control group and TLR4-/-group mice were injected with 1 mL of Castor oil for each mouse.

RNA interference
In total, three different sets of small interfering (si)RNA sequences against TLR4 were designed and synthesized by Guangzhou RiboBio Co., Ltd. (Guangzhou, China). The sequences of the three siRNA sequences used were: i) GCTATAGCTTCACCAATTT; ii) GCAGCAGGTCGAAT TGTAT; and iii) CCTAGAACATGTGGATCTT. A total of 2 × 10 5 MMQ cells/well were seeded into a 6-well plate and transfected with the different TLR4 siRNA separately. The cells were transfected using Lipofectamine® 2000 (Invitrogen;Thermo Fisher Scienti c, Inc.) according to the manufacturer's protocol. The RNA or protein was extracted 48 h later.

Enzyme-linked immunosorbent assay (ELISA)
The blood samples of mice were collected from the orbit. The serum was then isolated by centrifugation immediately. For the PRL ELISA, 50 µL of animal serum and 100 µL HRP labeled detection antibody were used to coat a 96-well plate 1h at 37°C. After being washed 5 times with 0.1% Tween (PBS solution). Samples were incubated with the 50 µL substrate A and B at 37°C for 15 min and then add stop solution for 50 µL, and 450 nm was set as the absorbance wavelength. The PRL ELISA kit was purchased from LunChangShuo Biotech Co.#SU-B20246(Xiamen, China).

Western blotting method
The pituitary glands of mice were obtained and stored at −80°C. The samples were placed in protein lysis buffer with protease inhibitors (Roche Diagnostics, Basel, Switzerland) to lyse the tissue protein in the homogenizing tubes and were subsequently homogenized using the tissue homogenizer (Wuhan Servicebio Technology Co., Ltd.). The protein concentrations of animal tissues or cell lysates were determined based on the Bradford protein assay. Equal amounts of proteins were separated with 10% or 12% SDS-PAGE, electro-transferred on nitrocellulose membranes and then blocked with 5% milk for 1h.

Immuno uorescence assay
The tissue specimens of animal or human pituitary glands were routinely formalin-xed, para nembedded and sectioned in 5µm pieces for immuno uorescence assays. Para n sections were dewaxed with water and then lled with ethylenediaminetetraacetic acid antigen repair buffer (pH 8.0) at 70°C for 20 min. After the sections were slightly dried, they were rinsed with owing water, and incubated with BSA for 30 min. Primary antibody against TLR4 (1: 1000, A nity Biosciences), ERβ (1: 2000, A nity Biosciences) and PRL (1: 1000, R&D Systems) were added and incubated with sections overnight at 4°C. After three PBS washes, the sections were incubated with secondary antibody for 50 min and then sealed with anti-uorescence quenching tablet. The uorescence images of sections were acquired under a Leica microscope and analyzed using Leica Application Suite X Software (Leica DFC450 C, Leica Microsystems Inc., Germany).

Immunohistochemistry
The human and mice pituitary gland or tumor specimens were paraformaldehyde-xed and para nembedded and were cut into 3 mm-thick sections. After xation onto glass slides, they were rehydrated in an ethanol gradient. Furthermore,3% of H 2 O 2 solution was added to devitalize the endogenous catalase, and 0.01 M of sodium citrate solution was added to extract the antigen. The slides were then stained with anti-TLR4 (1: 1000, A nity Biosciences), ERβ(1: 2000, A nity Biosciences), p38MAPK antibody from rabbit (A nity Biosciences, 1: 1000), and PRL (1: 1000, R&D Systems) antibody after blocked by 5 % of BSA. All tissues were visualized by incubation with goat anti-rabbit immunoglobulin G (IgG) (H+L)-HRP (1:100, Wuhan Servicebio, China). After being immunostained with a 3,3′-diaminobenzidine tetrahydrochloride immunohistochemistry color development kit (Wuhan Servicebio, China), the slides were xed with Mayer's hematoxylin. The stained tissues were photographed and analyzed using a Leica microscope (Leica DFC450 C, Leica Microsystems Inc., Germany) at 100× and 200× and 400× magni cations. The quanti cation of the expressions was assessed by manual counting of positive and negative cells per highpower visual eld.

Co-immunoprecipitation assay
Cultured MMQ cells were placed into 10cm dishes and treated with E2. The cells were then washed in phosphate-buffered saline (PBS) and resuspended in 1 ml of IP lysis buffer supplemented with phosphatase inhibitor and protease inhibitor cocktails (Wuhan Servicebio, China), and the cell lysates were ultrasonicated. After rotating at 4°C for 30 min, the cell lysates were collected and precleared by centrifugation at 12000 rpm at 4˚C for 10 min. Add 1.0µg IgG (the same species as the antibody used for IP experiment) and 20µL protein A/G beads to the supernatant of negative control (IgG) group (fully mixed before use). Add 20µL protein A/G beads directly to the experimental group, shake and incubate at 4℃ for 1h and centrifuged for 5min(2000g) at 4℃, supernatant was taken. For each pull-down, 5 µg of anti-ERβ Ab (dilution 1:3000,14007-1-AP; Proteintech) or anti-TLR4 Ab (dilution 1:500, SC-293072; Santa) was added to the normalized lysate, and the mixture was incubated overnight at 4˚C.Immune complexes were then precipitated with protein A/G plus agarose (Millipore). Immunoprecipitates were collected by centrifugation (2000g,4˚C,5min) and washed gently with lysis buffer. Immunoprecipitation samples were mixed with protein loading buffer for immunoblotting.

Statistical analysis
All quantitative data are presented as mean ± S.D. Statistical comparison were analyzed by the one-way ANOVA or Student's t-test using SPSS software (version 19.0). P value < 0.05 was considered statistically signi cant.

ERβ correlates with TLR4 expression positively in prolactinoma tissue
To identify the potential role of ERβ and TLR4 in the development and progression of prolactinoma, the pituitary glands of 20 prolactinoma patients and 10 patients with other pituitary diseases were collected and the expression of ERβ and TLR4 were evaluated by immunohistochemistry. Our results showed that expression of ERβ was negative in pituitary gland tissues while positive in tumor tissue, and that TLR4 was highly expressed in prolactinoma tissue than that of control. (Fig. 1A, C) Compared with pituitary gland tissues, the H-score of both ERβ(P<0.01) and TLR4 (P<0.01) were increased signi cantly in prolactinoma tissues. (Fig. 1B, D) Furthermore, the H-score of ERβ-positive tissues was highly correlated with the level of TLR4 overexpression in prolactinoma tissues (Spearman's correlation coe cient, r=0.873, P<0.001; Fig. 1F). TLR4 protein was found to be co-located with PRL protein in ve samples of human prolactinomas by immuno uorescence. After con rming that exposure to E2 induced endogenous interaction of ERβ and TLR4 and increased TLR4 protein levels, the PRL expression of the MMQ cell was assessed and mediated by exposure to E2, LPS or combined administration of E2+LPS. In addition, the role of TLR4 in prolactinoma by seting up siTLR4 and E2+siTLR4 group was observed. Compared with the E2-treated group, PRL protein expression decreased signi cantly(P<0.05), along with decreased ERβ (P<0.05), TLR4(P<0.01), Myd88(P<0.05), NF-κB p65(P<0.05) and p38MAPK(P<0.01) proteins in E2+siTLR4 cell groups detected via Western blotting. Our results suggest that TLR4 knockdown downregulated the TLR4/NF-κB/p38MAPK pathway activation induced by estradiol, and inhibited PRL protein expression in MMQ cells (P<0.05). Furthermore, Western blotting revealed that the cells treated with E2+LPS signi cantly increased PRL(P<0.05), ERβ(P<0.05), TLR4(P<0.05), Myd88(P<0.05), NF-κBp65 (P<0.05) and p38MAPK(P<0.05) protein expression, compared with the only E2 or LPS treated groups. (Fig. 2E and K) Taken together, our ndings indicated that activation of ERβ and TLR4 synergistically promoted the PRL expression and secretion in MMQ cells.

ERβ co-interacts with TLR4 protein in MMQ cell lines
The result showed that TLR4 protein co-immunoprecipitates with endogenous ERβ. After pre-precipitated TLR4 labeling, ERβ proteins were also expressed. (Fig. 3)

ES mice model was established successfully with the volume elevation of pituitary gland
The resultsshowed that the volume of the pituitary gland and tumor/body weight (Fig. 4A, B, P<0.01) were elevated signi cantly in estradiol-induced prolactinoma mice compared with WT mice, The ELISA results showed that serum PRL levels of prolactinomas model mice were signi cantly elevated compared with those of WT mice (Fig. 4C, P<0.01), indicating that the prolactinoma model was established successfully.

Discussion
The current theory on the pathogenesis of prolactinomas includes the effects of endocrine hormone, abnormal expression of genes, and microRNA irregular expression [30,31]. However, the exact pathogenesis of prolactinomas was still not clear. The gender difference in the occurrence of prolactinoma has long existed in the clinical, especially in the period of 20-30 years old, the incidence ratio of male and female even reached 1:10. The reason for this phenomenon may be the high estrogen level of women at this stage. [32] The role of estrogen is well established in promoting prolactinoma, DRD2 Knockout mice and estradiol-injected rats or mice are the most extensively acceptable prolactinoma animal models [10,28,29], Estradiol-injected mice was selected in the present study to clarify the role of estradiol in prolactinoma. ERs can regulate Toll-like receptor (TLR) signaling pathways in the immune system [33], Additionally, a previous study reported that estrogen promotes upregulation of TLR4/NF-κB/p38MAPK activation via ERβ [24]. Based on these previously results, we suppose that estrogen-activated ERβ/TLR4/NF-κB/p38MAPK pathway participated in the pathogenesis of prolactinomas.
TLR4 plays a fundamental role in pathogen recognition and activation of innate immunity, and is often overexpressed in malignant and tumour-in ltrating immune cells. TLR4 plays a crucial role in mesenchymal stem cell (MSC)induced inhibition of natural killer (NK) cell function. TLR4 ultimately activates the transcription factor NF-κB by linking to MyD88, which is required for the expression of cytokines, chemokines, and other stimulating molecules such as TNF-α, IL-1β, IL-6, IL-8, IL-12, and MIP 1α.
[34]Various stimulators induce activation of the p38MAPK, an important kinase involved in cell proliferation and apoptosis. [35] The proliferation-relatedproteinsB-cell-speci c Moloney murine leukemia virus insertion site1(BMI1) ,Bcl2, and Bax were determined in the pituitary gland via western blot. The BMI1 and Bcl2/Bax protein expression of estradiol-treated mice are increased signi cantly compared with WT mice. However, this effect reversed in estradiol-treated TLR4-/-mice. The results proved that estrogeninduced prolactinoma may be related to pituitary proliferation.
TLR4 is highly expressed in various kinds of tumor cells, and activation of TLR4 mediated p38MAPK promotes tumor growth and proliferation [36,37]. Therefore, TLR4 can be a therapeutic target. In this study, immunohistochemistry analysis showed that ERβ and TLR4 were overexpressed, and TLR4 was co-located with PRL protein in human prolactinoma tissues by immuno uorescence analysis. The high expression and positive correlation of TLR4 and ERβ are rst found in clinical samples of prolactinoma. Furthermore, TLR4 co-located with PRL, a marker of prolactinoma, laying a foundation for subsequent in vitro and in vivo studies.
Western blot results also showed that the expression of ERβ, TLR4 and PRL were increased in the pituitary tissues of estradiol-treated mice. Furthermore, TLR4 knockout signi cantly inhibited tumor overgrowth and PRL secretion, and ERβ, TLR4 and PRL protein expression in estradiol-induced mice. It was further veri ed that TLR4 is a potential target of prolactinomas. It is reported that estradiol and ERs regulate in ammatory pathways of innate immune cells, including dendritic cells and macrophages [33].As a classic protein regulating innate immunity and in ammation, TLR4 has many interactions with estrogen[38]. Therefore, excessive estrogen perhaps was a key cause that in uenced immune and in ammatory processes in the body. The TLR4 activation effect of estradiol have been proved in vitro and in vivo experiments in our study, thus we have su cient evidence to verify the role of TLR4 in estradiol-mediated prolactinoma. In order to verify the role of estradiol in the ERβ/TLR4/NF-κB/p38MAPK pathway in prolactinoma, the effects of inhibition and activation of ERβ on TLR4/NF-κB/p38MAPK pathway and PRL expression were investigated in MMQ cells in vitro. MMQ cells were treated with estradiol and fulvestrant, which can activate or inhibit the protein expression of ERβ/TLR4/NF-κ B/p38MAPK pathway and PRL expression. Meanwhile, the expression of TLR4/NF-κB/p38MAPK pathway was inhibited after the transfection knockdown of TLR4, and the expression of ERβ was also decreased, which indicate that there may be an interaction between TLR4 and ERβ. Then TLR4 protein directly binding to ERβ protein were found by immunoprecipitation, these results indicate that both TLR4 and ERβ play important roles in mediating the occurrence of prolactinoma. In addition, the expression of p38MAPK and PRL were increased signi cantly after treatment with LPS and E2 than with the single LPS or E2, which proves that the interaction of TLR4 and ERβ can synergistically regulate prolactin expression and tumor growth.
Our study rst reported the role of the TLR4/NF-κB/p38MAPK pathway in prolactinoma and TLR4 knockout inhibited the proliferation and secretion of prolactin in prolactinoma mice effectively. The limitation of this study is that only para n specimens of human prolactinoma tissues were obtained, and more detailed clinical data were not obtained, such as the relationship between tumor size and TLR4 positive expression in tissues. Moreover, drugs targeting inhibition of TLR4/p38MAPK in pituitary tissue for prolactinoma have not been developed, further research is needed. in conclusion, previous studies of our group demonstrated that MAPK14 and NLRP3 knockout can inhibit the proliferation and secretion of prolactin in prolactinoma. Therefore, this study focused on exploring the role of their upstream protein TLR4 in prolactinoma and whether estradiol-mediated prolactinoma genesis is mediated by TLR4/NF-κB/p38MAPK pathway. We reveal the correlation between the interacting ERβ and TLR4 signaling pathways. In addition, we demonstrated that E2 or LPS promoted PRL expression and prolactin tumor proliferation through the p38MAPK pathway. Further research on drugs targeting estrogen receptor and TLR4/NF-κB/p38MAPK pathway is the focus of our research, and new drug development based on these targets may provide new treatment strategies for prolactinoma.

Conclusion
This study rstly demonstrated that estrogen activated TLR4/NF-κB/p38MAPK pathway and promoted the development of prolactinomas, which provided new therapeutic targets for drug study in future.

Declarations
Ethics approval and consent to participate All animal experiments were approved by the Ethics Committee of Tongren Hospital A liated to Wuhan University, The Third Hospital of Wuhan. All samples and data from participants were anonymized and their written informed consent was obtained. The present study was approved by the Institutional Research Ethics Committee of the Central Hospital of Wuhan, Tongji Medical College, Huazhong University of Science and Technology. This study was conducted in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication NO.85-23, revised 1996). All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study and supporting our ndings are included and can be found in the manuscript. The raw data can be provided by corresponding author on reasonable request.