Paeoniflorin Ameliorates Hyperprolactinemia-Induced Inhibition of Osteoblastogenesis by Suppressing the NF-κB Signaling Pathway

Hyperprolactinemia is a common endocrine disease in women of reproductive age. Research has shown that patients with hyperprolactinemia often have decreased bone mineral density and an increased risk of fractures. However, there is still a lack of effective treatments. Paeoniflorin, one of the primary bioactive components in peony, is widely used in traditional Chinese medicine. Research has shown that paeoniflorin promotes osteoblast differentiation. However, whether paeoniflorin plays a role in hyperprolactinemia-induced osteoblastogenesis inhibition is not yet clear. In this study, we investigated the effect of paeoniflorin on prolactin (PRL)-mediated inhibition of osteoblast function. Our results showed that prolactin significantly reduced the expression of alkaline phosphatase (ALP), Osterix, and runt-related transcription factor 2 (RUNX2) in MC3T3-E1 cells cultured in an osteoblast differentiation medium, suggesting that prolactin inhibited osteoblast function. After treatment with paeoniflorin (PF), the expression of these osteoblast markers was upregulated. In addition, our findings proved that paeoniflorin increased the absorbance values of ALP-positive cells and the areas of alizarin red S (ARS) deposition compared to those in the prolactin group, suggesting that paeoniflorin reversed the PRL-induced reduction in osteoblast differentiation. The PRL-induced activation of nuclear factor kappa B (NF-κB) was significantly reversed by paeoniflorin, indicating that paeoniflorin promoted osteoblast function by inhibiting the NF-κB signaling pathway. In summary, these results showed that paeoniflorin alleviated the inhibitory effect of prolactin on osteoblastogenesis by suppressing the NF-κB signaling pathway.


Introduction
Hyperprolactinemia (HPRL) is characterized by dysfunction of the hypothalamic-pituitary-reproductive axis and mainly refers to serum levels of PRL that exceed normal values (PRL > 25 ng/ml). e incidence rate of HPRL is 0.4% in the general population and as high as 9% to 17% in women with reproductive disorders. In premenopausal women, HPRL can lead to oligomenorrhea, amenorrhea, and anovulatory cycles of infertility, while in men, it leads to impotence and breast development [1]. Studies have shown that basal circulating PRL (7-10 ng/ml) is essential for maintaining normal bone growth and remodeling [2,3]. High physiological levels of PRL during pregnancy (75-100 ng/ml) and prolonged lactation (200-350 ng/ml) can lead to transient osteopenia and a negative calcium balance [4][5][6]. Although lactation-induced osteopenia can be reversed after weaning, there have been reports of severe osteoporosis and an increased risk of fracture [7,8]. Pathological conditions associated with chronic HPRL (up to 1000 ng/ml), such as prolactinoma and prolonged antipsychotic drug use, result in progressive osteopenia and osteoporosis [9][10][11]. us, the treatment of HPRL-induced bone loss is a subject requiring close clinical attention.
PRL-mediated suppression of estrogen synthesis has long been considered a mechanism by which HPRL induces bone loss [12,13]. e bone may be the direct target of PRL; however, osteoblasts express transcripts for PRL receptors (PRLRs) [14,15]. Since osteoclasts do not express PRLRs, HPRL-induced bone loss could be due to PRL-mediated inhibition of osteoblast function. Studies have shown that HPRL reduces the expression of the osteoblast gene Runx2, downregulates the mRNA of osteoprotegerin (OPG), and increases the protein expression of the receptor activator for nuclear factor-κB ligand (RANKL) [13,16,17]. erefore, improving the function of osteoblasts may be an effective way to inhibit HPRL-induced bone loss. Currently, parathyroid hormone (PTH) is the main factor that promotes bone formation. However, prolonged administration of PTH can result in hypercalcemia [18]. In addition, the persistent role of PTH in promoting osteoblasts is also a problem to be addressed. Consequently, it is critical to identify new treatments for HPRL-induced osteoporosis. Traditional Chinese medicine is widely used to prevent and treat various diseases and is well-known for its affordability, mild side effects, and safety. is type of treatment is the current research direction of most scholars.
Prolactin, a hormone produced by the pituitary gland, has multiple physiological functions, including immunoregulation. e immune-modulatory activities of PRL may arise from increasing nuclear transcription factors such as IRF-1 and NF-kB, which play a pivotal role in many immune functions [27]. Ochoa-Amaya et al. [28] found that shortterm hyperprolactinemia has proinflammatory effects. Olavarría et al. [29] also discovered the capacity of PRL to promote, through the NF-kB signaling pathway, the polarization of fish macrophages to a proinflammatory M1/ classically activated phenotype characterized by the production of ROS and proinflammatory cytokines. However, it is not clear whether PRL plays a role in osteoblastogenesis inhibition through the NF-kB signaling pathway.
In our study, we established models of PRL-induced inhibition of osteoblast function in vitro to observe the effect of PF on PRL-induced inhibition of osteoblastogenesis. Our findings suggested that PF ameliorated HPRL-induced osteoblastogenesis inhibition by promoting osteoblast differentiation by inhibiting NF-κB activity (Scheme 1). erefore, drugs containing PF as the main active ingredient may be an effective way to treat HPRL-induced osteoporosis.

Prolactin Dose-dependently Inhibited Osteoblast Differentiation In Vitro.
e cell counting kit-8 (CCK-8) assay results indicated that cell viability was unaffected by treatment with PRL at concentrations below 1 ng/ml for 1, 3, and 5 days (Figure 1(a)). To examine the role of PRL in osteoblast differentiation, ALP and ARS staining were used. As shown in Figures 1(b)-1(c), visualization of the cells under an optical microscope revealed that the size of the ALP-positive area decreased as the PRL concentration increased, while analysis of ALP-positive cells revealed that PRL inhibited the differentiation of osteoblasts at all concentrations (0.1-1 ng/ ml). ARS staining of mineralization showed that PRL reduced mineralization in vitro (Figure 1(b)). Quantification of the mineral nodules revealed that PRL suppressed mineralization associated with osteogenesis in a dose-dependent manner (Figure 1(d)).
Western blot (WB) analysis and quantitative real-time polymerase chain reaction (qRT-PCR) were used to assess whether PRL suppressed osteoblast-related protein and gene expression, respectively. MC3T3-E1 cells were cultured in an osteogenic induction medium and treated with various concentrations of PRL (0.1, 0.2, 0.4, and 1 ng/ml). e WB results verified that the expression of proteins, including ALP, Osterix, and RUNX2, was markedly downregulated by PRL (P < 0.05) (Figures 2(a)-2(f )). Moreover, the same trend was observed in the mRNA expression levels of ALP, Osterix, and RUNX2 (Figures 2(g)-2(j)). ese results demonstrated that PRL suppressed osteoblast differentiation in a dose-dependent manner.

Paeoniflorin Ameliorated Hyperprolactinemia-induced
Inhibition of Osteoblastogenesis. To analyze the effect of PF on osteoblast viability, a CCK-8 assay was performed. As shown in Figure 3(b), PF had no toxic effects on MC3T3-E1 cells at concentrations below 100 μM for 1, 3, and 5 days.
MC3T3-E1 cells were incubated in an osteogenic induction medium and treated with 1 ng/ml PRL and different concentrations of PF (10 and 100 µM) for 2 and 3 weeks. ALP and ARS staining showed that PF dose-dependently reversed the PRL-induced reduction in osteoblast differentiation (Figure 3(c)). e quantification of ALP-positive cells is shown in Figure 3(d), and the absorbance values in the PF groups that were treated with 10 and 100 µM were 0.2420 ± 0.0017 nm and 0.3210 ± 0.0245 nm, respectively. Compared with 0.1557 ± 0.0091 nm in the PRL group, these values were statistically significant. ARS-positive areas were quantified, and the absorbance value in the PRL group was 0.03733 ± 0.00643 nm. In comparison, the value increased to 0.09533 ± 0.00208 nm and 0.1207 ± 0.0067 nm after treatment with 10 and 100 µM PF, respectively (Figure 3(e)).

Paeoniflorin and the NF-κB Signaling Pathway.
e NF-κB signaling pathway is an essential molecular response to various types of cellular stress. Xu et al. [27] and Zheng et al. [28] confirmed that this signaling pathway plays a crucial role in the differentiation of osteoblasts. PRL has been shown to affect the NF-κB signaling pathway [29,30]. us, we further investigated whether PRL affected osteoblast function via the NF-κB signaling pathway. WB analysis showed that the phosphorylation of IκB-α and NF-κB was dosedependently enhanced by PRL (s Figure 5(a)-5(c)), suggesting that PRL suppressed osteoblastogenesis by activating the NF-κB signaling pathway.
Our findings indicated that PF could promote osteoblast differentiation. Considering that PRL inhibited osteoblast function through the NF-κB signaling pathway, we further explored whether PF rescued PRL-induced inhibition of osteoblast differentiation via this signaling pathway. MC3T3-E1 cells were incubated in the osteogenic induction medium and treated with 1 ng/ml PRL and different concentrations of PF (10 and 100 µM) for 3 days, and then, WB analysis was performed. As shown Figures 6(a)-6(c), p-NF-κB and p-IκB-α were dose-dependently suppressed by PF in the presence of PRL.

Discussion
HPRL is a common cause of female reproductive disorders. Both physiological and pathological HPRL can lead to osteoporosis, which seriously threatens bone health in women over time [4][5][6][7][31][32][33][34]. Currently, there is still a lack of safe and effective treatments for HPRL-induced osteoporosis. Osteoblasts are directly involved in bone formation and play a crucial role in HPRL-induced bone loss. us, osteoblasts are a therapeutic target in HPRL-induced osteoporosis. It has been well documented that osteoblast functions are suppressed in the presence of PRL [15,16]. Our results also confirmed that PRL inhibited osteoblast function at concentrations of 0.1 to 1 ng/ml. It has been reported that PRL decreases the protein or mRNA expression of osteogenic markers, including ALP, OCN, OPN, and RUNX2, to inhibit osteoblastogenesis [14,17]. Our study was consistent with these previous reports that the expressions of proteins and mRNA, including ALP, Osterix and RUNX2, were markedly downregulated by PRL, which demonstrated that PRL suppressed osteoblast differentiation. All these results proved that increasing osteoblast function can rescue PRLinduced osteoblastogenesis inhibition. PF is a small molecule compound with strong adaptability and many targets. It was reported that PF had protective effects on advanced oxidation protein product (AOPP)-induced oxidative injury in human umbilical vein endothelial cells (HUVECs) [35]. In addition, Chen et al. [36] showed that PF ameliorated α-naphthylisothiocyanate (ANIT)-induced cholestasis in rats. Wang et al. [37] showed that PF could reduce bone loss induced by hyperlipidemia. In this study, we established models of osteoblast differentiation inhibition induced by PRL in vitro. We found that PF treatment can promote osteoblast function in the presence of PRL. is result is consistent with a previous study showing that PF facilitates osteoblastogenesis and thus reduces ovariectomy-induced osteoporosis in ovariectomized mice [26]. Paeoniflorin, as a water-soluble monoterpene glycoside [38], cannot penetrate directly through the phospholipid bilayer. erefore, we speculate that PF is transported across the cell membrane through passive transport; however, the specific mechanism needs to be further studied.
NF-κB is a family of dimeric transcription factors that is fundamental to cellular differentiation, proliferation, and survival in almost all multicellular organisms. Interactions with the inhibitor protein IκB controls NF-κB. NF-κB can be activated by signal-induced degradation of the IκB-α protein. As IκB-α degrades, NF-κB is activated and translocates into the nucleus, where it binds to target DNA sites and activates the transcription of genes encoding proteins that control cell growth. Previous studies have shown that PRL affects the NF-κB signaling pathway [39,40]. Similarly, our findings revealed that PRL suppressed osteoblastogenesis by activating the NF-κB signaling pathway. Numerous studies have proven that the NF-κB signaling pathway is important in regulating osteogenesis [30,41,42]. Wang et al. [43] showed that scutellarin promoted osteoblast proliferation and function via the NF-κB/p65 signaling pathway. Liu et al. [44] proved that pilose antler peptide enhanced osteoblast differentiation by regulating the NF-kB pathway. Similarly, we found that PF can inhibit the phosphorylation of IkBα and p65 in osteoblast differentiation inhibition induced by PRL, which demonstrated that PF alleviated PRL-induced inhibition of osteoblast function by suppressing the NF-κB signaling pathway.
In summary, our study demonstrated that PF could promote osteoblast differentiation in the presence of PRL by suppressing the NF-κB signaling pathway and reduce PRLinduced osteoblastogenesis inhibition. erefore, PF may be a potential treatment for bone loss induced by HPRL.   International Journal of Endocrinology 1, 3, and 5 days of incubation with PRL (0, 0.1, 0.2, 0.4, 1, 10, or 100 ng/ml) or PF (0, 1, 10, 100, or 1000 μM), 10-μl CCK-8 was added to each well, and the cells were incubated at 37°C and 5% CO 2 for an additional 1 hour. e absorbance of each well was measured at 450 nm by a microplate reader (ELx800, Bio-Tek, USA). additional wash, the stained cells and mineral nodules were photographed and observed at 450 nm in a microplate reader.

Western Blot (WB)
Analysis. MC3T3-E1 cells were seeded at 5 × 10 5 cells/well in 6-well plates, incubated with the osteogenic induction medium and treated with or without PRL (1 ng/ml) and PF (10 μM or 100 μM) for the indicated times. en, the cells were lysed with a RIPA lysis buffer (P0013 B, Beyotime Biotechnology, Shanghai, China) containing a protease inhibitor cocktail (P1010, Beyotime Biotechnology, Shanghai, China). After protein quantification, 20-μg sample was loaded onto gels, separated by 4-20% SDS-PAGE (P0468 M, Beyotime Biotechnology, Shanghai, China), and transferred onto PVDF membranes (IPVH00010, Millipore). Beta-actin was detected separately when the molecular weight of the target protein, including ALP, RUNX2, p-NF-κB, and NF-κB, was largely different from it. When target proteins such as Osterix, p-IκB-α, and IκB-α with molecular weight close to that of β-actin were detected, β-actin was detected after stripping with a Stripping buffer (P0025, Beyotime Biotechnology, Shanghai, China) and reprobing. Next, the membranes were blocked in a QuickBlock Blocking Buffer for Western blot (P0252, BeyotimeBiotechnology, Shanghai, China) at room temperature for 15 minutes and incubated with rabbit primary antibodies at 4°C overnight. e antibodies used were as follows: ALP (1 :1000, ab65834, Abcam, UK), Osterix (1 :1000, ab209484, Abcam, UK), RUNX2 (1 :1000, ab23981, Abcam, UK), IκB-α (1 :1000, ab32518, Abcam, UK), NF-κB (1 : 2000, ab16502, Abcam, UK), β-actin (1 :1000, ab8227, Abcam, UK), p-IκB-α (1 :1000, 2859, Cell Signaling Technology, USA), and p-NF-κB (1 :1000, 3031, Cell Signaling Technology, USA). After being washed with TBST (CW0043S, CWBiotech) for 30 minutes, the membranes were incubated with secondary antibodies (1 :1000, 5571, Cell Signaling Technology, USA) for 1 hour at room temperature. Finally, the proteins were analyzed with a chemiluminescent HRP substrate (WBKLS0500, Millipore Corporation). e protein bands were visualized by ImageQuant LAS 500 (Ge Healthcare Bio-Sciences AB, Sweden), and the gray values were analyzed by ImageJ software (NIH, USA) and normalized to β-actin. e experiments were repeated at least 3 times. International Journal of Endocrinology (10 μM or 100 μM) for 2 days. e cells were washed with PBS for 3 times in ice and lysed with 1 ml of TRIzol reagent (15596018, Ambion, USA). After complete dissolving at room temperature, the samples were removed to a new 1.5 ml microcentrifuge tube. Next, 200-μl chloroform was added and shaked vigorously, and then, the mixture was incubated at room temperature for 10 min. Later, the tube was centrifuged at 12000 × g for 15 min, then three distinct phases were observed in the mixture. 400 μl of the upper colorless phase was transferred into 400-μl cold isopropanol in a separate tube. e tube was incubated at room temperature for 10 min after gently mixing by inverting the closed tube several times. e tube was centrifuged at 12000 × g for 15 min, and then, isopropanol was discarded. RNA was precipitated as a pellet in the bottom of the tube. 1-ml 75% cold ethanol was added into the tube. Afterwards, the tube was centrifuged at 7500 × g for 5 min at 4°C, and then, the ethanol was discarded. After that, RNA was dissolved in 20-μl RNase-free water, and the concentration was detected by a NanoDrop ™ One/OneC Microvolume UV-Vis Spectrophotometer ( ermo Fisher Scientific, USA). e ratio of the absorbance at 260 and 280 nm (A260/280) was used to assess RNA purity. e A260/A280 ratio at 1.8-2.0 is acceptable. en, reverse transcription was performed to convert the RNA to cDNA. First, cDNA was synthesized using the Prime Script RT Master Mix kit (G492, Abm, Canada). 4 μl of Prime Script RT Master Mix, dH2O, and 1 μg of UV-quantified RNA were mixed. e reactions were incubated at 25°C for 10 min, 42°C for 15 min, 85°C for 5 min, and 4°C for infinite time by a thermal cycler (Applied Biosystems, USA). Samples were transferred to ice or at −20°C for long-term storage.  GAPDH was used to normalize the expression of target genes. e sequences for the forward and reverse primers used in this study were as follows: ALP (forward: 5′-CCAACTCTTTTGTGCCAGAGA-3′, reverse: 5′-GGCTA-CATTGGTGTTGAGCTTTT-3′), Osterix (forward: 5′-GGA AAGGAGGCACAAAGAAGC-3′, reverse: 5′-CCCCTTAGG CACTAGGAGC-3′), RUNX2 (forward: 5′-GACTGTGGT-TACCGTCATGGC-3′, reverse: 5′-ACTTGGTTTTTCATAA CAGCGGA-3′), and GAPDH (forward: 5′-AGGTCGGTG TGAACGGATTTG-3′, reverse: 5′-GGGGTCGTTGATGG CAACA-3′). Data acquisition is performed automatically by the PCR equipment software. For each sample, we calculated the average Ct (Ct av ) of the three technical replicates for both the target genes and the endogenous reference gene as follows: Ct av � (Ct 1 + Ct 2 + Ct 3 )/3; Ct 1 , Ct 2 , and Ct 3 are the average values of the three technical replicates. e relative amount of target genes was calculated for each sample using the ΔCt method [45] as follows: ΔCt av � [Ct av (end)−Ct av (target gene)], where end is the endogenous gene used as reference. e fold change (FC) of the target genes DNA amount between different samples were calculated as follows: FC � 2 −ΔCtav . We measured the relative amount of target genes DNA present in a sample compared to that of the control group.
GraphPad Software Inc., USA). Multiple comparisons were performed by one-way analysis of variance (ANOVA) with Newman-Keuls posttests. P-values less than 0.05 were considered to be statistically significant.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.