Inhibition of Phosphodiesterase 5 Promotes the Aromatase-Mediated Estrogen Biosynthesis in Osteoblastic Cells by Activation of cGMP/PKG/SHP2 Pathway

Mechanical stimulation induces bone growth and remodeling by the secondary messenger, cyclic guanosine 3’, 5’-monophosphate (cGMP), in osteoblasts. However, the role of cGMP in the regulation of estrogen biosynthesis, whose deficiency is a major cause of osteoporosis, remains unclear. Here, we found that the prenylated flavonoids, 3-O-methoxymethyl-7-O-benzylicaritin (13), 7-O-benzylicaritin (14), and 4'-O-methyl-8-isopentylkaempferol (15), which were synthesized using icariin analogs, promoted estrogen biosynthesis in osteoblastic UMR106 cells, with calculated EC50 values of 1.53, 3.45, and 10.57 µM, respectively. 14 and 15 increased the expression level of the bone specific promoter I.4-driven aromatase, the only enzyme that catalyzes estrogen formation by using androgens as substrates, in osteoblastic cells. 14 inhibited phosphodiesterase 5 (PDE5), stimulated intracellular cGMP level and promoted osteoblast cell differentiation. Inhibition of cGMP dependent-protein kinase G (PKG) abolished the stimulatory effect of 14 on estrogen biosynthesis and osteoblast cell differentiation. Further, PKG activation by 14 stimulated the activity of SHP2 (Src homology 2 domain-containing tyrosine phosphatase 2), thereby activating Src and ERK (extracellular signal-regulated kinase) signaling and increasing ERK-dependent aromatase expression in osteoblasts. Our findings reveal a previously unknown role of cGMP in the regulation of estrogen biosynthesis in the bone. These results support the further development of 14 as a PKG-activating drug to mimic the anabolic effects of mechanical stimulation of bone in the treatment of osteoporosis.


INTRODUCTION
Osteoporosis is a major global public health problem caused by the reduced estrogen level in postmenopausal women (1). In humans, aromatase cytochrome P450 (CYP19A1) catalyzes the formation of estrogens from C19 androgens (2). The aromatase expression at various sites is regulated by tissue-specific promoters through the alternative splicing mechanisms (3). In bone, class I cytokines such as TGF-b1, IL-1b, and TNF-a drive aromatase expression by the usage of promoter I.4 (4). Aromatase activity is a key factor in skeletal development and mineralization, and is crucial to estrogen production in the bone (5). Aromatase activity may decline with an increase in during aging, and the contribution of such decline to age-related bone loss is similar in magnitude to that of sex steroid deficiency in both women and men (6,7). Therefore, agonists of aromatase expression or activity in the bone would be a new therapeutic means for preventing and treating osteoporosis.
Mechanical stimulation is a primary determinant of bone growth and remodeling, through generating the shear stress that stimulates osteoblasts and osteocytes and enhances their anabolic activity. In mechanically stimulated osteoblasts the NO/cGMP/ PKG signaling pathway activates Erk-1/2 and a proliferative response through the recruitment of PKGII, Src, and SHP-1/2 into an integrin b3-containing mechanosome membrane complex (8). NO, which is increased by estrogen exposure, also mediates estrogen-stimulated human and rodent osteoblast proliferation and differentiation (9). Src kinase has also been found to regulate aromatase activity by directly phosphorylating aromatase or indirectly regulating aromatase expression through the MAPK pathway (10). In osteocyte, 17b-estradiol is found to prevent the bone loss by increasing its survival through NO/ cGMP-mediated stimulation of Akt and Akt-and PKGdependent phosphorylation of the pro-apoptotic Bcl-2 protein BAD (11). However, the role of cGMP in the regulation of estrogen biosynthesis in osteoblasts is still not well understood. An inhibitor of PDE5, which is responsible for cGMP degradation, has been found to increase aromatase expression and estrogen biosynthesis in human adipocytes and ovarian granulosa cells (12,13). Thus, it will be of interest to investigate the crosstalk of cGMP signaling on aromatase expression in osteoblastic cells, which may provide new insights into the underlying mechanism of cGMP-mediated signaling in osteoblast proliferation and differentiation.
It is unclear whether natural medicinal plants exert their antiosteoporotic effects by modulating estrogen biosynthesis in the bone. Previously we found that icariin from Epimedium brevicornum, a widely used antiosteoporotic medicinal plant, promotes the production of estrogen in human ovarian granulosa cells and osteoblastic cells, with an underlying mechanism that remains unclear (14). Icariin and its analogs are found to be the inhibitors of PDE5 (15). We showed that the PDE5 inhibitors, sildenafil and icariin analogs, promote aromatase expression in human ovarian granulosalike KGN cells by activating the cAMP/CREB pathway (13). Thus, further investigating the effect and mechanism of icariin analogs on aromatase regulation in bone tissue will be important for developing new therapeutic means to prevent and treat osteoporosis.

Chemicals and Reagent
The 18 icariin analogs were synthesized and identified as described previously (16, 17, Figure S1). The compounds were dissolved in DMSO (Sigma-Aldrich, Shanghai, China) and stored at -20°C. Testosterone was purchased from Sigma-Aldrich. The magnetic particle-based 17b-estradiol enzyme-linked immunosorbent assay (ELISA) kit was purchased from Bio-Ekon Biotechnology (Beijing, China

Cell-Based Estrogen Biosynthesis Assay
The assay was conducted as described previously (19). The UMR106 cells or MC3T3-E1 cells were seeded overnight in 24well plates. After that the medium was replaced with serum-free medium, and the cells were pretreated for 24 h with the test chemicals. Testosterone (10 nM) was then added to each well, the cells were incubated for an additional 48 h. The magnetic particle-based ELISA kit was used to quantify the 17b-estradiol in the culture medium according to the manufacturer's instructions (Bio-Ekon Biotechnology). The results were normalized to the total cellular protein content, and expressed as percentages of the control. The BCA protein assay kit was used for protein determination (Bestbio, Shanghai, China).

Western Blotting
Immunoblotting was performed as described (19). Cells cultured were harvested in RIPA buffer supplemented with a protease inhibitor cocktail (Sigma). Proteins lysate was loaded and separated on a sodium dodecyl sulfate-polyacrylamide gel electrophoresis. After that, the proteins were blotted onto nitrocellulose membranes and then incubated with each specific antibody, then enhanced chemiluminescence detection (Amersham Biosciences, Piscataway, NJ, USA).

Measurement of Intracellular cGMP Level
UMR106 cells seeded in 6-well plates overnight were treated with 14 and sildenafil for the indicated time. The pre-cooled PBS buffer (120-150 mL) was added in 1×10 6 cells to keep the cells suspended. The cells were lysed with the repeated freeze-thaw process. After centrifugation for 10 min at 1500 × g at 2-8°C, the supernatants were collected to carry out the assay. The cGMP concentration was determined with a commercial cGMP enzyme immunoassay kit (Elabscience, Wuhan, China). Thereafter, the results were measured with Thermo Scientific Verioskan Flash Multimode Reader at a wavelength of 450 nm ± 2 nm.
Alkaline Phosphatase (ALP) Activity Assay ALP activity was performed as described (20). UMR-106 cells were seeded in serum-free medium in a 24-well plate overnight and treated with the test compounds for 48 h. After that a kit using para-nitrophenyl phosphate as substrate was used to assay the cell lysate. The OD value was measured at 405 nm with Thermo Scientific Verioskan Flash Multimode Reader. The results were expressed as percentages of the control and normalized on a protein basis.

SHP2 Activity Assay
SHP2 was immunoprecipitated and its activity is assayed as described previously (21). UMR106 cells were lysed in RIPA buffer that contained a complete protease inhibitor cocktail after treatment with 14 and SHP2 inhibitor (NSC-87877) for 30 min. The lysates were incubated on ice for 10 min and centrifuged at 20,000 × g for 15 min at 4°C. SHP2 antibody was incubated with cleared lysates overnight at 4°C with agitation, followed by the incubation with the Protein A/G agarose (Santa Cruz Biotech, TX, USA). The immunoprecipitates were resuspended gently in reaction buffer (100 mL) and transferred to a 96-well plate. The DiFMUP was used as the substrate of SHP2 to measure its activity with a plate reader (Thermo Scientific Varioskan ® Flash).

Cellular Thermal Shift Assay
Cellular thermal shift assay was conducted as described previously (21). HEK293T cells were collected in PBS supplemented with protease inhibitor cocktail. The freezethawed cell lysates were centrifuged at 20,000 g for 20 min at 4°C, diluted with PBS and divided into two aliquots; one aliquot was treated with DMSO while the other was treated with 14 (100 mM). For temperature response experiments, 50 µL of lysate was transferred to PCR tubes and heated for 3 min to various temperatures. After that the cell lysates were centrifuged at 20,000 g for 20 min at 4°C to separate the soluble fractions from the precipitates. The supernatants were dissolved in loading buffer and analyzed by western blotting. The dose effect of 14 on the stability of PDE5A or vinculin was evaluated in the same manner.

Measuring Recombinant Expressed PDE5 Activity
The activity of recombinant expressed PDE5 (Enzo Biochem, Madison, USA) was evaluated using the PDE-Glo ™ Phosphodiesterase assay (Promega Corporation, Madison, WI, USA). Aliquots of PDE-Glo ™ reaction buffer containing appropriate amounts of purified human recombinants PDE5A were added to a 96-well plate. After the addition of diluted compounds to each well, cGMP ™ solution was added to initiate the reaction. After an appropriate incubation, Kinase-Glo ® reagent was pipetted into each well and 10 min later luminescence was measured using a plate reader (Thermo Scientific Varioskan ® Flash).

Molecular Docking
The crystal structure of PDE5 [PDB code: 2H42] was obtained from the Protein Data Bank. During the process, all water molecules were removed, and hydrogen atoms were added to the protein molecule. Autodock 4 was used to predict the interactions between compounds and protein structure of PDE5 according to the binding energy with the default setting.

Cell Viability Assay
UMR 106 cells were plated at 0.5 × 10 4 cells/well in 96-well plates with 100 mL medium. The different concentrations of 13, 14, and 15 were used to treat the cultured cells for 24 h. After that the medium was added with 10 mL of the Alamar blue reagent and incubated for another 2-4 h with the measurement of the relative fluorescence intensity in each well.

Statistical Analysis
Statistical analysis was analyzed by GraphPad Prism 6 (GraphPad, La Jolla, CA, USA). The results are expressed as mean ± standard error of the mean (S.E.M.) of three independent experiments with individual values. Data were compared by oneway ANOVA followed by Dunnett's post hoc test. A p-value of less than 0.05 was considered to indicate a significant difference relative to the control.

Effect of Icariin Analogs on Estrogen Biosynthesis
To search for small molecules that modulate estrogen biosynthesis, we examined the effects of icariin analogs (16, 17, Figure S1) and their effects on estrogen biosynthesis in the rat osteoblast-like cell line, UMR106. The chemical structure of the icariin ( Figure 1A, compound 2) and its analogs are presented in Figure 1A. As shown in Figure 1B, testosterone supplementation significantly increased 17bestradiol production in UMR106 cells, which was further enhanced by dexamethasone treatment, thereby aligning with previous reports (22,23). To examine the effect of icariin and its analogs on 17b-estradiol biosynthesis, UMR106 cells were incubated for 24 h with different concentrations of the test compounds followed by a further 24-h incubation with testosterone. Among the 18 compounds, 11 could increase the production of 17b-estradiol; these 9 compounds were the (18). Compounds 13, 14, and 15 promoted 17b-estradiol biosynthesis in a concentrationdependent manner, with EC 50 values of 1.53, 3.45, and 10.57 µM, respectively ( Figure 1C). They also had no effect on the viability of UMR106 cells ( Figure 1D). These results indicate that the icariin analogs, such as compounds 13, 14, and 15, could potently promote estrogen biosynthesis in rat osteoblast-like cells.

Effect of Icariin Analogs on Aromatase Expression
To determine whether compounds 14 and 15 promoted 17bestradiol biosynthesis by affecting aromatase, we examined the mRNA and protein levels of aromatase in UMR106 cells treated with the selected compounds. Compounds 14 and 15 significantly increased aromatase transcript levels in a concentration-dependent manner ( Figure 2A). 14 increased 58% of the aromatase mRNA levels at 10 mM while 15 increased 60% of the aromatase mRNA levels at 25 mM compared in the DMSO-treated control cells. 14 and 15 also significantly increased the bone specific aromatase promoter I.4 transcript in a concentration-dependent manner (Figure 2A, Figure S2). 14 and 15 also significantly increased aromatase protein expression in UMR106 and MC3T3-E1 cells in a concentration-dependent manner ( Figures 2B-D). Furthermore, in aromatase-overexpressing HEK293A cells, letrozole, a specific inhibitor of aromatase enzymatic activity, significantly inhibited 17b-estradiol biosynthesis; however, 14 had no apparent effect on 17b-estradiol production compared to the DMSO-treated cells ( Figure 2E), western blot also showed 14 had no effect on aromatase protein expression ( Figure S3). These results excluding the probability that 14 directly modulates the enzymatic activity of aromatase protein. Actinomycin D, an RNA polymerase inhibitor, significantly suppressed 14-induced aromatase mRNA transcription ( Figure S4). These results indicate that 14 and 15 promoted estrogen biosynthesis by affecting aromatase at the transcriptional level.

Inhibitory Effect of 14 on PDE5A Activity
Previously, the icariin analogs were found to be potent inhibitors of PDE5 (15); thus, we opted to use recombinant-expressed PDE5A to examine whether 14 inhibits PDE5 activity. As shown in Figure 3A and Figure S5, 14 significantly inhibited PDE5 activity in a concentration-dependent manner with the IC 50 value of 9.914 mM, a finding similar to that obtained with the specific PDE5 inhibitor, sildenafil. 14 had no effect on the PDE5 expression in both UMR106 cells and MC3T3-E1 cells ( Figure S6). Thereafter, we proceeded to perform a cellular thermal shift assay to examine whether 14 directly interacts with PDE5A in cells (24). Compared to the DMSO control, the presence of 14 markedly increased the accumulation of PDE5A in the soluble fraction at the temperatures examined ( Figure 3B). We also tested the concentration-response of 14 on PDE5A stability at increased temperatures. An increase in 14 concentration resulted in a marked increase in PDE5A accumulation ( Figure 3C). We then examined the effect of 14 on the intracellular cGMP level. As shown in Figure 3D, similar to sildenafil, 14 significantly stimulated the intracellular cGMP level in UMR106 cells. These findings suggest that 14 directly interacts with PDE5 and inhibits its activity in cells.
Computer docking analysis was conducted to assess the binding sites in PDE5. Based on our results, 14 fitted well within the active site of PDE5 ( Figure 3E). The formation of hydrogen-bond (H-bond) and the hydrophobic interactions between 14 and PDE5 were evaluated. Two polar hydrogens in 14 are involved in its H-bonding with the amino acid Gln817, Ala767, Leu765, Tyr612 and His613, of PDE5 with a high glide energy of -11 kcal/mol -1 . 14 also formed hydrophobic interactions with the residues Ile768, Phe820, Met816, Leu804, Val782, Phe786 and Asn661 ( Figures 3F, G), which may contribute to its inhibition of PDE5. The binding sites of sildenafil (site A) and 14 (site B) were very close with binding sites near the zinc ions and magnesium ions. Sildenafil formed a hydrogen bond with the amino acid Gln817 (Figure 3H), thereby aligning with previous reports (25). These results suggest that 14 might be subjected to a nucleophilic attack in the PDE5 to inhibit its activity.

Effect of 14 on Osteoblastic Cell Differentiation
To demonstrate that increased estrogen biosynthesis by 14 may promote osteoblastic cell differentiation, we examined the mRNA expression of osteoblastic cell differentiation markers in UMR106 cells. 17b-Estradiol significantly increased mRNA levels of osteocalcin (OCN), osterix (Osx), alkaline phosphatase (ALP), and Runt-related transcription factor 2 (Runx2) ( Figures  4A-D), aligning with the findings of previous reports (26,27). Similar to sildenafil, 14 significantly increased the mRNA levels of OCN, Osx, ALP, and Runx2 ( Figures 4A-D). The bone formation/resorption balance can be observed from the ratio of OPG/RANKL expression, which is stimulated by 17b-estradiol (28). Similar to 17b-estradiol, both 14 and sildenafil increased the ratio of OPG/RANKL ( Figure 4E). Compared with the control, calcium deposition was increased after treatment with 14 in a dose dependent manner ( Figure S7). While 14 had no effect on BMP2 protein expression ( Figure S8). These results indicate that 14 can promote osteoblast formation and differentiation.

Effect of 14 on cGMP/PKG/Src/ERK Signaling
The stimulation of intracellular cGMP by the 14-induced PDE5 inhibition may activate PKG to increase aromatase expression. Therefore, we examined the role of PKG in the regulation of estrogen biosynthesis. PKG inhibition by the PKG inhibitor, KT5823 or Rp-8-pCPT-cGMPS, abolished the stimulatory effect of 14 and sildenafil on 17b-estradiol production ( Figure 5A and Figure S9). PKG inhibition also abolished the stimulatory effect of 14 and sildenafil on ALP activity, a well-known marker of osteoblast differentiation ( Figure 5B). These results suggest that PKG mediates the promotive effect of 14 on estrogen biosynthesis and differentiation in osteoblastic cells. As the cGMP-PKG signaling pathway activates Src and ERK in mechanically stimulated osteoblasts (8) activation as it changes the protein to an active conformation and enables autophosphorylation of Tyr 418 in the kinase domain activation loop (29). As shown in Figure 5C, 14 significantly promoted the Src phosphorylation on Tyr 529 and decreased the phosphorylation on Tyr 418 , indicating Src activation. Similarly, 14 also promoted phosphorylation of ERK in both UMR106 and MC3T3-E1 cells. We treated the cells with a ERK inhibitor and found that it completely abolished the stimulatory effect of 14 on aromatase expression compared with 14 alone (Figure 5D), indicating that 14 enhances activation of ERK pathway signaling, thereby supporting the finding that inhibition of PKG abolished the stimulatory effect of 14 on estrogen biosynthesis ( Figure 5A). These results suggest that 14 promotes osteoblast differentiation by activating the PKG/Src/ERK pathway.

Effect of 14 on SHP2 Activation
In osteoblasts, cGMP/PKG-induced Src activation is mediated by SHP-2 (8). Compared to the DMSO control, 14 significantly promoted SHP2 activity in the treated cells in a concentrationdependent manner, similar to sildenafil ( Figure 6A). SHP2 activity was also stimulated by 14 in a time-dependent manner ( Figure 6B). To further confirm the role of SHP2 in the 14enhanced Src/ERK pathway signaling, we examined the effect of 14 alone or in combination with the SHP2 inhibitor (NSC87877). Based on our findings, SHP2 inhibitor treatment completely eliminated the promotive effect of 14 on the phosphorylation of Src-pTy 418 and phosphorylation of ERK ( Figure 6C). Furthermore, SHP2 inhibitor treatment significantly decreased the stimulatory effect of 14 on aromatase expression in both UMR106 and MC3T3-E1 cells ( Figure 6D). These results indicate that 14 stimulates aromatase expression by activating SHP2.

DISCUSSION
Icariin is the most abundant bioactive flavonoid contained in E. brevicornum (30,31). Both icariin and E. brevicornum exhibit antiosteoporotic effects in vitro and in vivo by stimulating osteoblast proliferation; these findings support the wide use of E. brevicornum in many Traditional Chinese Medicine formulas for the treat bone fracture and prevent osteoporosis (32)(33)(34). Of the 18 icariin analogs examined in the present study, 11 could increase estrogen biosynthesis in rat osteoblast-like UMR106 cells. This was consistent with a previous report where structure-activity relationship analysis suggested that prenylation at the C-8 and C-6 position was essential for promoting the differentiation of primary osteoblasts (35). In this study, we found that the prenyl group at the C-8 position was more potent than the prenyl group at the C-6 position for promoting estrogen biosynthesis. In the adipose tissue and bone, aromatase expression is stimulated primarily by class I cytokines through promoter I.4 (5). Consistently, we found that 14 potently promoted estrogen biosynthesis by increasing promoter I.4-driven aromatase mRNA and protein expression in osteoblastic UMR-106 and MC3T3-1 cells. Previously, we found that 2phenylbenzo[b]furans might enhance estrogen biosynthesis via direct allosteric regulation of aromatase enzymatic activity (18,19). However, in this study, we found that 14 had no effect on Currently, several PDE5 inhibitors have been approved by the FDA for the treatment of erectile dysfunction and pulmonary arterial hypertension (36). PDE5 plays a key role in cGMP signaling; however, its role in estrogen biosynthesis in the bone has been rarely evaluated. In the present study, we found that the icariin analog, 14, a validated PDE5 inhibitor with IC 50 9.914 ± 0.3325 mM, promoted estrogen biosynthesis in UMR 106 and MC3T3-E1 cells by enhancing aromatase expression in a similar manner to icariin (14). Earlier studies also revealed the importance of Gln817, Tyr612, Phe786, and Ala783 amino acid in PDE5-inhibitor interaction (25). Here, we found that both sildenafil and compound 14 could bind to these amino acids in PDE5. Additionally, 14 also differently binds to other amino acids in PDE5. Thus, further investigation is required to determine whether these amino acids also regulate PDE5 activity. PDE5 inhibitors, such as tadalafil and sildenafil, have been found to stimulate aromatase expression in human adipocytes (12), further supporting the role of PDE5 in the regulation of promoter I.4-driven aromatase expression. cGMP plays a key role in osteoblast differentiation by activating PKG (37). Icariin analogs, which inhibits PDE5 activity, was found to promote osteoblast differentiation and exhibit antiosteoporotic effect in vivo (32)(33)(34)38), thereby aligning with our finding that 14 increased the expression of osteoblast differentiation markers. PDE5 inhibitors were also reported to exert beneficial effects on ovariectomy or glucocorticoid-induced osteoporosis in rats (39,40). Furthermore, PDE5 inhibition was found to reduce bone mass by suppressing canonical Wnt signaling, indicating that long-term treatment with PDE5 inhibitors at high dosage may cause bone catabolism (41). Therefore, the role of PDE5 in the regulation of bone homeostasis should be further investigated to develop PDE5 inhibitors as new antiosteoporotic therapeutics. Recently it is reported that PDE5 inhibitors could enhance osteoblastic bone formation by targeting PDE5A and reverse osteopenia in ovariectomy mice by an osteogenic mechanism (42,43). Therefore, our findings that PDE5 inhibitors promote estrogen biosynthesis provide new insights for the clinical benefits of PDE5 inhibitors in the treatment of osteoporosis. Moreover, the prenyl group contributes to higher osteogenic activity than do flavonoids possibly by modulating estrogen receptors (44,45). Thus, whether 14 exhibits its osteogenic activity by a dual-functional modulator of PDE5 and estrogen receptor needs further investigation.
Mechanical stimulation, such as exercise, can strengthen bones and reduce the risk of fractures (46). Compressive forces generated by weight bearing and locomotion induce small bone deformations and increase interstitial fluid flow, thereby promoting anabolic responses in osteoblasts through different signal transduction pathways, including calcium channels, Raf-MEK-ERK cascade, and nitric oxide (NO) (47). Recently, the NO-cGMP-PKG pathway was reported to regulate osteoblast proliferation and differentiation through the formation of an Srccontaining mechanosome (8). Consistent with this result, we found that 14 or sildenafil increased intracellular cGMP level and activated ERK and Src via PKG in osteoblasts. More interestingly, we found that PKG inhibition suppressed 14 or sildenafil-induced estrogen production in osteoblasts, which FIGURE 7 | Proposed model of the role of PDE5 in the regulation of aromatase in osteoblasts. 14 inhibits the activity of PDE5, thereby stimulating the intracellular cGMP level, which causes PKG activation. PKG activation by 14 stimulated the activity of SHP2, which subsequently activated Src and ERK signaling and increased ERK-dependent gene expression, including that of aromatase, in osteoblasts. could be justified by our finding that 14 or sildenafil stimulated the activity of SHP2 that was directly phosphorylated by PKG and was required to activate Src (8). ERK has been implied to promote osteoblast differentiation by regulating ALP activity, integrin synthesis, focal adhesion kinase, Runx2 phosphorylation, and transcriptional activity (48)(49)(50). In this study, we found that the inhibition of ERK also attenuated the stimulatory effect of 14 on aromatase expression, providing new insight into mechanical stimulation in osteoblast differentiation. ERK could phosphorylate the glucocorticoid receptor and modulate its transcriptional activity (51). Therefore, it will be of interest to further investigate whether ERK modulates the glucocorticoid receptor or other transcriptional factors to stimulate promoter I.4-driven aromatase expression in osteoblasts. 17b-Estradiol can rapidly enhance aromatase enzymatic activity by increasing aromatase protein phosphorylation in breast cancer cell lines, which is mediated by Src (10). Thus, 14-induced Src activation may also stimulate aromatase enzymatic activity to promote estrogen production, which should be further investigated. SHP-2 regulates cell survival and proliferation by the activation of the RAS-ERK signaling pathway (52). SHP2 is found to physically interact with the estrogen receptor, which is necessary for the synergistic and persistent activation of ERK by leptin and estrogen (53). cGMP/PKG-mediated SHP2 activation may also regulate the function of the estrogen receptor to exert its anabolic effect in osteoblasts. Therefore, a further investigation to determine whether mechanical stimulation also modulates local estrogen biosynthesis or estrogen receptor to exert its antiosteoporotic effect is warranted.

CONCLUSIONS
In summary, we found that the prenylated flavonoid 14 promotes osteoblast differentiation by activating the cGMP/ PKG/SHP2/Src/ERK cascade via PDE5 inhibition, thereby leading to the localized production of estrogen by stimulating aromatase expression (Figure 7). These data not only provide new insights into the role of estrogen biosynthesis in mechanical stimulation-induced osteoblast differentiation but also support the use of PDE5-inhibiting drugs to mimic the anabolic effects of mechanical bone stimulation in the treatment of osteoporosis.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

AUTHOR CONTRIBUTIONS
WW, D-yC, and FW wrote the manuscript. WW, KW, D-yC, X-kS, and Z-yZ conducted the biological experiments. Z-hZ conducted the molecular docking analysis. FL, Q-gM, and CW synthesized the compounds. AS, G-lZ, and FW supervised the study, designed the experiments, and revised the manuscript. All authors contributed to the article and approved the submitted version.