Effect of FGF/FGFR Signal in Fluid Shear Stress and Estrogen Regulating Bone Metabolism

Estrogen and fluid shear stress (FSS) play an important role in bone metabolism, displaying a synergistic effect. Fibroblast growth factors (FGFs) are also of great importance in bone development before and after birth. FGFs regulate downstream signaling pathway by activating FGF receptors (FGFRs) to control the expression of bone tissue cells. This paper conducts a brief discussion of the effect of FGF/FGFR as well as fluid shear stress and estrogen in bone metabolism regulation.


Introduction
Bone is a kind of tissue that is always in constant dynamic development. Bone metabolism depends upon balanced osteoblast osteogenesis ability and osteoclast bone resorption ability, which are subject to the influence of many factors, such as cytokines, mechanical load and hormone [1]. Estrogen can enhance osteogenesis [2], because it has the dual effect of hindering osteoclast bone resorption and promoting osteoblast osteogenesis [3,4]. Mechanical stress serves as another important factor influencing bone remodeling, for suitable mechanical stress can promote the growth of bone tissues. Cells are more sensitive to FSS than to pressure and stretch stress [5]. Flowing fluid forms FSS on the surface of bone tissue cells, such as osteocytes and osteoblasts. At present, the mechanism of FSS being turned into biochemical signal and introduced into nucleus through the surface of cytomembrane is regarded as stress transduction mechanism [6]. FGFs fall into the family of polypeptides, which control several crucial cellular processes. In bones, FGF/FGFR signals are important regulators for prenatal and postnatal bone development [7,8]. This paper mainly discusses the influence of the separate and joint effect of FGF/FGFR signals, estrogen and FSS upon bone metabolism as well as the effect of FGF/FGFR signals in FSS and estrogen regulating bone metabolism. Starting from the intracellular signal paths induced by FGF/FGFR signals, estrogen and FSS, this paper explores their interaction at molecular level, presenting a new approach for following explorations into bone regeneration in clinics and research.

Introduction of the Structures of FGF/FGFR, Estrogen and Estrogen Receptor
There are two types of FGFs, namely secretion type and intracellular type. Secretion type FGFs can be further divided into canonical type and endocrine type, in which the former contains five subfamilies, namely FGF1, FGF4, FGF7, FGF8 and FGF9, while the latter contains three members, namely FGF15/19, FGF21 and FGF23 (Table 1).
Secretion type FGFs mainly combine with FGFR as signal protein to transmit signals and activate multiple cellular pathways. Intracellular FGF is a kind of non-signal protein, mainly used as voltage-gated sodium channel and the cofactor of other molecules [9][10][11] (Figure 1).   [9,13,14]. FGF and FGFR binding also needs the participation of cofactors to enhance the bioactivity of FGF. HS and HSPG are currently regarded as the potent cofactors of canonical type FGF signaling pathway [15][16][17]. The cofactor of endocrine type FGFs is Klotho family [19,20]).
FGFR is a kind of tyrosinekinase receptor, and shares similar structure with other types of tyrosine kinase receptors, mainly consisting of ligand-binding domain, transmembrane domain and cytoplasmic domain [12][13][14]. FGFR family includes FGFR1, FGFR2, FGFR3 and FGFR4. Fgfr1-Fgfr3 generates two additional major splice variants of immunoglobulin-like domain III, referred to as IIIb and IIIc. Data is derived from receptor activation assays using BaF3 cells, L6 myoblasts or HEK293 cells transfected with individual splice variants of FGFRs or by direct binding studies [9] (Table 2) [9]).
There are three important processes for FGFR activation, i.e., ligand binding, receptor ligand dimerization and intracellular phosphorylation of receptors [15]. The binding of FGF and FGFR also needs the participation of cofactors to enhance the bioactivity of FGF. Heparan sulfate (HS) and heparan sulfate proteoglycan (HSPG) are currently regarded as potent cofactors for canonical FGF signaling pathways [9,[16][17][18]. The cofactor of endocrine FGFs is a kind of Klotho protein used to form FGF-FGFR-Klotho trimer [19,20] (Figure 1).
Estrogen is an important type of endogenous hormones in human physiology, including estrone (E), 17-estrodiol (E1), estriol (E3), etc. They are mainly produced in ovary and testis and scattered in all cells [21]. Estrogen features extremely important biological functions, including its significant regulatory effect on the metabolism of multiple organs like reproductive system, bone and cartilage system, and cardiocerebral vascular system [21,22].
There are two subtypes of estrogen receptor (ER), namely ERα and ERβ, both of which fall into nuclear receptor hormone superfamily receptor and share common structural framework, i.e., N terminal (I region), DNA binding region (II region) and C terminal ligand binding region (III region) [23][24][25][26][27]. Canonical ER pathway is the binding between estrogen and ER, inducing comformational changes of receptors and ultimately leading to homodimerization or heterodimerization (ERα-ERβ, ERβ-ERβ or ERα-ERβ). The homodimerization or heterodimerization generated are further to be transferred to nucleus to make receptor and chromatin in high affinity combined, thereby regulating the transcription of target genes [28][29][30] ( Figure 2). Figure 2: Estrogen-ER structure. (There are two subtypes of ER, namely ERα and ERβ, both of which fall into nuclear receptor hormone superfamily receptor and share common structural framework, i.e., I N terminal, II DNA binding region and III C terminal ligand binding region [23][24][25]. In I region, there is one active functional region-AF-1 of ER that does not depend upon ligand, participating in regulating the binding between ER and ligand and response gene transcription and containing a structural domain that could be recognized by estrogen agonist [26]. In III region, there is another active functional region AF-2, with which the ligand can also realize binding [27]).
There is ER expression in many cells, mainly distributed in uterus, ovary, mammary gland, bone tissue, urinary organs, neurocyte and immunocyte, liver and fat, etc. [23]. ER activation is mainly related to phosphorylation [31], in which there are many types of enzyme involved, including tyrosine kinase (TK), casein kinase (CK), etc. [32]. In addition, other signal systems like commonly-seen cell phosphorylases (PKA, PKC), intracellular signals like polypeptide growth factor, kinase, neurotransmitter and regulatory factor can also realize ER phosphorylation [31]. ER serves as an operation node for estrogen to play its role in biological activity inside cells, mainly displayed as cell growth, differentiation, aging and apoptosis [33,34]. There is expression of ERα and ERβ in bone tissues yet with difference in distribution. In cortical bone, ERα is the main subtype expressed, while in bone trabecula it is ERβ [35]. The activity and number of ER change with the level of estrogen, while with the increase of age or after menopause among females, the expression of ER drops [36].

Estrogen and FSS Regulating Bone Metabolism
Estrogen is one type of steroids with wide biological activity that balance the osteoblast osteogenesis ability and osteoclast bone resorption ability [22]. Lack of estrogen is prone to cause postmenopausal osteoporosis (OP) [36,37]. In 1941, Albright first adopted estrogen substitution therapy to treat OP [38]. Lack of estrogen may promote the resorption of experimental periodontitis alveolar bone. Estrogen substitution therapy can effectively prevent decreased alveolar bone height and lower the risk of agomphiasis [39]. Research has demonstrated that the icariin separated from epimedium can prevent loss of bone mass, for it can enhance the activity of ERK and JNK in MC3T3-E1 osteoblast yet without obvious influence upon P38. After estrogen receptor antagonist ICI182780 is applied, the division and mineralization abilities of MC3T3-E1 osteoblasts significantly drop, and the phosphorylation of ERK and JNK is also hindered. Therefore, it is deducted that estrogen-ER-ERK/JNK pathway can promote the division and mineralization of osteoblasts and ultimately osteogenesis [40]. Research has proved it is by estrogen-ER-ERK-Target pathway that estrogen promotes the autophagy of osteoblasts to hinder its apoptosis and ultimately facilitate osteogenesis [41].
FSS stimulated osteoblast proliferation and differentiated stress transduction mechanism involve multiple cell-dependent and independent signaling pathways, such as intracellular and extracellular Ca current [42,43], NOS-NO [44], Cox-PGI2-PGE2, PTX-Gi/o, etc. [6]. Many of them are involved activating one or more members of the MAPK family, including ERK1/2, JNK, P38 and/or interaction with integrin signaling pathway [6]. Integrin, an ECM-cell surface receptor, is connected with cytoskeleton and extracellular matrix, serving as an important mechanical transducer for FSS stimulation and being able to act as mechanoreceptors [ [50] also proved that FSS could uplift the phosphorylation level of ERK, JNK and P38. Moreover, much research found out that mechanical stimulation could enhance the expression of insulin-like growth factor (IGF), a reaction factor regulating osteogenesis with mechanical load [51,52]. Kahlert et al. [53] further proposed that IGF-ER signaling pathway needed the participation of ERK signal; ERα signal was at the upper stream and downstream of ERK signal and ER was activated through ligand binding with ERK1/2. In addition, signaling pathways like Wnt and Smad also played a crucial role in FSS induced osteogenesis [50,54]. There was a series of research highlighting the importance of Lrp5/Wnt/b-catenin pathway in loading osteogenesis reaction [55,56], the process of which needs ER [57][58][59]. Lrp5/Wnt/b-catenin pathway and IGF-ER signal were crosslinked through GSK-3b [44,60].
Many studies have demonstrated the importance of estrogen receptor in GSS-induced osteogenesis [57,[61][62][63][64][65]. The increased number of ER could enhance stress or estrogen-related signaling pathway transduction, and all these responses could be prevented by ER antaognists ICI 182, 780, hence proving that these pathways were ER mediated [61]. Research by Lee et al. [62] showed that when mechanical stress was imposed upon the ulna of mice after ERα knockout, the formation amount of was only 1/3 of the newborn mice at the same litter where ERα existed; and the osteoblast extracted from mice after ERα knockout had a weak response to mechanical stress [57]. Damien et al. [63] pointed out that among males and females, mechanical stress could promote osteoblast proliferation through ER. If ROS 17/2.8 cell lineage is exposed to estrogen for 5 min or monocyclic dynamic strain force was imposed for 5 min, it can be observed that both ERα phosphorylation and ERK1 activation increase, which however can be hindered by MAPK inhibitor U0126 and protein kinase A (PKA) inhibitor PKI. That is to say, osteocyte can Joldersma et al. [66] found that estrogen could enhance the generation of E2 (PGE2) in postmenopausal non-osteoporosis female osteoblasts induced by FSS. Bakker et al. [67] demonstrated the adjection of estrogen and FSS against the PGE2 and NO in osteocytes among females with osteoporosis (age 62-90). Yeh et al. [68] proved that estrogen could enhance FSS, strengthening the phosphorylation of human osteogenesis sample cells (MG63 cells), primary osteoblasts ERK and P38 and increasing the expression of c-fos and Cox-2. This process was realized by estrogen receptor increasing β1 integrin. These results suggest that estrogen may influence the response function of osteocytes to FSS and there is a synergistic effect between estrogen and FSS in osteogenesis. Li et al.
[50] discovered that estrogen and FSS both had the effect of promoting osteoblast proliferation differentiation and their synergistic effect superseded the independent effect of individual, so they had the synergistic effect ( Figure 3).
Recent data showed that FGF2 activated PKC could activate the activity of Runx2 in mesenchymal stem cells and osteoblasts [79,80]. ERK1/2 signals induced by FGF2 were found to promote the proliferation of osteoblast precursor cells and induce FGFR2 induced osteoblast differentiation [70,71]. PLC/PKC activation was also involved in FGF2 or FGFR2 activated osteoblast differentiation [79]. PI3K is another FGF mediated signal cascade reaction controlled by osteoblast cytopoiesis. For example, FGF or FGF4 activated PI3K-Akt pathway promoted the proliferation of osteoblast precursor cells [81]. FGF18-FGFR1/FGFR4 mediated signaling pathways could regulate the differentiation of osteogenesis [82,83]. Delayed ossification has been described in FGF18-deficient mice [84], suggesting that FGF18 controls osteogenesis. Recent data indicated that FGF18 played an important role in osteogenesis of mice [85]. The mice mesenchymal stem cells osteoblast differentiation mediated by FGF18 and induced by FGFR1/FGFR2 was also involved in PI3K and ERK1/2 signals [83] and the Akt activity increase caused by FGFR2 mutation could promote osteocyte differentiation [73,84,85]. FGF23/FGFR1 signaling pathways were related to bone mass and bone mineralization [86,87].
FGFRs mutation could cause two types of skeleton diseases, namely skeletal dysplasia and craniosynostosis. FGFR1 expression went throughout the bone development and postnatal bone reconstruction, fracture healing and osteogenesis in embryo. Missense mutation of FGFR1 (P252R) could cause Pfeiffer Syndrome [88]. Mice with important fragmentary mutation of FGFR2 scarcity or FGFR2 displayed bone hyperplasia drop and bone mineral density drop. And the absence of FGFR2 in mice could lead to dysfunction of mature osteoblasts [89,90]. Research showed that activated FGFR2 mutation could promote osteoblast differentiation and ultimately cause human craniocerebral disease through increasing Runx2 expression and osteogenic marker genes [91]. Through activated FGFR2 mutation, activated ERK-MAPK could increase Runx2 transcription activity and osteogenic marker gene expression [81]. Therefore, activated FGFR2 mutation could promote osteoblast differentiation of mesenchymal stem cells [80]. FGFR3 failure and FGFR3 activation both could lead to osteoporosis [92,93]. In mouse model with achondroplasia, chondrocyte specific FGFR3 activation was related to the increase of osteogenesis around cartilage, indicating partial osteogenesis increase [94], which might be indirectly mediated through BMP.
Wnt-dependent BMP signals could guide intracellular β-catenin distribution, featuring an important positive feedback regulation effect upon β-catenin [96]. FGF signal and BMP signal, apart from the Wnt/ β-catenin mediated indirect relations, also shared certain direct relations. FGF2 could lower the resistance of BMP to noggin and thereby promote BMP signal transduction in skull development [102]. In addition, FGF18 hindered the expression of noggin, hence increasing the BMP signals in cytoskeletons [103]. Endogenous FGF signals up-regulated BMP2 in osteoblasts and skull, and uplifted BMP2 induced ectopia osteogenesis in mice [104][105][106]. These studies highlighted the positive interaction of FGF and BMP signals in controlling osteogenesis.

Interaction of FGF/FGFR Signals with FSS and Estrogen Regulating Bone Metabolism
Li et al.
[50] after imposing estrogen, FSS or two stimulation factors upon MC3T3-E1 cells, found out that there was obvious change in FGFR1 expression, suggesting that FGFR1 played an important role in regulating the process of estrogen and FSS promoting osteocyte proliferation and differentiation. Further experiments, after intervention of estrogen and FSS upon MC3T3-E1 osteoblasts, saw significant increase of FGFR1 expression. Applying FGFR1 inhibitor PD166866+estrogen+FSS and comparing with control group, the division and differentiation abilities of osteoblasts were hindered, indicating that FGFR1 played a core role in osteoblast response to estrogen and FSS. When estrogen and FSS independently or jointly worked on osteoblasts, the phosphorylation level of ERK, JNK and P38 significantly improved. After applying FGFR1 inhibitor PD166866, the phosphorylation level of ERK was hindered, i.e., FGFR1-ERK signals played a key role in osteogenesis induced by estrogen and FSS. This process was the result of the intracellular signal cascading of FGFR, estrogen and FSS. It remains to be further studied whether estrogen and FSS directly work on FGFR.
It could be obtained from the above discussions of intracellular signaling pathways induced by FGF/FGFR signals, estrogen and FSS that MAPKs-ERK1/2 signaling pathway, Wnt/β-catenin signaling pathway, PI3K-AKT signaling pathway and TGFβ-Smads-BMP signaling pathway were cross-linked pathway of the three (Figure 3).
FGF/FGFR signal scan could promote Runx2 activity, while Runx2 could induce FGFR mutation, leading to increased expression of Sox2. The interaction between Sox2 and β-catenin hindered the activity of Wnt responsive plasmids and leaded to the downregulated expression of Wnt target genes [98]. And Wnt signals could also hinder Runx2 activity through P38 and ERK signals [23]. In addition, FGF genes are also the direct target genes of Wnt signals [96,[99][100][101].

TGFβ-Smad-BMP signaling pathway
TGFβ-Smad-BMP signaling pathway and Wnt/β-catenin signaling pathway had a synergistic effect, so that those which could work on Wnt signals could indirectly work on TGFβ-Smad-BMP signals [96]. In addition, FGF/FGFR also had a direct effect on TGFβ-Smad-BMP signaling pathway, and Noggin had a negative feedback regulation effect on BMP signaling pathways. FGF18 hindered the expression of noggin, hence increasing the BMP signals in cytoskeletons [102,103]. Meanwhile, Smads could also be activated by P38 and ERK signaling pathways [108] and P38 and ERK were also downstream factors of FGF/FGFR [9].
Based on the above, a great deal of research has demonstrated that the independent effect of FGF/FGFR, estrogen and FSS and the joint effect of estrogen and FSS play an important role in bone development. In addition, this paper also discusses at the level of molecule that MAPKs-ERK1/2 signaling pathway, Wnt/β-catenin pathway, PI3K-AKT signaling pathway and TGFβ-Smads-BMP signaling pathway are the cross-linked pathways of FGF/FGFR, estrogen and FSS. There is also research shown that FGFR1-ERK signals play a key role in FSS and estrogen induced osteogenesis. In addition, it remains to be studied whether there are other direct or indirect relations between FGF/FGFR, estrogen and FSS.