A Review on the Molecular Mechanisms of Action of Natural Products in Preventing Bone Diseases

The drugs used for treating bone diseases (BDs), at present, elicit hazardous side effects that include certain types of cancers and strokes, hence the ongoing quest for the discovery of alternatives with little or no side effects. Natural products (NPs), mainly of plant origin, have shown compelling promise in the treatments of BDs, with little or no side effects. However, the paucity in knowledge of the mechanisms behind their activities on bone remodeling has remained a hindrance to NPs’ adoption. This review discusses the pathological development of some BDs, the NP-targeted components, and the actions exerted on bone remodeling signaling pathways (e.g., Receptor Activator of Nuclear Factor κ B-ligand (RANKL)/monocyte/macrophage colony-stimulating factor (M-CSF)/osteoprotegerin (OPG), mitogen-activated protein kinase (MAPK)s/c-Jun N-terminal kinase (JNK)/nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), Kelch-like ECH-associated protein 1 (Keap-1)/nuclear factor erythroid 2–related factor 2 (Nrf2)/Heme Oxygenase-1 (HO-1), Bone Morphogenetic Protein 2 (BMP2)-Wnt/β-catenin, PhosphatidylInositol 3-Kinase (PI3K)/protein kinase B (Akt)/Glycogen Synthase Kinase 3 Beta (GSK3β), and other signaling pathways). Although majority of the studies on the osteoprotective properties of NPs against BDs were conducted ex vivo and mostly on animals, the use of NPs for treating human BDs and the prospects for future development remain promising.


Introduction
Bone diseases (BD) are characterized by inflammation, fractures, pains, and overall morbidity that hinder mobility and reduce the ability to indulge in various physical activities. These outcomes negatively impact the individual's mental health conditions and/or those of their caretakers (often family members); as such, the need for the treatment of BDs cannot be overemphasized. The treatment of BDs involves the use of various medications and therapies such as antiresorptive and anabolic therapies. Most commonly used are bisphosphonates (e.g., alendronate, risedronate, ibandronate and zoledronic acid) [1] and selective estrogen receptor modulators (SERMs; e.g., tamoxifen, raloxifene, and toremifene) [2]. Although not completely understood, the mechanism of action of NPs of plant origins such as fruits, onion, garlic, Curcuma longa L. (Zingiberaceae), and legumes have all been reported to elicit positive outcomes on bone health [20,21]. For instance, a 12-week study on ovariectomized (OVX) rats fed legume-added diets containing either soybeans, mung bean, cowpeas, or azuki beans reported that bone mass density (BMD) for spine and femur increased in all legume-fed rats when compared to the OVX rats fed non-legume containing diets (control) [22]. The authors also reported significant increases in the expression of RANKL and OPG with increases in the serum calcium and phosphate ratios for all OVX rats fed legume-enriched diets compared to the control [22]. In humans, the potential impacts of dietary proteins on the reduction of BDs was recently highlighted [23] and supported by two systematic reviews and metaanalyses studies conducted independently, which collectively reported that dietary protein intake improved femoral neck and total hip BMD, showed improvements on the lumbar spine, and significantly reduced cases of hip fractures [24,25]. In previous reviews, the pro-osteoanabolic and anti-osteocatabolic properties of extracts from different plant species were thoroughly discussed [26][27][28][29], motivating several new studies on the application of herbal materials in halting bone resorption in vitro and in vivo [30][31][32][33][34][35][36]. In this review, we aim to discuss the cellular and molecular mechanisms behind the healthpromoting benefits of selected plant-derived and purely characterized NPs. Some health and safety concerns over the use of NPs are also briefly discussed, and our conclusions highlight some prospects for future studies. From reputable scientific databases and search engines such as PubMed, Google Scholar, and ScienceDirect, recent literature sources (2015-January 2022) on natural products with bone health-promoting effects and their mechanisms of action were retrieved. Search terms included "osteomodulatory",

Pathological Mechanism of Development of Some Bone Diseases
Bone formation or modeling is achieved through a tightly regulated cascade of complex homeostatic cell-cell interactions between osteoclasts and osteoblasts. Osteoblasts are specialized cells of mesenchymal origin that are implicated in the synthesis of bone matrix and the regulation of bone mineralization [37]. Osteoblast activities starts after their differentiation from their parental mesenchymal lineage, a process mediated by RUNX2 (runt-related transcription factor 2) and osterix. RUNX2 and osterix are transcription factors whose absence in osterix-null mice and RUNX2-null mice lines lead to the non-production of osteoblasts or bones tissues, an indication that RUNX2 and osterix are central to the differentiation of osteoblasts [38,39]. One of the most important functions of bone is the hematopoiesis process, which takes place in the bone marrow [40]. Its homeostasis is maintained by a complex unit that mainly involves monocytes/macrophages, from which specialized cells called osteoclasts originate, which adhere to the bone matrix, thus causing the secretion of acid and lytic enzymes that degrade the bone matrix. The differentiation of osteoclasts from their parental lineage is achieved upon the stimulation of RANKL and the M-CSF [41]. The binding of M-CSF to the colony-stimulating factor-1 receptor (c-Fms) activates the distinct signaling pathways that ensure the survival of osteoclasts for the precursor and mature cells. However, RANKL has been identified as the primary osteoclast Figure 2. Molecular mechanism of bone resorption and the signaling pathways involved. SASPsenescence-associated secretory phenotype; SOST-sclerostin; DMP-1-dentin matrix acidic phosphoprotein 1; HIF-1-hypoxia-inducible factor 1; NOX-1/2-NADPH oxidase 1/2; cyt c-cytochrome C; ROS-reactive oxygen species; AGE/RAGE-advanced glycated end-products/receptor; HMGB1-high mobility group box 1; PTH-parathyroid hormone; AMPK-AMP-dependent kinase, PGC-1α-peroxisome proliferator-activated receptor gamma coactivator 1-alpha, tumor necrosis factor alpha; NF-ĸb-nuclear factor kappa B; TNF-α-tumor necrosis factor-alpha; p38 MAPK-mitogen-activated protein kinase; PGE2-prostaglandin E2, FasL-Fas ligand; IL-1b/6interleukin 1 beta; IL-6-interleukin 6; VEGF-A-vascular endothelial growth factor alpha; CTSKcathepsin K; IFN-γ-interferon gamma; CTR-calcitonin receptor, PI3K-phosphoinositide 3-kinase; Akt-protein kinase-B; GSK3β-glycogen synthase kinase-3β; JNK-c-Jun N-terminal kinase; ERK-extracellular signal-regulated kinase; tumor necrosis factor alpha; NF-ĸb-nuclear factor kappa B; TNF-α-tumor necrosis factor-alpha; p38 MAPK-mitogen-activated protein kinase.
Other evidence of the involvement of the immune system is represented by B-cells, which regulate the RANK/RANKL/OPG axis, and which increase the production of RANKL in postmenopausal women [53,54]. Interestingly, new information has emerged on the interaction between homeostasis of bone metabolism and intestinal flora [55]. The set of microorganisms living in the human digestive tract influences the homeostasis of the gastrointestinal tract and the extra-gastrointestinal tract; in fact, the production and absorption of nutrients can be modulated. Germ-free mice have shown an increase in bone mass, highlighting a relationship between bone homeostasis and microbiota [56]. Moreover, the use of probiotics or antibiotics affects bone health [57]. Numerous treatments for osteoporosis have been developed, such as antiresorptive agents, while bone stimulators are all experimental and have not been approved by the Food and Drug Administration [58]. The general recommendation for all patients with osteoporosis is to ingest a physiological level of calcium and vitamin D, perform an appropriate exercise program, and adopt fall prevention.
A disequilibrium of bone turnover could lead to another skeletal dysplasia named osteopetrosis, a condition characterized by increased bone volume due to a dysfunction of the osteoclast activity or defects in the osteoclast formation. This condition produces deformities and structural fragility that is the cause of frequent fractures. In addition, affected patients have low blood cell production and loss of cranial nerve function, causing blindness, deafness, and facial nerve palsy [59]. According to the way in which the disease is inherited, it is classified into three different types: autosomal dominant, autosomal recessive, and X-linked recessive. The most common is the autosomal dominant form, which is characterized by mild symptoms that occur in late childhood through adulthood. The autosomal recessive form appears soon after birth and often shortens life expectancy. The X-linked form of osteopetrosis is extremely rare, with only a few reported cases [60]. The genes that cause the disease have been classified into the osteoclast-rich group and the osteoclast-poor group, where the latter includes mutations in RANKL and RANK genes [59,61]. A rare form of osteoclast-poor osteopetrosis depends on a mutation inside the RANKL gene that causes a deficiency of osteoclasts in bone tissue [62]; in particular, mutation on conserved residues alters RANKL homotrimerization with a potential impact on its function [63]. Given the nature of the disease, it may be useful to use genetic testing to confirm the diagnosis and to provide information on major organ involvement [64].

Cellular and Molecular Mechanisms of Bone Health-Promoting Properties of Some Natural Products
The maintenance of bone homoeostasis is dependent on the continuous removal of old and damaged bone tissues and the formation of new ones. These processes are linked to the activities of osteoclasts and osteoblasts, which are regulated by a cascade of interactions involving their respective signaling pathways [65,66]. This section presents the molecular mechanisms behind the actions of some NPs against bone malfunction initiated experimentally and their corresponding signaling pathways. As the major pathways associated with bone remodeling and pathogenesis of most bone diseases [42,[67][68][69][70][71], the osteoprotective properties of NPs through the modulation of RANKL/M-CSF/OPG, MAPKs/JNK/NF-κB, Keap-1/Nrf2/HO-1, BMP2-Wnt/β-catenin, PI3K/Akt/GSK3β, and other signaling pathways are discussed in this section. Recently, the application of kaempferol, artemisinin, and its analogues as potential osteoprotective agents as demonstrated in various models of BDs has been succinctly reviewed [72,73]; consequently, these compounds were excluded from the present work.

Targeting the RANKL/M-CSF/OPG and MAPKs/JNK/NF-κB Signaling Pathways
Binding of RANKL and M-CSF to their respective receptors, RANK and c-Fms, initiates a network of biochemical events that induce osteoclastogenesis (OCG) and survival of osteoclasts [42]. OCG, which is the proliferation, differentiation, and maturation of osteoclasts, is an important event in the pathogenesis of bone resorption, OP, and other BDs [74]. These occur when the gene expression profile of OPG is downregulated [75]. Consequently, the molecules that inhibit both RANKL-and M-CSF-originated suppressions of OPG and also halt the clonal expansion of osteoclasts are the preferred candidates for osteolytic disease prevention and management [76]. Moreover, the induction of OCG by RANKL/M-CSF/OPG signaling occurs due to the upregulation of TNF receptor-associated factor-6 (TRAF6). TRAF6 activates a cascade of downstream events such as the extracellular signal-regulated kinase (ERK)/JNK/MAPK signaling, NF-kB/nuclear factor of activated T Cells 1 (NFATc1) signaling, Ca 2+ signaling, and the PI3K/Akt/GSK3β signaling pathways, among others [77,78]. These signaling pathways lead to bone components' degradation, whether through the CFos or Ca 2+ signaling-mediated activation of the nuclear factor of NFATc-1, the transcription factor that upregulates the expression of several OC marker genes. Indeed, inflammation plays a significant role in OCG; thus, inflammation inducers such as lipopolysaccharides (LPS) are used in the experimental induction of OCG. LPS is a bacterial antigen that upregulates messenger ribonucleic acid (mRNA) expression of nuclear factor kappa B (NF-kB), which promotes the release of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) and activates OCG [79,80]. Similarly, OCG induction can be achieved through hypercalcemia using chemicals such as 1α,25-dihydroxy-vitamin D3 [1α,25(OH) 2 D 3 ] and prostaglandin E 2 with a concomitant upregulation of RANKL [81].
The ongoing quest to identify NPs from diverse sources for the treatment and management of BDs has revealed prospects for herbs and fruits, which are rich in polyphenols and other phytochemicals [82]. For instance, the effects of alisol-B (1), a phytosteroid from Alisma orientale Juzepczuk, was investigated on 1α,25(OH) 2 D 3 -induced OCG in mouse bone marrow cells (BMMs) and primary osteoblasts [83]. The authors reported that alisol-B downregulated the level of phosphorylated c-Jun N-terminal kinase (JNK) and mRNA expression of NFATc1 and CFos, the major transcription factors needed for OCG and the consequent resorption of bones. Alisol-B 23-acetate was reported to block the mobilization of calcium ion (Ca 2+ ), and its capability to reduce serum Ca 2+ level and prevention of bone loss by over 50%, 95%, and 100% at 0.5, 1, and 5 µM, respectively, in a hypercalcemia mouse model after intragastric ingestion [84] indicates that alisol-B is biostable and bioavailable.
Desoxyrhapontigenin (2), a stilbene derived from Rheum undulatum, was shown to prevent OCG in RANKL-exposed BMMs by halting the activation of extracellular signalregulated kinase (ERK) and suppressing the gene expression of c-Fos and NFATc1 [85]. In LPS-initiated bone loss in a mouse model, the stilbene also inhibited OCG by over 90% at 10 µM compared to 70% by resveratrol at the same concentration, and reduced bone resorption at 3, 10, and 30 µM by over 60%, 80%, and 95% by downregulating gene expression of matrix metalloproteinase-9 (MMP-9), tartrate-resistant acid phosphatase (TRAP), and cathepsin K (CTSK), which are target genes for NFATc1 and markers for OC abundance. Moreover, resveratrol activated the osteogenic transcription factor core-binding factor alpha 1 (Cbfa1) and NAD-dependent acetylase enzyme, Sirtuin 1 (Sirt-1), leading to the activation of differentiation and maturation of osteoblasts and osteogenesis. Furthermore, curcumin can reduce osteoarthritic environment-induced inflammation, extracellular matrix degradation, and chondrocyte apoptosis by targeting the NF-kB-Sox9 signaling axis [86]. In addition to the inhibition of RANKL-generated NFATc1, CTSK, and TRAP activation, another natural riboflavin-derived compound, lumichrome (3), suppresses calcium oscillation and activation of NF-κB and MAPKs needed for downstream events of the RANKL/M-CSF signaling pathway in RAW264.7 mouse macrophage cells (MMCs) and MC3T3-E1 pre-osteoblasts (MCPs) [87]. This anti-OCG property of lumichrome was confirmed in OVX mice, where the natural product inhibited bone loss by an approximately 2-fold decrease in osteoclast formation at 10 µM. Considering the role of estrogen in maintaining bone health, such as the promotion of type 1 collagen synthesis, osteoblast survival, and osteoclast apoptosis, and the inhibition of OCG [16,88], the depletion of estrogens in postmenopausal women is associated with several bone dysfunctions such as OP, osteoarthritis, and bone fractures [66,89]. Experimentally, estrogen deficiency is induced in female rodents by ovariectomy to mimic menopause-related estrogen depletion in aging women [90]. This animal model is widely adopted in the study of the potency of molecules for promoting bone health and preventing age-related bone degeneration [91].
OVX mice Prevented loss of bone mineral density (BMD).

OVX mice
Prevented OVX-induced bone loss and titanium particle-induced osteolysis reducing expression level of TRAP, V-ATPase, NFATc1, CTR, and CTSK between 25% and 90% at 0.4 µM compared to control, but increased TN, TT, and BV/TV levels by at least 20% at 0.5-2.5 mg/kg.
OVX mouse Reduced OCs and TS levels by 10-fold while increasing the expression of BVD and TN by over 50% at 10 mg/kg for 6 weeks. [95]

LPS-induced osteolysis in mouse
Prevented LPS-induced bone loss in a dose-dependent manner at 5-10 mg/kg/d for one week. [99]

Kavain (12)
Piper methysticum OVX mouse Inhibited bone loss and OCG by 50% and elevated protein levels of BVD and TN at 10 mg/kg i.p., 3x/week for 6 weeks.

RANKL-induced OCG in BMMs
At 5 µM, inhibited OCG via reduction in NFATc1, CFos, p38, ERK, TRAP, OSCAR, and ATP6v0d2 expression and F-actin ring formation by more than 80% and increased IκBα expression by over 60% after 7 days of treatment.

LPS-induced mice
Inhibited LPS-induced bone erosion by 90% via reduction in F-actin ring formation, lacunar resorption pits formation, and OcS/BS by over 30% after 8 weeks of treatment. [94]

LPS-induced osteolysis in mice
Inhibited bone loss by reducing F-actin ring formation and resorption indices by 90% and 75%, respectively, and increasing TN, TT, and BV/TV by 5% each at 10 mg/kg for 7 days. [111]
Titanium particle-induced osteolysis in mice Inhibited bone loss reducing bone level of NP and PP, with concomitant increases in TN, TT, and BV/TV levels by at least 85% at 15 µg/kg for 10 days. [113]

LPS-induced bone erosion in mice
Inhibited bone loss by increasing bone levels of TN, TT, and BV/TV by 85% each, while decreasing the levels of F-actin ring formation, resorption indices, and OC over 50% at 10 mg/kg for one week. [114] RANKL-induced OCG in BMMs Inhibited OCG by suppressing the gene expression of Syk, PLCγ2, Gab2, ERK, NFATc1, CFos, TRAF6, c-Src, and CTSK by 30% and above at 10 µM.

Targeting Keap-1/Nrf2/HO-1 Signaling Pathway
As intracellular signal molecules, reactive oxygen species (ROS) activate RANKLmediated OCG and enhance the intracellular oxidative stress [116]. Hence, the reduction of ROS levels in bone tissue is a good target for osteoprotection [117]. When dissociated from its repressor, Kelch-like ECH-associated protein 1 (Keap-1), nuclear factor E2-related factor 2 (Nrf2) is translocated to the nucleus where its interaction with antioxidant response element (ARE) increases the gene expression for antioxidant enzymes that protect cells from ROSinduced cellular injury [118][119][120]. The enhanced expression of cytoprotective enzymes such as γ-glutamylcysteine synthetase (GCS), heme oxygenase-1 (HO-1), and NAD(P)H:quinone reductase (NQO1) as mediated by Nrf2 and the consequent suppression of OS has been shown to promote bone health [121,122]. In addition to the inhibition of OCG, other strategies to improve bone health involve protecting osteoblasts against externally and internally generated damages. In a recent investigation, a novel Keap-1 inhibitor, iKeap1, was shown to protect osteoblast from chemically generated oxidative assault and apoptosis by improving Nrf2-mediated expression of cytoprotective factors [123]. Treatment with a low dose of glucocorticoids, including dexamethasone (DEX), cortisone, and prednisolone, over a short time was shown to regulate bone remodeling and maintain homeostasis of the bone tissues [124]. However, administering the same drugs at a higher dosage over a long time elicited OP [125] and osteonecrosis via upregulation of glycogen synthase kinase-3β (GSK3β) gene expression and its enzyme activity [126]. Therefore, an overdose of DEX is helpful in the experimental induction of OP. Gastrodin (30), a phytochemical isolated from Gastrodia elata, was shown to protect murine osteoblastic cells (MC3T3-E1) from DEXgenerated cellular assaults by improving cellular viability by 30% and the expression of osteogenic genes (RUNX2, osterix, bone morphogenetic protein-2 (BMP-2), and osteocalcin (OCN), and ALP activity [127], with over 0.15-, 0.3-, 2-, 0.2-, and 0.4-fold increases at 5 µM when compared to DEX-untreated group). In DEX-exposed rats, gastrodin prevented bone mineral loss, mitochondrial membrane dysfunction, and endoplasmic reticulum stress by suppressing apoptosis over 25% and 37.5% at 1 and 5 µM by reducing the levels of apoptosis-inducing factor, bax, cytochrome-C, and caspase-3 (by 50% at 5 µM compared to that DEX-induced untreated group), while activating Nrf2 signaling pathways [127]. The ability of gastrodin to offer protection against DEX-induced OP after oral ingestion is an indication of its bioavailability and biostability; these two important properties make it a suitable candidate for drug development towards OP prevention. Based on these findings, natural products such as gastrodin, icariin, and chlorogenic acid with antioxidant ability via the binding to and translocation of Nrf to the nucleus to upregulate the expression of antioxidant molecules will alleviate ROS-mediated bone loss. Besides gastrodin, there are other NPs such as icariin (31) [128], chlorogenic acid (32) [129], and 4-phenyl butyric acid (33) [130] that attenuate DEX-induced alteration in structure and function of bone tissues, potentially via modulation of the Keap-1/Nrf2/HO-1 signaling pathway (Figure 3). remodeling and maintain homeostasis of the bone tissues [124]. However, administering the same drugs at a higher dosage over a long time elicited OP [125] and osteonecrosis via upregulation of glycogen synthase kinase-3β (GSK3β) gene expression and its enzyme activity [126]. Therefore, an overdose of DEX is helpful in the experimental induction of OP. Gastrodin (30), a phytochemical isolated from Gastrodia elata, was shown to protect murine osteoblastic cells (MC3T3-E1) from DEX-generated cellular assaults by improving cellular viability by 30% and the expression of osteogenic genes (RUNX2, osterix, bone morphogenetic protein-2 (BMP-2), and osteocalcin (OCN), and ALP activity [127], with over 0.15-, 0.3-, 2-, 0.2-, and 0.4-fold increases at 5 µM when compared to DEX-untreated group). In DEX-exposed rats, gastrodin prevented bone mineral loss, mitochondrial membrane dysfunction, and endoplasmic reticulum stress by suppressing apoptosis over 25% and 37.5% at 1 and 5 µM by reducing the levels of apoptosis-inducing factor, bax, cytochrome-C, and caspase-3 (by 50% at 5 µM compared to that DEX-induced untreated group), while activating Nrf2 signaling pathways [127]. The ability of gastrodin to offer protection against DEX-induced OP after oral ingestion is an indication of its bioavailability and biostability; these two important properties make it a suitable candidate for drug development towards OP prevention. Based on these findings, natural products such as gastrodin, icariin, and chlorogenic acid with antioxidant ability via the binding to and translocation of Nrf to the nucleus to upregulate the expression of antioxidant molecules will alleviate ROS-mediated bone loss. Besides gastrodin, there are other NPs such as icariin (31) [128], chlorogenic acid (32) [129], and 4-phenyl butyric acid (33) [130] that attenuate DEX-induced alteration in structure and function of bone tissues, potentially via modulation of the Keap-1/Nrf2/HO-1 signaling pathway (Figure 3).  Bone loss caused by oxidative stress due to an imbalance in the antioxidant and free radicals in the system can be alleviated by natural products with antioxidant ability. Natural products promote the cytosol-to-nuclear translocation of Nrf2 (when detached from its repressor, Keap-1, during increased ROS generation) to interact with its operator, ARE, to induce the expression of antioxidant enzymes. Increased availability of antioxidant enzyme will hence scavenge the ROS and prevent ROS-mediated bone loss. Abbreviations: GCS-γ-glutamylcysteine synthetase; HO-1-heme oxygenase-1; NQO1-NAD(P)H:quinone reductase; Keap-1-Kelch-like ECH-associated protein 1; Nrf2-nuclear factor E2-related factor 2; ARE-antioxidant response element; GPx-glutathione peroxidase; GST-glutathione-S-transferase; CAT-catalase; DEX-dexamethasone.

Targeting BMP2-Wnt/β-Catenin Signaling Pathway
RUNX2 is a human protein that regulates the gene expression of the extracellular matrix protein encoded by the RUNX2 gene. The gene of RUNX2 protein is expressed early in the mesenchyme of an embryo; RUNX2 modulates the generation of skeletal components before the bone tissues are formed. The BMP-2/Wnt/β-catenin signaling pathway regulates RUNX2 activity [87]; extracellular bone morphogenetic protein-2 (BMP-2) activate serinethreonine kinase receptors I and II, causing the phosphorylation and translocation of SMAD (SMA, "small" worm phenotype, and MAD, "mothers against decapentaplegic") proteins into the nucleus to upregulate the expression of osteogenic genes such as RUNX2 [131] and osteocalcin [132]. Similarly, BMP activates the movement of β-catenin into the nucleus to initiate osteoblast development via the canonical Wnt/β-catenin signaling pathway that elevates osteoblastogenesis (OBG) and bone production [133,134]. The secreted Wnt proteins bind to the frizzled receptor and co-receptors (LRP5/6) to inhibit adenomatosis polyposis coli (Apc)-GSK3β-actin complex and prevent the phosphorylation-mediated breakdown of β-catenin by the proteasome. Therefore, β-catenin builds up in the cytosol, and is transported into the nucleus where it binds to its promoter regions, TCF and LEF, to stimulate the transcription of target genes such as the RUNX2 gene, C-myc, cyclin D1, and c-jun. The nuclear activity of β-catenin includes the expression of osteogenic proteins such as osteocalcin (OCN), alkaline phosphatase (ALP), and bone sialoprotein (BSP) [134,135]. When in a dephosphorylated state, its breakdown in proteosomes is induced by active GSK3β phosphorylates β-catenin, which leads to a reduction in its cytosolic level [136,137]. Hence, GSK3β is a regulator of BMP-2/Wnt/β-catenin signaling pathways at the Wnt/βcatenin axis.
The most prevalent therapeutic strategies for treating OP include the administration of estrogens, bisphosphonates, vitamin D analogues, and RANKL inhibitors [138]. Although effective in the management of OP, these drugs increase the risk of obesity and diabetes and metabolic syndrome, in general [139,140]. Consequently, NPs with little or no side effects are of interest in the research on the prevention and treatment of OP and related diseases. Low-molecular-weight compounds of natural origin have been reported to boost osteoblastogenesis (OBG) in several in vitro studies [141,142], including those that act via Wnt/β-catenin signaling pathway [137,143]. Notably, albiflorin (34) and paeoniflorin (35) isolated from Paeonia lactiflora enhanced OBG in MC3T3-E1 cells by upregulating the mRNA expression of Wnt (Wnt10b), β-catenin (ctnnb1), LRP5, LRP6, Dvl2, and cyclin D1 (Ccnd1) [144], suggesting the involvement of BMP2-Wnt/β-catenin signaling pathway. Other NPs that promote bone health either by inhibiting OCG or promoting OBG and fracture healing via BMP2-Wnt/β-catenin signaling pathway include quercetin (36), a flavonoid widely found in fruits and vegetables [145]; polydatin (37), a precursor of resveratrol [134,146,147]; and 3,5-dicaffeoyl-epi-quinic acid (38) from Atriplex gmelinii [140]. In general, NPs such as albiflorin and paeoniflorin with the ability to target BMP-2/Wnt/βcatenin signaling pathway to upregulate the RUNX2 gene required for bone health are potential agents to ameliorate BDs such as OP.

Targeting PI3K/Akt/GSK3β Signaling Pathway
The phosphoinositide 3-kinase/protein kinase-B (PI3K/Akt) signaling pathway is well known due to its multifunctionality in biological processes, including maintenance of glucose and bone homeostasis [148]. Studies on cell proliferation and death have always focused on the PI3K/Akt/GSK-3β signaling pathway and the BMP-2/Wnt/β-catenin signaling pathway [149]. The PI3K, an intracellular phosphatidylinositol kinase with several catalytic subunits, is activated by the interaction between an extracellular signal molecule and activated Ras protein receptor, leading to the phosphorylation of PIP2 to PIP3, which activates Akt by phosphorylation [150]. The active Akt, in turn, phosphorylates the ser-9 of GSK3β and inactivates it, making β-catenin available to play roles in the upregulation of expression of genes that code for osteogenic proteins [151]. DEX has been shown to elevate ROS production, which can lead to apoptosis through the mitochondrial caspase apoptosis pathway [152]. Furthermore, the accumulation of ROS causes oxidative stress in osteoblasts, activating the JNK pathway, which inactivates the Akt pathway and promotes osteoblast death [153]. Vinpocetin (39), a small chemical derived from the leaves of Phyllostachys pubescens, attenuates DEX-induced rat osteoblast apoptosis by 60% and 65% at 5 and 10 µM than that caused by DEX and inhibits the progression of osteonecrosis of the femoral head (ONFH) by lowering DEX-induced elevated levels of ROS by over 90% at 10 µM and activating the Akt pathway via Akt phosphorylation by over 2-fold at 10 µM. Vinpocetin had no effect on osteoclast differentiation and prevented DEX-stimulated protein upregulation, including cleaved-caspase3 and Bax, while dramatically suppressing Bcl-xl and Bcl2 downregulation [154] by over 2-fold each.

Targeting Rev-Erbs Signaling Pathway
The nuclear receptors, Rev-erbs, consisting of Rev-erbα and Rev-erbβ, corresponding to nuclear receptor subfamily 1, group D, members (NR1D) 1 and 2, play multiple physiological roles, such as the regulation of the circadian rhythm, bone health, and inflammatory processes [162,163]. Circadian clocks have been shown to regulate bone remodeling through the controlling of formation and resorption of bones [164]. The role of Rev-erbs in maintaining bone homeostasis has been demonstrated in knockout mouse where deletion of Rev-erbα gene increased inflammation, mitochondrial dysfunction, and oxidative stress, and promoted OCG [163]. However, the activation of Rev-erbβ inhibits OCG and prevented chemical-induced loss in bone mass [165,166]. Specifically, the activation of Reverbα prevents RANKL-generated podosome belt synthesis and prevents osteoclast-induced bone resorption, thereby suppressing OVX-induced bone loss. In addition, the activation of Rev-erbα, using its agonists SR9009 and GSK4112, abrogates OCG, as characterized by the suppression of NF-κB, NFATc1, and CFos and their downstream effectors, TRAP, MMP9, and CTSK [165], and inhibited LPS-generated inflammation [167]. Therefore, Rev-erbα exerts a crucial role in maintaining healthy bone status by inhibiting OCG while promoting osteogenesis. NPs have been shown to alleviate oxidative stress-induced diseases, including OP, and metabolic disorders by regulating the Rev-erbα/miR-882 pathway. For instance, berberine (45), isolated from Rhizoma coptidis, was reported to be an agonist of REV-ERBα in a colitis model [168]. Nevertheless, the effects of berberine on Rev-erbα are both dose-and "time-of-the-day"-dependent [168,169]. Thus, a comprehensive understanding of the impacts of berberine and Rev-erbs interaction in relation to bone health will require further studies in the future. It would also be interesting in the future to examine the osteoprotective effects of natural modulators of Rev-erbs expression and activities. In sum, bone health is improved with natural products such as berberine to alleviate chemical and OVX bone loss and RANKL-generated and osteoclast-induced bone resorption via the modulation of the Rev-erb signaling pathway.

Targeting Calcium Ion Signaling Pathway
Calcium is one of the major and ubiquitous components of bone found mostly in the form of calcium hydroxyapatite (Ca 10 [PO 4 ] 6 [OH] 2 ) [170]. Calcium, to a very large extent, contributes to the homeostasis and regulation of bone metabolism and bone resorption via many elucidated and yet-to-be-unraveled pathways. Moreover, calcium forms one major regulatory component of the osteoblast and osteoclast and plays a role in determining the fate of mesenchymal cells in humans [171]. Several hormones, cytokines, and matrix-embedded factors play invaluable roles in the maintenance of the balance between extracellular calcium and intracellular calcium [172]. Calcium-sensing receptors (CaSR) are vital parathyroid, seven-transmembrane G-protein couple receptors, which are also present in the kidney, and function in the maintenance of homeostasis of bone by controlling the influx and release of extracellular calcium from bone cells [173]. An increase in Ca 2+ activates CaSR, which causes an influx of Ca 2+ into the osteoblast, thereby promoting bone remodeling. However, when there is a decrease in extracellular Ca 2+ , CaSR causes resorption of Ca 2+ from distal tubules of the kidney as well as fostering bone resorption by osteoclasts [174]. When the extracellular Ca 2+ is increased, maybe as a result of increased osteoclastic activities, CaSR stimulates the influx of Ca 2+ into osteoblasts by activating phospholipases C, A 2 , and D [172].
Primarily, phospholipase C fosters the formation of the second messenger (diacylglycerol (DAG) and inositol-1,4,5-trisphosphate (IP3)) from the cleavage of phosphatidylinositol 4,5-bisphosphate (PIP2) as well as the activation of phospholipases A2 and D, whose role in maintaining bone balance is poorly understood [175]. While the DAG remains membranebounded, IP 3 , which is soluble, diffuses out of the parathyroid cells and binds to calcium channel receptors (inositol trisphosphate receptor-IP3R) located in the endoplasmic reticulum, thereby fostering the release of Ca 2+ in the cytosol and activating other Ca 2+ -regulated signaling axes [175]. Alternatively, intracellular calcium, especially in osteoblast cells, exerts some form of regulation on the bone homeostasis process. Hormones such as vitamin D3, insulin-like growth factor (IGF), and parathyroid hormone (PTH), including several other cytokines and growth factors, modulate the bone remodeling process mainly by influencing the amount of intracellular calcium in the osteoblasts [174]. Most of these hormones facilitate the translocation of extracellular calcium into intracellular osteoblast cells via the L-type and non-L-type isoforms of voltage-gated calcium channels. Increases in intracellular Ca 2+ concentration consequently activate the CaMK-dependent pathway, which fosters an increased rate of cell entry to cell cycle and osteoblast proliferation [176]. Like the influence of calcium on extracellular CaSR-mediated and intracellular endocrine-mediated bone homeostasis, natural compounds and their synthetic counterparts act as CaSR agonists (calcimetics) or antagonists (calcilytics), and as such, regulate bone resorption-modeling balance [177]. This-subsection focuses on NPs that can improve bone health through the Ca 2+ signaling axis. Several studies have identified some phytochemicals and bioactive peptides to play vital roles in maintaining bone health through the regulation of CaSR activities and other Ca 2+ signaling pathway. Feng et al. [174] identified ligustroflavone (47) from Ligustrum lucidum with the ability to inhibit CaSR and improve the level of parathyroid hormones in HEK-293 cells and the serum of diabetic mice, respectively. In contrast, a similar study reported an upregulation of CaSR in human aortic endothelial cells administered with a dipeptide, γ-Glutamyl valine (γ-EV), from edible beans [173]. Although the report was directed towards the management of diabetes and intestinal inflammation, the CaSR mRNA, which showed suppression in one study and upregulation in the other, also plays a vital role in the CaSR axis of bone homeostasis [173,174]. In conclusion, NPs that modulate CaSR activities and other calcium signaling pathway could regulate OCs, osteoblasts, and extracellular calcium ion balance, resulting in bone remodeling.
Several other studies, as summarized in Table 3, have investigated some other natural products on other axes of Ca 2+ signaling, such as on type 1 collagen [178], calmodulin phosphor kinase II [176], L-type Ca 2+ -gated channel [179,180], and myosin light chain kinase [181]. There is a need for more focused studies on the management and improvement of bone health through the several Ca 2+ signaling axes. Table 3. Natural products targeting the calcium signaling axis with potent relevance to bone health.

Natural Product
Source Cell Line/Animal Model Specific Therapeutic Activity Ref.

Targeting Endogenous Molecules with Bone Promoting Properties
In addition to natural compounds sourced externally, targeting the upregulation of endogenous molecules can be an additive strategy for promoting bone health. For example, an extracellular matrix protein, ameloblastin (51), inhibited RANKL/M-CSF-induced OCG in MMCs by suppressing NFATc1 activation via inhibition of p38-and JNK-mediated upregulation of CFos [183]. Further supporting this report on irisin, an endogenous molecule from human dental bud was shown to promote bone formation in cultured MMCs [184]. Similarly, an endogenous methoxyindole, melatonin (N-acetyl-5-methoxy-tryptamine) (52), was recently shown to halt OCG in MMCs exposed to RANKL and M-CSF by downregulating MRNA expression of CTSK and its protein level via the upregulation of expression of nuclear receptor subfamily-1 group D member-1 (NR1D1, also known as Rev-erbα) with a concomitant downregulation of miR-882 [185]. An in-depth exploration of melatonin towards improving bone health using preclinical and clinical studies has been given elsewhere [186]. Other natural products such as maackiain, cycloastragenol, gamabufotalin, abrine, hymenialdisine, oroxylin A, demethylbelamcandaquinone B, and pitavastatin have been recently shown to promote bone health by enhancing the differentiation and maturation of endogenous molecules that inhibit bone resorption and/or promote osteoblastogenesis, and some of the compounds such as hymenialdisine, cycloastragenol, 12-deoxyphorbol 13-acetate, aerophobin-1, and pitavastatin have been demonstrated to be bioactive in both in vitro and in vivo experiments [187][188][189][190][191][192][193][194][195][196]. Nonetheless, further studies towards the enhancement of the protein levels of these endogenous osteoprotective molecules and their receptors hold promise towards finding novel interventions in the prevention, treatment, or management of BDs. Based on the above findings, compounds such as ameloblastin, abrine, oroxylin, and pitavastatin can potentially promote bone health by targeting the upregulation of endogenous molecules that can inhibit RANKL/M-CSF-induced OCG.

Targeting Vitamin D Receptor (VDR)
In an OPG-deficient murine model, intragastric ingestion of 1α,25-dihydroxyvitamin D 3 [1α,25(OH) 2 D 3 ] was shown to halt bone resorption and erosion, despite the elevated activation of RANKL/RANK signaling pathway [81]. However, 1α,25(OH) 2 D 3 prevented M-CSF-and RANKL-induced OCG in cultured MMCs via suppression of gene expression of CFos protein [81]. This implies that the activation of VDR abrogates OCG by preventing CFos-mediated activation of NFATc1 and that vitamin D and its analogues exert their osteoprotective properties by targeting CFos-induced OCG. EGCG (40) was shown to exhibit an anti-OCG property by preventing IL-1-induced bone loss via targeting the gene expression of CFos and its protein activation [197]. However, whether VDR is involved in the downregulation of CFos gene expression and the suppression of its capacity to induce OCG and bone loss by EGCG remains to be investigated. In sum, NPs' bone remodeling capacity occurs by activating VDR suppression of CFos protein expression, leading to the reduction in OCs. Figure 4 summarizes the major pathways by which NPs inhibit OCG and the associated bone loss, in addition to the Keap-1/Nrf2 signaling pathway, while Figure S1  capacity occurs by activating VDR suppression of CFos protein expression, leading to the reduction in OCs. Figure 4 summarizes the major pathways by which NPs inhibit OCG and the associated bone loss, in addition to the Keap-1/Nrf2 signaling pathway, while Figure S1

Clinical Trials of Natural Products for Bone Diseases
Despite the numerous interesting outcomes on the efficacy of natural products on bone health as reported from a variety of in vitro cell lines and in vivo animal studies, the findings cannot be validated as efficacious in humans without proper clinical studies. A few clinical studies have shown that some natural products from plants and animals have improved the health condition of osteoarthritis and osteopenia patients, especially in the postmenopausal population ( Table 4). Most of the studies adopted randomized, parallel interventional, double-blinded, placebo-controlled models for the trials. Generally, changes in certain bone parameters such as the increase or maintenance of BMD and bone formation marker proteins (N-terminal propeptide of type I collagen (P1NP), collagen type 1 cross-linked C-telopeptide, and bone-specific alkaline phosphatase (BSAP)) are important indicators of bone health improvement in patients with unhealthy bone conditions. Conversely, an increase in bone resorption markers (such as osteocalcin, carbo-terminal telopeptide of type I collagen (CTx), and TRAP) and a concomitant decrease in BMD are indications of osteopenia/osteoarthritis progression (Table 4).  (a) BRM (CTX) decrease in the group that consumed more DPs (∼42 g) after the crossover phase than the group that consumed fewer DPs (∼14 g) (a) A significant decrease (p = 0.006) [207] BSAP-bone-specific alkaline phosphatase; TRAP-tartrate-resistant acid phosphatase; OC-osteocalcin; PTH-parathyroid hormone; EG-experimental group; CG-control group BMD-bone mineral density; P1NP-N-terminal propeptide of type I collagen; CTx-carbo-terminal telopeptide of type I collagen; 25(OH)D-25-hydroxyvitamin D; IGF-1-insulin-like growth factor I; N/A-not applicable or not available; CT1CLCT-collagen type 1 cross-linked C-telopeptide; VAS-visual analogue scale; KLG-Kellgren-Lawrence grade; WOMAC score-Western Ontario McMaster Universities Score; QOL score-quality of life score.
In a study by Shen et al. [198], polyphenols from green tea administered as an oral formulation daily to 171 osteopenic postmenopausal women successfully led to a significant increase in the bone formation marker BSAP and a decrease in the bone resorption factors TRAP in a ratio of 103.6% when compared to the baseline (100%) [198]. Another study reported a similar increase in the BMD of L2-L4 lumbar spine vertebra (p < 0.05), femoral neck (p < 0.01), and trochanter (p < 0.01) of the treatment group administered 80 mg/d of isoflavone aglycones-rich red clover extract supplemented with calcium (1040 mg/d), vitamin D (25 µg/d), and magnesium (487 mg/d), when compared to the placebo control group administered only the supplement [202]. A similar outcome was obtained with hop rho iso-alpha acids (200 mg) and berberine sulphate trihydrate (100 mg) [200,201], collagen peptide from pork skin and bovine bone [203,205], prenylflavonoids from Epimedium spp. [204], kefir-fermented milk peptides [206], and dried plum [207].
There is a need for more clinical studies on other natural products, especially from plant phytochemicals and bioactive peptides, to ensure that they are not just efficacious but also non-toxic (without adverse effects) to humans. Only a few clinical studies have investigated the adverse effects of natural products alongside their effects on bone health [204]. Moreover, most of the available clinical studies investigated the effects of natural products on bone by administering formulations orally to the volunteered human subjects; there are very sparse or no available studies that have administered their natural formulation through intramuscular or intravenous in human subjects. There is a possibility that better bone regeneration and improvement effects could be achieved through intramuscular or intravenous injection, as suggested from the results of in vivo animal studies (Table 4).

Safety Concerns of Applying Some Natural Products for Clinical Management of Bone Diseases and Limitations of Some of the Current Studies
Although many NPs possess bone health-promoting properties, such as the prevention of bone resorption via anti-OCG or the improvement of the renewal of new bone cells and its protectants via OBG and upregulation of OPG expressions, the potential of some NPs to induce OCG should be approached with caution. For instance, intragastric injection of psoralen, a derived compound from Psoralea corylifolia L., when administered at 20 mg/kg for four weeks was reported to promote fracture healing in a rat transverse tibial fracture model. In contrast, the same compound (psoralen) was reported to enhance rat bone resorption and OCG in MCPs and BMMs via the activation of the ERK signaling pathway [208]. In molecular biology, psoralen is utilized as a mutagen. Psoralen plus ultra violet radiation (UVA) (PUVA) therapy is a combination of psoralen and UVA that can be used to treat hyperproliferative skin conditions including psoriasis and some types of skin cancer [209]. Unfortunately, the use of PUVA increases the risk of skin cancer [210]. Similarly, a recent study demonstrated that puerarin isolated from Puerariae radix alleviates hyperhomocycteinemia by antagonizing Rev-erbα, a receptor whose activation is known to protect the bones [211]. These findings confounded the roles of psoralen and puerarin on bone health; thus, they would not be suitable candidates for drug development in relation to fracture and hyperhomocycteinemia management. Some of the studies reviewed also have limitations in terms of the clinical usefulness of the study design. For example, a study by Jinhua et al. [128] reported that the administration of icariin at 250 mg/kg/d for 60 days protected against DEX-induced OP in mice. However, at lower doses (10 and 20 mg/kg/d), Jang et al. [98] reported that the administration of protocatechuic acid for 12 weeks protected against trabecular bone loss in OVX mice. The study designs of the above researchers as well as their findings are not translatable in clinical practice due to either remarkably high doses that will create safety concerns and/or very long duration of treatment, making these NPs competitively low compared to current drugs for managing bone diseases. Future researchers should adopt study designs that are clinically feasible, such as administering low doses and shorter treatment duration. For instance, at 1 and 5 mg/kg for 7 days and at 5 and 10 mg/kg/d for 7 days, garcinol and sarsasapogenin, respectively, were reported to attenuate LPS-induced calvarial osteolysis mice [99,159]. The design of the above studies is translatable into clinical practice with minimal concern for drug adherence and toxicity.
Some natural bioactive chemicals are produced by plants and marine organisms to protect them against predators due to their poisonous nature [212,213]. With the growing interest in medicinal plants, there is a need for more rigorous scientific research into their usefulness and toxicity. Natural products, e.g., lectins such as concanvalin A, found in plants such as cereal and beans were discovered to have toxic effects, including disrupting digestion and causing nutrient deficiencies; they can also elicit IgG and IgM antibodies, leading to food allergies [214,215]. These compounds can bind to erythrocytes simultaneously with immune factors, causing hemagglutination and anemia. In experimental animals, lectins impair host resistance to infection, induce failure to flourish, and even cause death [216]. In addition to organ toxicities [217] and gastrointestinal tract discomfort [218], a number of these plants and their chemicals induce genotoxicity, which could result in carcinogenesis [219]. Similarly, a marine glycoprotein known as shrimp tropomyosin was demonstrated to induce serious allergic reaction in an animal model, which can be abolished by glycation [220] It is therefore important to be cautious in selecting only natural products which are validated to be safe and patients are not allergic to, as the consumption of natural products whose safety is not substantiated could pose risks to human health. Efforts should be made to certify both the short-and long-term safety profile of NPs before their administration as drugs.

Conclusions and Prospects
Bones disease is a global health concern, and as such, there is an ongoing quest to identify the most effective treatments with little or no side effects. Current medications for treatments cause side effects that lead to more serious health conditions such as cancer. Thus, the search for alternative treatments has revealed the efficacy of NPs, many of which are of plant origins known to exhibit little or no side effects; moreover, some of these, such as soybeans, are part of the human diet. Thus, natural products' preventive benefits in bone diseases such as osteoporosis might be a viable alternative, or at least serve as complementary therapeutics to conventional therapy to minimize side effects. However, the effects of combining NPs and conventional drugs on pharmacokinetics and their safety profile need to be clearly-established prior to adoption. In this review, we described the mechanisms behind the actions of some NPs towards the treatment of BDs. However, these are not exhaustive. There is need for future studies to focus on discovering more NPs and their corresponding mechanisms of action. Moreover, many of the studies on BDs have been conducted in other animals such as mice and rats; thus, it would be a welcome development if human studies on bone could be improved, especially studies using plants that are part of the human diet. The concern over dosage has remained one of the hinderances in the wider adoption of NPs in relation to the treatment of diseases. Although this is valid and perhaps justifiable, more emphasis should be laid on the consumption of diets rich in these NPs, especially for BD patients and those prone to developing such diseases.
Indeed, many of the NPs with potential to halt, reduce, or treat bones diseases originate from plants, but they are produced in small quantities, which hinders their adoption on a commercial level. However, plant breeding and the relevant technologies have improved over the years, such that the production of various phytochemicals can be manipulated through multiple strategies. Therefore, future research studies on NPs of plant origin in relation to bone diseases treatment should incorporate some plant breeding techniques with potential to improve the accumulation of the products of interest. Taken together, this review illustrates that the future of NPs in the treatment of BDs is promising but achieving this will rely on future research.