Physiologically and pathologically, bone volume fraction depends mainly on the rate of bone formation by osteoblasts and the rate of resorption by osteoclasts. In most pathological bone diseases such as osteoporosis, excessive bone resorption by osteoclasts is the main cause of bone loss. At present, three main categories of drugs are used for osteoporosis treatment: anti-resorptive agents (inhibits osteoclasts), bone-forming agents (boost osteoblasts), and dual-effect drugs (both promote bone formation and inhibit osteoclasts) (1). Although they have shown good clinical efficacy, their side effects cannot be ignored (2–5). For example, bisphosphonates are a type of anti-bone resorption agent that increases the survival and activity of osteoclasts mainly by inhibiting the mevalonate biosynthetic pathway or by binding to a non-hydrolyzable analogue of ATP (6). As a first-line drug for osteoporosis, bisphosphonates have demonstrated some beneficial effects, including increasing bone density and lowering the risk of fracture (7), particularly hip fracture (8). Bisphosphonates are commonly well tolerated, but side effects can occur in up to 10% of patients, mainly including arthralgia, myalgia, and gastrointestinal discomfort (9). Other negative effects including uveitis (10), atypical femoral fractures (AFFs) (11), and osteonecrosis of the jaw (ONJ) (12), are relatively uncommon. Selective estrogen receptor modulators (SERMs) and anti-RANKL monoclonal antibodies are the other two classes of anti-resorptive drugs. SERMs include raloxifene and bazedoxifene, both of which have been shown to effectively prevent bone loss and reduce bone turnover (13). However, in a three-year clinical trial, the incidence of vasodilatation (hot flashes), leg cramps and venous thromboembolic events was significantly higher in the bazedoxifene and raloxifene groups compared with the placebo group (14). The latter is represented by denosumab, a human monoclonal antibody against RANKL, a key bone resorption mediator (15). Postmenopausal women treated with the drug showed a remarkable long-term reduction in the risk of fracture of up to 10 years (16). But, after treatment was stopped, the rate of vertebral fractures rose to the equivalent in untreated people (2). As a potent bone-forming drug, teriparatide is a good candidate for improving bone microstructure (17). However, it is only used in patients with grievous osteoporosis because it requires daily subcutaneous injection and is significantly more expensive than other osteoporosis medications (18). Romosozumab is a monoclonal antibody that binds sclerostin and has a dual regulatory effect of promoting bone formation and suppressing bone resorption (19). According to the findings of the current study, romosozumab is contraindicated for individuals with a recent history of a cardiovascular incident and should only be used with caution in patients with a high cardiovascular risk (5). The safety and affordability of the drug are key concerns for patients. Therefore, how to innovate and optimize the safety and efficacy of therapeutic drugs, while keeping them affordable to osteoporosis patients is a practical clinical problem that needs to be addressed.

In recent years, probiotics have become a research hotspot. Probiotics are described as living microorganisms, which, in sufficient quantities, provide health benefits to the host (20). Probiotics have been used as preventive and curative therapy for multiple illnesses, including diabetes (21, 22), hypothyroidism (23), Hashimoto’s thyroiditis (24), and osteoporosis (25, 26). In particular, their effects on osteoporosis are even more far-reaching (25, 27, 28). Probiotic preparations are live bacterial preparations composed of probiotics or probiotic growth-promoting substances that confer health benefits to the host, using microbiological principles (29). Among them, Bifidobacterium, Lactobacillus, Escherichia, Enterococcus, Bacillus and Streptococcus are the most commonly bacteria used in probiotic preparations (30–32). In particular, Bifidobacterium inhibits osteoclast formation to ameliorate osteoporosis. Several research and clinical studies have demonstrated that, probiotics, despite their disease-prevention and treatment effects, are not absolutely safe or without side effects (33). Probiotics may be an occasional risk factor for sepsis (34). However, in general, the benefits of probiotics outweigh the disadvantages, especially the lower incidence of adverse events in Bifidobacterium therapy (34).

RANKL/RANK/OPG has been extensively corroborated as a classical pathway for regulating osteoclast formation, but its relationship with the TNF-α/NF-κB signaling pathway has elicited great interest in recent years. Bifidobacterium, a probiotic of the intestine, has profound effects on the TNF-α/NF-κB inflammatory pathway. Its secretions, including short-chain fatty acids, equol, and exopolysaccharides have distinct effects on the aforementioned inflammatory pathway. Among them, short-chain fatty acids and equol have been extensively demonstrated to exert inhibitory effects on osteoclast formation independent of this signaling pathway. This review provides an overview of the specific mechanisms of Bifidobacterium inhibition of the TNF-α/NF-κB inflammatory pathway to affect osteoclast formation. Through this review, we attempt to provide researchers with new insights into potential targets for the development of effective therapies for osteoporosis.

TNF-α/NF-κB is a well-known inflammatory signaling pathway (35) that is implicated in the development of endocrine system illnesses, particularly osteoporosis (36, 37). The tumor necrosis factor (TNF) superfamily molecules are mostly produced by macrophages (38). Among them, TNF-α is a vigorous pro-inflammatory cytokine with a crucial role in immune function, inflammation, and regulation of cell growth, differentiation, and apoptosis (39). TNF-α requires cell surface receptors tumor necrosis factor receptor 1 (TNFR1) and tumor necrosis factor receptor 2 (TNFR2) to exert its biological effects (40). Whereas TNFR1 and TNFR2 have extracellular domains enriched with cysteine, their intracellular domains are structurally very different. Notably, TNFR1 contains a conserved 80-amino acid sequence called the cytoplasmic “death domain,” which produces a characteristic fold (41). Through this death structural domain, TNFR1 can sequentially recruit tumor necrosis factor receptor-associated death domain protein (TRADD), TNFR-associated factor 2 (TRAF2), receptor-interacting protein, and nuclear factor-κB (NF-κB) kinase inhibitor (IKK), thereby activating NF-κB (40). Contrarily, TNFR2 does not have a cytoplasmic death region sequence and recruits TNFR-associated factor 1 (TRAF1) and TNFR-associated factor 2 (TRAF2), but not TRADD (42). Despite this difference, the signaling cascades downstream of TNFR1-TRADD-TRAF2 and TNFR2-TRAF2 are similar.

NF-κB is a homodimeric and heterodimeric complex composed of five members of the Rel family, including NF-kB1 (p50), NF-kB2 (p52), RelA (p65), RelB, and c-Rel (43). These factors regulate the expression of several genes involved in immune response and numerous other cellular processes, including growth, development, and apoptosis (44). In most unstimulated cells, IkB proteins are maintained inactive in the cytoplasm by interacting with NF-kB dimers (45). Kappa B inhibitor kinase (IKK) is a heterotrimeric enzyme made up of the kinase subunits IKKa and IKKb as well as the regulatory subunit IKKγ/NEMO (46). When activated, IKK phosphorylates and degrades two important serine residues in the N-terminal regulatory domain of the NF-κB inhibitor IkB, releasing NF-κB (47). After its release, NF-κB generates cytokines such as p50 and p65, which promote its translocation to the nucleus to revive transcription (47). When NF-κB p50 and p52 are expressed, RANKL-RANK induces osteoclastogenesis (36). NF-κB promotes the activation of c-Fos (48), a member of the Fos gene family, which together with Jun proteins make up the AP-1 family of heterodimeric transcription factors (49). Without c-Fos, osteoclasts cannot develop (50). Boyce et al. demonstrated that c-Fos primarily generates and interacts with NFATc1 to initiate a transcriptional regulatory cascade, which results in upregulation of several target genes involved in osteoclast development and function (51). The TNF-α/NF-κB signaling pathway is in Figure 1 .

RANKL/RANK/OPG is the predominant signaling pathway regulating osteoclast differentiation (52). The differentiation of osteoclasts primarily involves the fusion of monocytes to multinucleated osteoclasts in response to three cytokines: macrophage colony-stimulating factor (M-CSF), nuclear factor-κB (NF-κB) ligand receptor activator (RANKL) and osteoprotegerin (OPG) (25). M-CSF (also known as colony-stimulating factor-1) regulates mononuclear phagocyte production, through a process mediated by CSF-1 receptor (CSF-1R), which is encoded by the c-FMS proto-oncogene (53). As a dimeric cytokine, it modulates the formation of many different types of cells, such as trophoblasts, macrophages, and osteoclasts (54). In the early stage of osteoclast development, M-CSF binds to c-FMS expressed on precursor cells activating their proliferation (53). A type II membrane protein with close homology to the TNFSF members TRAIL, FasL, and TNF-a (55), RANKL is expressed on osteoblasts as a membrane-associated cytokine (56). In vitro, RANKL activates mature osteoclasts in a dose-dependent manner, but can activate pre-existing osteoclasts quickly to cause bone resorption in vivo (57, 58). Additionally, it has been demonstrated that M-CSF and RANKL promote the differentiation of osteoclast precursor cells into mature and functional osteoclasts (56, 59). Collectively, CSF-1 and RANKL can stimulate the expression of genes that characterize the osteoclast lineage, including those that encode tartrate-resistant acid phosphatase, cathepsin K, and calcitonin receptor, resulting in the maturation of osteoclasts (60). Through intercellular contacts, osteoblasts express RANKL, which is recognized and binds to osteoclast precursors, which then develop into osteoclasts in the presence of M-CSF (61). Meanwhile, M-CSF strongly promotes the binding of RANK to RANKL and the formation of osteoclasts (62).

Osteoprotegerin (OPG) is a cytokine receptor protein produced by osteoblasts (63). It acts as a decoy receptor by binding to RANKL to block its interaction with its functional receptor RANK, thereby inhibiting osteoclast formation (64). OPG has also been found to cause osteoclast pseudopod disassembly and safeguard the bone cortex via pathways like the Ca-p38-MAPK signaling pathway, inhibit RANKL binding to RANK, prevent osteoblast-induced osteoclast precursor cell differentiation, and control osteoclast function (65). Li et al. demonstrated that cytokines (such as OPG and RANKL) directly interact with bone regulating proteins to enhance bone homeostasis (66). Another study confirmed that dexamethasone-induced osteoporosis can be improved by restoring OPG expression by decreasing RANKL expression (67). Hence, the ratio of OPG/RANKL determines the degree of bone resorption and the course of bone metabolism.

Inflammation is closely linked to osteoporosis. TNF-α is a potent pro-inflammatory cytokine (39), and IL-6 is a “classical” bone resorption pro-inflammatory cytokine (68). It has been demonstrated that IL-1, a pro-inflammatory cytokine, promotes osteoclast production, which in turn stimulates bone resorption (69). TNF-α stimulates inflammatory cytokine mRNA transcription, which results in the production of IL-6 (70). It has been discovered that recombinant human tumor necrosis factor (rTNF-α) naturally induces IL-1 in the body (71). Furthermore, IL-1 can induce the expression of TNF-α via an autocrine mechanism (72) and IL-1 induces the production of IL-6 (73). Thus, inflammatory cytokines can not only promote bone resorption alone, but their mutual activation can enhance the activation of TNF-α/NF-κB signaling pathway, activate osteoclast-related genes, and enhance bone resorption, which can be seriously detrimental to osteoporotic patients. The cytokines TNF-α, IL-6, and IL-1 cause a significant augmentation of osteoclasts and a suppression of osteoblast activity when RANKL is present (74). IL-17 is an another pro-inflammatory cytokine that promotes bone resorption via upregulating RANKL (75). IL-6 trans-signaling directly increases RANKL on fibroblast-like synovial cells and is involved in the induction of RANKL by TNF and IL-17 (76). In addition, IL-6 and TNF-α can synergistically activate NF-κB (77). Ciucci et al. further ascertained that bone marrow CD4⁺ T cells belong to a distinct subpopulation of osteoclastic T cells termed Th17 TNF-α (+) cells that can generate IL-17 and TNF-α (78). These cells move to the bone marrow amid chronic inflammation, where they facilitate the recruitment of inflammatory monocytes (mainly osteoclast progenitors) (78). In an inflammatory event, immune system cells, such as T cells, B cells, macrophages, and dendritic cells, become activated and release inflammatory cytokines, which are among the most crucial mediators in bone immunology (74). Activated T cells are particularly significant mediators because they increase the production of the so-called bone resorbing cytokines, including TNF- and RANKL (74). Thus, the formation of osteoclasts is closely associated with chronic inflammation. In addition to stimulating osteoclast formation through the NF-κB signaling pathway, TNF-α can also mediate RANK ligands activation of osteoclast formation via an autocrine mechanism (79). The combination of TNF-α and RANKL greatly stimulated osteoclast formation and significantly up-regulated osteoclast mRNA markers (37).

To sum up, inhibiting the TNF-α/NF-κB inflammatory pathway can impede the formation of osteoclasts. Multiple studies (80, 81) have demonstrated a strong link between inflammation and osteoclast formation. The important role of the NK-kB transcription factor family in inflammation and innate immunity has also been elucidated (82, 83). Inhibition of osteoclast formation via the NF-kB pathway has also been reported. For example, preparations of Zanthoxylum piperitum (84), Sophorae flos ( 85), and Bajijiasu ( 86), were shown to inhibit the RANKL-induced NF-κB/NFATc1 pathway in osteoblasts to hinder bone resorption. As a pro-inflammatory cytokine, TNF-α promotes the production of osteoclasts by activating the NF-κB pathway, synergizing RANKL cytokines, and facilitating and enhancing RANK-RANKL binding. In periodontitis, down-regulating TNF-α, alveolar bone loss was delayed (87). In addition, Yao et al. explained the regulation of TNF-α-induced osteoclast formation (88). The effect of RANKL/RANK/OPG on osteoclasts has also been confirmed by many investigators (52, 89, 90). TNF-α/NF-κB signaling pathway and RANKL/RANK/OPG signaling pathway induced-osteoclast have been shown in Figure 2 .

In recent years, the impact of probiotics on bone health has become a hotspot of research. Current research suggests that probiotics regulate bone metabolism through different mechanisms, including intestinal barrier permeability (91), metabolite production (92), the immune response (93), and inflammation (94). Postmenopausal bone mineral density loss, which is associated with estrogen deficiency (95), was effectively reduced by probiotic supplementation, which also improved bone turnover in 78 postmenopausal patients with osteoporosis for more than a year (96). In another trial, probiotics significantly increased total hip bone density compared with placebo and modulated gut microbiota in postmenopausal women (97). In response to the aforementioned pathways, recent reports show that probiotics, particularly Bifidobacterium, have more benefits for bone health ( Table 1 ).

The most prevalent phyla of the human gut microbiota are Firmicutes, Bacteroidetes, and Actinobacteria (105). The phyla Firmicutes and Bacteroidetes collectively account for 90% of colonic microbiota (60–75% and 30–40%, respectively) (106, 107). By comparison, the phylum Actinomycetes is a smaller but essential component for preserving intestinal homeostasis (108). Bifidobacterium, which belongs to the phylum Actinobacteria (105), was first isolated from the feces of healthy breastfed infants by a French pediatrician Tissier and named Bifidobacterium because of its commonly bifurcated ends (109). Bifidobacterium can produce short-chain fatty acids (SCFA) (110), equol (Eq) (111), exopolysaccharides (112) and many other substances that can affect osteoclast formation, which is shown in Figure 3 .

Osteoporosis has become a major global public health problem, with a significant economic burden on health care systems. It is a bone disease that is characterized by low bone mass and microstructural degradation, which promote bone fragility and, consequently, increase the fracture risk (158). In previous studies (25, 159), the therapeutic potential of probiotics in bone health has been demonstrated, including their positive effect on osteoporosis. However, few studies have examined the relationship between specific genera of bacteria and osteoporosis, and certainly even fewer articles have discussed their specific mechanisms of action. Therefore, in this paper, we sought to fill this gap by reviewing the mechanisms of Bifidobacterium inhibition of the TNF-α/NF-κB inflammatory pathway to prevent osteoclast formation. In germ-free mice, higher bone mass is linked to changes in the immunological state, which is reflected by decreased expression of inflammatory cytokines in bone (25). Bifidobacterium can be used to inhibit the formation of osteoclasts by altering the inflammatory immune status of bones through its secretory products.

SCFAs can directly inhibit the classical pathway of inflammation, the TNF-α/NF-κB signaling pathway, to strongly inhibit inflammation. It can also promote the secretion of Tregs cells, regulate inflammation through the immune system, increase the expression of intestinal tight junction proteins, block the invasion of harmful substances through the intestinal mucosa, and effectively control inflammatory response. In the immune system, regulatory T cells (Treg cells) expressing the transcription factor Foxp3 have been shown to act as inhibitors of inflammatory response in the gut, and helper T cells 17 (Th17) are pro-inflammatory cells (160). Under inflammatory conditions, Foxp3 expression on Treg cells is lost, and this allows transdifferentiate of the cells into Th17 cells (161). The balance between Th17 cells and Treg cells influences the pathogenesis of osteoporosis. Increased Th17 cell frequency has been linked to the occurrence of bone resorption (162). Th17 also secretes high quantities of IL-17, NF-kB ligand receptor activator (RANKL), and TNF as well as low levels of interferon gamma, making it the most osteolytic subpopulation of T CD4⁺ cells (IFNγ) (163, 164). A subpopulation of immune cells known as Treg cells inhibits the differentiation and functionality of Th17 cells (165). In addition, Treg cells can inhibit OC differentiation and bone resorption by releasing TGF-β1 and IL-10 (166). It has been reported that enrichment of SCFA-producing probiotics downregulates intestinal epithelial permeability and restores the Treg/Th17 cell ratio (27). Butyric acid, in particular, is implicated in the control of Treg/Th17 balance and prevents the formation of inflammation in colonic mucosa (167). This protective effect on the intestinal mucosa contributes to the reduction in inflammation and formation of osteoclasts. These effects are mediated by the TNF-α/NF-κB inflammatory pathway. For instance, Th17 secrete high levels of NF-kB ligand receptor activator (RANKL) and TNF, and Treg inhibits Th17 cells thereby indirectly regulating the TNF-α/NF-κB signaling inflammatory pathway. Treg also inhibits osteoclast formation, and further studies are needed to expand our understanding on this. High concentration of SCFAs can inhibit bone growth. By blocking osteoblast-specific factors, high dose of sodium butyrate prevents the differentiation and mineralization of the ROS17/2.8 rat osteoblast line (168). Therefore, further investigations are advocated to determine the optimal concentration of SCFAs for the treatment of osteoporosis. On the side, butyrate also has the ability to modulate bone anabolic metabolism through Treg cell-mediated generation of Wnt10b from CD8+ T cells (169). This demonstrates the ability of short-chain fatty acids in bone formation, and perhaps Bifidobacterium could be drugs that have the dual effect of stimulating bone formation and inhibiting bone resorption.

There have been several reports about the effect of SCFAs on osteoclast formation (128, 136, 139). However, few scholars have explored the effect of equol and exopolysaccharides on osteoclast formation. Currently, few microorganisms have been identified to produce equol, and Bifidobacterium is one of them (111). Equol can inhibits the TNF-α/NF-κB pathway to exert anti-inflammatory effects and reduce the pro-osteoclastogenic function of inflammatory cytokines (145–148, 170). Ovariectomy model (OVX) mice showed significantly reduced bone mineral density (BMD) and bone mineral content (BMC) compared with sham-operated animals, and 0.5 mg/d Eq treatment preserved bone mass (149). Furthermore, it has stronger estrogenic activity and may be a potential agent for treatment of postmenopausal osteoporosis (171). Equol is produced in the gastrointestinal system by soy glycosides, however, its metabolism in humans differs among individuals (172). This may alter the efficacy of the drug leading to different responses in various patients with osteoporosis, but this concept has not been sufficiently studied. Therefore, further clinical evaluation and analysis is still needed. In vitro experiments have demonstrated that EPS can potentially prevent osteoclast formation, however, the optimal way to use EPS in humans that guarantees stable inhibition of osteoclast formation needs to be further explored (154).

Currently, Bifidobacterium is now an important product in the market. Compared with other microbial workhorses, engineered Bifidobacterium’s produces numerous bioproducts with additional benefit while using fewer resources (173). The Bifid shunt, which generates higher number of CoA predecessors for the bioproduction of polyketide products and fatty acid biosynthesis, is one of the crucial metabolic processes in Bifidobacterium (174). Although clostridia can also produce SCFAs and Eq (111, 175), it is thought to cause pathogenicity (176). Safety assessment of Bifidobacterium species identified only 2 cases of mild functional intestinal obstruction (177) and 1 case of sepsis (178), and these results demonstrate that probiotic preparations possess pathogenic risks. Therefore, there is need to balance between the risk and the cost-benefit and safety in the clinical treatment of patients to reduce the incidence of adverse events.

Inflammatory bowel disease (IBD) has been linked to increased risk of bone mineral loss and osteoporosis (179). Bifidobacterium lactis BL-99 can be used to prevent the development of osteoporosis in patients with ulcerative colitis (UC) by shaping the intestinal flora and inhibiting the production of inflammatory cytokines (180). Probiotics have shown potential for the treatment of IBD, and therefore, the authors suggest that fecal transplants could be used in the future to regulate intestinal flora to improve the symptoms of IBD (181).. Studies have demonstrated that extensive changes occur in the structure of the intestinal flora of rats after ovariectomy (182). Compared with healthy individuals, patients with osteoporosis or bone loss showed significant changes in gut microbial species (183). This suggested a correlation between osteoporosis and the constitution as well as functionality of the intestinal flora. It is possible that fecal colony transplantation technique may be an effective treatment for patients with osteoporosis. However, FMT is processed by collecting therapeutic stools from normal individuals and its treatment success depends largely on the quality of the donor’s gut microbes. Therefore, appropriate selection of donors is crucial. Nevertheless, there are many challenges affecting the adoption of this technology (184). Further animal experiments and clinical studies are needed to clarify this.

In this review, we summarized the mechanisms by which Bifidobacterium bifidum regulates osteoclast formation by inhibiting the TNF-α/NF-κB inflammatory pathway. Its effects are mediated by its secreted products, including short-chain fatty acids, equol, and exopolysaccharides.

The role of probiotics in osteoporosis is increasingly being studied. Activation of inflammatory factors associated with osteoporosis such as TNF-α, NF-κB, IL-1, IL-6, and IL-17 has been to be involved in the physiology of pro-osteoclast formation. Inhibition of the TNF-α/NF-κB signaling pathway prevented nuclear transfer of NF-κB and blocked the transcription of regulatory proteins associated with osteoclasts. Bifidobacterium secretions block the initiation of inflammation and inhibits osteoclast formation to improve osteoporosis symptoms. In addition, the Bifidobacterium secretions can regulate RANKL/RANK/OPG, the most important signaling pathway of osteoclasts, via the TNF-α/NF-κB signaling pathway, which block the synergistic effect of TNF-α on RANKL and reduce the binding of RANK to RANKL to regulate the formation of osteoclasts. In this way, it prevents bone loss caused by the bone resorption induced by osteoclasts. Although clinically effective osteoporosis treatment drugs are available, their safety and efficacy are not satisfactory, especially in patients with severe osteoporosis. Therefore, there is a need to actively search for more effective treatments with fewer side effects and more cost-effective. The data described in this review demonstrated that Bifidobacterium might be a good treatment agent.

YW and YY conceived the study and finished drafting the article. LW and YC polished up the article. XH, LS, and HC processed the table and pictures. QC provided guidance and resolved disagreements. All authors approved it for publication.