IFN-β belongs to type I interferons. The human body produces three known types of interferons: type I, type II, and type III (55). Type I interferons are mainly IFN-α and IFN-β, secreted by innate immune cells. Type II interferons, IFN-γ, is mainly produced by activated T cells. Type III interferons include IFN-λ, whose known distribution and function are minimal. Type I interferons are mainly produced by surface or internal receptors of innate immune cells (TLRs, NLRs, RLRs, and cGAS) binding to specific antigens from outside or inside the host (56). IFN receptors (IFNAR) are a class of heterodimers located on the cell membrane and consist of two subunits, IFNAR1 and IFNAR2, and widely distributed, including monocytes, macrophages, B cells, T cells, epithelial cells, endothelial cells, and tumor cells (57). Ligand receptor binding activates downstream protein kinases JAK1 and TYK2, and kinase activation activates cytoplasmic transcription factors STAT1 and STAT2, forming a dimer that enters the nucleus to assist IRF9 in transcribing some downstream effector genes (56). Type I interferons can play a biological role in antiviral and immunomodulatory, inhibiting specific cell growth and proliferation and killing tumor cells (58, 59). Therefore, interferon therapy has been used to treat common viral diseases such as hepatitis (60) and various cancers (61).

When osteoclasts are induced to produce IFN-β, the binding of IFN-β to its bioreceptor activates ISGF-3 (formed by the aggregation of STAT1, STAT2, and IRF9) via the classical JAK/STAT pathway, initiating a signal transduction cascade (62). Then, c-Fos will be inhibited, leading to the inhibition of osteoclast production and activity (11). Thereby, IFN-β forms negative feedback of its own (Figure 3). Thus, IFN-β also plays a vital role in regulating bone homeostasis. In addition, osteoclasts express iNOS and release NO, and the NO produced by this pathway also acts as a negative feedback signal to limit RANKL-stimulated osteoclastogenesis (63). In iNOS-deficient bone marrow cells, RANKL-induced NO production was inhibited, leading to an increase in the number of terminally differentiated osteoblasts (64). Direct administration of IFN-β in RAW264.7 cells stimulated iNOS expression in the absence of RANKL, thereby upregulating NO expression. NO, like IFN-β, inhibited osteoclast differentiation. These results suggest that IFN-β may be a key mediator of iNOS-derived NO induction by RANKL in developing osteoclasts and that iNOS can mediate the inhibitory effect of IFN-β on osteoclasts (65).

Another interaction of IFN-β involved in the regulation of bone homeostasis is 4-1BBL with 4-1BB. 4-1BB, also known as CD137, is similar to RANK and is a member of the same TNF receptor family, encoded by the TNFRSF9 gene. Upon its activation, antigen-presenting cells, such as dendritic cells, B cells, and macrophages, express 4-1BBL (66). Osteoclast precursors can express 4-1BB and 4- BBL after exposure to RANKL (67). In BMMs co-stimulated by M-CSF and RANKL, 4-1BBL mRNAs are upregulated (68). In the animal model, 4-1BB knockout mice also showed increased bone mass compared to the wild group. The number of osteoclasts was significantly reduced in the presence of immobilized recombinant 4-1BB (4-1BB-Fc). 4-1BB can induce the binding activity of IRF3, and IRF3 is activated by 4-1BB stimulation, which induces IFN-β (11, 69, 70). It is not difficult to speculate that the decrease in osteoclast activity caused by 4-1BB should be due to the inhibition of c-Fos expression by IFN-β.

As a critical signal transduction molecule involved in the innate immune response, STING, triggered by cytoplasmic DNA from pathogens and hosts, can induce type I interferon and pro-inflammatory cytokine secretion, defend against viral and intracellular bacterial infections, and regulate the spontaneous anti-tumor immune response in vivo. Targeted STING is a new tool for immunotherapy. In addition to immune or oncological diseases, the role of STING in bone metabolic diseases has been the focus of attention in recent years. DNase II is a nuclease that degrades dsDNA. Lack of DNase II causes DNA accumulation in cells and produces several cytokines, including type I IFN (71). Mice lacking DNase II and IFNAR were able to develop distal aggressive inflammatory arthritis (72). However, this arthritis was eliminated in the absence of STING (71, 73). Surprisingly, the arthritis model (DNase II^-/-/IFNAR^-/- double-knockout mice) showed aberrant accumulation of bone in both long bones and the spleen at sites of local DNA accumulation (13). STING deficiency inhibited bone accumulation (13), revealing a potential role of the STING pathway in bone, although the exact mechanism is unclear.

Patients with advanced cancer often suffer from severe pain due to bone metastases and bone destruction with osteolytic lesions (74). Agonists of the immunomodulator STING have significantly protected against pain (75), bone destruction (13, 19), and local tumor burden (76). One of its effects is alleviating cancerous bone pain by regulating osteoclast function in the tumor microenvironment to prevent local bone destruction, which depends on host-intrinsic STING/IFN-β signaling. Bone metastases in patients with cancer produce osteolytic bone destruction due to tumor-induced osteoclast formation and activation (77). Bone loss was significantly reduced in Lewis lung carcinoma (LLC) or breast cancer mice treated with DMXAA and ADU-S100, two different STING agonists, similar to zoledronic acid (ZA) (18). In contrast, the reversal effect of DMXAA on bone loss was eliminated in the STING knockout group of mice. Thus, the inhibitory effect of STING agonists on bone resorption is dependent on STING. Systemic administration of STING agonists also promotes a robust IFN-I response in the systemic and bone cancer tumor microenvironment. DMXAA treatment does not prevent bone destruction in IFNAR1-deficient mice. IFNAR is required for IFN signaling (57). Thus, the protective effects of STING agonists against cancerous bone destruction require IFN-I signaling (18).

STING, also known as Tmem173, has been shown to inhibit osteoclast differentiation and activity by regulating IFN-β production. It inhibits the expression of osteoclast-specific genes and related enzymes and downregulates the activation of osteoclast-specific transcription factors (20). CDNs are symbiotic bacterial-derived second messengers in the intestine that regulates bacterial survival, colonization, and biofilm formation and have immunomodulatory activity by inducing type I interferon expression by macrophages through the STING signaling pathway (34, 78). CDNs dose-dependently inhibit M-CSF and RANKL-induced differentiation of bone marrow macrophages to osteoclasts and induce phosphorylation of TBK1 and IRF3, representative features of STING activation (19). In contrast, inhibition of osteoclast differentiation was reversed in STING knockdown BMMs. These suggest that the STING signaling pathway plays a crucial role in CDNs-mediated inhibition of osteoclast differentiation. In addition, CDNs increased the expression of IFN-β, a member of the IFN-I family, which has also been identified as a typical negative regulator of RANKL-induced osteoclast differentiation (79, 80). RANKL induces IFN-β expression via c-Fos (81). In turn, IFN-β binds to IFNAR on the membrane, which activates ISGF-3 and prevents RANKL-induced c-Fos expression from inhibiting osteoclast differentiation (11). The inhibitory effect of CDNs on osteoclast differentiation was absent in the presence of antibodies blocking IFNAR, and no inhibitory effect was observed in knockout IFNAR macrophages (19). These also confirm that CDNs induce phosphorylation of STAT1, which mediates IFNAR signaling. Experiments performed with a mouse cranial implant model showed that CDNs inhibit RANKL-induced bone resorption (19). These results suggest STING induces IFN-β to inhibit osteoclast differentiation and bone resorption.

It is well known that IFN-I response is a weapon against viruses. IFN-I is induced during STING-mediated immune responses. IFN-I can also be stimulated by the osteoclast-specific gene c-Fos and ultimately inhibits osteoclast production and activation through IFNAR transmission (11, 82). STING regulates the inhibitory effect of IFN-I on osteoclasts, and the knockdown of STING reverses this effect (79). Knockdown of STING can offset this effect (19). Therefore, targeting STING/IFN-β to increase IFN-β expression and thereby inhibit osteoclast bone resorption is expected to be a new approach to treating osteoporosis. In addition, interferon therapy has been used in the clinic. And IFN-β has a relatively good clinical tolerability and safety profile. However, it faces several tests when using IFN-β to treat bone metabolic diseases, including osteoporosis. First, as with other protein drugs, treatment with interferon results in the production of neutralizing antibodies in the patient (83). Second, effective delivery of the drug to the bone microenvironment is another challenge in using IFN-β for treating bone metabolic diseases (84). Targeting STING to increase the level of interferon in the body may solve these problems faced by treatment with interferon alone. In addition, the inhibitory effect of IFN-β on osteoclasts is mainly due to its negative feedback mechanism. However, the action of IFN-β is also inhibited by another kind of negative feedback. Suppressors of cytokine signaling (SOCS)-1 and SOCS-3 in response to RANKL can act as inhibitory factors that significantly inhibit IFN-β signaling (85). Thus IFN-β-mediated inhibition of osteoclastogenesis has a potential counteracting pathway. It may be the same challenge for targeting STING/IFN-β with interferon therapy alone.

NF-κB is one of the best-characterized transcription factors that regulate inflammation and innate and adaptive immune responses (110). Activation of NF-κB signaling leads to the production of various inflammatory cytokines, chemokines, adhesion molecules, transcription factors, and antimicrobial effector molecules that initiate and mimic inflammatory responses and coordinate the immediate host response to pathogens and tissue damage. The NF-κB transcription factor family includes five members p50 (NF-κB1), p52 (NF-κB2), RelA (p65), RelB, and c-Rel (111). All NF-κB subunits have a structurally conserved N-terminal sequence spanning 300 amino acid residues called the Rel homology structural domain (RHD) (112). The RHD is responsible for DNA binding, dimerization, and nuclear translocation of NF-κB subunits, which can divide into three structural components - the N-terminal structural domain (NTD), the dimerization structural domain (DD), and the nuclear localization sequence (NLS) polypeptide - all of which mediate the various activities of the RHD and subsequent NF-κB signaling (110, 113, 114). Subunits RelA, RelB, and c-Rel are produced as mature proteins. In contrast, the p50 and p52 subunits are produced by the precursor proteins (113).

In the resting state, NF-κB subunits bind to IκB proteins, inhibiting their activity and maintaining NF-κB subunits in an inactive state (115). In turn, IκB kinase (IKK) can degrade these inhibitory proteins (116). Once activated by upstream signaling cascades, phosphorylated IKK degrades the IκB protein and releases the subunits of the NF-κB. Then these subunits go into the nucleus as dimers and participate in the transcription of various target genes (10). For example, the functional subunits p65 and p50 enter the nucleus and bind to target genes, producing large amounts of inflammatory mediators, and the gene products further activate NF-кB, causing an expanded cascade of uncontrolled inflammatory responses. More and more NF-κB target genes have been identified (117), including various cytokines such as interleukin (IL) and TNF, interferons, and antiapoptotic proteins, such as BIRC2, BIRC3, and BCL2L1.

Bone homeostasis is necessary for the maintenance of normal bone function. Bone homeostasis is maintained by bone remodeling mediated by osteoblasts and osteoclasts, which are responsible for bone formation and resorption (48). Osteoblasts and osteoclasts are the essential cells that regulate bone homeostasis. NF-κB is the master transcription factor that regulates the inflammatory response and bone remodeling process (118, 119). Chronic inflammation induces excess pro-inflammatory cytokines, disrupting homeostasis (120). These result in abnormal bone remodeling, including osteosclerotic and osteolytic lesions (121).

Pro-inflammatory cytokines driven by NF-κB are powerful signals to regulate bone homeostasis (122). Elevated expression of TNF, IL-1, IL-6, and IL-7 has been found in various chronic inflammatory bone diseases, including osteoarthritis (123), osteoporosis (124), and periodontal disease (125). These pro-inflammatory cytokines are all produced by macrophages, lymphocytes, osteoblasts, and bone marrow stromal cells under the regulation of NF-κB and stimulate NF-κB signaling in target cells, which further serves to amplify inflammation (126). Osteoclasts are specialized cells of the monocyte-macrophage lineage responsible for bone resorption. In contrast, osteoblasts are differentiated from mesenchymal stem cells to osteogenesis and are responsible for establishing new bone. NF-κB has an essential role in osteoblasts and osteoclasts, thus affecting bone regulation.

Excessive accumulation of ROS leading to redox imbalance and overactive osteoclasts is associated with the progression of osteoporosis. The process of osteoclastogenesis is accompanied by the production of ROS, which plays a role in osteoclast differentiation and bone resorption (52). In addition, ROS-induced mtDNA release induces inflammation through the activation of cGAS/STING (148). ROS can induce NF-κB through the RANKL/RANK cascade reaction, which is further involved in osteoclastogenesis (149). In addition to promoting type I interferon production in the innate immune response (150), STING can also act as an NF-κB upstream of NF-κB, stimulating its production.

In the past, it was generally considered that NF-κB activation via STING is exclusively dependent on TBK1. However, studies have now demonstrated that TBK1 is dispensable for NF-κB, although TBK1 and its kinase activity are essential for STING-dependent IRF3 activation and INF-I. In fact, TBK1 and IKK redundantly drive NF-κB activation when the IFN-I reaction is triggered by TBK1 and its homolog, IKK (151). Inhibition by TBK1/IKK kinase indicates that IRF3 activation highly depends on TBK1 kinase activity. In contrast, NF-κB sensitivity to TBK1/IKK kinase inhibition is significantly reduced, and TBK1 was dispensable for NF-κB activation downstream of STING in vitro and in vivo ( 151). So, NF-κB production can be activated by STING through a non-classical pathway, a process that is independent of IFN-β (44). In addition, non-classical STING signaling activates NF-κB pathways mainly through K63-mediated ubiquitination and has no effect on IFN-I (38).

CDNs can inhibit osteoclast differentiation by inducing IFN-β through STING signaling (19), suggesting that activation of STING can inhibit osteoclast differentiation through IFN-β. A sustained activation state of STING can cause a series of inflammatory responses in the organism. NF-κB, acting as a pro-inflammatory gene, is involved in osteoclastogenesis as another downstream of STING. In contrast to NF-κB, nuclear factor erythroid2-related factor 2 (Nrf2), a critical antioxidant molecule, has been shown to inhibit osteoclast formation and bone resorption by reducing ROS (152). In addition, Nrf2 negatively regulates STING signaling (153).

RTA-408 was found to act as an activator of Nrf2 that inhibits STING expression and subsequent NF-κB activation but does not affect IFN-β expression (21). RTA-408 inhibits RANKL-induced K63 ubiquitination of STING by suppressing the interaction between STING and the E3 ubiquitin ligase TRAF6. As a downstream of STING, NF-κB was also inhibited by RTA-408, mainly by suppressing IκBα protein degradation, preventing p65 from translocating to the nucleus and thus rendering NF-κB inactive. Overexpression of STING rescued the inhibitory effect of RTA-408 on NF-κB signaling and osteoclastogenesis. In vivo experiments showed that RTA-408 attenuated osteoclastogenesis-induced bone loss in C57BL/6 mice by inhibiting STING-mediated NF-κB (21). Thus, inhibition of STING-dependent NF-κB signaling could inhibit osteoclastogenesis and reduce bone loss. Targeting STING/NF-κB may be a promising pathway for the future treatment of osteoporosis.

Osteogenesis is linked to angiogenesis (156). The close spatial and temporal link between osteogenesis and angiogenesis has been termed “angiogenesis-osteogenesis coupling” (157). Type H vessels are located near the epiphyseal growth plate, the epiphyseal periosteum, and the endosteum. Type H vessels are densely surrounded by osteoprogenitors expressing Osterix, a potent promoter of bone formation (158). These osteoprogenitor cells can differentiate into osteoblasts and osteocytes. Under aging conditions, osteoblasts are significantly reduced in the long bones of mice (159), which is associated with a decrease in Type H vessels and reduced bone mass (160). The abundance of Type H vessels is an essential indicator of bone loss in elderly subjects and patients with osteoporosis (161).

PDGF-BB is a chemotactic and mitogenic factor of the PDGF family, produced by hematopoietic stem cells (162). It is essential for promoting the migration, proliferation, and differentiation of various mesenchymal cell types, such as endothelial progenitor cells and mesenchymal stem cells, to promote angiogenesis and osteogenesis (163, 164). PDGF-BB enhanced type H vessels and bone formation during bone plastination and remodeling. The concentration of PDGF-BB was decreased in the OVX mouse model (165). Pharmacological stimulation was able to secrete PDGF-BB to stimulate H-type angiogenesis, thereby promoting osteogenesis to prevent bone loss in OVX mice (166). Glucocorticoids reduce vascularity and blood flow to the bone, causing osteonecrosis and bone loss (167, 168). Glucocorticoid-induced osteoporosis (GIO) is also common osteoporosis. In GIO mouse models, glucocorticoids inhibit PDGF-BB secretion by pre-osteoblasts, inhibiting Type H vessels and reducing osteogenic capacity (169). And L-235, a cathepsin K inhibitor, prevents bone loss by inhibiting osteoclast-inducing bone resorption while maintaining PDGF-BB secreted by preosteoclasts preserving Type H vessels (169).

HIF-1α is a transcription factor that mediates the cellular response to an altered oxygen environment and controls angiogenesis (170).HIF-1α plays a crucial role in bone formation, regeneration plays a key role in bone formation and regeneration, and its expression and activity are regulated by hypoxia Oxygen is required for the high metabolic demand of osteoblasts. Therefore, osteoblasts and nearby ECs may increase HIF-1α expression during relative hypoxia during osteogenesis (155). And ECs express HIF-1α at high levels in young mice, which decreases with age and is associated with a decrease in ECs and age-dependent bone loss. Activation of hypoxic signaling in ECs increased the number of Type H vessels and enhanced angiogenesis and osteogenesis (171). EC-specific deletion of HIF-1α resulted in a significant decrease in osteoblast producers and was associated with reduced trabecular formation. Thus, HIF-1α signaling is vital in regulating type H vascular abundance and couples angiogenesis to osteogenesis (172). Tetramethylpyrazine activates the AMPK/mTOR/HIF-1α signaling pathway to induce type H vessel angiogenesis and improve bone homeostasis in aging mice (173). This provides an additional therapeutic target for the treatment of age-related osteoporosis.

Type H vessels also play an indelible role in bone remodeling. Osteoblasts, osteoclasts, and periosteal cells interact with vascular endothelial cells. The STING signaling pathway may act directly on these cells, thereby affecting the angiogenic process. Several studies have shown an association between STING and angiogenesis. STING is expressed in endothelial cells, and cGAMP leads to the activation of cGAS-STING in endothelial cells (174, 175). STING-associated vascular disease (SAVI), with onset in infancy, is an autoinflammatory disease caused by mutations in STING function, which can cause vascular and pulmonary syndromes and cause systemic inflammatory responses (176, 177).

Previous studies have shown that STING can affect angiogenesis in multiple ways. In the zebrafish xenograft model, exogenous administration of cGAMP can activate STING-dependent STAT3, leading to the inhibition of tumor vascular proliferation and migration (178). Palmitic acid (PA) induces mtDNA into the cytoplasm by inducing mitochondrial damage, activating the cGAS-STING-IRF3 signaling (179). Activation of the cGAS-STING-IRF3 pathway dysregulates the Hippo-YAP pathway and inhibits angiogenesis. In addition, STING-IRF3 can trigger endothelial inflammation in response to PA-induced mitochondrial damage (180). In retinal microangiopathy, mtDNA drives inflammation of microvascular endothelial cells via the cGAS-STING signaling pathway (181). Inflammation in the physiological state is a protective mechanism for tissue damage and the basis for tissue repair and regeneration. Nevertheless, an excessive inflammatory response can impair the integrity of the tissue and its function. The persistent inflammatory state of the vascular endothelium leads to impaired angiogenesis and poor bone healing, which affects bone reconstruction (182).

Activation of the STING signaling pathway impairs angiogenesis, including type H vessels. In addition, the prolonged inflammatory response stimulated by the STING pathway delays the bone healing process. Activation of STING inhibits angiogenesis both in vitro and in vivo and slows the bone healing process in vivo ( 12). Conversely, inhibition of STING accelerated bone healing by enhancing type H vessel formation during coupled osteogenesis (12). Therefore, targeting STING to enhance type H vessel formation and thus promote osteogenesis provides a new idea for the treatment of osteoporosis.