Macrophages amplify the inflammatory process by releasing proinflammatory factors to remove pathogens or other foreign bodies (9, 47). During tissue injury, the local cells which are infected by pathogens and undergo necrosis or pyroptosis could release PAMPs or damage-associated molecular patterns (DAMPs) that activate inflammatory signaling in macrophages and other resident cell populations. Activated cells recruit neutrophils, monocytes, and other inflammation-related cells into the tissue by a release series of cytokines. Once the acute injury is under control, macrophages supply nutrition to the tissues where they are located by decomposing remnants and secreting growth factors and mediators, exerting their function effectively in inhibiting inflammation (47). Some macrophages, characterized by producing of growth factors, including platelet-derived growth factor (PDGF), insulin-like growth factor 1 (IGF-1), and vascular endothelial growth factor-α (VEGF-α), are associated with tissue repair and help in promoting cell proliferation and vascular development and thus alleviating local hypoxia that occurs after injury (47). They also produce soluble mediators, such as transforming growth factor-β1 (TGF-β1) that induce local and recruited fibroblasts to differentiate into myofibroblasts, thereby synthesizing extracellular matrix components and promoting wound closure (69). In the final phase of tissue repairment, monocytes and macrophages present an anti-inflammatory phenotype (47, 70). It was found that these macrophages responded to inhibitory mediators such as IL-10 in the local microenvironment, eventually leading to relieving inflammation (71, 72).

Phagocytosis mainly belongs to the function of M1 macrophages, which goes along with the proinflammatory process, although M2 macrophages also demonstrate a weak function in phagocytosis (73, 74). Stimulated by the serum from Behçet’s disease (BD) patients, monocyte-derived macrophages (MDMs) could differentiate into M1 macrophages with enhanced phagocytic capacity (75). Subsequently, M1 macrophages display an enhanced capacity for the elimination of pathogens, which largely results from their increased production of superoxide, NO, and their derivatives (76). M2 MDMs also demonstrate the ability to phagocytose escherichia coli and cancer cells. Further research found that the phagocytosis in M2 macrophages mainly owes to their surface markers, such as CD14, CD206, and CD163 (77). However, compared with M0 and M2 macrophages, LPS-treated M1 macrophages exhibited an obviously higher ability for phagocytic activity (78).

Both autophagy and phagocytosis in macrophages are lysosomal-dependent catabolic processes, by which cells can engulf and deliver cargo to the lysosomes for digestion via forming transient vesicular structures (autophagosome and phagosome) (79). Acting as scavenger cells, macrophages could phagocytize cellular debris, invading pathogens, and other apoptotic cells (80, 81). Phagocytizing dead cells in macrophages contribute to diverse immune and inflammatory signals that could also trigger intracellular autophagy in macrophages (82, 83). Other studies have revealed more relations between autophagy and phagocytosis. Autophagy promotes phagocytosis and the clearance of pathogens via the NOD-like receptor family pyrin domain containing 3 (NLRP3) inflammasome in macrophages (84). As a novel function for autophagy proteins, the LC3-associated phagocytosis pathway (LAP) is closely associated with phagocytosis in macrophages, which exerts its role in blocking proinflammatory signals upon phagocytosis of dying cells and preventing the presentation of autoantigen to other cells (85, 86). Interestingly, autophagy-deficient macrophages may boost phagocytosis through increased scavenger receptor expression (87).

Regarding the functional changes in activated macrophages, vitamin D could restore anti-inflammatory M2 macrophages in an autophagy-dependent manner (88). Similarly, ubiquitin-specific protease 19 (USP19) could inhibit inflammatory responses and promote M2 polarization by increasing autophagy flux (89). Controversially, another study reported that inhibition of autophagy could drive macrophages to the M2 phenotype (90). In AL, a recent study revealed that by activating LAP, aluminum oxide nanoscale particles (Al-n) attenuated the macrophage M1 polarization and inhibited the secretion of inflammatory factors, leading to the prevention of the AL pathogenesis induced by particles in vivo (26). Although autophagy has been shown to be involved in the regulation of macrophage polarization, evidence regarding the regulatory mechanisms is underdeveloped.

Apoptosis is the sequential death of cells via the mechanisms called programmed cell death (PCD) (91), which is characterized by morphological changes in the cellular structures together with a series of enzyme-dependent biochemical processing (92). In the resolution phase of inflammation, the infiltrating leukocytes execute the acute innate response and undergo apoptosis, subsequently cleared by phagocytic macrophages. In this course, macrophages undergo reprogramming from inflammatory to anti-inflammatory, leading to the relief of inflammation (93).

Efficient clearance of early apoptotic cells requires macrophages polarizing into the M2c phenotype (94). The capacity of M2 macrophages to uptake apoptotic cells depends on several necessary molecules, such as Mer tyrosine kinase, Axl receptor tyrosine kinase, growth arrest-specific 6 (Gas-6) (94, 95). In addition, the macrophage itself could also undergo apoptosis induced by wear particles (96). Consequently, the increase in macrophage apoptosis limits the proinflammatory function of macrophages (97, 98), seemly to be another important mechanism regarding the delayed inflammatory response in AL. The caspase-3, a key mediator related to apoptosis, was detected in periprosthetic tissues in a mouse osteolysis model induced by UHMWPE particles (99, 100). Furthermore, apoptosis is associated with phenotypic changes in macrophages. Transcription factor Zhx2 deficiency could enrich the expression of M2 phenotype markers and as well promote the apoptosis of macrophages (101). Likewise, through the inhibition of M2-specific gene expression and apoptotic cell death, Delta-like-ligand 4 (DLL4) may prevent macrophage from polarizing into the M2 phenotype (102).

Different from apoptosis, pyroptosis is a proinflammatory type of PCD mediated by the gasdermin family, which usually leads to cell swelling and rapid rupture of plasma membranes, as well as the release of immunogenic cell contents, thereby exaggerating inflammatory status (103). In the aspect of the immune response, pyroptosis induced by the activation of pattern recognition receptors (PRRs) can stimulate inflammatory responses, independent of its effect in promoting cytokine induction (104). Pyroptosis is undoubtedly related to macrophage polarization. On one hand, by secreting exosomal cathepsin S, M1 macrophages induce pyroptosis in pancreatic acinar cells via the caspase1-mediated classical pyrolysis pathway, resulting in inflammation and pancreatic tissue damage (105). On the other hand, exosomal Mir-30D-5p of polymorphonuclear neutrophils (PMNs) is reported to induce M1 polarization by upregulating TNF-α, IL-1β, IL-6 and triggering pyroptosis in macrophages, leading to sepsis-associated acute lung injury (106).

Moreover, some mutual signals could mediate both M1 polarization and pyroptosis in macrophage, such as the caspase-1/GSDMD signaling pathway (107) and METTL3/MALAT1/PTBP1/USP8/TAK1 axis (108). The overexpression of brain and muscle Arnt-like protein 1 (BMAL1) also reduces the production of inflammasomes and pyroptosis in macrophages, as well as decreases the proportion of M1 phenotype via the TLR2/NF-κB pathway (109). A recent study has found that macrophages can activate the NLRP3 inflammasome and initiate subsequent pyroptosis to affect AL pathogenesis in mice model of cobalt-chromium-molybdenum (CoCrMo) alloy particles-induced osteolysis (110). As a result, wear particles not only induce M1 polarization and production of proinflammatory cytokines, but also boost inflammation by increasing pyroptosis in macrophages and inducing local tissue impairment.

In the bone-implant interface, macrophage polarized to the M1 phenotype is mainly stimulated by wear particles and endotoxin (55). Upon continuous stimulation, M1 macrophages cause tissue damage by strengthening local inflammation via the secretion of TNF-α, IL-1β, IL-6, and IL-8 and the reduction of IL-10 expression (15, 23, 24, 26, 54–58). Besides, other cytokines such as IFN-β (15) and various chemokines, including chemokine (C-X-C motif) ligand 9 (CXCL9), CXCL10, and CXCL11 (56) secreted by M1 macrophages also play a role in AL pathogenesis. These biological reactions are largely affected by the size and type of wear particles (23, 26). Endotoxin also contributes to macrophage differentiation into the M1 phenotype by stimulating TLRs on the cell surface. Particles with PAMPs adhering, such as LTA and LPS, could induce more production of the proinflammatory cytokine than those without endotoxin (30, 32). On the contrary, the removal of endotoxins from particles significantly reduces the cellular activity of macrophages (30) and inhibits OC differentiation (31).

It is well-established that both polyethylene (PE) and Ti particles can promote the expression of TLRs and various proinflammatory factors in macrophages (43, 125). A recent in vivo experiment also confirmed that alloy particles could induce significantly higher numbers of TLR-1, -4, and -6 positive cells in the synovial layer of joints (126). In the downstream of TLRs, NLRP3, ASC, caspase-1, and TNF-α and IL-1β were found to be positive through the colocalization with CD68 in the tissues around the revised prosthesis (43). The NF-κB, MAPK, and TAK1 pathways are involved in mediating signaling transduction downstream of TLRs (61–63, 127–129). In addition to TLRs, macrophages can engulf PMMA debris through macrophage receptors with collagenous structure (MARCO), which is a key pathogenic factor in promoting the phagocytosis of polymethyl methacrylate (PMMA) debris in aging macrophages (130).

As the main alternative hypothesis to PAMPs, DAMPs also activate TLRs during AL. The term “DAMPs” refer to self-molecules released by dying or damaged cells, which are defined as endogenous danger molecules due to they could activate the innate immune system by interacting with pattern recognition receptors (PRRs) (131). DAMPs are recognized by various membrane-bound receptors, including PRRs and non-PRRs, and also by intracellular sensors, notably through inflammasome (132, 133). Through binding to specific receptors, DAMPs activate the inflammatory process and recruit immune cells like neutrophils and monocytes. Therefore, after the clearance of DAMPs, the recruited leukocytes will change from a proinflammatory into a reparative program (133). Cobalt alloy particles could induce macrophage-associated inflammatory responses and bone loss through DAMPs rather than activated TLR4, due to the partially absent of metal-binding histidines in TLR4 (134). However, it has also been thought that the DAMPs generated in response to particles are insufficient to activate TLR2 or TLR4 in these cells (32).

Although DAMPs contribute to the host’s immune defense, they also promote pathological inflammatory responses. The DAMPs, such as high-mobility group box 1 (HMGB1), S100 proteins, and heat shock proteins (HSPs), are commonly known as regulatory molecules of inflammatory responses (131). HMGB1 is an ancient DNA-binding nucleoprotein. It can be passively released from dying cells or actively secreted by monocytes, macrophages, and myeloid dendritic cells (135, 136). HMGB1 could induce M1 polarization via TLR2, TLR4, and RAGE/NF−κB signaling pathways, leading to LPS−induced acute lung injury (124), meanwhile, it significantly yielded the expression of the M1 marker inducible nitric oxide synthase (iNOS) while decreasing the M2 marker IL-10 in macrophages (137). In vitro, HMGB1 silencing down-regulated the secretion of inflammatory cytokines in macrophages, which cannot be reversed by the exogenous HMGB1 (138). Interestingly, HMGB1 could also trigger M2 macrophage polarization via the TLR2/NOX2/autophagy axis (139). In the same aspect, loss of HMGB1 in macrophages can increase the differentiation of proinflammatory macrophages and enhance inflammatory response under specific conditions, not the otherwise (140). In this regard, HMGB1 may induce distinct macrophage phenotypes probably due to different redox isoforms (141). Besides HMGB1, the expression level of several members of HSPs is closely related to distinct stages of polarization in macrophages (142). Compared with unpolarized macrophages, a significant up-regulation of members of the HSP70 family (HSPA2 and HSPA8), as well as the HSP90 family (HSP90AA1) can be observed in M1 macrophages. On the other hand, changes in HSP expression were also observed in macrophages during the M2 polarization, although with only five transcripts being significantly modulated. Among them, DNAJB5, HSPA13, HSPBAP1 were upregulated, whereas HSPH1 and HSPB1 were down-regulated (142).

Recent studies have found that Ti, CoCrMo particles, and LPS could strongly induce inflammatory responses in macrophages, significantly increasing the production of TNF-α and IL-1β, rising RANKL/OPG ratio, and enhancing the OC activity (143–146). The current paradigm holds the view that the induction of OC differentiation by inflammatory cytokines is indirectly yielded under RANKL stimulation (147). Proinflammatory cytokines, such as IL-1β, TNF-α, IL-6, soluble IL-6 receptor, and IL-17, could all increase the production of RANKL from OBs. The effects of proinflammatory cytokines could be balanced by anti-inflammatory cytokines, such as IL-4 and IL-13, which could inhibit RANKL expression (147). In addition, a variety of wear particles could conduce to osteolysis via elevating M1 polarization (15, 24, 57, 58), since M1 macrophages are an important source of TNF-α production (11).

Although the concrete mechanism of TNF-α on OC differentiation is not fully understood, TNF-α signals promote OC differentiation by upregulating several proinflammatory target genes through the activation and nuclear translocation of NF-κB. One such target is RANK, which increases OC activity by mediating RANKL signaling (148). Interestingly, TNF-α produced by LPS/TLR4 signals can regulate OC generation in LPS-treated macrophages through the activation of RANKL signaling, whereas TNF-α in a RAW264.7 cells-based experiment demonstrates it may act as an autocrine/paracrine factor in promoting osteoclastogenesis, independent of RANKL signaling (149). In addition, the co-culture experiments of macrophages and MSCs (or OBs) revealed that PE and Ti particles could inhibit OB function, meanwhile promoting M1 polarization and osteoclastogenesis (150–152), proposing another mechanism for inflammation-induced osteolysis.

The classical M1/M2 system was based on experiments in vitro with different stimulation approaches. Subsequent studies revealed that the process of macrophage activation and polarization is much more complex and needed an additional subdivision of the M2 population (153). As a result, the M2 macrophages is further classified into four subtypes: alternative activated macrophages (M2a), type 2 macrophages (M2b), deactivated macrophages (M2c), and M2-like macrophages (M2d) (154). Among them, applied stimuli and the achieved transcriptional changes correspond to particular subtypes: 1) M2a macrophage induced by IL-4 or IL-13 is a profibrotic phenotype, 2) M2b macrophage is stimulated by immune complexes combined with Toll-like receptor or IL-1 receptor agonists, and 3) M2c macrophage is exposure to IL-10, TGF-β, or glucocorticoids, 4) M2d, activated by adenosines or IL-6 (154, 155). For the pathological study, although M2a/c macrophage is found to be beneficial in early inflammatory stages, they have been uncovered to impair tissue remodeling (156). Similarly, M2b macrophages have been thought to be relieving immune responses with minor damage to local tissue (157).

M2 macrophage has the effect of alleviating inflammation induced by wear particles, through increased expression of IL-10 and decreasing expression of IL-6 and TNF-α (24, 57, 58). In the presence of large amounts of TLR2 ligands, the anti-inflammatory activity of M2 macrophage is inhibited but without evident changes in cell surface markers (158). On the contrary, inhibition of TLR4 caused a shift from inflammatory M1 macrophages toward M2-dominant macrophages (159), suggesting that TLRs act as a switch that plays a role in the regulation of inflammation. In the downstream of TLRs, interleukin-1 receptor-associated kinase (IRAK) -m has been identified as an inhibitor of TLR signaling (160). One study found that knockdown of IRAK-m promoted M1 polarization and inhibited M2 polarization during the mycobacterium tuberculosis infection (161). But the discovery of IRAK-m-mediated local immunosuppression in periprosthetic tissues hints that macrophages have a self-protective mechanism, wherein the wear debris-mediated stimulation is inhibited to prevent overproduction of NK-κB-dependent proinflammatory cytokines and thus suppressing the deleterious host response in AL (162). However, the drawback is that the induction of IRAK-m overexpression triggered by wear debris also appears to conduce to the inhibition of LPS-induced TLR signaling, which leads to low-level biofilm-associated infection and chronic inflammation (162). Therefore, activation of IRAK-m is somehow affected by the local immune environment.

In addition, M2 polarization is regulated by cytokines in the surrounding, especially IL-10 and IL-4. IL-10 is an important anti-inflammatory cytokine (163). Recent studies have verified that IL-10 treatment significantly reduces iNOS expression and promotes CD163 overexpression, suggesting that IL-10 treatment could reprogram bone marrow-derived macrophages (BMDMs) to an M2 phenotype and regulate the process of M1/M2 polarization (164). In vivo experiments revealed that the extra addition of IL-10 partially reversed the inhibitory effect of PE particles on bone ingrowth (165). In the aspect of gene regulation, IL-10 not only decreased the activation of signal transducer and activator of transcription 1 (STAT1), NF-κB p65, and c-Jun N-terminal kinase 1 (JNK1) genes but also increased the expression of STAT3 (164). In addition to IL-10, IL-4 also has the ability to switch the M1 phenotype induced by Ti particles into the M2 phenotype (55, 166). Compared with non-activated macrophages, upon IL-4 stimulation, the shift from M1 to the anti-inflammatory M2 phenotype is more thorough (55).

Macrophages stimulated by cytokines, such as IL-4 and IL-13, have been confirmed to prevent the OC differentiation and inhibit the function of mature OCs (11, 147). In addition, M2 macrophages may promote OB differentiation by producing cytokines that are critical for osteogenesis, including BMP-2, TGF-β, and IGF-1 (11, 167, 168). Also, by interacting with MSCs, M2 macrophages create an anti-inflammatory environment that is conducive to osseointegration (14). A recent study revealed that preconditioning of murine MSCs with IFN-γ and IL-1β highly significant reduction of CD86 and iNOS protein in macrophages under M1 inducers (LPS + IFN-γ) and diminished TNF-α secretion (169). Additionally, CD86 and iNOS protein expression as well as NO and IL-10 secretion were markedly increased under M2a inducers (IL-4) (169). On the other hand, under the stimulation of IL-4, reduced expression of CD86 and iNOS, as well as increased secretion of nitric oxide (NO) and IL-10 could be discovered in macrophages (169). The secretion of IL-10 could be attributed to the phenotype of M2b macrophages which are generally suggested to be the main subtype of macrophages for inflammatory relief (170). In addition, macrophages stimulated by a conditioned medium from preconditioned MSCs (pre-MSC-CM) may display an overall increased phagocytic capacity (171, 172). Also, M2a macrophages are found to undergo reprogramming to an M2b/M2c phenotype after treatment with the pre-MSC-CM (171, 172), indicating an influence of MSCs behavior on the induction of macrophage polarization.

TLRs have been intensively studied in innate immunity regarding the recognition of PAMPs (182). It has been reported that TLR2 or TLR4 can recruit MyD88 and further bind to the IRAK to form a signal complex called Myddosome. The Myddosome-complex recruits the ubiquitin ligase TNF receptor-associated factor 6 (TRAF6), which triggers the TAK1 kinase signaling cascade pathway and ultimately leads to NF-κB nuclear translocation through the phosphorylation and activation of IκB kinase α/β (IKKα/β) (63, 183, 184).

TLRs are closely related to macrophage polarization (63). HMGB1 is an important mediator that induces M1 polarization through the activation of absent in melanoma 2 (AIM2) inflammasome, TLR2/4 and RAGE/NF-κB signaling pathways in macrophages (124). In addition, LPS can activate downstream TAK1, NF-κB, and MAPK signals through the TLR4/MyD88 pathway, which increases the expression of cyclooxygenase-2 (COX-2) and iNOS, promotes M1 polarization and leads to the up-regulated expression of cytokines TNF-α, IL-1β, IL-6, and PGE2 (61, 63). The particles with LTA adhering could increase the expression of proinflammatory factors by the activation of TLR2/NF-κB and MAPK pathways (32, 185). Previous studies have shown that a variety of TLRs, such as TLR2, which participate in local inflammatory and immune responses, are present in periprosthetic tissues after the revision of total hip arthroplasty (rTHA) (43).

NLRP3 inflammasome widely exists in macrophages, granulocytes, antigen-presenting cells (APC), and other immune cells. It consists of the NLRP3 (sensor protein), ASC (adapter protein), and caspase-1 (186). The initiation of NLRP3 inflammasome could be induced by multiple inflammatory stimuli, such as PAMPs. As the core protein of the NLRP3 inflammasome complex, NLRP3 senses endogenous DAMPs and microbial ligands. The production of ROS, low level of intracellular potassium (K⁺), and the release of lysosomal protease into the cytoplasm are all upstream mechanisms that trigger the activation of NLRP3 inflammasome in several conditions (187–189). NLRP3 inflammasome which is activated by pathologically stimulated macrophages could regulate the production of IL-18, IL-1β, and cleavage of caspase-1 (186, 188, 189).

On the other hand, studies have also found that NLRP3 inflammasome participates in macrophage polarization (89, 175, 190, 191). In a co-culture experiment, force-pre-treated human periodontal ligament cells (hPDLCs) promote M1 polarization and increase the secretion of IL-1β in macrophages via the activation of NLRP3 inflammasome (191). On the contrary, by inhibiting the activation of NLRP3 inflammasome, metformin induces M2 polarization in macrophages and promotes wound healing in rat dorsal skin (175). Moreover, USP19 could directly promote M2 polarization by suppressing the activation of NLRP3 inflammasome to interferon regulatory factor-4 (IRF-4) (89). Bruton’s tyrosine kinase (BTK) is a key factor in TLR4-related pathways, which could activate NF-κB signaling and promote p65 phosphorylation, ASC oligomerization, and caspase-1 activation (192). It has been recently found that BTK could promote TiAl6V4 alloy particles (TiPs)-induced inflammation in BMDMs by positively regulating NF-κB activation, NLRP3 inflammasome formation, and M1 polarization (193). On the contrary, NLRP3 silencing attenuates the promotive effect of conditioned exosomes on M1 polarization in a particles-induced osteolysis model (194).

Syk is a key component of TLRs in recognizing PAMPs and activating immune responses (195–199). Moreover, Syk plays a critical role in LPS-induced M1 polarization (200). In monocytes, pharmacological inhibition of Syk prevents LPS-induced TLR4 phosphorylation (199), suggesting that Syk may be involved in tyrosine phosphorylation of the TIR domain in TLR4. Recently, PMMA and HA particles were confirmed to augment the expression of CD86, and secretion of cytokines, such as TNF-α and IL-6, in a Syk- and MAPK-dependent (phosphorylation of ERK and p38) manner (56). Also, Syk is reported to be closely associated with the NLRP3 inflammasome-mediated proinflammatory cytokine release (201). In addition, Syk could control NLRP3 activation and IL-1β synthesis in macrophages in response to fungal infections (202).

In the resting state, NF-κB dimers are sequestered in the cytoplasm in an inactive state by the IκB family of proteins, including IκBα, IκBβ, IκBϵ, and the NF-κB precursors, p105 and p100. The IκB kinase (IKK) complex consists of kinases IKKα, IKKβ, and IKKγ. Upon receiving the activation signals, IκB protein is phosphorylated by IKK complexes, leading to proteasomal degradation of IκB. The released NF-κB dimers are then translocating into the nucleus, where they can bind to specific sites on DNA to regulate gene transcription (75).

LPS-induced M1 polarization is dependent on NF-κB p65 activation, and the treatment with IKKβ inhibitors reduces the mRNA expression of M1 markers in macrophages (203). Therefore, IKK inhibitors reduce LPS-induced expressions of IL-1, IL-6, IL-10, TNF-α, and IFN (204). Recent studies have revealed that diverse wear particles (e.g., Ti, TiPs, HA, PE, PMMA) could induce the macrophage to release proinflammatory cytokines and chemokines, which are accomplished by the activation of NF-κB signaling pathways and associated macrophage polarization (15, 26, 54, 205). Qiu et al. found that stimulation of LPS or Ti particles could up-regulate p-IKKβ, p-IκBα, and p-p65 and significantly increase the translocation of p65 into the nucleus in BMDMs. Meanwhile, these particles-induced inflammatory infiltrations increase the number of OCs (TRAP-positive cells) and thereby decrease bone mineral density, eventually leading to PPOL (15). Similarly, Gao et al. found that PMMA particles could up-regulate iNOS and decrease the production of Arginase-1 (Arg-1) and IL-10 by the activation of p65 nuclear translocation in macrophages (54).

Under the same settings, NF-κB can be activated by RANKL or LPS to augment particles-induced bone loss via the enhancement of osteoclastogenesis (206, 207). Upstream stimuli-triggered NF-κB signaling could act on transcription factors (c-fos and NFATc1) and regulate the expression of osteoclastogenesis-related genes, thus promoting OC differentiation and functions (206, 208, 209).

The MAPK signaling pathway takes part in cell differentiation, proliferation, and apoptosis (210). MAPK pathways are organized into three-tiered cascades consisting of three factors: MAPK, MAPK kinase, and MAPKK kinase. During the phosphorelay process, MAPKKKs which are serine/threonine protein kinases, phosphorylate and activate MAPKKs, and then dually phosphorylate the threonine and tyrosine residues of the conserved TXY motif that belongs to the activation loop of MAPKs, including extracellular signal-regulated kinase (ERK), JNK, and p38 (210, 211).

MAPK signaling pathway is closely associated with macrophage polarization (61, 212–214). The activation of the MAPK signaling pathway promotes polarization of M1 macrophages, expression of iNOS, and down-regulation of CD206, thereby mediating the secretion of cytokines, such as TNF-α, IL-1β, and IL-6 (214). On the other hand, inhibiting the phosphorylation of JNK, ERK, and p38 in MAPK pathways may switch repolarized M1 to reprogrammed M2 phenotype (213, 214). Studies elucidate that the MAPK signaling pathway is located downstream of TLRs, wherein LPS could activate the MAPK pathway through TLR4/MyD88 pathway to induce M1 polarization and participate in the inflammatory response (61, 62, 215). Further, PMMA and hydroxyapatite (HA) particles are reported to result in M1 polarization by increasing phosphorylation of ERK and p38 (56). Similar to NF-κB signaling, MAPK also mediates wear particles-induced OC differentiation and the following osteolysis (206).

STAT proteins are potent cytoplasmic transcription factors wherein several family members have participated in macrophage polarization and OC formation, including STAT1, STAT3, and STAT6 (216, 217). Growing studies reveal the involvement of the JAK/STAT pathway in regulating multiple biological events, such as innate and adaptive immunity, cell growth and differentiation, and programmed cell death (217). In the aspect of functioning, phosphorylated STATs promote monomeric dimerization through their SH2 domains and further translocation into the nucleus, where STATs regulate the transcription of target genes (216, 217).

Several studies have revealed the mediating role of STATs in macrophage polarization. PARP14 silencing or RBM4 knockdown can promote IFN-γ-induced signaling transduction via STAT1 activation, which leads to M1 polarization in macrophages (218, 219). Also, STAT1/6 pathway mediates M1/M2 polarization in macrophages after physalin D stimulation (220). By phosphorylating STAT6, protocatechuic acid (PCA) decreases the activation of NF-κB signaling and gives a bias towards M2 polarization over M1 polarization in macrophages (173). Conclusively, an obvious antagonism between STAT1 and STAT6 has been described to promote M1 and M2 cell polarization, respectively (220). In addition, the activation of JAK/STAT3 signaling facilitates macrophage transformation to M2c polarization (180). In an AL mouse model, compared with the pure stimulation of UHMWPE particles, the extra addition of IL-10 significantly decreases iNOS-positive cells and increases CD163-positive cells via reducing the transcription of STAT1, NF-κB p65, and JNK1, and promoting the expression of STAT6 (164).

Compared with STAT1 and STAT6, STAT3 is closely related to osteoclastogenesis. Therefore, inhibition of STAT3 can negatively affect RANKL-mediated OC formation and functions (221). An in vivo experiment demonstrated that treatment with TiPs pellet could promote the expression of STAT3 and production of RANKL in OBs, thereby stimulating OCs formation in particles-induced osteolysis models, whereas the activation of STAT3 also mediates nano-particles-induced IL-6-dependent inflammatory response in OBs (222).

The centration of Ca²⁺ in the cytoplasm ([Ca²⁺]c) is 20, 000 times lower than that outside the cell. This is achieved by the Ca²⁺-related transportation and exchange under the dependence on specific proteins. When activated by a variety of external stimuli, cells respond by an increase in the [Ca²⁺]c and trigger downstream signaling, in the form of Ca²⁺ spikes or oscillations (223). Ca²⁺ signaling participates in various biological processes, resulting from a complex switch between the activation and inactivation of Ca²⁺-permeable channels (224). The excitability of Ca²⁺ signaling is determined by intracellular Ca²⁺ oscillations, which are attributed by the extracellular Ca²⁺ influx, the effect of ITAM/PLCγ/IP₃ signaling on the release of Ca²⁺ from ER, and the capacity of collecting Ca²⁺ from the cytosol by SERCA. In addition, depending on the depletion of ER Ca²⁺ storage, extracellular Ca²⁺ entry could also be accomplished by the activation of STIM-mediated TRPC channels and Orai1 channels, which is the so-called SOCE mechanism (225, 226) ( Figure 3 ).

Ca²⁺ signaling is related to behavior changes in macrophages, such as polarization and phagocytosis, largely depending on Ca²⁺ uptake in mitochondria (227). Many studies have implicated that distinct Ca²⁺ entry channels determine the IFN-induced M1 polarization or IL-4-induced M2 polarization. Naive or M2 macrophages exhibit a robust Ca²⁺ entry that is dependent on the activity of Orai1 channels (228). As a result, blockade of Ca²⁺ entry inhibits NF-κB/STAT1 or STAT6 signaling events and consequently lowers cytokine production that is essential for M1 or M2 polarization in macrophages (228). In detail, Ca²⁺ influx facilitates M1 polarization, enabling the high productivity of proinflammatory mediators, such as cytokines and chemokines (229). Transient receptor canonical ion channel 1 (TRPC1)-mediated calcium entry seems to play a crucial role in M1 polarization, due to a non-selective TRPC1 current is found in macrophages with M1 phenotype (228, 230). Therefore, knockdown or blockade of the Kir2.1 channel significantly suppresses M1 polarization and promotes M2 polarization (231). Of note, a study found that transient receptor potential vanilloid 1 (TRPV1)-induced Ca²⁺ influx could promote the phosphorylation of Ca²⁺/calmodulin-dependent protein kinase II (CaMKII), but leads to the inhibition of M1 polarization (232), bringing uncertainty in elucidating regulatory mechanism between Ca²⁺ signaling and macrophage polarization.

Except for macrophage polarization, Ca²⁺ oscillations are the well-known mechanism in triggering RANKL-induced osteoclastogenesis and bone resorption (233, 234). The increase in the [Ca²⁺]c is a fundamental process for mediating OC biology, involving in OC proliferation, differentiation, and resorptive function. At the molecular level, cytosolic Ca²⁺ binds to calmodulin and subsequently activates calcineurin, leading to the activation of NFATc1 which is required for OC differentiation (235). Therefore, by blocking the Ca²⁺ entry channels, inhibition of the activity of Ca²⁺/calmodulin‐dependent protein kinase IV (CaMKIV) and calcineurin will lead to a reduction in the nuclear translocation of c-Fos and NFATc1, ultimately resulting in the suppression of osteoclastogenesis (236). Similarly, directly interfering with intracellular Ca²⁺ oscillations would also have the same negative impact on osteoclastogenesis (237). More content associated with Ca²⁺ oscillations in OC biology has been reviewed in the article by Okada H, et al. (234).