Some immune-inflammatory diseases, such as arthritis, suggest a correlation between mechanical force and the inflammatory response. Arthritis is characterized by inflammation that localized to the joints (such as the knee joint) when exposed to prolonged mechanical force. Therefore, mechanical force may be a factor that determines the transition of inflammation from systemic autoimmunity to local inflammation. According to Cambré’s research, MSCs in the mechanosensitive region of joints could sense mechanical stimulation and convert mechanical signals into chemical signals to trigger local inflammation and bone destruction, which ultimately led to the occurrence of arthritis (Cambre et al., 2018). Therefore, mechanical stimulation plays an essential role in inflammatory response (Hao et al., 2015).

Bone regeneration involves multiple stages and cell interactions. The formation of fracture hematoma and the subsequent acute inflammatory phase are key steps to determine the success of bone regeneration. The acute inflammatory phase begins with the activation of neutrophils that secrete inflammatory factors and chemokines to recruit monocytes and macrophages (Xing et al., 2010). In addition to cleaning up the necrotic tissue, macrophages secrete inflammatory cytokines and chemokines (such as TNF-α, IL-1β, IL-6, and CCL2) to recruit MSCs. Subsequently, MSC-rich granulation tissue replaces the hematoma. Immediately after, MSCs are stimulated by various factors in the environment to trigger osteogenic differentiation by either endochondral ossification or intramembranous ossification. Therefore, it is evident that an appropriate acute inflammatory phase duration is important for bone regeneration. The interactions between MSCs and macrophages dynamically regulate this phase. Macrophages have two phenotypes. The first is classically activated M1 macrophages, which function in initiating and sustaining inflammation, and the second is alternatively activated M2 macrophages, which function in resolving inflammation. The transformation of macrophages from M1 to M2 is known as macrophage polarization (Pugin et al., 1998). The main method of anti-inflammation in bone regeneration at this stage is the early polarization of pro-inflammatory M1 macrophages to anti-inflammatory M2 macrophages, which serves to promote the resolution of inflammation and the osteogenesis process.

Following the occurrence of a fracture, MSCs are mobilized into the peripheral blood by transforming growth factor-beta (TGF-β) (Wan et al., 2012). MSCs are exposed to fluid shear stresses, one of which is wall shear stress (WSS). WSS can stimulate MSCs to produce antioxidant and anti-inflammatory mediators. Additionally, the application of WSS to MSCs facilitates the recruitment of chemokines, including prostaglandin E2 (PEG2) and cyclooxygenase-2 (COX2), to inhibit the synthesis of tumor necrosis factor α (TNF-α) generated by immune cells and the inflammatory response (Diaz et al., 2017). The mechanism by which WSS stimulates MSCs to produce inflammatory mediators may be related to the FAK-COX2 signaling pathway. Lee et al.’s (2017) study showed that WSS could promote Ca²⁺ release and activate the Akt, MAPK and FAK signaling pathways of MSCs. When inhibited the above factors respectively, only FAK disrupted the induction of COX2 and decreased the production of MSC inflammatory mediators. Thus, the FAK-COX2 signaling pathway is significant for MSCs to respond to mechanical stimulation for immunomodulatory functions. In addition to shear stress, dynamic compression regulates the crosstalk between MSCs and macrophages. Zhang et al. (2021) developed an extracellular matrix-based hydroxyapatite scaffold fabricated by freeze-drying the ECM of compression-stimulated MSCs. This biofabricated scaffold could accelerate the polarization of macrophages from the pro-inflammatory M1 phenotype to the anti-inflammatory M2 phenotype to promote bone regeneration. These findings suggested that compression could promote the secretion of anti-inflammatory mediators in MSCs. However, recent studies have found that MSCs maintained their physiological levels through TNF-α endocytosis. Cyclic stretching promoted the proliferation and osteogenic differentiation of MSCs by TNFα endocytosis, which downregulated TNFα secretion in MSCs, rather than directly downregulating TNFα gene expression (Yu et al., 2021). Additionally, mechanical stimulation of adipose tissue modulates the anti-inflammatory properties of human adipose-derived stem cells (hADSCs) in adipose tissue. Carelli et al. compared the anti-inflammatory properties of hADSCs in mechanically stimulated adipose tissue and the control group. It was found that the anti-inflammatory effect of mechanical stimulated-hADSCs was superior to that of the control group hADSCs (Carelli et al., 2018). However, other studies have found that mechanical stimulation could promote inflammation and osteogenesis simultaneously, likely as a result of the MSC autocrine regulation of inflammatory factor secretion (Sumanasinghe et al., 2009; He et al., 2020).

Most studies have discussed the response of MSCs to mechanical stimulation during osteogenesis. A recent research found that macrophages were also capable of responding to mechanical stimulation (Dong et al., 2021). Mechanical stretch polarizes macrophages into the M2 phenotype that secrets inflammation-related cytokines, including IL10 and TGF-β, to regulate the local inflammatory microenvironment. Mechanical stimulation activates the YAP/BMP2 axis in macrophages to increase the expression of BMP2, which promotes the osteogenesis of MSCs. As an important component of the mechanical transduction pathway, YAP induces the polarization of M2 macrophages via Wnt5a and TGFβ1 (Feng et al., 2018). Schoenenberger et al. (2020) found that macrophages, as mechanosensitive cells, played an essential role in tendon repair. Mechanical stimulation was found to promote the transformation of macrophages to the M2 phenotype and subsequent tissue healing. These results suggested that future consideration might be given to exploring the role of mechanical stimulation in MSC and macrophage co-culture models.

Bone regeneration contains endochondral ossification and intramembranous ossification. Endochondral ossification is the process that stable cartilaginous soft callus first formed, followed by the formation of bone hard callus through vascular ingrowth and ossification centers (Gerstenfeld et al., 2003). The intramembranous osteogenesis process is accomplished by the differentiation of MSCs into osteoblasts at vascular-rich sites and the mineralization of osteoblasts to osteocytes. Thus, vascular formation, which is closely related to the osteogenesis process, is an important part of bone regeneration. During bone regeneration, MSCs and vascular endothelial cells (VECs) communicate with each other through paracrine mediators to promote osteogenesis (Li C.-J. et al., 2015). Mechanical stimulation is one of the biophysical factors that promote osteogenesis, and plays a crucial role in the crosstalk between MSCs and VECs.

Vascular endothelial growth factor (VEGF), a widely known signaling molecule regulating osteogenesis and vascularization, has been identified for its role in mechanical stimulation-induced osteogenesis. Charoenpanich et al. evaluated the effect of stretching on human MSC gene expression by microarray analysis. The results showed that stretching promoted the release of VEGF from human MSCs (Charoenpanich et al., 2014). Moreover, Jiang’s team found that stretching-stimulated VEGF secretion of MSCs not only promoted tube formation but also promoted VECs to release growth factors associated with bone formation, such as BMP-2 and IGF-1, which in turn regulated the osteogenesis process of MSCs (Jiang et al., 2018). In addition to stretching, dynamic compression can promote increased VEGF secretion in MSCs. Dynamic compression promotes VEGF secretion by upregulating YAP signaling activity in MSCs (Bandaru et al., 2020). In addition to VEGF, the expression level of placental growth factor (PlGF) in MSCs is related to the magnitude and duration of mechanical stimulation. PlGF has a variety of functions, including promoting osteogenesis and angiogenesis, and plays an essential role in the regulation of osteogenic-angiogenic interactions by mechanical stimulation (McCoy et al., 2013). It has also been found that mechanical stimulation can stimulate H vessel formation and VEGF secretion by downregulating exosomal miR-214-3p from MSCs (Wang et al., 2021).

In addition to the above in vitro studies, several in vivo experiments have investigated the effects of mechanical stimulation on bone regeneration and vascularization. Some studies have explored the effect of the initial application time of mechanical stimulation on vessel and bone formation. Boerckel et al., 2011 found that the application of mechanical loading in the early stage of bone defects could inhibit the growth of blood vessels into the defect area and lead to the failure of bone regeneration. In contrast, the application of mechanical loading delayed for 4 weeks could promote the reconstruction of blood vessel networks and bone regeneration (Boerckel et al., 2011). This result suggested that the effect of mechanical stimulation on vascularization and bone formation depended on the initial application time. McDermott et al. (2019) speculated that the difference was due to the different origin of the vessel forming at different times. Other studies have explored the effect of the loading application mode on revascularization. Claes et al., 2018 compared the effects of compression, stretching and shear stress on the vessel density in bone regeneration. As a result, the vessel density in the compression group was significantly higher than that in the other two groups, which suggested that compression was more beneficial to the bone regeneration process.

Osteocytes are mechanosensitive cells that reside in the lacunar-canalicular system (LCS) of cortical bone (Timmins and Wall, 1977). Recent studies have found several critical mechanical sensors of osteocytes, such as cilia, integrin, ion channels and G-protein-coupled receptors (Uda et al., 2017). Osteocytes regulate bone remodeling, mainly by sensing fluid shear stress caused by mechanical loading and regulating osteoblast-osteoclast communication (Dallas et al., 2013). As mentioned above, osteocytes play an important role in responding to mechanical stimulation. Osteocytes function as regulators influencing bone loss and formation by modulating osteoblast-osteoclast coupling. Osteocytes are of vital importance in the reconstruction of bone defects (Robling and Bonewald, 2020). Osteocytes regulate bone regeneration in both direct and indirect ways: secreting stimulators and inhibitors that affect osteoblast activity, and modulating osteoclast activity to regulate osteoblast behavior indirectly (Robling and Bonewald, 2020). However, due to limited research methods, the role of the osteocyte response to mechanical stimulation in bone regeneration has not been fully explored.

Osteocytes respond to external mechanical stimulation by secreting soluble factors that regulate MSC gene expression. Specific communications exist between osteoblasts and MSCs under mechanical stimulation conditions. (Hoey et al., 2011) found that conditional medium for mechanical stimulation of osteocytes upregulated osteogenic gene expression in MSCs, while no upregulation was seen in osteoblasts treated with the same. This suggested that mechanical stimulation played a vital role in the crosstalk between osteocytes and MSCs. Several studies have explored the paracrine mechanism by which osteocytes regulated MSC osteogenesis in response to mechanical stimulation. Du J. et al. (2020) suggested that mechanical regulating osteoblast-osteoclast coupling by promoting osteocyte secretion of leukemia inhibitory factor (LIF). Extracellular vesicles (EVs), as specific components of cell-cell and cell-matrix communication, also play an important role in osteocyte and MSC interactions under mechanical stimulation. Eichholz et al. (2020) comprehensively characterized the proteins secreted by osteoblasts after fluid shear stress through proteomics and found that proteins associated with EVs were significantly overexpressed. Moreover, culturing MSCs with the collected EVs resulted in MSC osteogenic differentiation, suggesting that mechanical stimulation promotes osteocytes to modulate MSC behavior via EVs. Peiying Lv’s team found that exosomes produced by osteocytes following mechanical stimulation also promoted the osteogenic differentiation of human periodontal ligament stem cells (PDLSC) (Lv et al., 2020).

The osteogenic differentiation of MSCs has been found to be correlated with the magnitude and frequency of mechanical stimulation. Stavenschi et al. explored the osteogenic effect of oscillatory fluid flow on MSCs of 1 Pa, 2 and 5 Pa. The results showed that the expression of osteogenic genes was significantly upregulated at the magnitude of 2 Pa and the frequency of 2 Hz (Stavenschi et al., 2017). The most effective shear stress for promoting MSC proliferation and osteogenesis has also been explored. Sanat’s research showed that MSCs exhibited a high cell proliferation rate when stimulated by intermittent flow at 1.09 mPa, while 10 mPa upregulated osteogenic gene expression (Dash et al., 2020). Xue and Cartmell, (2020) suggested the osteogenic effect of shear stress on MSCs in three-dimensional culture was different from that in the plate. Lower fluid shear stress (1–10 mPa) stimulated MSCs in the scaffold used to simulate a three-dimensional environment to promote the osteogenic differentiation, whereas 100–4,000 mPa was required when the MSCs were cultured in a plate (Xue and Cartmell, 2020).

Recent studies have shown that, compared to long-term continuous mechanical stimulation, a period of rest time during mechanical stimulation enhanced bone formation and improved the mechanical properties of bone (Robling et al., 2002; Saxon et al., 2005). Compared to continuous shear stress, intermittent application has been proven to maintain the mechanosensitivity of MSCs and osteocytes (Siller-Jackson et al., 2008; Liu et al., 2012). In addition to FSS, intermittent stretching can promote osteogenic differentiation of MSCs (Wang et al., 2017). Continuous cyclic mechanical tension (CCMT) has been found to downregulate Runx2 expression in MSCs and inhibit osteogenic differentiation (Shi et al., 2011). Another study compared the effects of intermittent compressive force (ICF) and continuous compressive force (CCF) on the behavior of PDLSCs. The results suggested that ICF upregulated TGFβ-1 and promoted the osteogenic differentiation of PDLSCs, whereas the osteogenic gene expression of the CCF group was unchanged (Manokawinchoke et al., 2019). Therefore, the intermittent mechanical stimulation mode is superior to continuous mechanical stimulation in terms of promoting bone regeneration.

At present, intermittent mechanical stimulation promotes osteogenesis in bone that needs the loading interval to recover its mechano-sensitivity to mechanical signals. Nardone et al. (2017) found that mechano-sensing switches (such as pFAK) were released from FAs into the cytoplasm during intermittent and activated YAP. This revocation of activation suggested that intermittent mechanical loading could activate integrin signaling downstream, which possibly explained the decreased mechano-sensitivity of bone tissue caused by continuous mechanical stimulation. Additionally, the mechanical environment is capable of modulating nuclear properties, and mechanical sensitivity may also be related to the nuclear biophysical properties (Heo et al., 2016). The nuclei in dynamic loading induced MSCs to stiffen and become resistant to deformation, which sensitizes MSCs to mechanical stimulated calcium signaling and differentiated marker expression (Heo et al., 2016). Thus, the nucleus plays an essential role in modulating cellular mechano-sensation during differentiation. There is limited research on the mechanisms by which MSCs respond to intermittent and continuous stimulation. However, studies on the mechanism of osteoblasts could provide hints for future research. It was suggested that mechano-sensitivity is primarily associated with actin stress fibers. Gardinier et al. suggested that osteoblasts responded to FSS through actin stress fiber formation (ASFF), and ASFF led to increased cell stiffness and decreased mechano-sensitivity (Gardinier et al., 2014). LIM kinase 2 (LIMK2) is a gene related to the reorganization of the cytoskeleton. Several studies found that inhibiting the LIMK2 increased the sensitivity of ERK1/2 to fluid shear stress and promoted the gene expression of c-fos to enhance the mechanical sensitivity of osteoblasts (Zhang et al., 2009; Xiang et al., 2012). These results suggested that the mechanism by which MSCs respond to intermittent mechanical stimulation may also be related to the actin stress fiber and cytoskeleton.

Delayed mechanical stimulation has a positive effect on osteogenesis by promoting angiogenesis. The formation of blood vessels is closely related to endochondral ossification, in which MSCs first aggregate and differentiate into hyaline cartilage to form the cartilage model. Following growth of blood vessels, the cartilage is gradually replaced by bone tissue. Joe’s laboratory investigated the effects of immediate and delayed mechanical loading on vascular ingrowth in bone regeneration (Boerckel et al., 2011; Ruehle et al., 2020). The results showed that load initiation was a key determinant of vascular network formation. Immediate loading significantly inhibited the growth of blood vessels into the bone defect area, which led to fracture nonunion. In contrast, delayed loading allowed the growth of vessels into the defect and induced vascular remodeling. The study by Anna showed that the bone accumulation rate was significantly elevated by 4-week delayed mechanical loading application, which coincided with chondrocyte hypertrophy and endochondral transition (McDermott et al., 2019). They concluded that a 4-week delay in mechanical loading better mimicked the process of endochondral ossification.

The mechanical stimulation applied to cells in two-dimensional (2D) environments is unidirectional. However, mechanical stimulation is multidirectional in physiological environments. The behavior of mechanical loading-induced cells is altered by the dimensionalities of their environments. Thus, the response of cells to external mechanical forces in three-dimensional (3D) environments that mimic the physiological environments in vivo needs to be explored. Li and his team found that long-term compression loading induced maturation of α5-integrin-based adhesions to form 3D-matrix adhesions (3DMAs) in the 3D environment (Li et al., 2020). In contrast to the FA formed in the 2D environment, the composition and morphology of 3DMAs are found only in native tissues and cell-derived matrices, suggesting that dimensionality influences the behavior of cells under mechanical stimulation. However, the exact mechanism by which external mechanical forces regulate cell fate in different dimensionalities remains unclear.

Exploring the effects of mechanical stimulation on cells in a 3D environment involves not only mechanical stimulation of cells but also the properties of materials. In 2D conditions, mechanical stimulation is applied to the cells directly. However, in a 3D environment, the force is first applied to the substrate which then transmits the mechanical signals to the cell through the deformation generated by the stimulation (Steinmetz et al., 2015). Thus, the process by which the mechanical signals are transmitted to cells contains two steps: the deformation of scaffolds produced by mechanical stimulation and the cell sensing and responding to the deformation. The ECM is not a linearly elastic material and has complex mechanical properties, including viscoelasticity, mechanical plasticity and nonlinear elasticity (Chaudhuri et al., 2020). The ECM responds to external mechanical stimulation by remodeling the stress fiber network, such as by changing the structure of the fiber network and forming bonds between the fibers (Loebel et al., 2019). The effect of mechanical stimulation on cells is related to the interactions between ECM properties and cells, which suggests that the mechanical properties of scaffold materials are of vital importance for cell differentiation in the 3D environment.

Materials that mimic the mechanical properties of ECM have been explored. Davidson and his team developed a multifiber hydrogel network with force-responsive characteristics (Davidson et al., 2020). In this network, the fibers form covalent bonds under mechanical loading, and the interactions of the fiber increase material stiffness and plastic deformation. Davidson’s design mimics the physiological process of ECM remodeling under mechanical stimulation, providing a model for exploring the effects of mechanical stimulation on cells in 3D environments (Davidson et al., 2020). Mechanical stimulation in a 3D environment fabricates the layered scaffolds with gradient mechanical properties. Horner et al. (2019) designed a 3D electrospinning scaffold with a tissue gradient that generates spatially controlled strain gradients in a scaffold depth-dependent manner under dynamic loading. MSCs in the greater compressive strain areas upregulate osteogenic gene expression, while chondrogenic markers are upregulated in the high local compressive strain areas. The formation of the mechanical gradient was maintained only under the application of dynamic loading. This study shows that regulating the local mechanical microenvironment provides a strategy to recapitulate the gradient structure of osteochondral tissues (Table 1).

Explorations of optimal mechanical parameters are significant for further application of mechanical stimulation in bone tissue engineering as presented above. Shear stress from 1.09 mPa to 5 Pa was applied to MSCs in previous studies, and 10–15 mPa were proved to have a superior osteogenic effect. Stretching or compression resulted 10% strain was discovered promoting osteogenesis. In comparison with continuous mechanical stimulation, the intermittent application is more efficient in inducing osteogenic differentiation via maintaining the mechanosensitivity of MSCs to mechanical signals. Therefore, mechanical stimulations are recommended to be performed with appropriate intervals. Application of delayed mechanical stimulation was reported to be an ideal option for facilitating angiogenesis in bone remodeling, which indicates that future researches should take the mechanical stimulation application time into consideration. Besides, in contrast to the 2D environment, MSCs showed a more bionic behavior in response to external mechanical stimulation in 3D environment that mimics physiological environments. Thus, 3D environment is recommended for the mechanical stimulation application.