Osteoblasts and adipocytes both originate from BMSCs (32). The imbalance between osteogenesis and adipogenesis in the bone marrow niche is an important factor contributing to fatty infiltration. Wingless-related integration site (Wnt) and PPARγ are important signaling pathways for osteogenesis and adipogenesis, respectively (33). Wnt signaling reduces the expression and activity of PPARγ and CCAAT/C/EBPα in BMSCs, which inhibits adipocyte formation. Similarly, adipogenesis in BMSCs requires activation of PPARγ signaling and inhibition of osteogenic pathways such as Wnt and Notch (18). This is a complex regulatory network and the relevant cytokines that regulate differentiation are shown in the following table. ( Table 2 ).

Osteoblasts show markedly impaired differentiation and function under fatty infiltration. High FFAs levels significantly inhibited osteoblast differentiation in human (44). As a natural ligand for PPARγ, FFAs also induce PPARγ activation in BMSCs, preosteoblasts and osteoblasts, and convert osteoblasts into adipocytes by inhibiting Runx2 activity (45). Duque et al. found that lipotoxicity affects the expression of important osteogenic genes such as ALP, Runx2, OCN, OSX, and BSP2, as well as bone mineralization capacity, with a significant decrease in alkaline phosphatase (ALP) and alizarin red staining (46).

Large amounts of osteoblast apoptosis can be observed in the fatty infiltration environment, a process also known as lipoapoptosis. There are two main pathways for apoptosis to occur: an extrinsic pathway that requires transmembrane death receptor mediation, and an intrinsic apoptosis that is initiated through mitochondrial mediation, while osteoblast lipoapoptosis is mediated by both the extrinsic and intrinsic pathways. Reactive oxygen species (ROS) levels were significantly upregulated when osteoblasts were co-cultured with adipocytes. The ROS-activated ERK/P38 signaling pathway contributes to osteoblast apoptosis, which is significantly ameliorated when the pathway is inhibited (47). Furthermore, the low survival rate of osteoblasts with high apoptotic levels, associated with activation of intrinsic apoptotic caspases 3/7. PA-induced human-derived osteoblasts activate the Fas/Jun kinase (JNK) apoptotic pathway and express Bak and Bax, proteins that permeabilize the mitochondrial outer membrane, prompting mitochondria to release cytochrome C into the cytoplasm to trigger apoptosis, and osteoblasts can be rescued from PA-induced lipoapoptosis when the JNK inhibitor SP600125 is added to the culture medium (48).

An article on new sources of receptor activator for nuclear factor-κB ligand (RANKL) was published in the journal EMB in 2021, demonstrating by single-cell transcriptome sequencing that adipose lineage cells in mice bone marrow are capable of secreting RANKL (49). Therefore, the increase in MAT will inevitably affect RANKL secretion under fatty infiltration.

The secretion of RANKL in bone tissue is mainly done by osteoblasts, osteocytes, proliferative chondrocytes, B cells, and newly discovered intramedullary adipocytes (50). The nuclear factor kappa-B (NF-κB) pathway involved in RANKL is thought to be a key factor in the regulation of osteoclastogenesis and bone metabolism, in which RANKL binds to nuclear factor-kappaB (RANK) on the membrane surface of osteoclasts to drive osteoclast differentiation, cell survival, migration, and attachment (51). Excessive enhancement of bone resorption results in structural damage manifested by decreased bone mass and increased bone fragility. The adipogenic transcription factors C/EBPβ and C/EBPδ were demonstrated to bind to the RANKL promoter and stimulate the transcriptional expression of RANKL, a process that is accompanied by the downregulation of osteoprotegerin (OPG) which blocks the binding of RANKL and RANK (6). However, the mechanism of excessive RANKL production by adipocytes is not well clarified yet, which still needs further research. Inhibition of excessive RANKL secretion by adipocytes may be an important direction for the treatment of osteoporosis. In addition, low-level inflammation caused by the release of cytokines such as monocyte chemotactic protein-1 (MCP-1), also known as C-C motif chemokine ligand 2 (CCL2), tumor necrosis factor-α-associated apoptosis-inducing ligand (TRAIL), IL-6, and TNF-α under fatty infiltration enhances RANKL-induced NF-κB signaling, leading to an exaggerated response and inflammatory osteolysis (52). For instance, TNF-α is able to directly regulate the RANKL-induced signaling pathway, reduce the production of OPG and increase the concentration of RANKL (53). Even some studies have suggested that the presence of PA is sufficient to induce osteoclast differentiation and promote bone resorption regardless of the presence of RANKL in mice (54). So, whether it is the direct stimulation of osteoclasts or the alteration of the bone marrow niche caused by fatty infiltration, there is a promotion of bone resorption, resulting in a disruption of bone homeostasis.

Fatty infiltration disrupts the normal secretory function of murine long bone osteocytes, and several cytokines, including RANKL, Dickkopf1 (DKK1), and sclerostin, are affected by lipotoxicity. When PA-induced osteocytes were compared with normal culture groups, the fatty infiltration group showed significantly impaired cell viability, lipoapoptosis, and abnormal expression of RANKL, DKK1, and sclerostin (55). Lipotoxicity stimulates osteocytes to secrete sclerostin to activate FAS, a member of the TNF Superfamily Receptors, triggering osteoblasts and osteocytes apoptosis and inhibiting bone morphogenetic protein-induced osteoblastogenesis (56). Dkk1, a negative regulator of bone formation secreted by osteocytes, affects bone mass by inhibiting the Wnt pathway, which has a key role in osteogenesis and bone homeostasis (57). In addition, apoptosis of osteoblasts itself can stimulate osteoclast aggregation and thus enhance bone resorption, further contributing to bone loss.

Interestingly, PA-induced significant increase in lysosomal vital functional protein microtubule-associated protein light chain 3-II (LC3-II) in osteocytes indicated that lysosomal activity was indeed enhanced, but the co-localization of LC3-II and long terminal repeat (LTR) was found to be low after using the lysosomal tracker LTR (55), which means that the enhanced lysosomal activity did not lead to efficacious autophagy. This is at variance with some of the studies, but it is understandable that either the loss of osteocytes due to enhanced autophagy or the inability to clear impaired cells to the accumulation of malignant metabolites due to impaired autophagy can disrupt internal homeostasis and thus affect normal bone metabolism.

In general, the change in the bone marrow niche is closely linked to the raised level of MAT, and the presence of MAT no longer acts as a filler for bone loss, but can secrete relevant factors to influence the overall bone metabolism. Early adipocytes do have some advantageous functions, but with the superimposed effects of aging and various pathological factors, the negative effects of fatty infiltration will severely damage the bone marrow niche, leading to altered differentiation of BMSCs and damage to the above-mentioned intramedullary osteogenic lineage, ultimately resulting in bone loss. ( Figure 1 ).

Bone marrow adipocyte lineage cells are capable of expressing dipeptidyl peptidase 4 (DPP4), a protease associated with diabetes treatment (58), which is negatively correlated with bone mass, but no clear mechanism has been proposed (59). MAT also inhibits B-lymphopoiesis by stimulating the production of myeloid-derived suppressor cells and Granulocyte-macrophage colony stimulating factor (GM-CSF) in mice, after which the decrease in OPG production by B cells disrupts OPG/RANKL-mediated bone homeostasis (60).

In addition, numerous studies have found an important relationship between MAT and hematopoietic or blood vessels in the bone marrow. Shafat (61) and Mattiucci et al. (62) found MAT to actively transfer FFA to hematopoietic stem cells (HSC) as well as to secrete CXCL12 in vitro cultures, which plays an important role in the mobilization and maintenance of quiescence for HSCs. Zhong et al. (63) discovered a type of marrow adipogenic lineage precursor cell in mouse bone marrow capable of differentiating into adipocytes and increasing with aging. This cell type functions to maintain the marrow vasculature system and to prevent BMSCs from differentiate into osteoblasts. When it was selectively removed from mice, sinusoidal vessels were destroyed and the bone trabeculae were increased (64). Some scholars have suggested that the increase in adipocytes has a negative effect on hematopoiesis by disrupting the original cellular components within the bone marrow niche and intercalating within the hematopoietic milieu, which interferes with the connections between HSCs and other cells (31). Nevertheless, there are a limited number of studies addressing the exploration of the mechanisms of fatty infiltration on hematopoiesis and blood vessels, representing a crucial avenue that warrants further investigation in the future.

FA accumulated around mitochondria is susceptible to ROS-induced lipid peroxidation and these lipid peroxides may have toxicity effects on mitochondrial mtDNA, RNA, and proteins, leading to mitochondrial dysfunction (72).

ROS-induced endoplasmic reticulum stress is an important cause of skeletal muscle damage. Palmitate treatment in cultured human myotubes induces ER stress markers and decreases cell viability in an in vitro study (73). In contrast, oleate treatment (74) or PPARβ/δ activation (75) inhibits ER stress that could ameliorate PA-induced lipotoxicity. Firstly, endoplasmic reticulum stress induces lipolysis and exacerbates fatty infiltration by activating cyclic adenosine monophosphate/protein kinase A (cAMP/PKA) and extracellular signal-related kinase 1 and 2 (ERK1/2) signaling in adipocytes (76). Secondly, the endoplasmic reticulum is unable to provide sufficient protein due to its own impaired function, resulting in a reduced source of myogenesis, and the abnormal modified protein produced during the impairment activates both Caspase3 and fibrillar apoptosis. Thirdly, additional studies have demonstrated that PA-induced endoplasmic reticulum stress will also increase endoplasmic reticulum Ca²⁺ depletion, leading to further skeletal muscle damage from abnormal sphingolipid metabolism and palmitoylation of proteins. In addition, excessive accumulation of ROS in myblast cells activates the NF-κB pathway, leading to an increase in S100B, which induces myoblast transition into brown adipocytes, and negatively regulates miR-133, which has promyogenic and anti-adipogenic effects, through the NF-κB/YY1/miR-133 axis, accompanied by an increase in the brown adipogenic transcription regulator, PRDM-16 (77).

Interestingly, we found that type II fibers in skeletal muscle have stronger oxidative stress. The mitochondrial H2O2 produced/O2 consumed is 2 to 3 times higher in type IIB than normal for basal respiration supported by complex I or complex II substrates. The potential enhanced mitochondrial ROS production capacity of type II fibers (78), coupled with their own special superoxide topology that limits rapid dismutation of O2−•, renders type II fibers more susceptible to oxidative stress than type I fibers (71). This is an important reason for the high sensitivity of tissues to ROS and the poor resistance to lipid stress.

Adipocytes have the function of secreting a variety of cytokines, including pro-inflammatory factors, chemokines, and coagulation factors. Impaired adipose triglyceride lipase (ATGL) and PPARα-mediated lipid signaling pathways were found in skeletal muscle adipocytes in mice with sarcopenia, which resulted in upregulation of pro-inflammatory cytokines such as TNF-α and IL-6 (79). Low-level inflammation caused by fatty infiltration affects redox homeostasis, induces differentiation of adipose mesenchymal stem cells into macrophages and promotes further expression of inflammatory cytokines such as IL-6, TNF-α, IL-1β and MCP-1 (80). Synergy between macrophages and fatty acids also leads to skeletal muscle insulin function impairment (81). The release of certain cytokines enhances the signaling cascade and activates the NF-κB pathway, a pathway whose prolonged activation leads to severe muscle atrophy (82), and reduced insulin sensitivity affecting muscle protein synthesis (83). In addition, TNF-α acts as a strong promoter of TAG hydrolysis in adipocytes, inhibiting the antilipolytic effects of insulin and adenosine by inhibiting insulin receptor signaling and G protein-coupled adenosine receptors signaling, or interacting with the lipid-binding protein perilipin to directly stimulate lipolysis and exacerbate fatty infiltration (65). Impaired activation at the type II neuromuscular junction, excitation-contraction uncoupling, and disruption of the muscle fiber cross-bridge cycle can be observed under inflammation. Fatty infiltration contributes to inflammation, which in turn stimulates the continued accumulation of adipocytes, both of which persist in a vicious cycle and ultimately cause damage to skeletal muscle structure and function.

Fatty infiltration, pro-inflammation, oxidative stress, and the accumulation of toxic lipids ceramide and diacylglycerol can induce insulin resistance. While insulin resistance inhibits insulin-mediated antilipolytic capacity, increased plasma levels of FFA provide the precursor conditions for fatty infiltration (76). The release of chemokines such as adipose tissue-activated pattern recognition receptor (PPR) and MCP-1 that recruit macrophages and thus produce pro-inflammatory effects also induces insulin resistance. Insulin sensitivity is strongly negatively correlated with fatty acyl-CoA (FA-CoA) content, and increased FA-CoA due to insulin resistance exacerbates mitochondrial dysfunction via interfering with mitochondrial adenosine triphosphate synthesis through inhibition of the electron transport chain and reduction of the inner mitochondrial membrane potential (84). Increased ROS and local inflammation due to mitochondrial dysfunction leads to activation of JNK, IκB kinase (IKK), protein kinase C (PKC), and p38-mitogen-activated protein kinase (p38-MAPK) stress signaling pathways that inhibit the insulin PI3K/mTOR signaling pathway (85). In addition, the accumulation of lipid metabolites ceramide and diglycerides in skeletal muscle impairs muscle glucose uptake inducing insulin resistance through Phosphoinositide 3-kinase (PI3K) and Insulin Receptor Substrate-1 (IRS-1) signaling by activating protein kinase C (PKCs).

Insulin resistance affects glucose and amino acid transport through the PI3K/Akt/mTOR signaling pathway. PI3K promotes phosphorylation of the protein kinase AKT, and phosphorylated AKT activates TBC1D4 and TBC1D1 to enhance glucose uptake and mediate mTOR activation to affect protein synthesis. Meanwhile, AKT inhibition of Forkhead box O (FOXO) transcription factors activity affects E3 ubiquitin ligase expression, which is responsible for regulating muscular atrophy, and thereby interfers muscle function (82). It was also found that WNT/GSK/β-catenin, a key pathway regulating adipogenesis in FAPs, is stimulated by insulin signaling, and when GSK3 is pharmacologically blocked, it stabilizes β-catenin and inhibits PPARγ expression, thereby blocking adipogenesis in FAPs and limiting steatosis in vivo (86). In addition, ceramide and other sphingolipids can promote TNF-α and Caspace3/7-mediated apoptosis (87). In general, fatty infiltration-induced insulin resistance can impair skeletal muscle function in several ways.

In the fatty infiltration environment, impaired muscle fiber integrity, excitation-contraction uncoupling, and an imbalance of calcium homeostasis can be observed due to disruption of extracellular matrix remodeling (ECM), thus leading to skeletal muscle dysfunction and atrophy.

Excess FFA caused by fatty infiltration can disrupt myogenic differentiation and myotube formation via the AMPKα/HDAC4/miR-206 pathway in mice (88). Lipotoxicity can also induce the transformation of calcium cyclin isoforms on sarcoplasmic fibers, and excess lipids induce changes in phospholipid composition that affect Sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) activity and thus lead to impaired sarcoplasmic reticulum function. In addition, extracellular matrix remodeling and collagen deposition are triggered by lipotoxicity. Extracellular matrix is associated with high fat dietary intake and has an important role in skeletal muscle development, motor endplate function, and glucose metabolism (89), while increased collagen content in skeletal muscle increases the risk of insulin resistance (90). The two pathological changes impair the function of the sarcoplasmic reticulum and also affect the structural integrity of the sarcomeres, fibers, and neuromuscular junctions (71).

Fatty infiltration also affects the damage repair process of muscle fibers, especially MuSCs. Fatty infiltration leads to atrophy and a decrease in the number of MuSCs in type II fibers, suggesting that the stressful environment generated by fatty infiltration can limit the activation, proliferation, and differentiation of MuSCs, thereby diminishing their repair capacity. PAX7, an important marker gene of MuSCs, regulates the myogenic differentiation genes Myod and Myf5, while Myf5 and MyoD are the first two genes expressed in the gene family of the myogenic regulatory factors MRFs. The lack of both Myod and Myf5 results in a complete failure of muscle growth to initiate (91). Another study found that fatty infiltration induced differentiation of mesenchymal stem cells into macrophages, and Interferon-gamma (IFN-γ) secreted by macrophages also had important effects on skeletal muscle damage repair. INF-γ is required in acute muscle injury but is detrimental in chronic injury such as fatty infiltration (92). Increased INF-γ inhibits skeletal muscle damage repair in mice mediated by ERK1/2 and MAPK38 pathways. Moreover, under persistent inflammation, INF-γ impairs the regenerative process by suppressing myogenin expression, but this process can be neutralized by regulatory T cells, providing a potential direction for later therapy (93). In addition, FAPs also affect MuSCs and thus interfere with skeletal muscle damage repair. FAPs help maintain homeostasis within skeletal muscle and assists MuSCs in resolving limited damage in healthy muscle. However, changes in FAPs differentiation tendencies due to the disruption of niche cues can disrupt this balance (94). FAPs themselves have an important role in the development of fatty infiltration which may be the main resident muscle stem cell population involved in the development of ectopic obesity in human skeletal muscle. In a mouse regeneration model, only progenitor cells expressing the cell surface receptor platelet-derived growth factor receptor alpha (PDGFRα) were able to differentiate into adipocytes after glycerol-induced muscle injury, while in the muscle injury state, PDGFRα was expressed only by FAPs, demonstrating the strong lipogenic capacity of FAPs in vivo (95). Paracrine factors and soluble molecules secreted by FAPs can regulate the fate of MuSCs. A mouse study found that loss of FAPs-derived WNT1 Induced Signaling Pathway Protein 1 (WISP1) during aging affects the expansion and asymmetric commitment of MuSCs through the Akt signaling pathway (96). Alessio et al. found that inhibition of the key WNT/GSK/β-catenin pathway of FAPs adipogenesis in mice skeletal muscle using an inhibitor was effective in stimulating the differentiation of MuSCs into mature myotubes, thereby improving the pro-myogenic effects of FAPs (97). With a shift of FAPs more towards lipogenesis, the depletion of FAPs can’t maintain muscle mass and support MuSCs proliferation and differentiation during muscle regeneration (98).

Fatty infiltration affects intercellular signaling pathways within skeletal muscle regulating myofiber growth and death. Muscle mass is determined by the balance between protein synthesis and degradation. Skeletal muscle proteolysis is mainly mediated by the ubiquitin-proteasome system (UPS) and the autophagy-lysosomal pathway (ALP) under pathological conditions. Muscle atrophy F-box and MuRF1 are two muscle-specific E3 ubiquitin ligases in UPS system, which are specifically expressed and target specific protein substrates for degradation (99). FOXOs activate both UPS and ALP to ensure that the loss of different cellular components during muscle atrophy is coordinated. Of which, FOXO3 is a key regulator of autophagy as well as controls the expression of other autophagy-related genes (100). Protein synthesis is mainly regulated by the Insulin-like growth factor-1 (IGF-1)/PI3K/Akt/mTOR signaling axis (101). mTOR is an essential factor mediating skeletal muscle growth, the activation of which inhibits UPS and ALP. Nutrients, mechanical and neural signals activate mTOR, while inflammatory cytokines and unfolded proteins inhibit mTOR activity (102). Insulin/IGF-1, which are upstream of mTOR, indirectly regulate muscle growth. Under fatty infiltration, many key inflammatory mediators are involved in skeletal muscle atrophy. High levels of lipid metabolites such as TNF-α, IL-6 and IGF-1 activate MuRF-1 and MAFBx/Atrogin-1, key transcription factors of muscle atrophy, through the IGF/Akt-1 pathway. In additional, TNF-α regulates apoptosis by inhibiting the Akt/mTOR pathway (103). Its lead to prolonged activation of the NF-κB pathway can also lead to severe muscle atrophy.

Additional pathways involved in fatty infiltration are AMP-activated protein kinases (AMPKs), mitogen-activated protein kinases (MAPKs) and histone deacetylases (HDACs), etc. AMPK is a positive regulator involved in glucose transport, mitochondrial function, and fatty acid oxidation, as well as a negative regulator of inflammation, oxidative stress, ceramide and DAG production, and mTOR-related pathways. MAPKs are serine/threonine-protein kinases, and recent studies have demonstrated that the MAPKs family has a central role in the regulation of adipose tissue function. ERK1/2, JNKs, and p38s, which are involved in endoplasmic reticulum stress and insulin resistance mentioned above, are all important members of this family (104). Moreover, HDACs mediated FOXO nuclear localization and activation and enhanced FOXO activity and function through deacetylation, while HDACs regulated the expression of the Atrogin-1 (105). Fatty infiltration involves the activation of multiple signaling pathways within skeletal muscle, affecting skeletal muscle growth, lipolysis, and regulating the expression of genes associated with skeletal muscle atrophy.

Lipotoxicity impairs mitochondrial function, increases ROS release, persistent low level chronic inflammation and insulin resistance will disrupt fibrin homeostasis and affect skeletal muscle generation, apoptosis and damage repair, and these major functional lesions will in turn further stimulate lipid accumulation, creating a vicious cycle. The structural and functional specificity of type II fibers is such that they differ markedly from type I fibers in their level of response to fatty infiltration ( Figure 2 ).