The endogenous production of MG primarily depends on the glycolytic pathway [9]. During glycolysis, glucose is gone through a series of enzymatic reactions converting into glyceraldehyde-3-phosphate (G3P) and dihydroxyacetone phosphate (DHAP). These two triose phosphates undergo non-enzymatic β-elimination, losing a phosphate group to form MG. In addition to the glycolytic pathway, MG can be generated through several other routes, including glucose auto-oxidation, degradation of glycated proteins, catabolism of ketone bodies during diabetic ketoacidosis, threonine catabolism, lipid peroxidation, and the metabolic activity of gut microbiota [14][15]. Under physiological conditions, the conversion of glucose to MG is inefficient, and the body possesses robust MG clearance mechanisms. Consequently, MG is maintained at a low, dynamic equilibrium without exerting significant cytotoxicity.

The metabolic clearance of MG relies primarily on the glyoxalase system. This system consists of two key enzymes, including glyoxalase I (GLO I) and glyoxalase II (GLO II). MG undergoes a non-enzymatic reaction with glutathione (GSH) to form a hemithioacetal. Subsequently, GLO I catalyzes the conversion of this hemithioacetal to S-D-lactoylglutathione. Finally, GLO II hydrolyzes S-D-lactoylglutathione to D-lactic acid, regenerating GSH and completing the detoxification process [16]. Moreover, the body possesses other auxiliary metabolic pathways that participate in MG clearance, such as aldose reductase (ALR), aldehyde dehydrogenases (ALDHs), and the Parkinson’s disease-associated protein DJ-1 [17][18]. These enzymes contribute to MG clearance by converting it into non-toxic or low-toxicity products via distinct mechanisms, further ensuring the capacity for MG clearance.

Under diabetic hyperglycemic conditions, the balance between MG production and clearance is disrupted, leading to its abnormal accumulation in cells and tissues. This accumulation is critically involved in the pathogenesis of diabetic chronic complications, including DOP. MG levels in the plasma and tissues (such as bone, kidney, and liver) of diabetic patients are 2 to 6 times higher than in healthy individuals [19][20]. The increase in MG arises from a combination of increased production and impaired clearance. This mutual reinforcement establishes a vicious cycle that further exacerbates MG-induced toxicity.

2.2.1. Aberrant MG Generation

The elevated rate of MG generation under hyperglycemic conditions underlies its increased production, driving its subsequent accumulation. Under physiological conditions, the amount of MG precursors (G3P and DHAP) generated from glycolysis is limited, thereby maintaining low MG production. Under diabetic hyperglycemic conditions, elevated intracellular glucose levels result in excessive activation of the glycolytic pathway, leading to a marked increase in glycolytic flux. This results in the substantial accumulation of triose phosphate substrates like G3P and DHAP [14]. These precursors are rapidly converted to MG via non-enzymatic β-elimination, driving a marked increase in MG production and resulting in levels that far exceed the normal physiological range. Moreover, hyperglycemia also activates additional MG-generating pathways, including glucose auto-oxidation and the degradation of glycated proteins, which in turn further elevate MG production and promote its accumulation.

2.2.2. Impaired MG Metabolism

The hyperglycemic environment significantly compromises the body’s capacity to metabolize and clear MG, contributing to its accumulation. This metabolic impairment is mediated by several molecular mechanisms. First, hyperglycemia-induced oxidative stress can directly inhibit the expression and activity of GLO I. Reduced GLO I activity directly compromises the system’s ability to metabolize and clear MG, preventing its timely conversion to non-toxic products [21]. The hyperglycemic environment enhances oxidative stress, leading to significant consumption and decreased levels of GSH. GSH deficiency limits the non-enzymatic reaction between MG and GSH, reducing the efficiency of MG detoxification [22]. Under hyperglycemic conditions, advanced glycation end products (AGEs) accumulate and further suppress glyoxalase system function via feedback regulation. This establishes a vicious cycle in which MG accumulation drives AGE formation, and subsequent glyoxalase inhibition further promotes MG accumulation [23]. Hyperglycemia also suppresses the activity of other auxiliary MG-metabolizing enzymes, such as ALR and ALDHs [20][24][25], further compromising overall MG clearance.

The cellular toxicity of MG is primarily attributed to its high electrophilicity. This property enables it to undergo non-enzymatic nucleophilic addition reactions with various intracellular biological macromolecules, including proteins and nucleic acids. By modifying the structure and function of these macromolecules, MG triggers cellular damage and dysfunction, which can be categorized into three main aspects.

2.3.1. MG-Induced Protein Modification

MG primarily reacts with arginine residues, and to a lesser extent with lysine and cysteine residues [26]. These reactions form covalent adducts that generate various AGEs, such as argpyrimidine and methylglyoxal-lysine dimer (MOLD) [27][28]. These modifications directly alter the spatial conformation and function of proteins, resulting in loss of enzymatic activity, abnormal receptor binding capacity, and protein cross-linking and aggregation. Collectively, these changes impair normal cellular physiological functions [29][30].

2.3.2. MG-Induced Nucleic Acid Modification

MG can react with guanine bases in DNA through nucleophilic addition, forming DNA adducts such as MG-guanine adducts. These modifications can lead to DNA strand breaks, nucleotide transversions, DNA-DNA cross-links, and DNA-protein cross-links, thereby inducing gene mutations, chromosomal damage, and ultimately leading to apoptosis or abnormal cell proliferation [31][32]. Furthermore, MG modification of RNA can also affect gene transcription and translation processes, further exacerbating cellular dysfunction [33].

MG can significantly increase intracellular ROS levels through both direct and indirect pathways, a core mechanism underlying its induction of oxidative stress.

3.1.1. Direct Pathways

The production and metabolism of MG are directly accompanied by the generation of ROS. The conversion of MG to pyruvate catalyzed by glyoxal oxidase is accompanied by the generation of H₂O₂ [34]. Similarly, the conversion of aminoacetone to MG by semicarbazide-sensitive amine oxidase (SSAO) also releases H₂O₂ in rat aortic vascular smooth muscle cells [35]. MG can induce conformational changes in hemoglobin in human umbilical vein endothelial cells (HUVECs), promoting hemoglobin glycation. The resulting hemoglobin-advanced glycation end products (Hb-AGEs) significantly increase ROS production in HUVECs [36]. These generated ROS can directly attack intracellular biomacromolecules, initiating oxidative damage.

3.1.2. Indirect Pathways

MG can also indirectly increase ROS accumulation by activating intracellular ROS-generating systems. A primary mechanism is the activation of the NADPH oxidase (NOX) family [37]. MG can induce the expression of inducible nitric oxide synthase (iNOS) in macrophages, which generates nitric oxide (NO), a reactive nitrogen species (RNS) that triggers nitrosative stress, a distinct but often concurrent form of redox imbalance [38]. Moreover, MG can bind to the receptor for advanced glycation end products (RAGE) on the cell surface, increasing oxidative stress and inflammation, contributing to various diseases [39]-[41]. MG can upregulate iNOS expression via p38 MAPK-NF κB pathway, stimulating iNOS activation and generating ROS through the activation of ERK and JNK in endothelial cell [42]. Additionally, MG promotes the generation of nitric oxide (NO), which rapidly reacts with superoxide anion ( · O 2 − ) to form the potent oxidant peroxynitrite (ONOO⁻). ONOO⁻ has stronger oxidizing properties than ROS, further exacerbating intracellular oxidative damage by attacking DNA, proteins, and lipids, leading to cellular dysfunction [38].

MG directly depletes intracellular GSH, which is the most abundant non-enzymatic antioxidant in cells. The strong nucleophilicity of the thiol (-SH) group of GSH facilitates its nucleophilic addition with MG to form a hemithioacetal, leading to significant GSH consumption [43]. MG directly depletes GSH and inhibits the activity of key enzymes involved in its synthesis, such as γ-glutamylcysteine ligase (γ-GCL). Consequently, the cellular antioxidant defense system is weakened, leading to exacerbated oxidative stress and mitochondrial dysfunction, which ultimately cause cell injury [44][45]. GLO I is a key enzyme for MG detoxification, and its activity directly impacts MG clearance efficiency. Excess MG can also inhibit GLO I activity. When GLO I activity is suppressed or deficient, MG clearance is compromised, leading to its accumulation. This accumulation subsequently triggers the AGEs-RAGE-NF-κB signaling axis, culminating in inflammation and oxidative stress [46][47]. By inactivating GLO I, MG disrupts its primary detoxification pathway, leading to MG accumulation. This accumulation subsequently induces oxidative stress and mitochondrial dysfunction in osteoblastic MC3T3-E1 cells [48]. MG inhibits the nuclear translocation and transcriptional activity of Nrf2. This suppresses the Nrf2-mediated antioxidant defense system. As a result, the expression of antioxidant enzymes, including heme oxygenase-1 (HO-1), is downregulated. This further diminishes the cellular antioxidant capacity, leading to increased ROS production and exacerbating oxidative stress injury in myotubes [49].

3.3.1. AGEs-RAGE Axis and Inflammatory Signaling Pathway

AGEs formed by the non-enzymatic reaction of MG with intracellular proteins can bind to RAGE on the cell surface, activating the AGEs-RAGE axis and subsequently regulating multiple downstream signaling pathways. This is a key mechanism by which MG induces oxidative stress injury in bone cells [50][51]. RAGE is a pattern recognition receptor widely expressed on the surface of various cells, including osteoblasts, osteoclasts, and bone marrow stromal cells [52][53]. Binding of AGEs to RAGE can promote ROS generation by activating the NOX family, while simultaneously activating NF-κB and the MAPK cascade, initiating a vicious cycle of inflammatory response and oxidative stress [54][55]. MG can activate the MAPK signaling pathway via the AGEs-RAGE axis, specifically activating the p38 and JNK signaling pathways, inducing the upregulation of osteogenic differentiation-related proteins such as RUNX2 and OPG [56]. Soluble RAGE (sRAGE) is a splice variant of RAGE that functions as a competitive inhibitor. It binds to AGEs in competition with membrane-bound RAGE, thereby blocking activation of the AGEs-RAGE axis and reducing oxidative stress and inflammation. Vitamin D treatment may possibly be beneficial to reduce AGE levels and to augment sRAGE levels in populations with vitamin D-deficient situations [57]. Studies have shown that in T1DM patients the sRAGE/RAGE ratio was significantly decrease, leading to overactivation of the RAGE signaling pathway and further exacerbating cellular oxidative stress injury on osteoblastic MC3T3-E1 cells [58][59].

3.3.2. Cell Death Signaling Pathways

MG-induced oxidative stress can induce bone cell apoptosis through both the mitochondrial pathway and the extrinsic death receptor pathway, a critical mechanism underlying MG-induced dysfunction in bone cells. MG increases the Bax/Bcl-2 ratio, activates cleaved caspase-3, and inhibits the PI3K/Akt and Nrf2/HO-1 antioxidant signaling pathways in endothelial cell. Consequently, massive ROS production occurs, amplifying oxidative stress injury in a cascade and ultimately leading to apoptosis [60]. MG can also induce inflammatory pyroptosis by activating caspase-3, which subsequently cleaves gasdermin E (GSDME). Concurrently, by promoting the cleavage of non-muscle myosin heavy chain IIA (NMMHC IIA) and affecting the phosphorylation state of myosin light chain 2 (MLC2), MG further exacerbates cytoskeletal damage and vascular dysfunction in HUVECs [61]. Through ROS generation, MG activates the MAPK p38 signaling pathway and upregulates the transcription of the E3 ubiquitin ligase FBXO32, which catalyzes the ubiquitination and degradation of the transcription factor KLF4. This inhibits the transcription of the macrophage phagocytic receptor MerTK, ultimately impairing the ability of macrophages to clear apoptotic cells and hindering the resolution of inflammation and tissue repair [62].

Mitochondria are central to cellular energy metabolism and are both primary sites of ROS generation and primary targets of ROS attack. MG-induced mitochondrial dysfunction manifests primarily as impaired mitochondrial respiratory chain function, altered mitochondrial dynamics, abnormal mitophagy, and impaired mitochondrial biogenesis. This dysfunction leads to the release of cytochrome c (cytC) from the mitochondrial matrix into the cytoplasm, where it binds to apoptotic protease activating factor 1 (Apaf-1), activating caspase-9 and subsequently downstream caspase-3, ultimately triggering synaptic apoptosis [63]. MG induces mitochondrial dysfunction by downregulating the PI3K-Akt and Nrf2-HO-1/NQO-1 antioxidant signaling pathways, decreasing mitochondrial membrane potential, and increasing lactate dehydrogenase (LDH) release. This activation of apoptotic markers such as p-P38, cleaved caspase-9/3, and Bax further leads to PC-12 cells apoptosis [64]. MG also reprograms metabolism in myoblasts, inhibiting the tricarboxylic acid (TCA) cycle and promoting fatty acid synthesis, while inducing mitochondrial morphological abnormalities. This results in a significant reduction in ATP production from mitochondria, triggering a cascade of oxidative stress responses and ultimately leading to myoblast cellular dysfunction and injury [65]. MG suppresses the expression of the mitochondrial antioxidant enzymes IDH2 and MGST1 in epithelial cells, while simultaneously downregulating the fusion protein MFN2 and upregulating the fission protein FIS1, leading to an imbalance in mitochondrial fusion and fission. This dynamic imbalance further exacerbates the loss of mitochondrial membrane potential and increases ROS production, creating a vicious cycle of oxidative stress and mitochondrial dysfunction [66]. MG induces excessive mitochondrial fission by downregulating the fusion proteins OPA1 and MFN1, leading to mitochondrial network fragmentation and subsequent dysfunction, thereby causing cellular injury. Enhancing GLO1 expression can restore OPA1 and MFN1 levels, reversing mitochondrial fission and mitigating endothelial cells’ damage [67]. Increased mitochondrial fission and decreased fusion not only impair mitochondrial energy metabolism but also enhance the release of mitochondrial ROS (mtROS), further exacerbating cellular oxidative stress injury in a vicious cycle. Additionally, MG activates Parkin-1-mediated mitophagy. However, this excessive mitophagy disrupts mitochondrial homeostasis, leading to the downregulation of tight junction proteins (occludin, claudin-5, ZO-1) and compromising the barrier integrity of brain endothelial cells [68]. MG can also downregulate the expression of SIRT1 and PGC-1α, key regulators of mitochondrial biogenesis. This impaired biogenesis leads to a reduction in both the number and quality of mitochondria, while simultaneously weakening the activity of the endogenous antioxidant enzyme system (SOD, GPX). This creates an imbalance between ROS production and clearance, ultimately resulting in mitochondrial damage and cellular dysfunction [69].

Bone homeostasis refers to the dynamic equilibrium maintained by osteoblast-mediated bone formation and osteoclast-mediated bone resorption, ensuring the stability of bone mass and microarchitecture. This balance is fundamental to normal skeletal physiology. Osteoblasts and osteoclasts, as the two key effector cells in this process, must maintain functional balance to preserve normal bone mass and structure [70][71]. In the pathological process of DOP, the hyperglycemic environment and its secondary metabolic disturbances (such as MG accumulation, oxidative stress, and inflammation) disrupt this balance by suppressing osteoblast number and function while enhancing osteoclast differentiation and activity, leading to progressive bone loss and increased fracture risk [72].

Both type 1 (T1DM) and type 2 (T2DM) diabetes increase fracture risk, but via distinct bone phenotypes. T1DM typically exhibits low bone turnover and reduced bone mineral density (BMD), whereas T2DM often shows normal or high BMD with impaired bone quality [73]. Clinical and translational evidence directly links MG/AGE burden to adverse bone outcomes in both types. Elevated MG and AGE levels correlate with lower BMD in T1DM, reflecting low bone turnover and reduced bone formation. Bone biopsy studies in T1DM patients show increased AGE deposition in collagen, which contributes to reduced bone toughness and increased fragility [74]. The femoral cortical bone specimens from T1DM patients with >50 years disease duration exhibits 17% higher AGE levels, 30% lower bone toughness and decreased mineral crystallinity [75]. In patients with T2DM, skin autofluorescence-derived accumulation of AGEs is positively associated with low bone density/osteoporosis and major osteoporotic fractures. Elevated AGE levels also correlate with increased bone turnover markers (PINP, CTX), indicating that AGEs accumulation is an important risk factor for bone damage in T2DM [76]. The MG serum levels and gene expression of RAGE in PBMCs are higher in T2DM patients than control groups [77][78]. Clinical translational evidence reveals that MG/AGE burden independently correlates with low bone mineral density in T1DM and with fracture risk in T2DM, implying that MG plays a pathogenic role in both diabetic populations, yet the underlying mechanisms differ.

Osteoblasts, as the core effector cells for bone formation, are central to the insufficient bone formation seen in DOP. MG-induced oxidative stress can damage osteoblasts through multiple specific mechanisms, primarily by inhibiting osteoblast differentiation and mineralization, disrupting the cytoskeleton and adhesion, and promoting apoptosis. These effects collectively lead to impaired bone formation [79][80]. MG treatment causes a significant decrease in mitochondrial membrane potential, reduced ATP production, and increased levels of mtROS and intracellular Ca²⁺ in osteoblasts. The increase in mtROS further damages the mitochondrial respiratory chain. CypD, a key component of the mitochondrial membrane potential, plays a crucial regulatory role in this process. PGAM5 acts as an upstream signaling molecule for CypD. MG can activate PGAM5, which in turn regulates CypD activity, promoting mPTP opening and leading to mitochondrial dysfunction and osteoblast injury [81]. MG induces the generation of intracellular ROS and mitochondrial superoxide in osteoblasts, while also promoting cardiolipin peroxidation. Additionally, it reduces GSH levels, inhibits GLO I and HO-1 activities, and downregulates Nrf2 expression. Collectively, these effects trigger mitochondrial dysfunction and inflammation, ultimately resulting in oxidative stress injury [59]. These findings suggest that targeting mitochondrial function may be an effective strategy to inhibit MG-induced osteoblast damage. Furthermore, treatment of MC3T3-E1 osteoblasts with MG dose-dependently reduces cell viability, significantly inhibits the expression of osteogenic differentiation markers (ALP, OCN), and decreases the formation of mineralized nodules, indicating that MG directly suppresses osteoblast proliferation and differentiation and impairs their mineralization function [23].

Current research suggests that MG exerts a bidirectional regulatory effect on RANKL-induced osteoclast differentiation, which may be closely related to factors such as MG concentration, cell type, and exposure time. This regulation also involves various specific mechanisms, including mitochondrial dysfunction and metabolic reprogramming, ultimately influencing osteoclast differentiation and bone resorption activity [82]. Under hyperglycemic conditions, increased MG production simultaneously inhibits the expression of osteoblast markers (ALP, OCN, OPG) and downregulates the osteoclast-associated receptor (OSCAR), thereby impairing the differentiation function of both osteoblasts and osteoclasts and leading to delayed bone defect healing [83]. It has been reported that treatment of RAW264.7 cells with 300 - 600 µM MG for 4 days during RANKL-induced osteoclastogenesis significantly inhibited TRAP activity and osteoclast differentiation. MG significantly inhibits RANKL-induced osteoclast differentiation by inhibiting GLO I activity, leading to impaired MG metabolism, inducing mitochondrial dysfunction, and decreasing intracellular calcium concentration [84]. However, other studies have shown that treatment of RAW264.7 cells with a lower concentration of MG (50 - 200 µM) for 3 days enhanced osteoclast differentiation via JNK activation and upregulating the expression of osteoclast marker genes CTSK, OSCAR, and TRACP5, suggesting that MG mediates osteoclast activation via the JNK pathway, thereby contributing to the bone remodeling imbalance in diabetes-related osteoporosis [85]. This bidirectional effect may be closely related to MG concentration, cell type (e.g., osteoclast precursors vs. mature osteoclasts), and exposure time. High-dose MG (e.g., 300 - 600 µM) may inhibit osteoclast differentiation by inducing excessive oxidative stress, while moderate oxidative stress induced by low-dose MG (e.g., 50 - 200 µM) may act as a signal promoting osteoclast differentiation. Additionally, differences in cell models (RAW264.7 vs. bone marrow-derived macrophages), differentiation stage (early vs. late osteoclastogenesis), and exposure timing (continuous vs. intermittent) may also contribute to the observed discrepancies. The specific molecular mechanisms underlying this bidirectional regulation are not yet fully understood and require further investigation.

The adhesion capacity and cytoskeletal integrity of osteoblasts are fundamental to their bone-forming function, as osteoblasts need to adhere to the bone matrix surface to synthesize and mineralize it. MG can also modify type I collagen in the bone matrix, altering the biomechanical properties of the bone and increasing its fragility [86][87]. Type I collagen is the main component of the bone matrix, and its structural integrity directly affects the mechanical strength of the skeleton. The non-enzymatic reaction of MG with type I collagen to form AGEs can lead to collagen cross-linking and aggregation, altering its physicochemical properties. This not only reduces the mechanical strength of the bone but can also diminish the adhesion capacity of osteoblasts to the bone matrix proteins, further inhibiting the bone formation process and exacerbating bone loss and structural damage [88].

Since oxidative stress is a central role in MG-induced bone cell injury, antioxidants represent a potential strategy for intervening in DOP. N-acetylcysteine (NAC), a classic antioxidant and precursor of GSH, can significantly increase intracellular GSH levels, enhancing cellular antioxidant capacity [89]. NAC effectively ameliorates MG-induced mitochondrial dysfunction, osteoblast apoptosis, and impaired osteogenic differentiation. Studies show that NAC pretreatment significantly reduces MG-induced ROS levels in osteoblasts, restores mitochondrial membrane potential and ATP production, decreases the number of TUNEL-positive cells, and upregulates the expression of osteogenic marker genes like ALP and OCN, thereby improving osteoblast differentiation and mineralization function [81]. Trolox, a water-soluble analog of vitamin E, possesses potent antioxidant activity and can mitigate cellular damage by scavenging ROS and inhibiting oxidative stress [90]. Additionally, natural antioxidants like vitamin C and vitamin E protect bone cell function by alleviating oxidative stress injury [91][92], offering simple and feasible strategies for DOP intervention.

Many natural small molecule compounds, characterized by their multi-target effects and low toxicity, exhibit significant protective effects against MG-induced bone cell injury, representing a potential source for DOP drug development. Glabridin, a major active component of licorice, possesses strong antioxidant and anti-inflammatory properties. Glabridin alleviates oxidative stress-induced osteoblast apoptosis by activating the Akt signaling pathway, subsequently phosphorylating and regulating downstream GSK-3β and NF-κB p65 proteins [93]. Crocin, a major active component of saffron, exhibits significant antioxidant and mitochondrial protective effects. Crocin significantly restores succinate dehydrogenase (SDH) activity, reverses the loss of mitochondrial membrane potential and mitochondrial swelling, while reducing ROS generation, malondialdehyde (MDA) levels, and maintaining GSH content. By preserving mitochondrial redox homeostasis and energy metabolism, crocin protects cells from oxidative stress injury [94].

Mitochondrial dysfunction is a central link in MG-induced oxidative stress injury in bone cells, making mitochondrial-targeted therapy a new direction for DOP intervention. Cyclosporin A (CsA) is a specific inhibitor of the mPTP. By blocking mPTP opening, CsA protects mitochondrial membrane potential, reduces cytochrome c release, and inhibits mitochondrial pathway-mediated apoptosis, thereby preventing MG-induced osteoblast toxicity. Studies have shown that CsA pretreatment significantly restores MG-induced mitochondrial function in osteoblasts, reduces ROS production, increases cell viability, and improves osteoblast differentiation function [81]. Furthermore, mitochondrial-targeted antioxidants, such as MitoQ, can specifically target mitochondria, scavenge mtROS, and protect mitochondrial function, effectively inhibiting osteoclast differentiation and osteogenic mineralization function [95][96], providing new avenues for DOP intervention.

Since abnormal MG accumulation is a prerequisite for MG-induced bone cell injury, directly clearing excess MG in the body using MG scavengers to reduce its toxic effects on bone cells is a potential strategy. Aminoguanidine is a classic MG scavenger that specifically reacts with MG to form non-toxic adducts, reducing MG accumulation in the body. It can reverse MG-induced ROS generation, mitochondrial dysfunction, and cytotoxicity in osteoblasts, block MG-triggered oxidative stress and cellular dysfunction, upregulate the expression of osteogenic marker genes like ALP and OCN, and improve osteoblast differentiation and mineralization function, thereby restoring the osteogenic differentiation capacity of BMSCs [23].