The primary function of mitochondria is the production of ATP through oxidative phosphorylation. Oxidative phosphorylation is a process during which successive redox reactions are performed alongside enzymes that form an electron transport chain (ETC) in the inner mitochondrial membrane, a result of which is the formation of a proton gradient and the generation of energy during the reverse proton transfer, which is stored in the form of ATP. Electron carriers (nicotinamide adenine dinucleotide hydrogene (NADH) and flavin adenine dinucleotide dihydrogene (FADH2)) are formed in the mitochondrial matrix during the tricarboxylic acid (TCA) cycle and precede oxidative phosphorylation [7]. These carriers transfer electrons to the ETC, consisting of large protein complexes—enzymes (I–IV)—and two additional carriers: Coenzyme Q and cytochrome C. Complex I is the first and largest of the complexes, consisting of 45 main subunits. Complexes I and II mediate the transfer of two electrons from electron carriers NADH/FADH2, respectively, to coenzyme Q, which carries mobile electrons. The latter can also receive electrons from the breakdown of many nutrients, for example, during the oxidation of fatty acids. Complex III is an adapter that receives two electrons from the reduced coenzyme Q and transfers one electron to cytochrome C. Complex IV completes the respiratory chain by accepting electrons from cytochrome C to reduce oxygen to water completely [8]. Redox reactions in the ETC lead to conformational changes in protein complexes, which facilitate the transfer of protons from the matrix into the intermembrane space to produce a transmembrane potential [9]. As a result of the proton gradient, the energy in the transmembrane potential is used by ATP synthase to form ATP from adenosine diphosphate (ADP). In addition to the formation of ATP, additional processes depend on the transmembrane potential, including the importation of mitochondrial proteins through the inner mitochondrial membrane [7].

Glutamine is considered the most abundant amino acid in nature. The process of glutamine uptake by cells occurs with the help of special transmembrane transporter proteins [10]. Once in cells, glutamine is used in translational processes to synthesize proteins and nucleotides, creating a chemical potential for the uptake of other metabolites. [11]. Thus, the essential amino acid leucine is absorbed by the cell through the LAT1 protein carrier, while the secretion of glutamine into the intercellular space is also required for the transfer to occur [12].

The greatest amount of glutamine inside the cell is transformed into glutamate and ammonium ions during a reaction involving the enzyme glutaminase. Part of the mitochondrial glutamate is exported to the cytosol, where it attaches to the glutamate pool, and the other part of the glutamate will be further converted to $\alpha$-ketoglutarate by glutamate dehydrogenase, which releases ammonia. This interconversion between glutamate and $\alpha$-ketoglutarate is catalyzed by the enzyme aspartate aminotransferase (AST). However, instead of releasing ammonia, this reaction leads to the transfer of the amino group from glutamate to oxaloacetate and the synthesis of aspartate, which is the amino acid necessary for nucleotide synthesis [10].

Fatty acid (FA) oxidation is a mitochondrial aerobic process that breaks down fatty acids into acetyl-CoA. Fatty acids are involved in this pathway as CoA derivatives through nicotinamide adenine dinucleotide (NAD) and flavin adenine dinucleotide (FAD) [13]. The importation of mitochondrial FAs is a defining step in fatty acid oxidation (FAO) and demonstrates how metabolic compartmentalization adapts to the cellular state. Since long-chain FAs cannot cross mitochondrial membranes, the mitochondria have developed a complex set of reactions and transport activities that allow FAs to gain access to the mitochondrial $\beta$-oxidation mechanism. The outer mitochondrial membrane (OMM) enzyme carnitine palmitoyl transferase 1 (CPT1) forms acylcarnitines from acyl-CoA fatty acids [10]. These acylcarnitines are translocated into the mitochondria, via the carnitine acylcarnitine translocase (SLC25A20), to the inner mitochondrial membrane (IMM). CPT2 releases fatty acids from carnitine, initiating free fatty acids (FFAs). Acetyl-CoA from the oxidation of fatty acids (OFAs) is used in the TCA and for synthesizing aspartate and nucleotides [14].

FFAs are extremely important in cellular homeostasis since FAs form the main energy substrate of cytotoxic T cells, necessary for their survival and the functioning of CD8${}^{+}$ T cells; in contrast to the effector cells involved in innate immunity, which use mainly glucose and glutamine as an energy source [15]. Similarly, FFA is activated in insulin resistance as a compensatory mechanism for the lack of glucose in cells [16].

Pyruvate is formed from a number of sources, depending on nutrient and tissue availability, including the main source of glucose and lactate. In healthy tissue, the fate of pyruvate depends on the availability of oxygen and the respiratory capacity of mitochondria [17]. In a state of sufficient oxygen levels, pyruvate is synthesized during glycolysis and passes through the IMM via the mitochondrial pyruvate transporter (MPC) [18]. Pyruvate is further catabolized within the mitochondria in the TCA cycle. During hypoxia, mitochondrial respiration is suppressed, causing cells to adaptively transfer electrons to pyruvate through lactate dehydrogenase (LDH), forming lactate in the cytosol [10]. This pathway is involved in the muscles, the intestines, and the kidneys during exercise. Within mitochondria, pyruvate is introduced into the TCA cycle through reactions involving two enzymes: the pyruvate dehydrogenase complex (PDC), which is responsible for the synthesis of acetyl-CoA, and pyruvate carboxylase (PC), which catalyzes oxaloacetate [19].

Increased glucose production in the liver is one of the reasons that patients with T2DM develop hyperglycemia. Insulin controls glucose production in the liver and promotes glucose utilization in the skeletal muscles [20]. In patients with T2DM, there is an increase in gluconeogenesis and, to a lesser extent, glycogenolysis in the liver. Factors contributing to the acceleration of glucogenesis are the increased production of lactate, alanine, and glycerol, which are gluconeogenic precursors, as well as hyperglucagonemia, accompanied by an increase in fatty acid oxidation. In addition, disruption of postprandial glucose homeostasis is associated with reduced suppression of hepatic glucose production following carbohydrate intake [21].

An increase in glycogenolysis and a decrease in hepatic glucose uptake by glucagon result in a hyperglycemic phenotype, which is determined by insulin deficiency and insulin resistance (IR). In healthy individuals, fasting plasma glucose levels remain unchanged throughout the day. This constant glucose level is highly dependent on the balance between glucose produced in the liver and glucose absorbed by peripheral tissues [21]. In T2DM, fasting hyperglycemia correlates to a lesser extent with increased hepatic glucose production due to decreased hepatic sensitivity to insulin. However, this is largely due to a reduction in insulin secretion and an increase in glucagon secretion. Although basal levels of immunoreactive insulin and glucagon may be normal in T2DM patients, testing the islet function at appropriate glucose levels reveals various abnormalities in insulin and glucagon secretion due to decreased $\beta$-cell secretory capacity and impaired ability of glucose to suppress glucagon release [22]. There is a decrease in the efficiency of glucose uptake by peripheral tissues, which results from a combination of reduced insulin secretion and a defect in the cellular action of insulin [21]. Impaired insulin secretion induced by glucose in the first phase is accompanied by fasting hyperglycemia, and progressive insufficiency of pancreatic $\beta$-cell function is accompanied by an increase in glucose levels to maintain basal and second-phase insulin production. Ultimately, glucose directly regulates the synthesis and secretion of insulin and weakens all other signaling pathways that occur in pancreatic cells, which include metabolic reactions involving other substrates, hormone synthesis, etc. [23].

The dysfunction of $\beta$-cells is associated with a disruption in the first phase of insulin production, which occurs during glucose stimulation and is a precursor to the development of glucose intolerance in T2DM [24]. The occurrence of the insulin response is closely related to the transmembrane transport of glucose and its interaction with the sensors. The next step is where the sensor complex initiates an increase in glucokinase activity following the stabilization of the glucokinase structure, preventing its degradation. Glucose transport in the $\beta$-cells of patients with T2DM is significantly reduced, which leads to a shift in the control of insulin secretion from glucokinase to the glucose transport system [25]. This defect is eliminated by sulfonylurea [23].

In the later stages of the disease, there is a disturbance in the second phase of insulin secretion, although this disturbance can be reduced by introducing strict glycemic control. This phenomenon is called desensitization or glucose toxicity of $\beta$-cells and may result from the accumulation of glycogen in $\beta$-cells during sustained hyperglycemia [26].

Other defects in the $\beta$-cell function in T2DM include impaired glucose potentiation in response to non-glucose agents that initiate insulin secretion, asynchronous insulin release, and decreased frequency of the response leading to insulin synthesis from proinsulin [23].

Metabolic changes in patients with T2DM are characterized by a high level of glucogenesis, yet a decrease in glycogenolysis. In patients with IR, there is a disruption in glucose transport, as well as the transmission of insulin signals in tissues, which is accompanied by the production of inflammatory markers in the adipose tissue [27]. The mediators that transmit signals to the $\beta$-cells in response to IR include compounds such as fatty acids, lipid hormones, and gut hormone glucagon-like peptide-1 (GLP-1). To minimize the disturbances in glucose and lipid metabolism, a reciprocal relationship is required between IR and insulin sensitivity and secretion. However, this connection is disrupted by a reduction in the ability to influence these $\beta$-cell signals to trigger insulin secretion, which leads to further development of dysglycemia [28].

In addition to dysglycemia, the concomitant conditions of IR are dyslipidemia, hyperinsulinemia, obesity, and arterial hypertension, which means IR is a key sign of the development of metabolic syndrome and CVD. Moreover, it is believed that genetic predisposition worsens the disease in patients with T2DM. One of the most important etiological causes of T2DM is central obesity, which is thought to be exacerbated by genetic predisposition; however, it is worth noting that diet and exercise can reduce the effects of IR [29].

The main regulators of glucose metabolism in skeletal muscles are hexokinase, glycogen synthase, and the insulin-dependent glucose transporter GLUT4. At the same time, the emerging disorders of glycogen production in skeletal muscles play an important role in the development of IR in T2DM patients [20]. Impairment of the ability of skeletal muscles to bind circulating insulin is one of the factors in the development of IR, which can develop 10–20 years before the immediate diagnosis of T2DM [21].

Insulin-resistant patients may have overt type 2 diabetes when pancreatic $\beta$-cells are unable to produce enough insulin to maintain normoglycemia. $\beta$-cells in the pancreas of T2DM patients cannot adequately perceive glucose, which contributes to impaired insulin secretion. The main factor leading to the development of mitochondrial dysfunction in pancreatic $\beta$-cells is oxidative stress [30].

Diabetes is characterized by persistently elevated blood sugar levels, which promote increased production of reactive oxygen species (ROS) through various processes, including glucose auto-oxidation, polyol pathway activation, and the formation of advanced glycation end products (AGEs) [31]. As soon as glucose arrives in the cells, it is oxidized by either the glycolytic pathway or the pentose phosphate pathway, which produces biosynthetic molecules and nicotinamide adenine dinucleotide phosphate (NADPH). ROS are produced during glycolysis, although the antioxidant defense mechanism in the cell normally effectively neutralizes them. Conversely, overly high blood glucose levels cause an increase in radical generation, which inhibits antioxidant mechanisms and damages DNA. Glyceraldehyde-3-phosphate dehydrogenase (GAPD) is inactivated by an increase in the concentration of DNA repair enzymes such as poly-ADP-ribose polymerase-1 (PARP-1). This causes an accumulation of metabolites, including glyceraldehyde-3-phosphate (GAP), glucose 6-phosphate (G-6-P), and fructose 6-phosphate (F-6-P). Increased concentrations of these metabolites have an impact on how oxidative stress progresses: GAP activates protein kinase C (PKC), whereby G-6-P and F-6-P can follow the polyol route, and auto-oxidation of GAP and G-6-P can result in the generation of AGE precursors [32]. Notably, as the quantity of glucose increases, the hexokinase enzyme gets saturated and cannot catalyze the synthesis of G-6-P. As a result, aldose reductase converts glucose to sorbitol, which sorbitol dehydrogenase (SDH) then further converts to fructose. Excess NADPH is used in this process, which provides GPx with a substrate to make glutathione (GSH) [33]. Thus, oxidative stress is also facilitated by the suppression of antioxidant enzymes in this pathway. Furthermore, increased fructose synthesis resulting from the activation of SDH levels in hyperglycemic circumstances leads to PKC activation and oxidative stress.

Elevated levels of ROS activate the mitochondrial protein uncoupling protein 2 (UCP2), which leads to the leakage of protons through the inner mitochondrial membrane. This, in turn, leads to a decrease in ATP production, which is necessary for insulin secretion [30]. On the other hand, ROS causes damage to the phospholipids in the mitochondrial membrane, which increases its permeability and promotes the leakage of cytochrome C into the cytosol, ultimately activating the apoptosis of pancreatic $\beta$-cells [30]. In addition to pancreatic $\beta$-cells, mitochondrial dysfunction has also been noted in other tissues and organs, as shown in the study and will be described below [34].

Impaired mitochondrial function has been demonstrated in the muscles with impaired insulin resistance in type 2 diabetic patients [35]. It is believed that defective mitochondrial fatty acid metabolism in skeletal muscle affects insulin signaling pathways, leading to insulin resistance [36]. Disruption of mitochondrial fatty acid $\beta$-oxidation, alone or in combination with increased plasma FFA delivery, increases levels of intracellular fatty acid metabolites, such as fatty acyl-CoA, diacylglycerol, and ceramide. Metabolites formed as a result of such metabolic disorders can activate serine/threonine kinases, including protein kinase C, which leads to increased phosphorylation of IRS-1 [36]. As a result, the activity of the insulin receptor tyrosine kinase on IRS-1 and the activity of insulin-stimulated phosphatidylinositol-3-kinase are inhibited, which ultimately leads to the activity of insulin-stimulated protein kinase B–AKT becoming inhibited. This, in turn, blocks GLUT4 translocation and promotes a further decrease in glycogen production [37].

The liver plays a critical role in the development of insulin resistance in T2DM [38]. Some studies indicate that the resulting mitochondrial dysfunction in liver cells may directly cause hepatic insulin resistance [39]. For example, a decrease in the level of mitochondrial $\beta$-oxidation of fatty acids in the liver and skeletal muscle leads to the accumulation of intracellular fatty acid metabolites. It should be noted that similar results were observed either with an increase in hepatic de novo lipogenesis or an increase in the intake of FFA from the plasma. However, metabolites negatively affected intracellular insulin signaling, which led to a decrease in the insulin stimulation of glycogen synthesis and an increase in gluconeogenesis in the liver [36].

Adipose tissue is an endocrine organ that is important in energy metabolism [40]. The main regulators of adipose tissue metabolism are compounds called adipocytokines, the best known of which are adiponectin, TNF-$\alpha$, leptin, and resistin [36]. Adiponectin is known to have an insulin-sensitizing effect. However, in patients with T2DM, there is a sharp decrease in the plasma concentration of adiponectin, in contrast to other adipocytokines [37]. A study showed that plasma and adipose tissue levels of adiponectin were significantly reduced in obese mice; simultaneously, a related decrease in the number of mitochondria was noted, as well as an increase in those possessing energy dysfunction [41]. Rosiglitazone, which is a peroxisome proliferator-activated receptor-$\gamma$ (PPAR$\gamma$) agonist, halted the decline in plasma adiponectin concentration in an obese mouse model alongside having a positive effect on the qualitative and quantitative mitochondrial parameters in adipocytes. These data suggest that mitochondrial dysfunction in adipose tissue leads to decreased plasma adiponectin levels in individuals with T2DM. A number of studies in animal models have shown that the activity of fatty acid oxidation in brown adipose tissue plays a vital role in regulating animal weight, thermogenesis, and energy balance [36]. It was believed that the presence of Brown Adipose Tissue (BAT) only plays a role in newborns and small mammals. However, positron emission tomography and computed tomography (PET-CT) studies have shown that adults have active BAT [42]. Based on this finding, mitochondrial dysfunction in BAT seems to be associated with impaired thermogenesis and energy metabolism processes, which contribute to the development of T2DM and the progression of IR in adults [36]. Additionally, a genetic predisposition is considered in the development of mitochondrial dysfunction in adipocytes in T2DM and has been shown previously [43]. The pathogenesis of T2DM based on mitochondrial dysfunction in different tissues is shown in Fig. 1.

Fig. 1.

Pathogenesis of T2DM based on mitochondrial dysfunction in different tissues. T2DM, type 2 diabetes mellitus; ROS, reactive oxygen species; ATP, adenosine triphosphate; FA, fatty acid; FFA, free fatty acid.

The occurrence of mitochondrial dysfunction can affect the pathogenesis of diabetes mellitus in different tissues differently, which leads to the onset of certain adverse events. Thus, in the pancreas, mitochondrial dysfunction is one of the factors leading to impaired insulin secretion and increased apoptosis of $\beta$-cells. In the liver and skeletal muscles, glycogen synthesis decreases, increasing fatty acid metabolite accumulation. In adipose tissue, there is an increase in insulin resistance.

Mitochondrial dysfunction can be directly related to a change in the state of some factors immediately associated with mitochondrial function. Hence, a previous study [44] analyzed a culture of monocyte cells taken from patients with T2DM and found alterations in both the morphology of the mitochondria—a decrease in their volume resulting from an increase in mitochondrial division; their functional state associated with the hyperpolarization of the mitochondrial membrane. Chronic hyperglycemia and insulin resistance are thought to alter the mitochondrial structure and function, leading to an increase in small, hyperpolarized mitochondria that produce elevated levels of ROS, resulting in reduced ATP synthesis productivity and increased activation of inflammation. In another study [45], it was shown that even the treatment of myocardial mitochondria in patients with T2DM with diazoxide (which is an activator of potassium channels) did not lead to the depolarization of the mitochondrial membrane, in contrast to the negative control, which could possibly be associated with mitochondrial potassium channel dysfunction.

Based on the fact that ATP production is a necessary source of energy, including for the secretion of insulin, and is provided by the work of five large protein complexes consisting of 90 subunits, 13 of which are encoded by the mitochondrial gene, the depletion of mtDNA in $\beta$-cells can have a direct impact on mitochondrial dysfunction and pathogenesis of type 2 diabetes [46]. A previous study [47] demonstrated that mitochondrial haplogroups correlated better than nuclear haplogroups with insulin requirements. Moreover, it has been shown that a number of mitochondrial DNA mutations that lead to mitochondrial dysfunction may be associated with the development of type 2 diabetes. Such mutations mainly target mitochondrial tRNA-coding genes, as several clinical studies show [48, 49, 50].

A number of studies have shown that the degree of DNA methylation in the nuclear-encoded ETC and OXPHOS genes, cytochrome c oxidase polypeptide 7A1, NADH dehydrogenase 1 beta-subcomplex subunit 6, and PPARGC1A negatively correlates with changes in gene expression in insulin-resistant skeletal muscles [51]. Another study showed that mitochondrial DNA can also be methylated in a manner similar to prokaryotic DNA methylation [52]. In a rat model with type 2 diabetes, it was found that severe retinopathy was associated with an increase in the proportion of methylated mitochondrial DNA [53]. Additional epigenetic changes may be associated with non-coding RNAs. Thus, the study [54] showed that some LncRNAs can cause changes in mitochondrial homeostasis in diabetic retinopathy.

One of the broadest classes of mitochondria-targeted therapeutic agents includes compounds whose activity is associated with modulating the activity of ETC components. Among them, imeglimin is the most promising. Imeglimin is a newly developed antidiabetic compound that restores mitochondrial function in various organs of patients, including skeletal muscle, liver, and pancreas [55]. However, the exact mechanism of action of imeglimin on mitochondria remains unknown. In a number of studies on hepatocytes, it was found that imeglimin acts as a competitive inhibitor of ETC complex I and helps restore complex III activity, thus, improving the efficiency of oxidative phosphorylation [56]. Improvement in mitochondrial function under the action of imeglimin has also been confirmed in studies demonstrating the prevention of mitochondrial permeability-induced cell death in endothelial and islet cell models [57].

Another potential mitochondrial target for treating metabolic disorders is the mitochondrial pyruvate transporter (MPC), which mediates the importation of pyruvate into the mitochondrial matrix [58]. Impaired pyruvate uptake in the mitochondria is associated with mitochondrial dysfunction and is noted in the pathogenesis of T2DM.

Thiazolidinediones (TZDs), also called glitazones, are a class of peroxisomal proliferation-activated receptor (PPAR)-$\gamma$ agonists that can increase insulin sensitivity and inhibit the action of MPC [45]. These are currently used to treat T2DM either instead of metformin or in combination. Furthermore, pioglitazone is a promising TZD variant that affects mitochondrial dysfunction in T2DM and can restore mitochondrial function in patients with T2DM [59].

Recently, significant research has focused on developing molecules capable of enhancing the transcription of nuclear and mitochondrial genes encoding subunits in the ETC chain to improve the efficiency of oxidative phosphorylation. A promising strategy is targeting peroxisomal proliferator-activated receptors (PPARs), which are a group of transcription factors associated with regulating various cellular functions, including mitochondrial metabolism and energy homeostasis. Among the downstream PPAR targets is the PPAR-$\gamma$-1$\alpha$ (PGC-1$\alpha$) coactivator, which, after activation, binds to transcription factors involved in mitochondrial biogenesis and stimulates them [60]. Presently, several PPAR agonists are being tested as potential therapeutic agents for re-storing mitochondrial function.

Sirtuin 1 (Sirt1) is another potential target for modulating PGC-1$\alpha$ activity and restoring mitochondrial function. Sirt1 is a nicotinamide adenine dinucleotide (NAD${}^{+}$)-dependent histone deacetylase that plays a significant role in metabolism, promoting mitochondrial biogenesis and renewal [61]. Sirt1 activity can be changed through the action of allosteric modulators— cysteine-rich domain-containing proteins (STAC). In addition, Sirt1 activation can be increased using NAD${}^{+}$ precursors, also called NAD${}^{+}$ stimulating molecules (NBMs), which target enzymes that promote NAD${}^{+}$ synthesis or inhibit its degradation [62].

Since oxidative stress has been shown to play a role in the pathogenesis of T2DM, as already described in the previous paragraphs, the administration of antioxidants can potentially be used to treat T2DM. To date, several clinical studies have explored the use of antioxidants in treating T2DM, such as quercetin, resveratrol, and vitamin C. However, the results of these studies were found to be contradictory, which may be due to the difficulty in assessing the antioxidant capacity in insulin-sensitive tissues, such as skeletal muscle and adipose tissue, meaning it requires more research [63]. At the same time, pharmaceuticals aimed at increasing antioxidant protection in cells, such as NADPH oxidase inhibitors, xanthine oxidase inhibitors, and superoxide dismutase (SOD) mimetics, may also be promising for the treatment of T2DM [63]. Additionally, mitochondrial-targeting antioxidants, such as MitoQ and thyron, are worth investigating since they can directly penetrate the mitochondria and bind ROS at their site of origin, thus, preventing the development of oxidative stress at its inception [64].

A generalized scheme of therapeutic strategies aimed at restoring mitochondrial function in T2DM is presented in Table 1.