Oxidative Stress, Endothelial Dysfunction, and N-Acetylcysteine in Type 2 Diabetes Mellitus

Abstract

Significance:

Cardiovascular diseases (CVDs) remain the leading cause of morbidity and mortality globally. Endothelial dysfunction is closely associated with the development and progression of CVDs. Patients with diabetes mellitus (DM) especially type 2 DM (T2DM) exhibit a significant endothelial cell (EC) dysfunction with substantially increased risk for CVDs.

Recent Advances:

Excessive reactive oxygen species (ROS) and oxidative stress are important contributing factors to EC dysfunction and subsequent CVDs. ROS production is significantly increased in DM and is critically involved in the development of endothelial dysfunction in diabetic patients. In this review, efforts are made to discuss the role of excessive ROS and oxidative stress in the pathogenesis of endothelial dysfunction and the mechanisms for excessive ROS production and oxidative stress in T2DM.

Critical Issues:

Although studies with diabetic animal models have shown that targeting ROS with traditional antioxidant vitamins C and E or other antioxidant supplements provides promising beneficial effects on endothelial function, the cardiovascular outcomes of clinical studies with these antioxidant supplements have been inconsistent in diabetic patients.

Future Directions:

Preclinical and limited clinical data suggest that N-acetylcysteine (NAC) treatment may improve endothelial function in diabetic patients. However, well-designed clinical studies are needed to determine if NAC supplementation would effectively preserve endothelial function and improve the clinical outcomes of diabetic patients with reduced cardiovascular morbidity and mortality. With better understanding on the mechanisms of ROS generation and ROS-mediated endothelial damages/dysfunction, it is anticipated that new selective ROS-modulating agents and effective personalized strategies will be developed for the management of endothelial dysfunction in DM.

Introduction

Cardiovascular diseases (CVDs) represent a major challenge to public health with significant morbidity and mortality in the world. Diabetes mellitus (DM) especially type 2 DM (T2DM) is an independent risk factor for CVDs, including atherosclerosis and stroke (Poznyak et al., 2020; Tsao et al., 2023). It is estimated that the prevalence of coronary heart disease in patients with T2DM is doubled over the nondiabetic subjects. About one third of diabetic patients struggle with disabling CVDs that account for >50% of all deaths in patients with DM (Fan, 2017; Kaze et al., 2021). Yet, the number of individuals with DM has been steadily increasing globally for the past couple of decades (Fan, 2017).

Diabetic patients have a twofold increase in the risk of hospitalization for heart failure and are often associated with hypertension (HTN), dyslipidemia, and other important risk factors of CVDs (Ji et al., 2013; Lam et al., 2018; Poznyak et al., 2020). Thus, compared with blood glucose control alone, a combined intervention to effectively target multiple risk factors significantly reduces the risk of CVDs and cardiovascular mortality in T2DM (Gaede et al., 2008; Ray et al., 2009; Tsao et al., 2023).

CVDs in diabetic patients can be broadly divided into microvascular diseases and macrovascular disorders. Microvascular complications of diabetes involve small blood vasculature and typically include retinopathy, nephropathy, and erectile dysfunction (Fan, 2017; Kaze et al., 2021; Trebatický et al., 2019). Macrovascular diseases in DM (especially T2DM) affect large vasculature and include atherosclerotic cardiovascular diseases (ASCVDs) such as coronary heart disease, cerebrovascular disease, and peripheral artery disease (Fan, 2017; Kaze et al., 2021). Among the diabetic cardiovascular complications, macrovascular diseases are the primary contributors to the morbidity and mortality in diabetic patients.

Of note, erectile dysfunction is considered an important and early presentation of microvascular diseases and ASCVD, and may serve as a marker of early subclinical coronary artery disease (CAD) and as an independent predictor of future adverse cardiovascular events in diabetic men (Trebatický et al., 2019). Endothelial dysfunction is closely related to the development and progression of CVDs, including ASCVD (Bkaily and Jacques, 2023). Studies with animal models and human subjects have demonstrated that a significant endothelial dysfunction is present in DM and proportionally associated with the clinical outcomes for diabetic patients (Clyne, 2021; Knapp et al., 2019).

Oxidative stress and chronic inflammation are important mechanisms for endothelial dysfunction and the development and progression of CVDs (Maruhashi and Higashi, 2021; Takeda et al., 2020). It is known that the levels of oxidative stress and chronic inflammation are both significantly increased in DM through multiple mechanisms, including (but not limited to) increased level of advanced glycation end-products (AGEs), enhanced activations of protein kinase C (PKC) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), reduced level of nitric oxide (NO), increased expression of lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1), increased productions of inflammatory cytokines (such as tumor necrosis factor alpha, TNF-α) and reactive oxygen species (ROS), and metabolic abnormalities (Gora et al., 2021; Iacobini et al., 2021), leading to increased oxidative stress and inflammation and subsequent endothelial cell (EC) dysfunction and diabetic cardiovascular complications.

The outcomes of therapies with traditional antioxidants including vitamins C and E have been inconsistent regarding the potential beneficial effects against diabetic cardiovascular complications in patients (Hor et al., 2018; Lee et al., 2004; Song et al., 2009). N-acetylcysteine (NAC) is an FDA-approved drug for the treatment of acetaminophen overdose.

Administration of NAC decreases the levels of ischemic ROS production and plasma TNF-α as well as inflammatory cell infiltration in ischemic tissue, and significantly improves the recovery of ischemic limb in T2DM mice (Cui et al., 2015b; Xu et al., 2021; Zhu et al., 2022a; Zhu et al., 2021). NAC has been shown to preserve endothelial function and maintain the population of circulating endothelial progenitor cells (EPCs, an important group of cells that are critical to endothelial function) (Liu et al., 2022; Toborek et al., 1995).

In this review, we summarize the role of oxidative stress in the development of endothelial dysfunction and related mechanisms in DM and discuss the potential beneficial effect of NAC on endothelial function in DM. Although the risk for CVDs is also significantly increased with early development of significant endothelial dysfunction in patients with type 1 DM (T1DM) (Harjutsalo et al., 2021; Lespagnol et al., 2020; Sousa et al., 2019), our efforts are focused on T2DM since the majority (>90%) of diabetic patients have T2DM.

Endothelial Dysfunction in Diabetes

Vascular endothelium is a single monolayer structure between blood and the vascular wall anatomically. More importantly, vascular endothelium is an important source of a variety of potent vasoactive substances, including growth factors such as vascular endothelial growth factor (VEGF), NO, prostaglandins, thromboxane A₂, endothelin-1 (ET-1), cytokines, adhesion molecules, and other signaling molecules such as ROS (Bkaily and Jacques, 2023; Xu et al., 2021).

ECs also express a large number of receptors that are essential for many important and diverse functions, including (but not limited to) cell metabolism, homeostasis, angiogenesis, growth, blood pressure (BP) control, and regulations of redox reaction, immune response, and inflammation, as well as responses to stress and external stimulations such as hypoxia and infections for the entire body (Bkaily and Jacques, 2023; Xu et al., 2021). Functional and healthy ECs are critical for the structural and functional integrity of vasculature and prevention of disease development and progression such as atherosclerosis (Bkaily and Jacques, 2023; Xu et al., 2021).

Endothelial dysfunction is defined as a pathological condition of vascular endothelium with interruptions or impairment of its structural and functional integrity, and is associated with the development and progression of many disease conditions, including CVDs such as HTN, stroke, and atherosclerosis (Bkaily and Jacques, 2023; Maruhashi and Higashi, 2021; Takeda et al., 2020; Viigimaa et al., 2020; Xu et al., 2021).

Studies with both animal models and human subjects have clearly demonstrated that a significant endothelial dysfunction is present in DM (Maruhashi and Higashi, 2021; Takeda et al., 2020; Viigimaa et al., 2020). In fact, endothelial dysfunction occurs before the development of vascular complications in DM clinically and is closely associated with the clinical outcomes for patients with DM (Maruhashi and Higashi, 2021; Takeda et al., 2020; Viigimaa et al., 2020).

The mechanisms for endothelial dysfunction in DM are very complex and have not yet fully defined, including (but not limited to) (1) oxidative stress and excessive ROS (Iacobini et al., 2021; Mansour et al., 2023; Takeda et al., 2020); (2) chronic inflammation with increased levels of proinflammatory molecules (NF-κB, interleukin [IL]-1β, IL-6, IL-18, and TNF-α) and M1 macrophage (Gora et al., 2021); (3) mitochondrial dysfunction with abnormal metabolism and decreased network extent, punctate mitochondria, and increased fission-1 protein and dynamin-related protein-1 (Shenouda et al., 2011); (4) insulin resistance (IR) with downregulation of phosphoinositide 3-kinase (PI3-K)/protein kinase B (Akt)/endothelial nitric oxide synthase (eNOS) pathway and upregulation of mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK)/ET-1 pathway (Maruhashi and Higashi, 2021); (5) decreased NO availability, resulting in decreased vascular endothelial function and increased leukocyte attachment and platelet aggregation (Maruhashi and Higashi, 2021); (6) decreases in the survival, proliferation, mobilization, paracrine function, and differentiation of circulating EPCs (Hu et al., 2018); (7) microRNAs (miRNAs) (Qadir et al., 2019; Vezza et al., 2021); and (8) hyperosmolar stress (Madonna et al., 2017), as shown in Figure 1.

FIG. 1.

Mechanisms for endothelial dysfunction in DM. A significant endothelial dysfunction appears early in DM. Multiple factors significantly contribute to the pathogenesis of endothelial dysfunction, including (but not limited to) chronic inflammation, mitochondrial dysfunction, microRNAs, IR, and decreased NO availability, decreased numbers and/or function of circulating EPCs, and hyperosmolar stress. These factors either individually or work together to impair endothelial function in DM through increased production of ROS or other mechanisms or combinations. ↑, increase; ↓, decrease; Akt, protein kinase B; CHOP, C/EBP homologous protein; DM, diabetes mellitus; eNOS, endothelial nitric oxide synthase; EPCs, endothelial progenitor cells; ERK, extracellular signal-regulated kinase; ET-1, endothelin-1; IL, interleukin; IR, insulin resistance; MAPK, mitogen-activated protein kinase; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; NO, nitric oxide; PI3-K, phosphoinositide 3-kinase; ROS, reactive oxygen species; SIRT1, sirtuin 1; TNF-α, tumor necrosis factor alpha; VSMCs, vascular smooth muscle cells.

Oxidative Stress and Endothelial Dysfunction in Diabetes

Oxidative stress and excessive ROS are important mechanisms for damages and dysfunction of cells (including ECs) and organ systems as well as the development and progression of a variety of diseases such as CVDs (Batty et al., 2022). It is known that DM (both type 1 and type 2) is associated with significantly increased levels of oxidative stress and ROS, including hydrogen peroxide (H₂O₂) and superoxide (O₂⁻) (Iacobini et al., 2021), which have been strongly implicated as critical factors for the pathogenesis of vascular complications in DM (Batty et al., 2022; Iacobini et al., 2021).

The pathogenesis for excessive ROS in DM is complex, and multiple mechanisms have been proposed to significantly contribute to the increases in oxidative stress and ROS in DM (Batty et al., 2022; Iacobini et al., 2021; Rehman and Akash, 2017), including (but not limited to) (1) hyperglycemia (Papachristoforou et al., 2020); (2) production of AGEs (Cepas et al., 2020); (3) IR (Araújo et al., 2016); (4) mitochondrial dysfunction (Nishikawa et al., 2007); (5) hyperlipidemia (Garcia-Fuentes et al., 2010); (6) decreased levels of antioxidant enzymes (Dunn et al., 2010; Levy et al., 2019); and (7) miRNAs (Qadir et al., 2019; Vezza et al., 2021), as shown in Figure 2.

FIG. 2.

Mechanisms for excessive ROS and increased oxidative stress in DM. ROS production is significantly increased in DM, leading to endothelial dysfunction and subsequent diabetic cardiovascular complications. The mechanisms for excessive ROS in DM are very complex and multifactorial, including (but not limited to) IR, AGEs, hyperglycemia, mitochondrial dysfunction, microRNAs, decreased expressions of antioxidant enzymes, and hyperlipidemia, through a variety of different pathways. ↑, increase; ↓, decrease; AGEs, advanced glycation end-products; FADH2, 1,5-dihydroflavin adenine dinucleotide; FFA, free fatty acids; Gpx, glutathione peroxidase; NADH, nicotinamide adenine dinucleotide; NADPH, nicotinamide adenine dinucleotide phosphate; NOX, nicotinamide adenine dinucleotide phosphate oxidase; ox-LDL, oxidized low-density lipoprotein cholesterol; PKC, protein kinase C; PON, paraoxonase; PPARγ, peroxisome proliferator-activated receptor-γ; RAGE, receptor for AGE; ROCK, rhoA-rho kinase; SOD, superoxide dismutase; TRX, thioredoxin.

There are extensive interactions between these mechanisms to enhance ROS generation in DM, and thus it is very challenging to determine the temporal sequence in which the excess ROS origins from and their relative contributions to ROS overproduction and the development and progression of endothelial dysfunction from early to late stage of disease in DM.

Excessive ROS and oxidative stress result in structural and functional impairment of ECs, as shown in Figure 3. In DM, excessive ROS enhances the generation of low molecular weight hyaluronan fragments (LMW-HA) and the degradation of glycocalyx, and activates protease-activated receptor (PAR) in ECs through hyaluronic acid binding protein 2 (HABP2) and ligates CD44v10 (variant 10), leading to an increased production of action stress fiber and damages to EC function through RhoA-rho kinase (ROCK) pathway (Broekhuizen et al., 2010; Lennon and Singleton, 2011; Mambetsariev et al., 2010; Singleton et al., 2006; Wheeler-Jones et al., 2010).

FIG. 3.

Mechanisms for structural and functional impairment of ECs induced by excessive ROS and oxidative stress. Excessive ROS is an important part of pathogenesis of impaired endothelial structure and function in DM. Overproduction of ROS results in degradation of glycocalyx and actin stress fiber in ECs, leading to the destruction of EC structure with increased permeability through multiple cellular signaling pathways (left panel). Impaired ECs along with increased ROS result in a significant attenuation of NO and PGI2 levels, as well as increased level of EDCFs through respective pathways, rendering a significant reduction of the function of VSMCs (right panel). AA, arachidonic acid; BH4, tetrahydrobiopterin; COX-2, cyclooxygenases-2; EC, endothelial cell; EDCFs, endothelium-derived contracting factors; LMW-HA, low molecular weight hyaluronan fragments; LOOH, lipid peroxide; NLRP3, pyrin domain-containing protein 3; ONOO⁻, peroxynitrite anion; PAR, protease-activated receptor; PARP-1, poly (ADP-ribosome) polymerase 1; PGI2, prostacyclin. The illustrations of LMW-HA, Glycocalyx, CD44v10, and Actin Stress Fibers were adopted from a figure in Lennon and Singleton (2011).

Excessive ROS triggers EC pyroptosis via pyrin domain-containing protein 3 (NLRP3) inflammasome complex activation, parthanatos due to extensive DNA damage and poly (ADP-ribosome) polymerase 1 (PARP-1) activity, and ferroptosis secondary to increased iron and lipid peroxide (Zheng et al., 2022).

Vascular repair and angiogenesis are an important function of ECs in response to vascular injuries and ischemia or hypoxia (Howangyin and Silvestre, 2014). In DM, the vascular repair process is significantly impaired through multiple mechanisms, including decreased hypoxia-induced angiogenesis, reductions in endothelial mobilization and recruitment, impaired activation of Akt and eNOS, and decreased VEGFR2 expression, as well as attenuated angiogenic potential of bone marrow-derived progenitor cells (Howangyin and Silvestre, 2014).

It is important to point out that diabetic patients exhibit a significantly increased susceptibility to acute ischemia-reperfusion injury (IRI), including acute myocardial infarction (AMI) (Bertoluci and Rocha, 2017). Large randomized clinical trials or the data from large national registries have also shown that the clinical outcomes are less desirable for diabetic patients with acute myocardial IRI such as AMI, percutaneous coronary artery intervention (PCI), or coronary artery bypass graft surgery (CABG) as compared with nondiabetic subjects.

DM (especially T2DM) has been identified as an independent risk factor for major adverse cardiovascular events (MACE) including death and myocardial infarction for patients with acute coronary syndromes, including unstable angina and AMI or revascularization procedures such as PCI and CABG (Alserius et al., 2006; Donahoe et al., 2007; Mathew et al., 2004). Although the reason(s) for less optimal outcomes for diabetic patients with acute IRI has yet to be fully defined, hyperglycemia is considered an important contributing factor for the increased risk of MACE in DM (Penna et al., 2020).

In addition, excessive ROS formation and oxidative stress are critically involved in acute cardiac IRI, and DM could certainly exacerbate oxidative stress during acute cardiac IRI through modulation of the expressions of Bcl-2-associated athanogene 3/NF-E2 p45-related factor 2/heme oxygenase-1 signaling molecules (Chien et al., 2020), thus aggravating myocardial IRI and subsequent cardiovascular adverse events in patients with DM.

Excessive ROS also interrupts the balance between vasoactive substances generated from ECs. NO is a major vasodilator in ECs synthesized by eNOS. In diabetes, eNOS is uncoupled by excessive ROS, leading to generation of more superoxide and subsequent formation of peroxynitrite radical with NO, and causing more uncoupled eNOS (Batty et al., 2022; Gray et al., 2016; Hong et al., 2019; Li and Förstermann, 2013).

A reduction of NO impairs the function of vascular smooth muscle cells (VSMCs), increases platelet aggregation, and enhances leukocyte adhesion to ECs, and promotes VSMC proliferation (Batty et al., 2022). In T2DM, excessive ROS also increases lipid oxidation and decreases lipoxin A4 (LXA4) and arachidonic acid (AA) in the plasma of human subjects, thus attenuating the production of prostacyclin (PGI2, another important vasodilator) since AA is the source of PGI2 and facilitates eNOS expression (Batty et al., 2022; Bkaily and Jacques, 2023; Chiu et al., 2014; Das, 2019; Gundala et al., 2018; Kaviarasan et al., 2015).

ET-1 is one of the potent vasoconstrictors generated in vascular ECs and is increased in T2DM. Hyperglycemia activates PKC via increased diacylglycerol (DAG), resulting in a reduced expression of eNOS and an enhanced expression of ET-1 (Maruhashi and Higashi, 2021). IR and the compensatory hyperinsulinemia leads to enhanced activities of MAPK/ERK/ET-1 pathway, whereas the PI3-K/Akt/eNOS pathway is impaired due to a decreased expression of insulin receptor substrate (IRS) (Kubota et al., 2013; Maruhashi and Higashi, 2021).

Hyperglycemia-induced excess ROS increases cyclooxygenases-2 (COX-2) expression, facilitating the formation of endothelium-derived contracting factors (EDCFs) from AA (Wang et al., 2022), as shown in Figure 3.

Hyperglycemia and oxidative stress in diabetes

Elevated blood glucose level is one of the diagnostic criteria for DM. Acute high concentration of blood glucose leads to the entrance of a large amount of glucose into ECs through glucose transporters (GLUT), and results in mitochondrial dysfunction or uncoupling of eNOS via polyol pathway, thus increasing ROS production and activating PKC and actin-myosin contraction with resultant EC barrier dysfunction (Fiorentino et al., 2013; Maruhashi and Higashi, 2021; Swärd and Rippe, 2012).

The polyol pathway is critically involved in the conversion of glucose to fructose and at the same time it oxidizes NADPH (nicotinamide adenine dinucleotide phosphate) to NADP⁺ and produces NADH (nicotinamide adenine dinucleotide) from NAD⁺. A decrease in NADPH weakens the antioxidant systems and recycling of oxidized glutathione, whereas NADH is transported into mitochondria and oxidized with the production of excess ROS (Ceriello et al., 1996; Giacco and Brownlee, 2010; Papachristoforou et al., 2020).

The polyol pathway is significantly enhanced in uncontrolled diabetes, thus increasing ROS formation. Hyperglycemia also activates PKC isoforms by the lipid second messenger DAG, leading to a decrease in eNOS expression and increases in NF-κB and NADPH oxidase (NOX) with increased ROS production (Brownlee, 2001; Ha et al., 2002; Papachristoforou et al., 2020; Steinberg, 2008; Xia et al., 1994).

High glucose levels enhance the production of UDP-N-acetylglucosamine (UDP-GlcNAc) through increased activities of the hexosamine pathway, facilitating the formation of O-linked glycoproteins. O-GlcNAc modifications of serine or threonine residues could inhibit eNOS activity in arteries and increase ROS production in DM (Buse, 2006; Musicki et al., 2005; Papachristoforou et al., 2020; Wells et al., 2001).

AGEs and oxidative stress in diabetes

A significant amount of AGEs is generated through nonenzymatic glycation and oxidation of proteins or lipids in DM (Wautier et al., 1996). AGEs are accumulated on long-life proteins, and act as potent oxidative agents with production of excess ROS, leading to tissue inflammation via activation of the receptor for AGE (RAGE) signaling cascade (Johnson et al., 2021; Schmidt et al., 1996; Wautier et al., 2001; Wautier et al., 1996; Yan et al., 1994). AGEs are believed to be associated with the development of diabetic vascular complications (Johnson et al., 2021).

Sustained hyperglycemia promotes the formation of AGEs in diabetes, and an increased level of superoxide could further promote the formation of AGEs and enhance the expression of AGE receptors in different cell types, including vascular endothelium, facilitating the interactions with their ligands, ultimately activating proinflammatory pathways with long-lasting effects on organ systems, including vasculatures (Cepas et al., 2020).

NOX4 is a potent oxidizing enzyme that is closely involved in ROS production. Activation of NOX and resultant oxidative stress by the AGE-RAGE signaling cascade impair endothelial barrier and lead to vascular hyperpermeability in diabetes (Wautier et al., 2001; Wautier et al., 1996; Yan et al., 1994). AGEs also promote the release of proinflammatory factors such as TNF-α (Zhang et al., 2009), ILs 1β, 6, and 18, monocyte chemoattractant protein 1 (MCP-1), and C-reactive protein (CRP) (Gora et al., 2021; Lagrand et al., 1997; Son, 2012; Sproston and Ashworth, 2018), through activation of NF-κB and excessive ROS (Cepas et al., 2020).

Other mechanisms for AGEs-induced ROS production include dysfunction of glycated intracellular proteins and extracellular matrix. AGEs could accelerate the binding of glycated plasma proteins to RAGE on ECs, inducing excessive ROS production through activation of NOXs and damage of mitochondrial respiratory chain (Cepas et al., 2020; Papachristoforou et al., 2020; Rodiño-Janeiro et al., 2015; Wautier et al., 2001).

IR and oxidative stress in diabetes

Insulin and insulin receptor-mediated signaling play a key role in the regulation of glucose metabolism and the development and progression of DM. Glucose transporter 4 (GLUT4) is a critical glucose transporter in insulin-sensitive cells (mainly skeletal muscle cells and adipocytes) (Hurrle and Hsu, 2017). GLUT4 expression increases in the insulin-sensitive cells in response to insulin stimulations normally, thus increasing cellular uptake of glucose from the bloodstream to maintain the physiological dynamics of glucose and lipid metabolisms.

One of the important features for T2DM is IR with elevated levels of blood glucose and insulin. Sustained high level of insulin beyond the optimal physiological concentrations paradoxically downregulates the expression of GLUT4 in the insulin-sensitive cells, thus decreasing the cellular responses to insulin stimulation and interrupting glucose uptake. This results in a sustained high level of blood glucose with increased oxidative stress and substantial metabolic abnormalities, including an increase in fatty acid uptake through mediation of CD36, leading to activation of the potent ROS-producing enzyme NOX4 and thus generating excessive ROS (Hurrle and Hsu, 2017; Prandi et al., 2023).

DM is also associated with a leptin deficiency, decreased AMP-activated protein kinase (AMPK) activation, and increased activity of mammalian target of rapamycin (m-TOR) signaling pathway, leading to inflammation and oxidative stress via activation of NF-κB signaling and releases of inflammatory mediators such as ILs, interferons (INF-s), and TNF-α (Lashgari et al., 2023; Prandi et al., 2023).

There are extensive interactions between oxidative stress and IR. On one hand, excess ROS and associated oxidative stress could cause significant damages to a variety of macromolecules such as receptors and other critical signaling molecules such as insulin receptors and some of the key molecules in insulin-receptor-mediated pathways, directly contributing to the reduction and/or dysfunction of insulin receptors and/or related signaling molecules in insulin-sensitive cells and thus the development and progression of IR (Hurrle and Hsu, 2017; Tangvarasittichai, 2015).

On the other hand, IR promotes ROS production and oxidative stress through multiple mechanisms, including (but not limited to) an increase of NOX4, activation of NF-κB signaling, mitochondrial dysfunction, a reduction of antioxidant systems such as peroxisome proliferator-activated receptor-γ (PPARγ) and Cu²⁺, Zn²⁺-superoxide dismutase (CuZn-SOD), and oxidative modifications of lipids, nucleic acids, and proteins (Araújo et al., 2016; Garcia-Fuentes et al., 2010; Inoue et al., 2001; Onyango, 2018; Ricote et al., 1998), reenforcing the vicious cycle of oxidative stress and IR in DM.

Of note, insulin receptor and its messenger RNA are primarily expressed in metabolically active cells such as hepatocytes, skeletal muscle cells, and adipocytes, as well as brain cells, although it is also present on vascular ECs (Vicent et al., 2003). It is known that not all cells behave the same with glucose metabolism and in response to insulin. However, IR has significant impact on glucose metabolism and oxidative stress in these cells, and thus collectively contributing to increased level of oxidative stress in DM. It could be ideal, but challenging, to determine how these cells respond individually to IR and contribute to oxidative stress and subsequent endothelial dysfunction in DM.

Mitochondrial dysfunction and oxidative stress in diabetes

Mitochondria consists of mitochondrial matrix, inner membrane, and intermembrane space, and is critical for energy metabolism and the regulation of ROS production (Batty et al., 2022). Mitochondria are the major source of energy generation and ROS in cells. Physiologically, low levels of mitochondrial ROS (mtROS) play important roles in many signaling pathways and are necessary for regulating gene expression and stress responses (Batty et al., 2022; Shadel and Horvath, 2015).

However, in diabetes, high glucose flux drives the glucose metabolism into glycolysis pathway and produces more pyruvate, which generates abundant electron donors NADH and 1,5-dihydroflavin adenine dinucleotide (FADH2) in mitochondrial tricarboxylic acid (TCA) cycle. The increased electron donors provide excessive electrons into the mitochondrial electron transport chain (ETC), which creates the voltage gradient needed for ATP synthesis, above the capacity of ETC. In this situation, the overflowing electrons are directed to oxygen, resulting in overproduction of their natural byproduct of superoxide and H₂O₂ with increased level of oxidative stress (Batty et al., 2022; Brownlee, 2005; Iacobini et al., 2021; Nishikawa et al., 2007).

It is known that mitochondria critically rely on antioxidants for their structural integrity and optimal function. Glutathione (GSH) is an important intracellular antioxidant against oxidative stress and a reduction of intracellular GSH level could lead to significant mitochondrial injury and/or dysfunction. The intracellular level of reduced form of glutathione has been reported to be significantly decreased in DM, whereas the intracellular level of glutathione disulfide (GSSG, oxidized form of GSH) increased (Murakami et al., 1989), thus resulting in significant mitochondrial dysfunction and dysregulation of energy production and ROS generation.

In addition, mitochondrial dysfunction is associated with IR, further increasing ROS production and oxidative stress in DM. Another important antioxidant system in mitochondria is manganese-SOD (Mn-SOD or SOD2). There is a significant mitochondrial SOD polymorphism especially the Ala16Val polymorphism that may be associated with DM (Pourvali et al., 2016).

It has been shown that the variability of concentration/activity of SODs, including SOD2 in plasma is altered in the subjects with T2DM (Lewandowski et al., 2021), and the protein level of Mn-SOD2 in polymorphonuclear leukocytes was significantly decreased in T1DM patients with microvascular complications (Wegner et al., 2014), potentially contributing to increased ROS production and diabetic complications.

Hyperlipidemia and oxidative stress in diabetes

DM is a metabolic disorder in nature and is associated with significant metabolic abnormalities, including hyperlipidemia (Ji et al., 2013; Lam et al., 2018; Poznyak et al., 2020). ASCVD, including coronary heart disease, cerebrovascular disease, and peripheral vascular disease, is significantly accelerated by dyslipidemia, hyperglycemia, and IR in diabetic patients (Poznyak et al., 2020). Specifically, oxidized low-density lipoprotein cholesterol (ox-LDL) is a critical component in hyperlipidemic states and a potent oxidative agent, and plays an important role in the pathogenesis of endothelial dysfunction and atherosclerosis (Poznyak et al., 2020).

In diabetes, the increases in inflammatory factors and oxidative stress enhance the production of more atherogenic low-density lipoprotein (LDL) through glycation and oxidation (Poznyak et al., 2020), and increased expression of LOX-1, leading to the accumulation of ox-LDL in the atherosclerotic plaques (Kattoor et al., 2019; Poznyak et al., 2020). In T2DM, diabetic dyslipidemia (DD) commonly occurs and is one of the important risk factors for ASCVD (Athyros et al., 2018).

Lipid aberrations in DM include an increase in triglycerides (TGs), a decrease in high-density lipoprotein (HDL), and increases in glycated apolipoproteins and small-dense LDL, contributing to vascular dysfunction (Arca et al., 2012; Athyros et al., 2018; Taskinen, 2003; Tziomalos et al., 2009). Increased fatty acid and excessive generation of very-low-density lipoprotein (VLDL) in the liver are significant in DM with the presence of IR (Taskinen and Borén, 2015). Hypertriglyceridemia and increased level of free fatty acids (FFA) significantly enhance ROS production and oxidative stress via activation of NOX4 and impairment of antioxidant defenses as well as generation of ox-LDL (Garcia-Fuentes et al., 2010).

Altered expressions of redox enzymes and oxidative stress in diabetes

There is a delicate balance between ROS formation and clearance in biological systems. An imbalance between ROS formation and antioxidant defenses favors oxidative stress, tissue damage, and endothelial dysfunction, and ultimately vascular dysfunction (Batty et al., 2022; Poznyak et al., 2020). ROS can be generated through enzyme-mediated and nonenzymatic mechanisms, and both are significantly increased in DM (Batty et al., 2022; Poznyak et al., 2020).

Many enzymes are involved in maintaining the delicate redox balance, and thus the normal function of cells and organ systems. The enzymes that are critically associated with ROS generation and oxidative stress are NOX and uncoupled eNOS, although other enzymes such as xanthine dehydrogenase/oxidase (XO), heme protein myeloperoxidase (MPO), lipoxygenase, cyclooxygenase (COX), and glucose oxidase may also be involved in some degree (Sies and Jones, 2020).

In contrast, antioxidant enzymes, including SODs, glutathione peroxidase (Gpx)-1, and catalase, are critical components that clear excess ROS and prevent cell damage (Batty et al., 2022; Poznyak et al., 2020). SODs are important enzymes against O₂^•− and consist of three isoforms: cytoplasmic Cu/ZnSOD (SOD1), mitochondrial MnSOD (SOD2), and extracellular Cu/ZnSOD (SOD3). SODs catalyze the conversion of O₂^•− to H₂O₂, which, in turn, is reduced to water by catalase, or Gpx, or peroxiredoxins (Batty et al., 2022; Yang et al., 2004).

The normal redox balance is interrupted in DM with ROS overproduction and increased level of oxidative stress. The ratio of intracellular GSH and GSSG is important to maintaining a normal redox balance, and thus optimal cell function including mitochondrial function. In DM especially T2DM, the intracellular GSH and GSSG ratio is significantly decreased due to depletion of intracellular GSH and a significant increase in the intracellular level of GSSG (Murakami et al., 1989), favoring oxidative stress.

Other oxidative stress biomarkers such as lipid hyperperoxides (LHPs), H₂O₂, superoxide, and ox-LDL are also significantly increased in diabetes (Batty et al., 2022). A few isoforms of NOX such as NOX1, 4, and 5 are highly expressed in DM, resulting in excessive ROS (especially H₂O₂) generation (Iacobini et al., 2021; Poznyak et al., 2020). The activities and/or expressions of other enzymes, including COX, xanthine oxidase, and MPO, are also reported to be significantly increased in DM, contributing to oxidative stress (Virdis et al., 2013). In contrast, the expression levels of antioxidant enzymes, including SOD-1, Gpx-1, and catalase, were significantly decreased in T2DM (Zhu et al., 2022a).

Other antioxidant enzymes, including paraoxonase (PON) and thioredoxin (TRX) system, may also be compromised in DM. PON isoform is found in vascular walls and critically involved in the clearance of mtROS. In T2DM patients, gene polymorphisms of PON are related to a decreased PON activity, homeostatic model assessment for IR, lipid metabolism disorder, and nephropathy (Levy et al., 2019). Both thioredoxin-interacting protein (TXNIP) and TRX1 are important components of the antioxidant TRX. However, in hyperglycemic condition, overexpression of TXNIP in VSMCs and ECs suppresses TRX1 activity and increases ROS production (Dunn et al., 2010).

miRNAs and oxidative stress in diabetes

miRNAs are a large class of endogenous, single-stranded, and short nonprotein-coding RNA molecules of 19–25 nucleotides and are closely associated with a wide spectrum of biological functions (including the regulation of endothelial function, glucose and lipid metabolism, immune responses, and inflammation) and the development and progression of a variety of disease conditions, including CVDs such as endothelial dysfunction, atherosclerosis, cardiac hypertrophy, cardiac inflammation, fibrosis, and arrythmia (Laggerbauer and Engelhardt, 2022). Recently, miRNAs have been considered as therapeutic targets for the treatment of CVDs (Laggerbauer and Engelhardt, 2022).

Many miRNAs such as miRNA-21, miRNA-182, and miRNA-377 also play important roles in the pathogenesis of DM (both type 1 and type 2) and diabetic complications, and the circulating miRNAs have been serving as important biomarkers for diagnosis and prediction of DM and its complications including cardiovascular complications (Jiménez-Lucena et al., 2018; Mishra et al., 2023).

In addition, miRNAs are critically involved in the modulation of redox signaling pathways (such as NOX4, NF-κB, SODs, ROCK2, and PPAR alpha) and the expressions of various important inflammatory cytokines (including TNF-α, IL-1β, and IL-6) and chemokines through many diverse mechanisms (Qadir et al., 2019; Vezza et al., 2021), thus significantly contributing to excessive ROS generation and oxidative stress and subsequent endothelial dysfunction in DM, as nicely summarized in a few recent review articles (Laggerbauer and Engelhardt, 2022; Mishra et al., 2023; Qadir et al., 2019; Vezza et al., 2021).

Of note, some miRNAs such as miRNA-200, miRNA-34, and miRNA-204 may directly impair endothelial function in DM by targeting the expressions of sirtuin 1 (SIRT1), C/EBP homologous protein (CHOP), and other genes that are important for endothelial function (Qadir et al., 2019; Vezza et al., 2021). However, it is certainly beyond the scope of this review to summarize the roles of individual miRNAs in ROS production and oxidative stress as well as endothelial dysfunction in DM.

Chronic Inflammation and Diabetes

There is a close relationship between inflammation and DM with increased levels of a variety of inflammatory molecules, including vascular cell adhesion molecule (VCAM)-1, MCP-1, CRP, TNF-α, IL-1β, 6, 17, 23, and 18, and excessive ROS, as shown in Table 1. In DM, ECs are activated due to high level of glucose, excessive ROS, decreased NO, IR, and mitochondrial dysfunction, triggering immune cell activation and accumulation as well as release of proinflammatory molecules, leading to diabetic vascular inflammation and endothelial dysfunction (Takeda et al., 2020).

Table 1.

Key Inflammatory Molecules and Their Effects on Endothelial Cells in Diabetes Mellitus

Inflammatory molecules (References)	Effects on ECs
VCAM-1 (Gora et al., 2021; Poznyak et al., 2020)	Promoting monocytes adhesion to vascular ECs
MCP-1 (Gora et al., 2021; Poznyak et al., 2020)	Promoting macrophage infiltration in ECs
CRP (Lagrand et al., 1997; Sproston and Ashworth, 2018)	Promoting local complement activation in atherosclerotic lesions
TNF-α (Batty et al., 2022; Gora et al., 2021; Nallasamy et al., 2021)	NF-κB activation, promoting endothelial apoptosis, and proliferation of VSMCs
ILs (IL-1β, 6, 17, 23, 18) (Quevedo-Martínez et al., 2021; Velikova et al., 2021)	Proinflammatory cytokines that were increased in DM trigger endothelial dysfunction

CRP, C-reactive protein; DM, diabetes mellitus; ECs, endothelial cells; IL, interleukin; MCP-1, monocyte chemoattractant protein 1; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; TNF-α, tumor necrosis factor alpha; VCAM, vascular cell adhesion molecule; VSMCs, vascular smooth muscle cells.

Hyperglycemia-induced ROS overproduction promotes inflammation through multiple signaling pathways, including NF-κB, hypoxia-inducible factor-1α (HIF-1α) (Batty et al., 2022), and NLRP3 (Gora et al., 2021), as shown in Figure 4. Activations of these pathways enhance the expressions of intracellular adhesion molecule 1 (ICAM-1), VCAM-1, and E-selectin in ECs, thus facilitating the accumulation and activation of immune active cells to the surface of ECs (Batty et al., 2022; Gora et al., 2021; Son, 2012), and reenforcing the generation of more proinflammatory cytokines and chemokines such as TNF-α (Zhang et al., 2009), IL-1β, 6, and 18, MCP-1, and CRP (Gora et al., 2021; Lagrand et al., 1997; Son, 2012; Sproston and Ashworth, 2018).

FIG. 4.

Mechanisms for vascular endothelium inflammation induced by ROS. Chronic inflammation plays an important role in the pathogenesis of oxidative stress, endothelial dysfunction, and diabetic cardiovascular complications. Hyperglycemia generates excessive ROS and activates ECs, leading to increases in NF-κB-mediated signaling and expressions of HIF-1α and proinflammatory genes, and decreases in thrombomodulin and NO. Increases in LDL oxidation and expressions of ICAM-1, VCAM-1, and E-selectin promote the accumulation and/or recruitment of T lymphocytes and monocytes to vascular endothelium, triggering the release of inflammatory molecules such as TNF-α, IL-1β, CRP, and IL-6. Activated ECs also activate platelets and promote the release of vWF. ↑, increase; ↓, decrease; CRP, C-reactive protein; HIF-1α, hypoxia-inducible factor-1α; ICAM-1, intracellular adhesion molecule 1; LDL, low-density lipoprotein; MCP-1, monocyte chemoattractant protein 1; VCAM, vascular cell adhesion molecule; vWF, von-Willebrand factor.

Inflammation transforms the activated ECs to a prothrombotic phenotype via TNF-α, CD40, and other inflammatory cytokines (Bavendiek et al., 2002; Theofilis et al., 2021). In addition, ox-LDL and inflammatory molecules downregulate the expression of thrombomodulin on the surface of ECs, resulting in a prothrombotic state (Ishii et al., 2003). ROS-induced decrease in NO availability also impairs the ability of ECs to attenuate platelet aggregation (Hong et al., 2019; Theofilis et al., 2021), whereas inflammatory cytokines and excessive ROS increase the release of von-Willebrand factor (vWF) from ECs, leading to platelet activation and thrombus formation (Bernardo et al., 2004; Bkaily and Jacques, 2023; Theofilis et al., 2021).

Diabetes-associated chronic inflammation significantly increases the accumulations of circulating monocytes to atherosclerotic lesions (Poznyak et al., 2020) and decreases the macrophage subtype M2 population (an important source of anti-inflammatory cytokines) (Howangyin and Silvestre, 2014). Vascular ECs and macrophages communicate with each other, and a variety of bioactive molecules secreted from ECs directly modulate the polarization of macrophages (Takeda et al., 2020).

In diabetes, an increase in TNF-α and a decrease in NO can result in endothelial dysfunction and promote macrophage phenotypic changes from anti-inflammatory M2 toward proinflammatory M1 as well as increased productions of inflammatory cytokines (Takeda et al., 2020). From therapeutic perspective, targeting inflammation has been considered an important strategy for effective glucose control and management of diabetic cardiovascular complications (Agrawal and Kant, 2014; Goldfine and Shoelson, 2017; Kuryłowicz and Koźniewski, 2020; Li et al., 2023).

DM is an important risk factor for atherosclerosis (a leading contributing factor to cardiovascular morbidity and mortality), and both conditions are now considered inflammatory diseases pathophysiologically. It is also known that DM is associated with a significantly increased severity of coronavirus disease 2019 (COVID-19), caused by infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), with worse outcomes including higher mortality (Lim et al., 2021; Lv et al., 2022).

All these three conditions share at least three features (but not limited to) in their pathogenesis: (1) significant EC dysfunction; (2) increased levels of inflammation with excessive inflammatory cytokines and ROS and oxidative stress through a variety of signaling pathways such as NF-κB activation and trained EC immunity both systematically and locally; and (3) metabolic abnormalities (especially glucose, lipid, and free AA metabolisms) (Agrawal and Kant, 2014; Goldfine and Shoelson, 2017; Kuryłowicz and Koźniewski, 2020; Li et al., 2023; Lim et al., 2021; Lv et al., 2022; Maruhashi and Higashi, 2021; Poznyak et al., 2020; Xu et al., 2021).

The dynamic nature of inflammation is a ubiquitous component that plays a key role in the initiation and development of endothelial dysfunction and associated cardiovascular complications through recruitments and activations of inflammatory cells such as monocytes, macrophages, and lymphocytes and releases of a variety of potent inflammatory cytokines, including TNF-α, IL-1β, and IL-6 (Chung et al., 2021; Corona et al., 2021; Kuryłowicz and Koźniewski, 2020; Lim et al., 2021; Lv et al., 2022; Marchio et al., 2019; Popa-Fotea et al., 2023; Shao et al., 2021).

It is certainly beyond the scope of this review to provide a detailed comparison regarding inflammation and oxidative stress associated with DM versus other metabolic CVDs and COVID-19. However, the shared features of inflammation and endothelial dysfunction between DM, metabolic CVDs, and COVID-19 could have important implications on the prevention and/or treatment of endothelial dysfunction and CVDs associated with DM, COVID-19, or other metabolic disorders.

Antioxidant Therapy for Diabetic Complications

Treatments with antioxidant vitamins and supplements with inconsistent clinical outcomes

There is no doubt that excessive ROS and oxidative stress play an important role in the pathogenesis of diabetic complications, including diabetic CVDs and diabetic nephropathy (DN). Thus, targeting ROS becomes a natural and attractive option for prevention and treatment of diabetic complications. Preclinical studies using both T1DM and T2DM animal models have shown that treatments with antioxidants, including vitamins C and E and NAC, improve endothelial function and attenuate diabetic complications such as limb ischemia in DM (Batty et al., 2022; Dludla et al., 2018; Iacobini et al., 2021; Siti et al., 2015; Zhu et al., 2022a).

Short-term intra-arterial infusion of vitamin C also enhances endothelial function in diabetic patients (both type 1 and type 2) (Batty et al., 2022; Das, 2019; Mason et al., 2019; Siti et al., 2015; Ting et al., 1996). However, the outcomes of clinical studies with treatments of vitamins C have failed to demonstrate a significant beneficial effect against diabetic complications, including diabetic CVDs or DN (Batty et al., 2022; Belch et al., 2008; Hor et al., 2018; Iacobini et al., 2021; Siti et al., 2015), as shown in Table 2.

Table 2.

The Cardiovascular Outcomes of Clinical Studies with Treatment of Antioxidant Vitamins or Supplements in Patients with Diabetes Mellitus

Study design	Intervention	No. of subjects		Treatment duration	Outcomes	References
		Treatment	Control
Randomized crossover study in T2DM	Vitamins C (1 g twice a day orally)	27	27	4 Months	Improved PBG and 24-h BG, decreased BP	Mason et al. (2019)
Multicenter randomized study in T1DM or T2DM	Antioxidant capsulea/Aspirin (100 mg a day orally)	320	318	4 Years	No benefits on cardiovascular events and mortality	Belch et al. (2008)
Randomized study in T2DM with Hp 2-2 genotypeb	Vitamins E (400 IU a day orally)	726	708	18 Months	Reduced adverse cardiovascular events	Milman et al. (2008)
Randomized double-blind and placebo-controlled study in T2DM	Tocopherol (500 mg/day orally)	37	18	6 Weeks	Increased BP, no effect on endothelial function	Ward et al. (2007)
A pilot study in T2DM	Lipoic acid (0.2 mM intra-arterially)	39	—	One time	Improved endothelial function	Heitzer et al. (2001)

^a The antioxidant capsule contained α-tocopherol (200 mg), ascorbic acid (100 mg), pyridoxine hydrochloride (25 mg), zinc sulfate (10 mg), nicotinamide (10 mg), lecithin (9.4 mg), and sodium selenite (0.8 mg).

^b The Hp 2-2 phenotype, type 2 diabetic patients with haptoglobin 2-2 genotype.

BG, blood glucose or glycemia; BP, blood pressure; PBG, postprandial blood glucose or glycemia; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus.

A randomized study with placebo control of 1434 T2DM individuals (55 years of age or older) with the Hp 2-2 genotype showed that vitamin E supplementation (400 U/day for 18 months) exhibited a significant cardiovascular benefit with reduction in the primary composite outcome (myocardial infarction, stroke, and cardiovascular death; 2.2% for vitamin E group vs. 4.7% for placebo) (Milman et al., 2008). In contrast, another small randomized double-blind and placebo-controlled trial of 58 subjects with T2DM demonstrated that treatment with alpha-tocopherol (500 mg/day for 6 weeks) or mixed tocopherols significantly increased systolic BP, diastolic BP, and pulse pressure as compared with placebo.

No difference in endothelium-dependent or endothelium-independent vasodilation was observed in the subjects receiving alpha-tocopherol versus placebo (Ward et al., 2007), as shown in Table 2. Limited data from clinical studies have suggested that vitamin E attenuated oxidative stress and lipid peroxidation in diabetic subjects (both type 1 and type 2) (Gupta et al., 2011; Pazdro and Burgess, 2010). More studies are needed to determine if vitamin E supplement improves endothelial function and cardiovascular outcomes in diabetic patients.

Lipoic acid, also known as α-lipoic acid and thioctic acid, is produced endogenously from octanoic acid, and is considered a natural antioxidant. A small study of 39 T2DM patients demonstrated that intra-arterial infusion of lipoic acid significantly improved acetylcholine-induced (endothelium-dependent) vasodilation and attenuated oxidative stress in diabetic patients (Heitzer et al., 2001). Flavonoids such as cocoa flavanols are another large group of natural compounds of plant origins with antioxidant and anti-inflammatory properties (Panche et al., 2016).

Supplements with flavonoids have been shown to preserve endothelial function and eNOS activity in human subjects with obesity, pre-HTN, or T2DM through multiple mechanisms, and improve cardiovascular health with a reduced risk of CVDs (Panche et al., 2016; Yi et al., 2023). There has been considerable research and development for new antioxidants such as Tempol (a SOD mimetic) that has been under phase-2/3 clinical studies for specifically targeted fields for respiratory diseases treatment, including asthma, respiratory syncytial virus, influenza, and COVID-19.

Tempol has been shown to decrease ROS level and attenuate oxidative stress in preclinical diabetic animal models (Taye et al., 2013); however, no clinical studies have been done in diabetic patients. NOX inhibitors are another group of molecules that have been under active investigation as potential new antioxidants. Preclinical animal studies have demonstrated that NOX inhibitors are effective on inhibiting ROS production. However, their translation to the clinical applications remains in the early stages (Zhang et al., 2020).

The reasons for the apparent differences in the outcomes between preclinical investigations and clinical studies with antioxidant therapies have not been well defined yet. One may argue that competition for the same transporter between glucose and dehydroascorbate (the oxidation product of ascorbic acid) in DM may render treatment with vitamin C less effective or failure. Under physiological conditions, dehydroascorbate enters the cells and can be reduced back to ascorbic acid intracellularly.

In DM, the transporters could be mainly occupied by high levels of glucose and thus dehydroascorbate does not enter the cell and could not be regenerated. However, blood glucose level in preclinical animal models is usually higher than (or as high as) that in most diabetic patients. Therefore, the competition for the same transporter between glucose and dehydroascorbate could be similar and might not explain the outcome differences between preclinical and clinical studies with vitamin C unless a significant difference in the transporters for glucose could exist between human and other species that were used in preclinical studies.

It could be possible that the pharmacokinetic/pharmacodynamic features might be different for the study antioxidants between preclinical animal models and the patients in clinical studies. In addition, it is known that diabetic patients are often associated with other significant medical conditions, including (but not limited to) HTN, dyslipidemia, atherosclerosis, obesity, diabetic cardiomyopathy (a condition with significant abnormalities on myocardial structure, function, and metabolism directly caused by diabetes without CAD, HTN, or valvular heart disease), diabetic cardiovascular autonomic neuropathy, and DN (Ji et al., 2013; Lam et al., 2018; Poznyak et al., 2020; Prandi et al., 2023; Spallone, 2019; Williams et al., 2022).

These important co-existing medical conditions are unusually not present in animal models, and yet could certainly create significant challenges to the interpretation of the data from clinical studies with patients and to the management of DM. Thus, in addition to optimal blood glucose control and targeting oxidative stress, a combination of other effective interventions for other conditions could significantly improve the outcomes of diabetic complications especially cardiovascular morbidity and mortality in DM (Gaede et al., 2008; Ray et al., 2009; Tsao et al., 2023).

Although both vitamin C and E are considered potent antioxidants with broad spectrum of health benefits including cardiovascular health (Doseděl et al., 2021; Liao et al., 2022; Mohd Zaffarin et al., 2020; Xiong et al., 2023), vitamin C has been shown to have biphasic effects and higher doses of vitamin C can act as a pro-oxidant (Aronovitch et al., 1987; Podmore et al., 1998; Tyuryaeva and Lyublinskaya, 2023).

Indeed, two recent prospectively harmonized randomized clinical studies in critically ill patients with COVID-19 receiving organ support have revealed that treatment of critically ill hospitalized COVID-19 patients with high dose of vitamin C vitamin C (50 mg/kg of body weight intravenously every 6 h for 96 h) had low probability of improving the primary composite outcomes of organ support–free days and hospital survival. In fact, the enrollments were terminated because the statistical triggers for harm and futility were met for the studies (Adhikari et al., 2023).

In contrast, NAC treatment significantly reduces the severity and mortality in hospitalized COVID-19 patients (Afaghi et al., 2023; Alam et al., 2023; Milara et al., 2022), very likely due to its immunomodulatory properties and anti-inflammatory effects (Milara et al., 2022; Tieu et al., 2023). Similarly, evidence from some clinical studies with different patient populations (including T2DM patients) has demonstrated that vitamin E (especially at high doses) can behave as a pro-oxidant with increased levels of oxidative markers such as oxidative DNA damage and lipid oxidation (Baltusnikiene et al., 2023; Pearson et al., 2006; Weinberg et al., 2001; Winterbone et al., 2007).

Thus, the clinical outcomes of therapies with vitamins C and E could be very variable and difficult to predict in patients. Further studies are needed to address the inconsistencies and sometimes controversies on the effects of these antioxidant vitamins in different clinical settings and more importantly to improve the clinical outcomes for patients. Recent developments in technologies especially single-cell RNA sequencing techniques could help identify specific pathways that are responsible for the generation of excessive ROS in ECs in disease conditions such as DM and/or responsiveness to antioxidant therapies in individual diabetic patients (Zhao et al., 2021).

The single-cell transcriptomic data of ECs could also provide important information for the development of a personalized and targeted antioxidant therapy for each patient for optimal clinical outcomes (Wu et al., 2022).

NAC: a promising agent worthy of further investigation

NAC can be synthesized by acylating the amino acid l-cysteine and it has been traditionally considered an antioxidant, although it is more like an anti-inflammatory agent and effectively attenuates ROS production and reduces inflammation (Šalamon et al., 2019; Samuni et al., 2013; Tenório et al., 2021; Tieu et al., 2023). NAC regulates cytokine synthesis and inhibits some important inflammation-related signaling pathways (such as NF-κB), leading to reduced ROS production and a decreased level of oxidative stress (Šalamon et al., 2019; Samuni et al., 2013; Tenório et al., 2021).

NAC is an FDA-approved drug for the treatment of acetaminophen overdose (paracetamol) and as a mucolytic agent in respiratory diseases, such as cystic fibrosis or chronic obstructive pulmonary disease (COPD) (Šalamon et al., 2019; Samuni et al., 2013; Sochman, 2002; Tenório et al., 2021). NAC has been recently used as the study drug in many clinical studies for a variety of disease conditions, including CAD, atrial fibrillation, heart failure, chronic renal diseases, and psychological and neurological disorders (Kretzschmar et al., 1996; Ozaydin et al., 2013; Pasupathy et al., 2017; Pereira et al., 2019; Salsano et al., 2005; Soleimani et al., 2018; Tepel et al., 2003; Tossios et al., 2003).

It is reported that treating ECs from porcine pulmonary arteries with NAC significantly increases the cellular glutathione level and partially prevents TNF-α-induced endothelial dysfunctions (Xu et al., 2021). NAC also attenuates aortic endothelial damage in ApoE^−/− mice with streptozotocin-induced diabetes and high-fat diet (HFD) in association with increased levels of phosphorylated Akt (p-Akt) and p-eNOS in aorta, as well as NO in serum (Toborek et al., 1995).

Treatment of human aortic ECs with NAC significantly attenuates TNF-α-induced ROS production and DNA-binding activities of activator protein-1 and NF-κB, as well as p65 Ser276 phosphorylation (Fang et al., 2021). Long-term NAC treatment of arterial ECs from patients with severe CAD delays senescence of diseased ECs (Yang et al., 2010). Intra-arterial infusion of NAC in healthy human subjects at a rate to achieve a blood concentration of 1 mM potentiates the effects of nitroglycerin (NTG) on vasodilation and enhances the biotransformation to an endothelium-derived relaxing factor equivalent nitrosothiol (Voghel et al., 2008).

Intracoronary infusion of NAC in patients with or without coronary atherosclerosis significantly potentiates acetylcholine-induced coronary and femoral vasodilation and sodium nitroprusside (SNP)-induced coronary vasodilation (Creager et al., 1997). NAC treatment effectively prevents excessive productions of ROS and inflammatory cytokines, including TNF-α, IL-6, and IL-1β (Tieu et al., 2023), and preserves the population of EPCs in both circulation and in bone marrow in mice with ambient fine particulate matter exposure (Liu et al., 2022).

Interestingly, NAC treatment also preserves eNOS protein expression in the heart and in aortic and mesenteric artery tissues and normalizes systemic NO bioavailability in diabetic rats (Xia et al., 2006).

ROS formation has been shown to be significantly increased in diabetic ischemic hindlimb with an impaired recovery of blood flow and mechanical function in association with a significant reduction in the protein expressions of antioxidant enzymes, including SOD-1, Gpx-1, and catalase, as well as p-Akt (Ser473) and eNOS (Ser1177).

NAC treatment in T2D mice effectively preserved the expressions of antioxidant enzymes (SOD-1, Gpx-1, and catalase), and attenuated ROS production in the diabetic ischemic limb together with reversed levels of phosphorylated IRS-1, Akt, and eNOS, and plasma TNF-α. Associated with these changes in signaling was improved post-ischemia recovery of blood perfusion and limb function in T2D mice with increased capillary density, as well as decreases in inflammatory cell infiltration and rate of lower-extremity amputation (Zhu et al., 2022a).

Other important mechanisms for the beneficial effects of NAC on the reduction of ROS levels in DM include (but not limited to) (1) blocking ROS production from AGEs (Zhu et al., 2021); (2) inhibition of the in vivo biotransformation of native LDL to ox-LDL (Cui et al., 2015b); (3) direct suppression of ROS production from ox-LDL (Cui et al., 2015a; Li et al., 2014; Lu et al., 2010); (4) increase of intracellular glutathione level, and thus protecting mitochondria and their function (Murakami et al., 1989); (5) attenuation of NF-κB signaling, thus decreasing inflammatory cytokine productions (Šalamon et al., 2019; Samuni et al., 2013; Tenório et al., 2021); (6) normalization of eNOS expression and NO bioavailability (Xia et al., 2006); and (7) preservation of M2 polarization in hyperlipidemic condition, thus reducing inflammation and oxidative stress (Zhu et al., 2022b), as shown in Figure 5.

FIG. 5.

Mechanisms for the beneficial effect of NAC on ROS in DM. NAC could attenuate ROS through multiple mechanisms, including (but not limited to) increases in Akt-mediated signaling and expressions of eNOS and antioxidant enzymes, improved IR and mitochondria function, decreases in the levels of ox-LDL, AGEs, NF-κB signaling, and TNF-α, and preserved level of M2 macrophages. +, increase; −, decrease; NAC, N-acetylcysteine; p-IRS-1, phosphorylated insulin receptor substrate-1.

Although NAC treatment has no significant protective effects against DN in patients with T2DM (Rasi Hashemi et al., 2012; Saklayen et al., 2010), a clinical study of a small number of patients with HTN and T2DM (12 patients received NAC treatment and 12 as controls) has shown that NAC treatment for 6 months decreased endothelial activation and reduced systolic BP (Martina et al., 2008).

A recent small pilot study of 10 patients has demonstrated that supplement of glycine and NAC (GlyNAC) for 2 weeks could significantly attenuate mitochondrial dysfunction with significant reductions in mitochondrial fatty-acid oxidation and mitochondrial glucose oxidation and improve IR with decreased levels of serum FFA in patients with T2DM (Sekhar, 2022), as shown in Table 3. A combined treatment of GlyNAC has been shown to improve endothelial function and IR in prediabetic older adults (Kumar et al., 2023).

Table 3.

The Outcomes of Clinical Studies with N-Acetylcysteine Treatment in Patients with Diabetes Mellitus

Study design	Intervention	No. of subjects		Treatment duration	Outcomes	References
		Treatment	Control
Randomized study in prediabetic older adults	GlyNACa	12	12	16 Weeks	Improved endothelial dysfunction and IR	Kumar et al. (2023)
A pilot study in patients with T2DM	GlyNACa	10	10	2 Weeks	Improved mitochondrial dysfunction and IR	Sekhar (2022)
Case–control study in patients with T2DM	NAC infusionb	12	14	50 min	Diminished vasoconstrictive response to aldosterone infusion	Finsen et al. (2021)
A pilot study in patients with T2DM	NAC (600/1200 mg twice daily)	13	—	4 Weeks	No improvement in glucose tolerance or β-cell function	Szkudlinska et al. (2016)
Randomized study in patients with T2DM and DN	NAC (600 mg twice daily)	35	35	2 Months	No change in urine protein levels	Rasi Hashemi et al. (2012)
Randomized study in patients with T2DM	NAC (1200 mg once daily)	14	—	8 Days	Reduced platelet-monocyte conjugation	Treweeke et al. (2012)
Randomized crossover study in patients with T2DM and nonhealing diabetic ulcers	NAC (600 mg twice daily +600 mg every 30 min during hyperbaric therapy)	25	25	5 Days before and during hyperbaric oxygen therapy	Improved outcome of hyperbaric oxygen therapy	Efrati et al. (2009)
Randomized study in patients with T2DM	ARG+NACc	12	12	6 Months	Improved endothelial function, decreased ROS	Martina et al. (2008)

^a GlyNAC, capsule with 100 mg/kg per day of glycine and 100 mg/kg per day of NAC.

^b NAC infusion, 125 mg/kg per hour for 20 min followed by 25 mg/kg per hour for another 30 min.

^c ARG+NAC, 1200 mg arginine once a day and 600 mg NAC twice a day.

DN, diabetic nephropathy; GlyNAC, glycine and NAC; IR, insulin resistance; NAC, N-acetylcysteine; ROS, reactive oxygen species.

Although NAC effectively prevented the vasoconstrictive response to aldosterone infusion in patients with T2DM (Finsen et al., 2021), NAC did not improve glucose tolerance or β-cell function in these patients (Szkudlinska et al., 2016). In addition, NAC has been reported to reduce platelet-monocyte conjugation in patients with T2DM (Treweeke et al., 2012), and improve the outcome of hyperbaric oxygen therapy in T2DM patients with nonhealing ulcers (Efrati et al., 2009).

However, all these clinical studies to evaluate the effect of NAC on oxidative stress and endothelial function in DM NAC were very small as detailed in Table 3. There has been no clinical study to determine if NAC treatment has significant beneficial effects on the outcomes of cardiovascular complications in diabetic patients.

There are multiple sources for excessive ROS in DM through different mechanisms. Clearly, therapy with one antioxidant such as vitamin C or E or their combination may not effectively block ROS production from all sources with potential biphasic response in diabetes. NAC, in contrast, may provide adequate suppression of ROS generation since it can attenuate ROS formation from multiple sources in DM as shown in Figure 5.

Thus, NAC treatment could lead to improved clinical outcomes with a decreased risk of cardiovascular complications in diabetic patients. Recently, ROS systems have been considered an integrated network for oxidative signal sensing and stress alarming in subcellular organelles such as mitochondria (Sun et al., 2020). This novel model in combination with single-cell transcriptomic data could identify specific redox signaling pathways in different subcellular organelles of ECs or other cell types to help developing novel and effective therapeutic strategies that target the specific pathways critical to excessive ROS in DM.

Inflammation and Trained EC Immunity

For the past decade, accumulating evidence has supported a novel concept that ECs function as dynamic innate immune cells and could develop innate immune memory with significant plasticity phenotypically in response to extracellular environmental changes (Drummer et al., 2021; Lu et al., 2019; Mai et al., 2013; Saaoud et al., 2023; Shao et al., 2020).

It has been shown that ECs can detect exogenous pathogens such as lipopolysaccharide and virus and harmful endogenous metabolic stress signals such as ox-LDL, FFAs, and excessive glucose in circulation, triggering inflammatory and/or immune responses with the production of proinflammatory cytokines and chemokines, including TNF-α and IL-6, as well as mobilization and recruitments of immune cells such as T cells and monocytes (Mai et al., 2013; Shao et al., 2020; Sun et al., 2020; Yin et al., 2013).

The data have demonstrated that ECs can be trained to develop lasting memory on immunity, thus enhancing their innate immune/proinflammatory responses through a variety of mechanisms, including metabolic and epigenetic reprogramming (Lu et al., 2019; Millán-Zambrano et al., 2022).

These novel findings could have significant therapeutic implications for disease conditions associated with inflammation and autoimmune disorders as well as their associated complications such as DM and atherosclerosis (Xu et al., 2023). Indeed, a recent study revealed that alternate day fasting significantly improved glucose metabolism and endothelial function in diabetic Lepr^db mice with reduced levels of oxidative stress biomarkers, very likely by blocking trained immunity-related metabolic pathways (Cui et al., 2022).

Another study recently reported that treatment of diabetic db/db mice with the agonistic analog of growth hormone-releasing hormone MR409 effectively inhibited the expression and production of trained immunity mediator ROS in vasculature with improved EC function and enhanced the expression of anti-calcifying protein Klotho with significant reduction in vascular calcium deposition (Ren et al., 2023). This new model could certainly promote the discovery of new pathways for ROS overproduction and novel therapeutic targets in DM and the development of new effective targeted treatment strategies for endothelial dysfunction and subsequent cardiovascular complications in DM.

Summary and Clinical Significances

In this review, we have discussed the role of ROS and oxidative stress in the development of EC dysfunction in DM and related mechanisms on excessive ROS generation. ROS overproduction and oxidative stress are critically involved in the development and progression of endothelial dysfunction and subsequent cardiovascular complications in DM. However, targeting ROS with antioxidant vitamins C and E or other antioxidant supplements has no significant protective effects against diabetic vascular complications. Preclinical and limited clinical data suggest that NAC treatment may improve endothelial function in diabetic patients.

However, well-designed clinical studies are needed to determine if NAC supplementation could significantly improve endothelial function and the clinical outcomes of diabetic patients with reduced cardiovascular morbidity and mortality. With better understanding of the mechanisms contributing to excessive ROS generation and ROS-mediated endothelial damages/dysfunction, it is predicted that development of new selective ROS-modulating agents will lead to effective personalized preventive and/or treatment strategies for endothelial dysfunction and subsequent cardiovascular complications such as atherosclerosis in DM.

Abbreviations Used

AA

arachidonic acid

AGE

advanced glycation end-product

Akt

protein kinase B or serine/threonine-specific protein kinases

AMI

acute myocardial infarction

ASCVD

atherosclerotic cardiovascular disease

BG

blood glucose or glycemia

BH4

tetrahydrobiopterin

BP

blood pressure

CABG

coronary artery bypass graft surgery

CAD

coronary artery disease

CHOP

C/EBP homologous protein

COVID-19

coronavirus disease 2019

COX

cyclooxygenase

CRP

C-reactive protein

CVD

cardiovascular disease

DAG

diacylglycerol

DM

diabetes mellitus

DN

diabetic nephropathy

EC

endothelial cell

EDCF

endothelium-derived contracting factor

eNOS

endothelial nitric oxide synthase

EPC

endothelial progenitor cell

ERK

extracellular signal-regulated kinase

ET-1

endothelin-1

ETC

electron transport chain

FADH2

1,5-dihydroflavin adenine dinucleotide

FFA

free fatty acid

GLUT

glucose transporters

GlyNAC

glycine and NAC

Gpx

glutathione peroxidase

GSH

glutathione

GSSG

glutathione disulfide

H₂O₂

hydrogen peroxide

HIF-1α

hypoxia-inducible factor-1α

HTN

hypertension

ICAM

intracellular adhesion molecule

IL

interleukin

IR

insulin resistance

IRI

ischemia-reperfusion injury

IRS

insulin receptor substrate

LDL

low-density lipoprotein

LMW-HA

low molecular weight hyaluronan fragments

LOOH

lipid peroxide

LOX-1

lectin-like oxidized low-density lipoprotein receptor-1

MACE

major adverse cardiovascular events

MAPK

mitogen-activated protein kinase

MCP-1

monocyte chemoattractant protein 1

miRNA

microRNA

Mn-SOD

manganese-superoxide dismutase

MPO

myeloperoxidase

mtROS

mitochondrial reactive oxygen species

NAC

N-acetylcysteine

NADH

nicotinamide adenine dinucleotide

NADPH

nicotinamide adenine dinucleotide phosphate

NF-κB

nuclear factor kappa-light-chain-enhancer of activated B cells

NLRP3

pyrin domain-containing protein 3

NO

nitric oxide

NOX

nicotinamide adenine dinucleotide phosphate oxidase

ONOO⁻

peroxynitrite anion

ox-LDL

oxidized low-density lipoprotein cholesterol

p-IRS-1

phosphorylated insulin receptor substrate-1

PAR

protease-activated receptor

PARP-1

poly (ADP-ribosome) polymerase 1

PBG

postprandial blood glucose or glycemia

PCI

percutaneous coronary artery intervention

PGI2

prostacyclin

PI3-K

phosphoinositide 3-kinase

PKC

protein kinase C

PON

paraoxonase

PPARγ

peroxisome proliferator-activated receptor-γ

RAGE

receptor for advanced glycation end-product

ROCK

RhoA-rho kinase

ROS

reactive oxygen species

SIRT1

sirtuin 1

SOD

superoxide dismutase

T1DM

type 1 diabetes mellitus

T2DM

type 2 diabetes mellitus

TNF-α

tumor necrosis factor alpha

TRX

thioredoxin

TXNIP

thioredoxin-interacting protein

VCAM

vascular cell adhesion molecule

VSMC

vascular smooth muscle cell

vWF

von-Willebrand factor

Authors' Contributions

Conceptualization by Z.L. and X.L.; validation by Z.L., X.L., and H.C.; resources by Z.L. and H.H.; data curations by X.L., A.L., J.C., J.Z., and H.H.; writing—original draft preparation by X.L.; writing—review and editing by Z.L., H.H., and H.C.; visualization by X.L. and J.Z.; supervision and funding acquisition by Z.L. All authors have read and agreed to publish the article.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was partially supported by U.S. NIH grants HL148196 and NS132279 (to Z.L.).