Oxidative stress as a potential target in acute kidney injury

Are reactive species important in maintaining homeostasis?

It is considered that RS are important for maintaining homeostasis when they are present in low concentrations, in certain compartments, at a certain moment. ROS and RNS are responsible for normal redox signaling, and therefore they represent a promotor for cell survival, growth, and proliferation (Ratliff et al., 2016).

In the kidney, the blood flow and the glomerular filtration rate are maintained at a constant level by renal autoregulation, despite all the physiological changes or pathological states that may take place in the body. Two key mechanisms are responsible for the renal autoregulation: the myogenic response and tubuloglomerular feedback. Studies show that both vasoreactivity and the myogenic response are influenced by the presence of RS such as nitric oxide (NO) and superoxide (O₂⁻) (Carlström, Wilcox & Arendshorst, 2015; Just, 1997; Loutzenhiser et al., 2006; Navar et al., 2008).

Nitric oxide is a vasodilator factor generated by endothelial NO synthase (eNOS), and it defends the renal cells (endothelial and mesangial cells) from fibrosis and apoptosis by inducing the expression of antioxidative genes and promoting a physiological renal hemodynamics (Ratliff et al., 2016). It is well known that NO is a major factor in modulating the myogenic response (Carlström, Wilcox & Arendshorst, 2015), with numerous studies showing that inhibition of NOS decreases renal blood flow and increases the renal vascular resistance in both pressor dosages (Baylis & Qiu, 1996; Majid, Williams & Navar, 1993) as well as in dosages that have no impact on the blood pressure (Carlström, Wilcox & Arendshorst, 2015; Deng & Baylis, 1993). Also, the infusion of NO donors (sodium nitroprusside) and NO precursors (L-arginine) seems to counteract renal vasoconstriction (Kiyomoto et al., 1992; Kumagai et al., 1994).

Intracellularly, low levels of NO may inhibit cytochrome-c oxidase, altering the generation of ROS in the mitochondria (Brown, 1995). When ROS produced by mitochondria increase moderately, they stabilize hypoxia-inducible factor (HIF) in endothelial cells, and stimulate the nuclear factor erythroid 2–related factor 2 (Nrf2) (Piantadosi & Suliman, 2006; Guzy et al., 2005). Both HIF and Nrf2 protect the renal tissues against OS, therefore ROS may exhibit renoprotective effects in certain concentrations (Ratliff et al., 2016; Nangaku & Eckardt, 2007). Nrf2 represents the main regulator of the response to OS (Ruiz et al., 2013).

Interestingly, one study that focused on the effects of endogenous O₂⁻ showed that it becomes a crucial factor in maintaining a physiological tone in the renal vasculature when it is induced by vascular NADPH oxidase (NOX) (Haque & Majid, 2004). This further demonstrates that RS might play a crucial role in maintaining homeostasis.

What reactive species are generated in AKI?

The main etiologies of AKI are linked to ischemia and hypoxia. Due to the decrease in the renal blood flow, the cellular nutrient and oxygen uptake is limited, leading do the development of inflammation and acute tubular necrosis (Basile, Anderson & Sutton, 2012). Renal ischemia and reperfusion represent major triggers of ROS with consequential damage in renal function and tissue integrity (Deng & Baylis, 1993). In sepsis-induced AKI there is an extensive immune response that is responsible for renal vasoconstriction, endothelial injury, and localized hypoxia, which ultimately triggers the formation of ROS (Andrades et al., 2011). Further, uremia is highly associated with an increase in circulating levels of indole and carbonyl compounds, which are able to upregulate systemic OS (Tanaka et al., 2017).

Some of the initial changes that are observed with AKI development are depletion of ATP and changes in the structure of the mitochondria, leading to alterations and dysfunction in the energetic metabolism (Kiyomoto et al., 1992; Tanaka et al., 2017; Tracz et al., 2007). Each type of renal cell presents a different number of mitochondria, which is considered to be one of the major sources of ROS (Duchen & Szabadkai, 2010), therefore the production of ROS varies between the renal structures. Mitochondria is also responsible for producing molecules that counteract the OS (Singh et al., 2019). In the kidney, NOX and the mitochondrial respiratory chain are considered to be the main sources of ROS production (Sureshbabu, Ryter & Choi, 2015).

One study evaluated the mitochondrial function, structure and redox state in rodents with induced nephrotoxic and ischemic AKI via in vivo exogenous and endogenous multi-photon imaging. The results showed changes in mitochondrial NADH levels and proton motive force, and upregulated levels of mitochondrial O₂, along with disjointed mitochondria. They concluded that mitochondria represents a major source of ROS, and that an alteration in mitochondrial function plays an important role in the early phase of renal ischemia (Hall et al., 2013; Tanaka et al., 2017). In contrast, after gentamycin exposure, the first alterations to appear were at lysosomal level in the renal epithelium, along with anomalies in brush border cells. The mitochondrial dysfunction, with changes in morphology of the mitochondria, alterations in NADH levels, RS and altered proton motive force occurred later in the development of nephrotoxic AKI (Tanaka et al., 2017; Deisseroth & Dounce, 1970). Mitochondria is, therefore, directly responsible for increased levels of OS in AKI.

Despite the abundance of studies conducted on this topic, the exact mechanism through which RS are generated in AKI remains unknown. However, OS might represent a potential therapeutic target. Table 1 summarizes the main ROS found in AKI.

The production of the superoxide anion O₂⁻ is the result of the one-electron reduction of oxygen in its molecular form (Noiri, Addabbo & Goligorsky, 2011). Superoxide can be generated by a large variety of oxidase enzymes, and can also be generated inside the mitochondria by components of the electron transport chain. O₂⁻ is mainly transported through anion channels, therefore the diffusion across different membranes is limited. Superoxide anion is a rather selective FR, which is able to form a non-radical ROS (H₂O₂) via dismutation. The reaction can take place spontaneously, but it can also be facilitated by enzymatic catalysis (Ratliff et al., 2016). The presence of the superoxide anion triggers a cascade of events, as its presence leads to the generation of other ROS. O₂⁻ may also be produced from xanthine by xanthine oxidase (XO), and from NADPH and NADH by various oxidase enzymes which are induced by an inflammatory response (Deng & Baylis, 1993). The most potent action of O₂⁻ is represented by the scavenging of NO. As the levels of superoxide anion increases, it is able to disrupt the iron-sulfur centers, and it may react with catecholamines (Ratliff et al., 2016). Local ischemia and cytokines generated in AKI induced by sepsis activate the endothelium of the renal vasculature and recruit cells from the immune system that are able to generate O₂⁻ via NOX (Kiyomoto et al., 1992).

Hydrogen peroxide (H₂O₂) can be generated by dismutation but also by oxidases which are able to directly reduce the molecular oxygen. H₂O₂ diffuses across different biological membranes in a similar way to water, which makes it able to express its oxidative properties in other cellular compartments and even in other cells. Peroxides are able to react with different molecules containing iron, leading to generation of additional ROS. When peroxides react with Fe²⁺ they generate the hydroxyl radical (HO⁻), known as the Fenton reaction. When peroxide is metabolized by specific heme peroxidase in the presence of other molecules such as chloride or nitrite, it can generate hypochlorous acid, and nitrogen dioxide (Ratliff et al., 2016; Dennis & Witting, 2017).

The hydroxyl radical (HO⁻) may be generated by the Fenton reaction. This radical alters and reacts with almost every cellular component, generating additional ROS (Brown, 1995). It produces lipid peroxidation with subsequent membrane damage and toxic compounds release, including aldehydes (Dennis & Witting, 2017).

Hypochlorous acid (HClO), a highly reactive lipid soluble molecule (Ratliff et al., 2016), is generated by phagocyte myeloperoxidase in inflammatory cells when local ischemia is present (Kiyomoto et al., 1992), and it may react with amines, producing chloroamines.

Peroxynitrite (ONOO⁻) is generated when superoxide anion reacts with NO. The spontaneous decomposition of peroxynitrite generates nitrogen dioxide (NO₂). Also, NO₂ is generated by heme peroxidase enzymes from the nitrate metabolism. These RS are able to activate a variety of signaling mechanisms by oxidizing or nitrating thiols, sometimes influencing processes that are also targeted by peroxide. When RNS react with unsaturated fatty acids (FA), the result is oxidized and nitrated FA which present a multitude of biological actions. ONOO⁻ seems to be a major inhibitor of the mitochondrial respiration chain, irreversibly disrupting the centers of iron-sulfur (Ratliff et al., 2016).

Nitric oxideNO represents a soluble gas that is able to bind to ferrous sites of heme. It is generated by NOS enzymes, and in normal amounts it is a part of physiological signaling processes: it binds to guanylate cyclase in order to regulate the production of cGMP, and it binds to cytochrome c oxidase to regulate the mitochondrial respiration. While it is mainly generated by regulating the activity of nNOS (NOS1) and eNOS (NOS3), an inflammatory status is able to induce iNOS (NOS2) expression (iNOS—inducible NO synthase; nNOS—neuronal NO synthase). This process leads to the generation of larger quantities of NO which consequently modulates the inflammatory processes. Larger levels of NO interact with bound iron, and produce NO-derived RNS that can nitrosate thiols. When the concentration of NO equalizes the levels of local antioxidant enzymes, it competes with them for the scavenging of O₂⁻, generating ONOO⁻(Ratliff et al., 2016; Dennis & Witting, 2017). However, it has been shown that when using agents to inhibit the global production of NO, including the NO production from constitutive eNOS, the effect on renal ischemia-reperfusion injury (IRI) was not renoprotective (Yaqoob, Edelstein & Schrier, 1996).

Another study showed that mice who were deficient in iNOS were resistant to renal IRI, concluding that a constant release of NO from NOS2 could be a pathogenetic factor (Ling et al., 1999). NOS2 is constitutively expressed in the renal tissue (Thomas et al., 2015), underlining that NO presents a physiological role in kidney functioning (Araujo & Welch, 2006). These mechanisms may potentially contribute to the development of renal disease.

Carbonyls, nitrated lipids, and unsaturated aldehydes are just some of the toxic reactive molecules that can be generated by ROS and RNS. These molecules are able to trigger signaling processes, at low levels of OS. As the level of OS increases, these molecules promote oxidative damage and cellular death (Ratliff et al., 2016; Dennis & Witting, 2017). Out of all possible etiologies for AKI, two of the most common are IRI and sepsis (Piantadosi & Suliman, 2006).

Ischemia-reperfusion injury induced AKI

Ischemia-reperfusion injury represents a major cause of AKI, and the main cause of delayed renal graft function, and renal graft loss after kidney transplantation (Bogdan, 2001). In IRI, the reperfusion phase is the crucial moment when the most IRI damage might occur. The initial event that takes place immediately after reperfusion is a sudden increase in superoxide anion production in the mitochondria which is released inside the cell, and represents the main trigger for the pathology that follows reperfusion (Zweier, Flaherty & Weisfeldt, 1987; Chouchani et al., 2016). In IRI, this mechanism plays a crucial role in initiating AKI, but also in maintaining it (Nath & Norby, 2000). Also, the early inflammatory response in IRI consists mainly of neutrophils, which generate ROS, and are recruited by ROS (Friedewald & Rabb, 2004).

One study that focused on the pathophysiology of renal IRI showed that pericytes express adhesion protein-1, a key protein that generates a local H₂O₂ gradient which further stimulates the neutrophil infiltration (Tanaka et al., 2017), proving that RS are directly involved in developing renal injury following ischemia.

Another study concluded that, following renal IRI, Nrf2-regulated cell defense genes presented significantly high levels in the kidneys of wild-type mice but not in Nrf2-knockout (^−/−) mice (Leonard et al., 2006; Liu et al., 2009). The same study showed that the loss of Nrf2 leads to an increase in severity of renal IRI, which also proves that OS plays a crucial role in the pathogenesis and outcome of renal IRI.

Tracz et al. (2007) and Piantadosi & Suliman (2006) showed that mice who were heme oxygenase-1 knockout (HO-1^−/−) had increased sensitivity to renal IRI, and expressed a higher inflammatory response and oxidative damage and an increase in mortality when compared to wild-type mice.

Another study demonstrated that mice deficient in transient receptor potential melastatin 2 (TRPM2) who were subjected to renal IRI showed resistance to OS and apoptosis (Gao et al., 2014). TRPM2 is a nonselective cation channel (for calcium, sodium, and potassium) activated by OS, ADP-ribose, tumor necrosis factor-α (TNF-α), and intracellular calcium, all of which are considered to be increased in kidney ischemia. Gao et al. showed that TRPM2-knockout mice are resistant to IRI, with a reduction in OS, NOX activity, and apoptosis in the kidney. Also, they obtained similar results by pharmacologically inhibiting TRPM2. The cationic channel was identified mainly in the proximal tubule epithelial cells, and Gao et al. showed that its effects are attributable to expression in parenchymal cells. Also, they showed that the activation of RAC1, a constituent of the NOX complex, was promoted by TRPM2 following renal ischemia. In contrast, inhibition of RAC1 in vivo reduced ischemic injury and OS. This further shows that OS and ischemic injury are closely interrelated.

Sepsis-induced AKI

In intensive care units (ICU), sepsis represents the major cause of AKI (Uchino et al., 2005). Of all the bacterial pathogens that induce AKI, Gram-negative bacteria represent a major threat due to the composition of their outer membrane. One of the main components of their outer membrane is represented by lipopolysaccharide (LPS), an endotoxin considered to be an important trigger of the inflammatory response in sepsis. These LPSs are recognized by Toll-like receptor (TLR), a transmembrane protein expressed in renal tubules. Once the LPS bind to TLRs, the immune system is triggered, and the acquired immunity develops an antigen-specific response (Medzhitov, 2001).

TLR4 binds specifically to the LPS ligand, and their binding triggers a cascade of events that leads to OS and a pro-inflammatory response, with new cytokines and mediators being further released. TNF-α and interleukin-1β (IL-1β) are just two of the newly generated molecules which consequently promote the H₂O₂ formation, with oxidative damage which further enhances the inflammatory response (Li & Engelhardt, 2006).

Many studies suggest that both sepsis and ischemia upregulate TLR4 (Wu et al., 2007; Pulskens et al., 2008; El-Achkar et al., 2006). Hato et al. (2015) studied the effect of endotoxin preconditioning and concluded that it confers renal epithelial protection in vivo, by preventing peroxisomal damage, abolishing OS, and tubule injury. OS was measured in both preconditioned and nonpreconditioned mice, with measurements performed after 4 h from LPS inoculation intraperitoneally. High levels of OS were determined in proximal tubules of NP wild-type mice, whereas in preconditioned wild-type mice OS was absent. TLR4^−/− mice showed no OS, which confirms that TLR4 plays a crucial role in LPS signaling pathway. Also, in preconditioned mice, KIM-1, and NGAL, markers for tubular injury, were significantly reduced, suggesting renal protection. This study proves that sepsis indeed upregulates TRL4, and leads to high levels of OS in renal proximal tubules, with subsequent oxidative damage.

TLR9 identifies bacterial DNA with unmethylated CpG motifs (Tsuji et al., 2016), but the TLR9 pathway is also activated by endogenous mitochondrial DNA (mtDNA) (Zhang, Itagaki & Hauser, 2010). Tsuji et al. (Gao et al., 2014) studied the role of mtDNA via TLR9 in septic AKI and showed that cytokine production, tubular mitochondrial disorders (ATP depletion), and splenic apoptosis are inflicted by the circulating mtDNA released in the early phase of sepsis. They used an animal model of sepsis in wild-type and TLR9^−/− mice, but they also injected mitochondrial debris intravenously. The results show that TLR9^−/− mice presented a lower bacteremia than wild-type mice, but with a higher level of leucocyte migration. Also, TLR9^−/− mice presented lower levels of IFN-γ and IL-12 than wild-type mice, with attenuated production of superoxide in the proximal tubule cells. When injected with mitochondrial debris containing a substantial amount of mtDNA, wild-type mice had increased plasma levels of IL-12, but not of IFN-γ and TNF-α, while the levels of IL-12 in TLR9^−/− mice did not increase as much. This study further shows that sepsis upregulates certain receptors that stimulate the production of RS and oxidative damage.

While sepsis remains ubiquitous, aforementioned studies suggest that the cascade of events that it triggers, including OS damage and inflammatory response, could be counteracted at receptor level. More studies are needed in order to gain a better perspective of how OS could be addressed in sepsis-induced AKI.