Abstract

1. Introduction

Superoxide dismutases (SODs) are metalloenzymes found in eukaryotes and some prokaryotes and as shown in Figure 1A, they are localized in the cytosol and the mitochondrial intermembrane (Cu, Zn-SOD or SOD1), the mitochondrial matrix and inner membrane (Mn-SOD or SOD2) [1], and extracellular compartment (Cu, Zn-SOD or SOD3) [2].

Open in a separate window

Figure 1

Superoxide dismutase enzymes. (A) Superoxide dismutases (SODs) are metalloenzymes constitutively expressed in eukaryotes: SOD1 is a Cu, Zn-SOD and is present in the cytosol and the mitochondrial intermembrane; SOD2 is a Mn-SOD localized in the matrix and inner membrane of mitochondria; SOD3 is a Cu, Zn-SOD expressed in the extracellular compartment. Nevertheless, all three forms catalyze the conversion of the superoxide anion free radical (^•O₂⁻) into hydrogen peroxide (H₂O₂). (B) In detail, SOD converts the ^•O₂⁻, generated in several cellular insults/metabolism, into H₂O₂ and molecular oxygen (O₂). The resulting H₂O₂ may undergo reduction to water via catalase (CAT), glutathione peroxidases (GPx), or thioredoxin (Trx)-dependent peroxiredoxin (Prx). Otherwise, H₂O₂ originates ^•OH via the Fenton reaction in the presence of Fe²⁺. ^•O₂⁻ may also react with ^•NO originating the oxidant and nitrating agent peroxynitrite (ONOO⁻), which further contributes to oxidative-stress damage. GSH = glutathione; GSSG = glutathione disulfide; TrxSH₂ = reduced thioredoxin; TrxS₂ = oxidized thioredoxin.

Since their discovery by Joe McCord and Irwin Fridovich [3], their role as a major antioxidant defensehas been firmly recognized [4]. The work by I. Fridovich and collaborators was crucial in defining the role of oxidant/antioxidant processes in ischemia/reperfusion-associated pathologies in humans and animal models [5,6].

SOD catalyzes the conversion of the superoxide anion free radical (^•O₂⁻) to hydrogen peroxide (H₂O₂) and molecular oxygen O₂ (Figure 1A,B). Subsequently, H₂O₂ is reduced to water by the catalase (CAT) enzyme, glutathione peroxidase (GPx), and/or thioredoxin (Trx)-dependent peroxiredoxin (Prx) enzymes (Figure 1B). H₂O₂ may also generate another reactive oxygen species (ROS), the hydroxide ion (^•HO) via the Fenton reaction in the presence of Fe²⁺ (Figure 1B).

H₂O₂ is an essential sensor in redox metabolism. Its levels are critical to oxidative stress: under physiological conditions, when H₂O₂ intracellular concentration are 1–10 nM, it mediates the stress response involved in the physiological and adaptive processes called oxidative eustress; higher concentrations (more than 100 nM) are responsible for the so-called oxidative distress, in which the evoked inflammatory response leads to cell damage [7,8]. Considering the endogenous antioxidant system involved in H₂O₂ production and removal, a parallel dual role, physiological and pathological, can also be recognized for all the enzymes involved. SOD activity may therefore have a double and opposite meaning [9]: firstly, it is an antioxidant enzyme when its activity is coordinated with either the CAT, GPx or Prx/Trx enzymes, which avoid H₂O₂ accumulation by neutralizing it into H₂O; secondly, SOD may act as a pro-oxidant as H₂O₂ can overaccumulate, leading to ROS overproduction and cell toxicity [7].

Accordingly, a bell-shaped dose-response curve describes the protective effects of SOD on isolated heart preparation, with low doses (up to 5 μg/mL in the perfusate) protecting, and high doses (50 μg/mL in the perfusate) exacerbating reoxygenation-induced injury [10]. However, when SOD activity increases, the enhanced levels of H₂O₂ trigger the upregulation of CAT [11] and/or GPx [12], with a final antioxidant balance as a compensatory and defense response strategy.

SODs are also involved, at least partially, in detoxification from the oxidant and nitrating agent peroxynitrite (ONOO⁻), which is formed from the reaction between ^•NO and ^•O₂⁻ (Figure 1B). ONOO^- rapidly forms reactive free radicals upon reaction with CO₂ [11]. SOD also prevents this detrimental event.

On this basis, it is universally recognized that SOD is the first line of defense against the toxicity of ^•O₂⁻ because catalyzing the dismutation of two molecules of ^•O₂⁻ to hydrogen H₂O₂ and O₂ limits the ^•O₂⁻ availability. Low and diminished SOD activity has been associated with a significant risk of oxidative stress, resulting in disease, such as hypertension, hypercholesterolemia, atherosclerosis, diabetes, heart failure, stroke and other cardiovascular diseases [12,13]. Therefore, it has been suggested that the antioxidant properties of SOD supplementation are useful in a variety of pathophysiological conditions, from protecting the immune system to the prevention of aging [14]. The consumption of natural sources of SOD, such as cabbage, Brussels sprouts, wheatgrass, barley grass and broccoli has been encouraged [15].

The use of SOD as a drug may be advantageous in terms of the quantity and duration of the pharmacological effect, compared to other antioxidants. Indeed, SOD supplementation may trigger the endogenous antioxidant machinery to neutralize a free radical excess without being consumed upon ROS detoxification. By contrast, non-enzymatic antioxidants, such as glutathione (GSH), are known to be depleted [16]. However, pharmacological treatment using exogenous SOD administration is not yet an established clinical practice, and usually dietary supplementation is pursued. Indeed, efficacy depends on the source of SOD. Although there is a lack of head-to-head studies, a study in rats has demonstrated that human and bovine SODconferred higher pharmacological activity that the rat enzyme [17].

Moreover, the treatment of human diseases with the human enzyme may not yield beneficial effects. Bovine SOD, known as orgotein, was usually preferred. However, it can be limited by its intramuscular administration, administration frequency (2~3 times weekly) [9], and possible toxicity, caused by the presence of 20% impurities (albumin and chymotrypsin are the primary contaminants), in the pharmaceutical preparation that may result in immediate hypersensitivity reactions [18], and other side effects, including allergy [16]. Orgotein, marketed for the treatment of a range of inflammatory diseases, was withdrawn from European countries [18], due to allergic reactions, and limited to veterinary use in the US.

Over time, plant-extracted SOD became the alternative. Cantaloupe-melon-(Cucumis melo L.C.)-derived SOD, SODB, which offers the advantage of a high SOD concentration (100 U/mg) and low contents of other antioxidants, such as CAT (10 U/mg) and GSH (1 U/mg), is one of the most commonly used [19,20]. However, the oral bioavailability of this form of SOD is still very low, according to the general pharmacokinetics principle of drugs, and this is because of its high molecular weight, which affects cellular uptake [21], and the low pH and high proteolytic activity in the digestive tract [22]. As natural SOD is an exogenous protein, we can hypothesize that it may induce antibody formation (anti-drug antibodies ADA). However, considerable experience with the infusion of proteins as drugs for therapeutic purposes has indicated that there is only a marginal reduction in their effect and no clinically demonstrated toxicity.

Thus, the use of SOD mimetics and new delivery systems to protect SOD are under investigation [23]. SOD mimetics are intended to overcome the limits of natural SOD enzymes. They have better pharmacokinetic properties and some pharmacodynamic differences, with negligible antigenicity potential. Indeed, SOD mimetics have a low molecular weight, more stability and a long-circulating half-life, guaranteeing a better pharmacokinetic profile. Moreover, they have a different dose–response curve; natural SOD displays a bell-shaped dose-dependent curve, while most SOD mimetics have a dose-proportional response [24]. Finally, their mechanism of action is far beyond that of ^•O₂⁻ scavenger activity alone, as discussed below.

This paper aimed to provide an extensive review of the possible uses of SOD in different human diseases and explore the current pitfalls in development processes to solve the bioavailability issues. Selection was based on orgotein indications and included neurological, cardiovascular, respiratory, gastrointestinal, renal, skin, metabolic and ocular diseases. We are aware that cancer is a meaningful field of application for SOD. However, we stress that oncology is far beyond our expertise and has been extensively reviewed in I. Batinic-Haberle and coll. (2018) [25], I. Batinic-Haberle and I. Spasojevic [26], and I. Batinic-Haberle and M. E. Tome [27]. We therefore carried out a PubMed query starting with the keywords “SOD”, “SOD mimetics”, and “SOD supplementation” that included papers published in the English language, between 2012 and 2020, on the potential therapeutic applications of SOD, including detoxification strategies.

2. Mechanism of SOD Induction and Inactivation

The three isoforms of SOD show differences in their protein structures, metal cofactor requirements, subcellular localization (Figure 1), and tissue distribution. Human SOD1 is an homodimer of 88 kDa that is encoded by a gene on chromosome 21q22 [28]. SOD2 is a smaller homotetramer protein of 32 kDa, encoded by a gene on chromosome 6q25.3 [29]. Finally, SOD3 is an homotetramer glycoprotein of 135 kDa encoded by a gene on chromosome 4 [30].

Some unique transcription factors that play specific regulatory roles have been described [31]. However, all three SOD isoforms share the presence of binding sites for several transcription factors, such as the Nuclear Factor (NF)-κB, the specificity protein (Sp)-1, CCAAT-Enhancer-Binding Proteins (C/EBP), and the activator proteins (AP)-1 and-2, which exert similar effects on the regulation of all three SOD genes [31,32,33]. A prominent role has been recognized for nuclear factor erythroid 2-related factor 2 (Nrf2). The first evidence of the relationship between SOD1 and Nrf2 dates back to 2005, when the presence of the SODG93A mutation was associated with a reduction in Nrf2 mRNA [34]. Nrf2 translocates to the nucleus from the cytoplasm following binding with the Kelch-like ECH-associated protein 1 (Keap1). Keap1 is a cysteine-rich protein that interacts with ROS and promotes both the nuclear translocation and the ubiquitination and degradation of Nrf2. In the nucleus, Nrf2 forms a complex with Maf (musculoaponeurotic fibrosarcoma) proteins. It binds the antioxidant responsive elements (AREs) [35] at the sequence located in the promoter region between −356 and −330 from the transcription start site of sod1 [36].

The Keap1/Nrf2 pathway regulates the expression of many antioxidant genes besides SODs, such as those encoding for CAT, GPx, NAD(P)H-quinone oxidoreductase 1, GSH-S-transferase, Prx, ferritin and heme oxygenase-1 (HO-1) [37]. Interestingly, the Keap1/Nrf2 pathway can be considered the effector of the SOD mimetic mechanism of action. Indeed, SOD mimeticsalter the cysteine oxidation/protein S-glutathionylation cycle. These compounds cause the oxidation of the thiols of the peptide cysteine of Keap1, thus inducing Nrf2 activation and leading to SOD overexpression [27].

The Keap1/Nrf2/HO-1 axis and its link to SOD expression have been well characterized, and are based on the complementary function of SOD and HO-1; the first produces H₂O₂ and the second catalyzes the rate-limiting step in the breakdown of heme to bilirubin [38], which is known to remove ROS, including ^•OH, singlet oxygen and ^•O₂ [39]. Accordingly, the subsequent induction of SOD2 and HO-1 has been identified as the mechanism by which the Nrf2-ARE inducer tert-butylhydroquinone protects mitochondria that are exposed to oxidative stress [40], and astrocytes that are damaged by lanthanum chloride [41]. Moreover, Nrf2/HO-1 has been demonstrated to confer protection from doxorubicin-induced mitochondrial damage by upregulating antioxidant genes, including SOD2 [42]. Similarly, cobalt protoporphyrin, a potent inducer of the HO-1 protein and activity, increased SOD3 expression in rat aorta, possibly via the activation of the mitogen-activated protein kinase (MAPK) pathway [43]. Nrf2 is a direct downstream target of MAPK, like ERK [44]. Accordingly, the Nrf2/ERK signaling pathway has been implicated in the upregulation of the gene expression of HO-1 and SOD1 by fucoidan, a sulfated polysaccharide found in edible brown algae [45]. However, in a study by M. Dell’Orco and coll. (2016), Nrf2 does not appear to be associated with SOD1 in human neuroblastoma SH-SY5Y cells that are exposed to H₂O₂ [46]. Considering the role of Keap1/Nrf2 in SOD expression, the Nrf2 activators, or Keap1 inhibitors [47], should be included between the SOD inducers. Among them, the peroxisome proliferator-activated receptor (PPAR)γ is particularly attractive. Indeed, it could regulate SOD expression both directly through its association with the PPAR responsive element of the SOD promoter region, and indirectly inducing the expression of Nrf2, HO-1, CAT, and GPx-3 [48]. In particular, between Nrf2 and PPARγ, a positive feedback loop reinforcing the antioxidant response is established: Nrf2 through the ARE region present on the PPARγ promoter may directly upregulate PPARγ expression and PPARγ may in turn regulate the Nrf2 interacting with a PPAR responsive element [49].

Another interesting axis in SOD transcriptional regulation can be found in the phosphoinositide 3-kinase (PI3K)/AKT/NF-κB/transcription factors of the forkhead box, class O (FOXO) axis, which has been reported to exert antioxidant effects by increasing SOD expression. Indeed, the PI3K/Akt pathway induces SOD1, SOD2 and SOD3 expression [50,51,52], as well as HO-1 [53,54]. The activation of the PI3K/AKT axis inversely regulates the distribution of NF-κB and FOXO transcription factors; FOXO factors are phosphorylated and displaced from the nucleus to the cytoplasm, while NF-κB translocates to the nucleus, activating antioxidant genes, including SODs [50]. Again, PPARγ can participate: it may increase FOXO activity through the activation of AKT and NF-κB transrepression [55]. Interestingly, the role of the NF-κB-SOD axis in homeostasis through the NF-κB p65 subunit translocation is well documented and has been implicated, for instance, in the endotoxin-induced stress [56]. However, a vicious loop can be identified between SOD and NF-κB: the IKKβ/NF-κB signaling pathway regulates SOD2 expression through p53, and p53 transcription is in turn dysregulated by SOD2, causing the upregulation of IKKβ. This loop may be detrimental to the progression of tumorigenesis. Indeed, SOD2 expression was positively associated with pathologic tumor stages and negatively correlated with overall survival in nasopharyngeal carcinoma [57] or lung adenocarcinoma [58].

In addition to transcriptional regulation, epigenetic and post-transcriptional regulation can also contribute. Epigenetic regulation is primarily associated with SOD expression and activity in cancer. The most documented epigenetic regulation involves the promoter methylation of the SOD2 gene [59]. It has recently been demonstrated that the deacetylation of histones at its promoter reduces sod3 expression in old lung fibroblasts. Accordingly, histone deacetylase inhibitors were able to preserve sod3 expression [60]. On the other hand, in THP-1, histone H3 and H4 acetylation regulates sod3 expression during differentiation, while DNA methylation is responsible for sod3 silencing in human peripheral blood mononuclear cells (PBMCs) [61]. Post-transcriptional regulation is responsible for the rapid modulation of SOD expression and includes: (i) phosphorylation; (ii) amino acid modification, such as lysine acylation (including sumoylation, ubiquitination and glycation); (iii) redox modifications, such as oxidation, glutathionylation and cysteinylation; (iv) s-acylation; and (v) nitration [62,63,64].

Apart from expression regulation, SOD activity also depends on the presence of the associated metals. These mechanisms have been extensively reviewed by Culotta et al. (2007), Fukai and Fukai (2011), and Hatori and Lutsenko (2016). Briefly, while SOD1 and SOD3 exist as apoenzymes that are activated post-transcriptionally by copper insertion (without new protein synthesis), metal insertion for SOD2 cannot occur post-translationally. Indeed, manganese insertion only occurs in newly synthesized SOD2, when the pre-sequence for mitochondrial targeting at the N-terminus is still present. Subsequently, manganese trafficking to SOD2 is driven by the Smf2p manganese transporter and Mtm1p, which are members of the mitochondrial carrier family of transporters. SOD2 is therefore imported into mitochondria and cleaved into the mature form. Conversely, SOD1 activation occurs post-transcriptionally via a 4-step process that involves the copper chaperone for SOD1 (CCS). CCS docks with and transfers copper to the disulfide-reduced SOD1. The disulfide is essential for both structural stabilization and functional activation, allowing the dimeric state to form [65,66]. Finally, SOD3 is loaded with copper via a copper chaperone antioxidant-1 (Atox1) pathway [67,68,69]. However, Atox1 is not sufficient, and the Menkes ATPase, ATP7A, is required to deliver the copper to SOD3 at the trans-Golgi network [66]. The activation of SOD leads to the conversion of ^•O₂^- to H₂O₂ and O₂, as described in the above section and depicted in Figure 1. However, SOD1 can also act as a transcription factor. Indeed, H₂O₂ induces SOD1 translocation to the nucleus following association with the Mec1/ATM effector Dun1/Cds1 kinase and phosphorylation. Once in the nucleus, SOD1 regulates the expression of various oxidative stress-responsive genes that are known to confer resistance to oxidative stress, DNA damage repair and replication stress relief [70]. Moreover, upon binding to DNA, SOD1 regulates the ROS-responsive expression of functional genes, including oncogenes and amyotrophic lateral sclerosis-linked genes [71]. Finally, SOD1 has also been reported to activate the muscarinic M1 receptor, thus inducing AKT and ERK phosphorylation in neuroblastoma SK-N-BE cells [72].

As SOD activity depends on the associated metals, it is reasonable to assume that any perturbation of the enzyme structure that causes their release is responsible for the inactivation of the enzyme. Accordingly, using a zebrafish model, it has been demonstrated that lead forms a complex with SOD1 via an electrostatic effect. Consequently, the metal enters the active channel of SOD, hindering substrate access. Therefore, copper and zinc are released from the SOD1 active site [73]. Moreover, it is well known that the reaction of peroxynitrite with the metal center of the enzyme is responsible for SOD inactivation. In particular, both SOD1 and SOD2 react directly with peroxynitrite; SOD1 is subjected to histidinyl radical formation [74], and SOD2 is subjected to tyrosine nitration [75].

3. The Role of SOD: What We Have Learned from Knock-Out (KO) Mice

SOD’s role in oxidative stress defense means that its role in other pathophysiological contexts is inferable. Accordingly, the use of SOD supplements or SOD mimetics in several potential therapeutic applications is currently under investigation. Each of these possible therapeutic indications for SOD is mainly based on the use of transgenic mice. Indeed, mice that lack either SOD1, SOD2 or SOD3 have helped us to understand the relative role of each isoform in fertility, mortality/survival and the development of specific diseases. The very first difference between SOD1, SOD2 and SOD3 is in terms of survival. Homozygous mice that lack SOD2 (SOD2^−/−), and not SOD1 or SOD3, show a dramatic phenotype that affects lifespan, with death occurring: (i) within the first 10 days with dilated cardiomyopathy, the accumulation of lipids in the liver and skeletal muscle, and metabolic acidosis [76]; or (ii) within the first 3 weeks with severe anemia, the degeneration of neurons in the basal ganglia and brainstem, and progressive motor disturbances, characterized by weakness, rapid fatigue and circling behavior [77]. Accordingly, the homozygous missense variant, c.542G > T, p.(Gly181Val), in SOD2 may lead to toxic increases in the levels of damaging oxygen radicals in the neonatal heart, which can result in rapidly developing heart failure and death [78]. As SOD2^−/− die in 2~3 weeks [76,77], heterozygous SOD2 (SOD2^+/−) mice and alternatively, conditional KO mice, in which deletion involves individual tissues, have been generated [79]. Thanks to these experimental models, it is clear that the contribution of SOD to homeostasis is tissue-specific: heart/muscle-specific SOD2 KO shows a reduced lifespan, with several electrophysiological abnormalities occurring [80]; T cell-specific SOD2 KO demonstrates a compensatory phenotype, in which other mechanisms may compensate for any loss of function; while liver-specific SOD2 KO does not show a phenotype, with the tissue appearing unaffected by SOD2 loss [79]. Platelet content and function were not affected by SOD2^+/− phenotype, with no difference being observed between KO and wild-type mice in the tail-bleeding or arterial-thrombosis indices. Similar results have also been obtained when comparing these two phenotypes for outcomes in both sepsis and autoimmune inflammatory arthritis models [81].

Interestingly, postnatal motor neuron SOD2 KO shows no signs of oxidative damage up to 1 year after birth. These data suggest that postnatal motor neurons are resistant to oxidative-stress damage, although the disorganization of the distal nerve axon occurs [82]. Mammary-gland development is also not affected by SOD deletion; postnatal mammary gland SOD2 KO mice show no changes in pre- and post-pregnancy developmental structures and mammary-gland function [83].

In SOD2^+/− animals, enzymatic activity is decreased by 30–80% depending on the specific tissue [84]. This defect has been correlated with an increase in oxidative damage to mitochondria, but not to cytosolic proteins or nuclear DNA [85]. At 6 months, SOD2^+/− mice show behavioral impairments involving learning and memory processes, and alterations in glutamatergic synaptic transmission with a decrease in the n-methyl-D-aspartate (NMDA) receptor [86]. A clear phenotype has also been recognized in SOD1 KO mice. In this case, homozygous KO females have reduced fertility due to an increase in embryonic lethality, although normal ovulation and conception were observed [87]. These mice are healthy, although they have reduced survival time (mean lifespans of 20.8 ± 0.7 compared to 29.8 ± 2.1 months for the wild-type counterpart), with a higher incidence (79% of KO animals) of hepatocellular carcinoma development [88]. SOD1 has long been linked to age-associated diseases because SOD1 deletion leads to different phenotypes that mimic accelerated aging [89]. For instance, SOD1^−/− senescent mice show the decreased production of both stimulated and non-stimulated tears due to several alterations in the lacrimal gland, including: the atrophy of acinar units; fibrosis; infiltration of T-cells, monocytes and neutrophils; increases in apoptotic cells; and signs of epithelial-mesenchymal transition [90]. At 1 year of age, SOD^−/− mice develop cortical lens opacity, and within 1 more year, they showed reduced GSH content at the lens level [91]. Accordingly, a study of 415 cataract patients has demonstrated an increased risk of cataracts in patients that are polymorphic for SOD1 due to a reduced capacity to scavenge superoxide radicals in lenses [92]. Moreover, serum SOD activity has been observed to be significantly reduced in 60 patients with newly diagnosed senile non-pathologic cataracts [93]. In contrast, the SOD2^+/− phenotype was not related to age-related cataract development [94], suggesting that SOD1 may have a more detrimental effect on ageing. SOD1 deletion is also associated with cochlear degeneration over time; null mice developed early age-related hearing loss with spiral ganglion cell degeneration at 7–9 months of age [95]. Notably, SOD2 has also been found to be involved in hearing loss. Indeed, SOD2^+/− mice have shown significant outer hair cell damage in cochlear turns, and their response to post-noise exposure (120 dB at 4 Hz for 4 h) at 7 and 14 days was worse than that of their wild-type counterparts [96].

Notably, SOD1 KO mice display other features of aging apart from age-related hearing loss, and these include frailty, which is a clinical syndrome highly prevalent in old age that presents at least three of the following criteria: unintentional weight loss; exhaustion; weakness: slow walking speed; and low physical activity [97]. SOD1^−/− mice exhibit weight loss, weakness, low physical activity and exhaustion, while inflammation and sarcopenia develop in parallel [98]. Again, a similar effect is evoked by SOD2 deletion, with SOD2^+/- mice showing a reduction in work-to-exhaustion that is correlated with whole-body oxygen consumption [99]. A loss of muscle mass and function is one of the most prominent aging phenotypes shown by SOD1^−/− mice [100]. The importance of SOD1 in motor neuron degeneration is also confirmed by the demonstrated association between SOD1 defects in skeletal muscle and amyotrophic lateral sclerosis (ALS). SOD1 mutation, leading to reduced enzyme activity, is one of the key pathological events in ALS [101], and mice that express the SOD^G93A mutation are the most commonly used model for this disease [102]. Other mutations of SOD1 have also been recognized in ALS, although their significance in development and penetrance differs. For instance, the SOD1 G93D mutation caused a slowly developing lower motor neuron disease with reduced penetrance [103]. On the other hand, the mutation c.271G > A, which leads to the substitution of asparagine with aspartate at position 90, seems to be associated with the rapid progression and a prominent pain syndrome [104]. Moreover, A. Canosa and coll. (2018) have reported the presence of a heterozygous novel frameshift SOD1 mutation (p.Ser108 LeufsTer15), which was predicted to cause premature protein truncation in a sporadic ALS patient. This mutation could have two different consequences: (i) less active SOD1; and (ii) a less charged protein with a higher propensity to aggregate. In both cases, the result would be an increase in oxidative damage [105].

Finally, SOD1^−/− mice are more susceptible to paraquat toxicity [87], and motor neuron loss after axonal injury [106].

By contrast, mice that lack SOD3 have normal development and remain healthy until at least 14 months of age without the compensatory induction of other SOD isoenzymes [107]. However, their survival time was significantly affected by exposure to >99% oxygen as severe lung edema developed [107]. These data, combined with the results of gene-array screening in SOD3^−/− mice [108], suggest that compensatory mechanisms occur, including the unbalance of the expression of genes involved in cell signaling, inflammation and gene transcription (37 are upregulated and nine downregulated) [108]. Like SOD1, SOD3 has also been implicated in some age-related dysfunctions. For instance, both SOD-3 and SOD1 appear to have functions in preserving corneal endothelial integrity in aging [109]. Indeed, SOD3^−/− mice have shown the early (starting from month 2) spontaneous age-related loss of endothelial cells in the cornea and increased susceptibility to acute inflammatory endothelial damage [110]. By comparison, the corneal endothelial cells in SOD1/3^−/− mice have shown more irregular morphology at an older age, suggesting they have a more vulnerable corneal endothelium [109].

SOD3^−/− mice of 22 months have displayed reduced transforming growth factor beta (TGF-β) levels and, consequently, a lower differentiation of fibroblasts into myofibroblasts, which results in delayed wound closure, reduced neovascularization and increased neutrophil recruitment. These results suggest that reduced levels of cutaneous SOD3 in aged mice may contribute to the impaired wound healing response in aged skin [111]. By contrast, only a slight increase in inflammatory variables and fibrosis were found in lungs from 2-year-old SOD3^−/− mice, compared to their wild-type counterparts [112]. However, the response of SOD3^−/− mice to ovalbumin (OVA) challenge resulted in severe allergic asthma [113]. Interestingly, SOD3^−/− mice seem to be more prone to developing injury at the inner retina and may be more susceptible to vitreoretinal diseases, including diabetic vitreoretinopathy. Indeed, SOD3^−/− mice present higher oxidative stress markers at the vitreoretinal interface and signaling abnormalities within the inner retina [114]. SOD3^−/− mice have recently been used to study the contribution of oxidative stress to proteinuric kidney diseases. A study by R.J. Tan and coll. (2015) has demonstrated that SOD3^−/− mice are more susceptible to renal injury in an Adriamycin-(ADR)-induced nephropathy model [115].

4. SOD as a Detoxification Strategy

Oxidative stress is the most common mechanism of xenobiotic toxicity. For instance, heavy metals, such as mercury, arsenic and lead, induce oxidative stress by promoting the production of ROS and reactive nitrogen species (RNS). These metals may replace the transition metals, such as Zn and Cu, which are required for SOD catalytic function, and inhibit their function [13]. Various chemicals can affect the balance between pro-oxidant challenge and antioxidant defenses by enhancing ROS and/or RNS formation and by depressing their removal [116].

Due to its role in limiting the formation of ROS and RNS and the consequent oxidative-stress damage, the availability of SOD as an antidote for xenobiotic toxicity would be a therapeutic advantage.

As SOD2^+/− mice have been used as an experimental model to investigate of the role of mitochondrial toxicity in troglitazone-induced liver injury [117], SOD2 has been postulated to be a key enzyme against the hepatotoxicity of some drugs and chemicals [118]. For instance, SOD2 is inactivated by protein nitration during paracetamol hepatotoxicity [119]. Furthermore, partial SOD2 deficiency and inactivation have been associated with increased liver injury [120,121,122]. It has therefore been hypothesized that increasing SOD2 expression/activity might have a beneficial effect. This strategy has been pursued using nitroxide mito-tempo, which is a compound that combines piperidine nitroxide (tempo or tempol) with triphenylphosphonium (TPP⁺), which is a membrane-permeant cation that accumulates within mitochondria thanks to membrane potential [123], tempol [124], and the Mn pyridoxyl ethyldiamine derivative (MnPLED) mangafodipir (MnDPDP) [125]. Mito-tempo and tempol are both nitroxides and their classification as SOD mimetics is controversial [24,126].

The promising results obtained in C57 BL/6 J mice with paracetamol overload (300 mg/kg i.p.) [127], and in BALB/c mice with paracetamol (1000 mg/kg i.p. or 500 mg/kg p.o.)-induced acute liver failure [125], have led to a successful evaluation of the safety and tolerability of another MnPLED SOD-mimetic, calmangafodipir [Ca₄Mn(DPDP)₅], in combination with n-acetylcysteine (the gold standard antidote for paracetamol toxicity) for paracetamol overdose in humans [128]. Thus far, calmangafodipir has been reported among the established and emerging therapies against paracetamol hepatotoxicity in a recent review [129].

Due to its beneficial effects on hepatotoxicity, SOD2 has also been proposed as an antidote against carbon tetrachloride (CCl₄) intoxication. The CCl₄ metabolic process in the liver gives rise to two active microsomal radicals or peroxides (CCl₃ or CCl₃OO) [130,131], via the cytochrome P450 pathway, thus causing lipid peroxidation and undermining the integrity of liver-cell membranes [132]. The administration of an SOD2 mimic (SOD2m) for 7 days has prevented the oxidative stress and inflammatory responses induced in the liver, by the exposure of mice to 0.05% CCl₄, within 24 h. Indeed, a SOD2m-treated group showed a significant decrease in two crucial liver-injury biomarkers: aspartate aminotransferase (AST); and alanine aminotransferase (ALT). Accordingly, a reduction in histologically evaluated liver damage was observed. Moreover, the levels of several pro-inflammatory mediators, including prostaglandin E₂ (PGE₂), cyclooxygenase-2 (COX-2), interleukin (IL)-6 and tumor necrosis factor-α (TNF-α), were reduced [133].

The correlation between SOD and alcohol intoxication is now well established. Homozygous mutations in the SOD2 gene have been associated with a major risk of developing severe alcoholic liver disease in humans [134]. Interestingly, a study on a Han-Chinese population (80 patients with alcoholic cirrhosis, 80 patients with alcoholic non-cirrhosis, 80 with viral hepatitis B-related cirrhosis and 165 healthy controls) has demonstrated that patients with alcoholic cirrhosis had a higher frequency of the SOD2 C/C and C/T genotypes than the other groups, suggesting that the SOD2 47T > C genetic variant is a risk factor for alcoholic cirrhosis susceptibility [135]. On the other hand, moderate ethanol consumption (7–9 g/kg body wt/day) in SOD1^−/− mice promotes the onset and progression of alcoholic liver injury via a decrease in SOD2 and an increase in peroxynitrite contents, protein carbonyls and lipid peroxidation [136]. Accordingly, the adenovirus-mediated expression of SOD1 has been observed to be effective in reducing early alcohol-induced liver injury in rats [137]. More recently, SOD1 encapsulated in poly-L-lysine (PLL50)-polyethylene glycol (PEG) and then cross-linked with a reducible cross-linker (nano-SOD) reduced the steatohepatitis induced by ethanol in mice that were fed an ethanol liquid diet (5% of ethanol) for 4 weeks [138].

Several studies have associated a downregulation in SOD activity, and the consequent oxidative stress, with the progression of chronic skin damage induced by UV-irradiation [139]. SOD1 has been shown to exert a protective effect on human keratinocytes exposed to UVB [140]. Transfecting human keratinocytes with the SOD1 expression vector was effective in reducing UVB-induced apoptosis [141]. Moreover, a study on B16F10 murine melanoma cells has demonstrated that SOD1 (1–1000 ng/mL) inhibits melanin production within 24 h in a dose-dependent manner [142]. Accordingly, the topical administration of 1000 ng/mL SOD1 to HRM-2 melanin-possessing hairless mice before UVB 190 mJ/cm² exposure decreased UVB-induced melanogenesis by blocking the aggravation of melanogenesis and thus potentially preventing melanoma development [142]. This evidence indicates the possible use of the exogenous supplementation or endogenous up-regulation of SOD to counteract UV-radiation-induced oxidative stress. An in vitro study demonstrated that the SOD mimetic belonging to the ethylenediamine chloride complex (EUK) family, EUK-134, increases human keratinocyte survival, after UVB-induced oxidative stress, via the indirect inhibition of the MAPK pathways [143]. Accordingly, the 30 U SOD/mL of the dried melon juice concentrate SODB, administered 24 h before UV exposure, has been seen to reduce keratinocyte apoptosis [139]. Moreover, the topical application of SOD, linked with the human immunodeficiency virus type 1 (HIV) transactivator of transcription (TAT) domain (TAT-SOD) at 300 U/cm², 1 h before UVB irradiation, was effective in preventing UVB-induced erythema formation and blood-flow rise in Fitzpatrick skin type II and III subjects [144].

Similarly, it has been suggested that SOD2 is important in preventing the damage caused by UV radiation-induced oxidative stress, which can lead to numerous ocular pathologies [145]. Interestingly, an ophthalmic carbopol 934-based gel formulation, containing recombinant SOD2 (rMnSOD) as an active ingredient, reduced the number of microvilli damaged both in conjunctiva and cornea epithelial cells from rabbit eyes exposed to UV radiation [146]. The protective role of SOD in ocular damage may also have therapeutic implications in methanol intoxication. Visual symptoms usually occur within 12–36 h after ingestion and can be ascribed to the inhibition of cytochrome oxidase activity and the prevention of mitochondrial oxygen production in the optic nerve by formic acid, a toxic methanol metabolite [147]. Indeed, HCO₂ can easily pass through the ganglion cell wall due to methanol-induced acidosis, leading to formate-oxidation reactions in the mitochondria and lysosome [148]. The optic nerve, retina and basal ganglia are the main tissues that are damaged by the increased oxidative-stress response [149]. The administration of tempol 2 h after methanol ingestion prevented the structural integrity of retinal ganglion cells in methanol-intoxicated rats [148]. Therefore, it is possible to hypothesize that SOD can be used as an antioxidant therapy for methanol-induced toxic optic neuropathy.

The ionizing radiation used in radiotherapy is known to trigger both ROS generation and the cytotoxic response, resulting in several different side effects, including fibrosis. When a deficiency in antioxidant enzymes is present, an increase in radio-sensitivity occurs [150]. The first observation of the beneficial effects of antioxidant therapy in preventing these events arrived in 1983, when a liposomal formulation of SOD was administered to two patients treated with high-dose pelvic radiotherapy, to reduce the fibrotic and necrosis response that occurred [151]. Since then, several publications have supported the role of SOD supplementation in radioprotection. The precise mechanisms responsible for the radioprotective effects of SOD are still unknown. Of the different possible forms of SOD, SOD2 is currently considered to be pivotal in protecting cells during exposure to ionizing radiation. Its importance has led to an investigation into the possible use of SOD activity in blood cells as a predictive biomarker for the selection of individualized irradiation therapy protocols. In an in vitro study of blood samples obtained from 32 breast-cancer patients, the activity of SOD after irradiation depended on initial SOD levels; these were decreased when initially high, and preserved when initially medium or low [152]. According to the authors, it is possible to consider patients with high basal levels of SOD to be poor responders, whereas patients with low basal levels may benefit from defense against the reactive free radicals produced after radiation. On the other hand, proton irradiation reduced SOD2 activity, while X-rays induced its overactivity [153]. This observation may be related to the bell-shaped dose-response curve observed following SOD administration. According to this, the optimization of concentration is essential in any application [9]. Therefore, SOD has been proposed as a strategy to prevent radiation-induced damage to different normal tissues. D. Leu and coll. (2017) have evaluated the effect of a lipophilic Mn porphyrin (MnP)-based SOD mimic, MnTnBuOE-2-PyP⁵⁺ (BMX-001), administered subcutaneously for one week before cranial irradiation and continued for one week afterward, in the radioprotection of hippocampal neurogenesis in a mouse model [154], and obtained promising results. Accordingly, MnTnHex-2-PyP⁵⁺, a similar SOD mimetic compound [155], delayed the onset of radiation-induced lung lesions, reduced respiratory-rate elevation and lessened the pathologic increases in lung weight in a model of radiation-induced lung injury in a non-human primate [156]. More recently, the MnP SOD mimetic AEOL 10150, also known as MnTDE-2-ImP⁵⁺, showed promising results in a whole thoracic lung irradiation model in nonhuman primates [157,158,159].

Moreover, the subcutaneous administration of bovine SOD (15 mg/kg) ameliorates radiation-induced lung injury in female rats by suppressing reactive oxygen species/reactive nitrogen species and ROS/RNS-dependent tissue damage [160]. Moreover, SOD3 administration has been tested in the treatment of radiation-induced pulmonary fibrosis. SOD3 has been recognized to be the main SOD form that is expressed in the lung, and is bound to the extracellular matrix [161]. The use of an association product that combines mesenchymal stromal cells (MSCs) with SOD3 was recognized as a promising strategy to counteract fibrotic processes: MSCs have already been reported to be effective in the early stages [162], but detrimental in the late stages [163] of pulmonary fibrosis, while SOD3 overexpression in the lung was recognized as being protective against the development of fibrosis [164]. The injection, 2 h post-irradiation, of SOD3-overexpressing MSC into mice that had been exposed to Cobalt-60 (20 Gy) was able to reduce collagen deposition, inhibit myofibroblast proliferation and reduce inflammatory cell infiltration,and consequently had an anti-fibrotic effect by preventing oxidative stress [165].

SOD had a generally beneficial effect on fibrotic response in a range of experimental settings. Melon-derived SOD has been administered in a gliadin oral formulation at 10,000 U/kg/day for 8 days to mice exposed to 25 Gy, 6 months before SOD treatment, and reduced the mean dermal thickness, which is predictive of radiation-induced fibrosis [166]. The same SOD formulation effectively reduced capsular fibrosis around silicone after implant surgery in an experimental model resembling breast-cancer treatment in rats [167]. However, the study failed to demonstrate that there was any beneficial effect in preventing or reducing radiation-induced fibrosis. These results are apparently in conflict with other previous studies that have had clear positive outcomes. However, the lower dose of SOD supplementation (500 mg/day for 3 weeks in the study [167] vs. 10,000 U/kg/day in the study [166]) and the use of different subcutaneously injected formulations [160], instead of oral administration, may account for these differences. The overall evidence for the use of SOD as a protective treatment in post-radiation fibrosis has led to at least two recently published clinical studies. However, the results obtained were not so comforting. The prospective study by K.C. Landeen and coll. (2018) [168] failed to demonstrate the effectiveness of topical SOD (280 U/g) at providing relief from the fibrosis of the head and neck area induced by radiation therapy in patients with a history of squamous cell carcinoma of the head and neck that had been treated with radiation. The study involved 68 adult patients, mostly males, and 86% had received radiation treatment at least 6 months before the initiation of the study. The improvement in the fibrosis score at 3 months was comparable in the SOD and placebo groups, suggesting that SOD had a marginal effect, compared to active physical therapy, in the post-treatment of neck fibrosis in patients with head and neck cancer [168]. Accordingly, the genetic association between SOD2 gene variations and radiation-induced soft-tissue toxicity has been reported in only one, monocentric, small-sample-size study [169]. On the other hand, a Phase 1b/2a study by C.M. Anderson and coll. (2018) [170] provided promising results regarding the effectiveness and safety of a cyclic polyamine SOD mimetic, avasopasem Mn or GC4419 (previously known as M40419, the enantiomer of M40403) at reducing the severe oral-mucositis that is induced by radiation-concurrent cisplatin in oral-cavity and oropharyngeal cancer. Patients (n = 46) with oral-cavity or oropharyngeal cancer, stages III–IVb, received fractionation intensity-modulated radiation therapy (once daily, Monday–Friday, at 2.0 to 2.2 Gy/d, to a cumulative tumor dose of between 60 and 72 Gy) with concurrent cisplatin (80–100 mg/m² every 3 weeks or 30–40 mg/m² weekly). GC4419 doses of 30 and 90 mg/day, administered throughout the chemoradiotherapy period, were the most effective and showed no particular safety concerns. These doses were therefore selected for the Phase 2b extension of the study [170].

5-fluorouracil is a chemotherapy agent known to cause severe mucositis and induce intestinal damage [171]. The administration of SOD was therefore also tested in a model of 5-fluorouracil-induced intestinal mucositis in mice. The study showed that Multi-modified Stable Anti-Oxidant Enzymes^® (MS-AOE^®), an rMnSOD obtained from a mutant high-temperature-resistant SOD strain, alleviates the mucositis caused by 5-fluorouracil, primarily in the first 3–5 days [172].

Interestingly, oral mucositis is not the only side effect of cisplatin therapy that can be treated with SOD. SODs have also been proposed for the treatment of cisplatin nephrotoxicity. Indeed, cisplatin nephrotoxicity has been associated with ROS production, DNA fragmentation and the activation of caspase enzymes, especially caspase-3 [173,174]. The administration of tempol prevented a decline in the kidney function of rats that developed nephrotoxicity following a single i.p. injection of cisplatin 6mg/kg [175]. Accordingly, rats treated with tempol showed an increase in kidney GSH content and SOD activity and a parallel decrease in kidney lipid peroxidation and NOx production [176].

Finally, a more recent example of SOD as a possible antidote has been proposed by Liu Z. and coll. (2020). The authors, using both an in vitro and an in vivo approach, demonstrated that bupivacaine induced the over-production of mitochondrial ROS, the activation of C-Jun n-terminal kinase (JNK), thus leading to SOD2 upregulation. On the other side, the SH-SY5Y cells transfected with SOD2 siRNA showed a higher susceptibility to bupivacaine, as demonstrated by the cell apoptosis increase. The SOD2 deletion induced mitochondrial ROS, malondialdehyde, and 8-hydroxydeoxyguanosine over-production, with a parallel decrease in the mitochondrial membrane potential. All these events were prevented by mito-tempo [177].

A summary of the proposed applications of SOD as a detoxification strategy, as discussed above, is provided in Table 1.

5. SOD as a Pharmacological Agent

The imbalance between oxidative-stress mediators and protective pathways, including SOD, has been recognized as a detrimental event in many pathophysiological disorders. This review highlights the most investigated applications of SOD as a therapeutic agent from 2012 to 2020, excluding the field of oncology (Table 2). Despite their differences in etiopathogenesis, oxidative stress has been recognized as a promoter of tissue damage. It can be argued that the generic antioxidant effects of SOD supplementation are beneficial in all of these conditions, from hypoxic damage and cardiovascular diseases to neurodegenerative disorders (Parkinson’s disease, Alzheimer’s disease, ALS), and metabolic diseases, including diabetes, its complications and obesity (Table 2).

However, it must be underlined that clinical evidence for this is limited, and consequently, real proof of efficacy is far from having been demonstrated. It is possible that the lack of clinical evidence of positive effects is, at least partially, due to the so-called “antioxidant paradox” [232], which is based on the cross-talk between oxidative stress and inflammation. These processes strictly influence each other and coexist in many pathological conditions. Therefore, a vicious circle is established: ROS and reactive nitrogen species (RNS) activate intracellular responses enhancing the expression of pro-inflammatory genes, and consequently, a number of pro-inflammatory mediators are released, and inflammatory cells are recruited. On the other hand, the inflammatory cells exaggerate the oxidative stress by producing ROS and RNS [232]. Several mediators participate in this vicious circle. Of these, a key role is played by the high-mobility group box protein 1 (HMGB1), a protein with a dual function: as a non-histone chromatin-binding protein involved in regulating transcription in the nucleus; and as a pro-inflammatory cytokine/chemokine when released into the extracellular space. Its relevance in oxidative stress-inflammation cross-talk is due to the extracellular form; ROS/RNS have been suggested to be both the cause and consequence of HMGB1 release [233]. Interestingly, a study on 86 patients with atrial fibrillation revealed a negative correlation between serum HMGB1 levels and SOD activity (r = −0.491, p < 0.05) [234]. Moreover, HMGB1 translocation and release are promoted by H₂O₂ in hepatocytes [235], primary human epidermal melanocytes [236], and neonatal rat cardiomyocytes [237]. Therefore, the increase in SOD activity and the parallel reduction in HMGB1 levels have been proposed as the mechanisms underlying the protective effects exerted by quercetin in a rat model of sepsis [238], the amelioration of the cisplatin-induced hepatotoxicity by the Ganoderma lucidum mushroom [239], and the anti-inflammatory effect of the midazolam–sufentanil combination [240]. Two cross-talking pathways are involved: Nrf2/HO-1 and the Toll-like receptor (TLR)/NF-κB axis [241]. Indeed, HMGB1 can suppress the Nrf2 pathway [236,242], as well as activating TLR-4, and thus activates NF-κB signaling [243,244]. Considering its crucial role in SOD induction, the Nrf2 pathway is an attractive target for different chronic diseases in which oxidative stress is involved [245,246]. Therefore, pharmacologic modulators of Nrf2 may exert significant antioxidant effects through indirect SOD targeting, such as by PPAR activation. Nrf2-driven PPARγ induction was demonstrated to be protective against the pulmonary oxidant injury [247]. The review by I. Dovinova and coll. (2020) highlights PPARγ as one effector of SOD1, SOD2, and SOD3 expression in spontaneously hypertensive rats [248] and how this event contributes to pioglitazone’s therapeutic effects, including the control of blood pressure [249]. Moreover, S. Agarwal (2017) reviewed PPARs as promising therapeutic targets for several neurodegenerative disorders such as Parkinson’s, Alzheimer’s and Huntington’s disease, and ALS. In all these conditions, the role of oxidative stress has been recognized. Therefore, PPARs may have a beneficial effect even modulating SOD2 expression [250].

6. SOD Sources

Different SOD-based compounds have been tested; from plant and animal extracts and SOD recombinant forms to SOD mimetics and SOD gene therapy (Table 3).

This heterogeneity stems from the need for an exogenous SOD with optimal pharmacokinetics properties. Exogenous SOD has relatively low bioavailability, especially when orally administered. Indeed, due to its enzymatic nature, exogenous SOD is digested and denatured in the stomach. Moreover, it should be noted that exogenous SOD has a high molecular weight, meaning that cellular uptake is limited, even when it is injected [21]. These aspects explain why SOD use is restricted to drug applications in animals, and to non-drug applications in humans (including supplements, cosmetics, food, agriculture and chemical industries) [303]. Although exogenous SOD administration has often proven problematic, a variety of innovative approaches are currently being explored [12]. SODB has been considered the gold standard for the dietary supplementation of SOD since 2000. However, its efficacy is affected by the low pH and high proteolytic activity in the digestive tract [16]. Research on designing formulations with SOD encapsulated in lipids and/or proteins has been performed to overcome the low bioavailability of natural SOD. Thus far, gliadin-SOD, nano-SOD and O-HTCC-SOD (Table 3) have been created. These products should protect the enzyme from degradation, but do not entirely solve the absorption problem caused by SOD’s high molecular weight [16]. Indeed, if the formulation scaffold cannot activate tight-junction promoting absorption, intestinal permeability is still limited. Therefore, another strategy has been pursued since the late 1970s; the development of synthetic antioxidant enzymes, SOD mimetics, that were developed to overcome the bioavailability problem of SOD supplementation. SOD mimetics are characterized by low-molecular weight (about 483 Da) and better intestinal permeability when administered orally, but this also grants a higher circulating half-life and lower antigenicity [24]. Approaches for the future of this field seem to include gene therapy to produce more antioxidants in the body, for instance, by creating stem cells that overexpress the SOD enzyme via genetic modifications. The development of SOD3-overexpressing MSCs is being investigated in this field. The aim here is to overcome the limits of MSC therapy, such as circumscribed survival and reduced immunomodulatory potential, using the benefits of SOD3 antioxidant and immunomodulatory activity [304]. Future studies will provide more in-depth knowledge of the feasibility of this strategy. In addition to these pharmaceutical approaches, several sources of exogenous SOD have been pursued. SOD was formerly obtained from the liver and serum of mammals such as pigs, horses, bulls and dogs [303]. Of these, bovine-derived SOD, known as orgotein, has been licensed as a veterinary product for use as a non-steroidal anti-inflammatory drug (ATC code M01AX14). Nowadays, if not of human origin, (recombinant human SOD), SOD is mostly derived from terrestrial and marine plants, microbial, cyanobacterial and chromista sources (Table 4). However, marine and terrestrial fungi, as well as yeasts, are also important sources of SOD.

The use of SODs from various sources reflects the need to emphasize different properties of different forms of the enzyme, with the different sources mainly used for specific applications; plant-derived SOD is mostly used for supplements and nutraceuticals, while SOD from marine source is used in cosmetics [305]. Plants have three isoforms of SOD: chloroplastic and cytosolic Cu, Zn-SOD; mitochondria Mn-SOD; and chloroplastic and plastidial Fe-SOD [21,306]. Cyanobacteria and marine creatures contain the Ni-SOD isoform. Cu, Zn-SOD, Ni-SOD and Fe-SOD are very sensitive to H₂O₂, but Cu, Zn-SOD and Ni-SOD are also sensitive to cyanide. Mn-SOD is insensitive to both H₂O₂ and cyanide [306]. Fe-SOD has also been found in prokaryotes, including marine bacteria, such as Photobacterium leiognathi and Photobacterium sepia, as well as in protozoans and the chloroplasts of algae such as Lingulodinium polyedrum [303]. The use of these SODs for large scale commercialization has often been limited by the minimal SOD content and the high cost of extraction methods. Therefore, the most used source of SOD has been Cucumis melo L.C., in which SOD is extracted from dried melon pulp with relatively high efficiency; 1 kg of freeze-dried concentrated melon juice [307], containing 90,000 U/g of SOD, is obtained from 15 kg of melon pulp, after filtration and concentration steps [20]. By comparison, other terrestrial plant sources of SOD have a low abundance of the enzyme. For instance, it is possible to extract 5–44 U/min/g fresh weight of SOD from sugarcane leaves [21]. However, several strategies have been developed to enhance SOD activity, more than 100-fold, and make these alternative sources effective drugs. For instance, Z. Hou and coll. (2019) have significantly improved the extraction of SOD from sea buckthorn and chestnut rose by adding purification steps, such as ammonium sulfate precipitation and anion exchange chromatography [308]. Furthermore, changing the salinity or adding heavy metals to the ground has been seen to provoke water deficiency, as does reducing or increasing the temperature, inducing an oxidative-stress response that led to SOD-content increases [21]. Moreover, culinary herbs such as Rosmarinus officinalis, Thymus officinalis and Salvia officinalis possess SOD mimetic activity that can even increase when cooked, or cooked and digested [309].

According to the literature sources reported herein, SOD mimetics can be divided into different classes according to their structure: cyclic polyamines; MnPLEDs; MnP; salen–Mn complexes; three metal-based compounds; and nitroxides (Table 5).

These structural differences can result in differing pharmacokinetic properties, including the route of administration and subsequent bioavailability. While the pharmacokinetics of MnPs have been widely investigated, a similar in-depth pharmacokinetic analysis is not available for other SOD mimetics, as reported by I. Batinic-Haberle and coll. (2018) [25]. Therefore, a proper pharmacokinetic comparison of the different SOD-based strategies is not possible.

Useful tools for classification can be found in the review by R. Bonetta (2018) [24]. A more extensive analysis of MnP compounds has been reported by I. Batinic-Haberle and coll. (2018) [25], I. Batinic-Haberle and I. Spasojevic [26] and I. Batinic-Haberle and M.E. Tome [27]. Briefly, the cyclic polyamine class differs in the metal, Fe²⁺, Mn²⁺ and even Cu²⁺, as well as in the polyamine moiety. However, they all have a dose-proportional response curve [310,311], instead of a bell-shaped dose-response curve, which is characteristic of the natural SOD enzyme [312]. Within this class, GC4419 has been extensively described. Developed by Galera Therapeutics, Inc. as a 1,4,7,10,13-pentazazcyclopentadecane derivative, it has already been tested in humans for the treatment of oral-mucositis as induced by radiation-concurrent cisplatin treatment [170]. MnPLEDs have several antioxidant properties, including the inhibition of SOD, GPx and CAT activity, as well as iron-binding, and consequently, the Fenton reaction. Accordingly, these compounds inhibit the formation of both ONOO⁻ and ^•OH and increase H₂O₂ detoxification [313]. MnP-based SOD mimetics combine the effect evoked by the Mn moiety on ^•O₂⁻ dismutation, via the reduction of Mn³⁺ to Mn²⁺ and its oxidization back to Mn³⁺, with the CAT activity that is attributed to the porphyrin radical cation’s ability to undergo oxidation to higher oxidation states, Mn⁴⁺ or Mn⁵⁺ [314]. Moreover, MnPs can produce H₂O₂ by first undergoing rapid one-electron reduction with endogenous or exogenous ascorbate or thiols, and then being re-oxidized by O₂ or ^•O₂⁻ [25]. MnPs offer several favorable features, including the absence of antigenicity, high stability that assures the integrity of the metal site, and low molecular weight [315]. Moreover, various delivery systems can reduce side effects such as acute hypotensive response observed with MnTnBuOE-2-PyP⁵⁺. Under this task, S.L. Schlichte (2020) developed a mesoporous silica nanoparticle and lipid bilayer nanoformulation of MnTnBuOE-2-PyP⁵⁺. The nanoformulation allows a slow and sustained release of the drug, thus reducing the acute reduction in renal sympathetic nerve activity induced by the injection of the free drug [316].

From a mechanistic point of view, this class is far beyond just being a radical scavenger as they add the reaction with H₂O₂, ^•O₂⁻ and ONOO⁻ to that with thiols. This last property is responsible for activating the Keap1/Nrf2 pathway, which is responsible for transcriptional activity, and SOD upregulation [25,27]. Salen-Mn complexes are thought to have SOD/CAT biomimetic activity. A multi-step process describes their mechanism of action: (i) interaction with ^•O₂⁻ reduces Mn³⁺ to Mn²⁺; (ii) Mn²⁺ is oxidized back to Mn³⁺ by ^•O₂⁻ consumption; (iii) salen-manganese is oxidized to salen-oxomanganese by H₂O₂; (iv) salen-oxomanganese is then reduced to salen-manganese by H₂O₂, liberating H₂O and O₂. Moreover, they have even been reported to scavenge RNS. The EUK compounds belong to this class. In particular, EUK-134 is a first-generation compound (it has a non-cyclized structure), while EUK-207 is a second-generation compound with greater stability due to this cyclized structure [24,317]. More recently, another class of SOD/CAT mimetics has been added; the metal-based compounds. This class has a conserved core, 1-[bis(pyridin-2-ylmethyl) amino]-3-chloropropan-2-ol (HPClNOL), that can be complexed with Fe³⁺, Mn²⁺ and even Cu²⁺. Collectively, this class possesses intrinsic ^•O₂^- and H₂O₂ scavenger activity. Compared to the salen–manganese complexes, the metal-based compounds have not been observed to affect the capacity of cells to synthesize neutral lipids and to compartmentalize them into lipid droplets. Cell-membrane integrity is thus maintained, hinting at possible higher efficacy against aging [286]. However, further studies that explore the real mechanism of action of these compounds must be performed to support this hypothesis. The classification of nitroxides as SOD mimetics is more controversial. Some authors, like S. Miriyala and coll. (2012), have highlighted the inability of nitroxides to catalytically scavenge superoxide [126]. Their mechanism includes reducing hydroxylamine within mitochondria [315], where these compounds display a weak and pH-dependent SOD-like activity [24]. According to this mechanism, some authors include nitroxide among SOD-mimetic compounds. However, this remains a controversial proposal [24,176,188,220,318]. The discussion includes the classification of tempol and mito-tempo as SOD mimetics. Tempol acts as a redox-cycling nitroxide water-soluble SOD mimetic [24,176,188,220,318], and shares the activation of the PI3K/Akt/Nrf2 pathway with other SOD mimetics [124,319,320]. Accordingly, the combination of tempol with the TPP+ moiety, resulting in mito-tempo, is accepted as a SOD mimetic [127,321]. According to the mechanistic interpretation of the SOD mimetic based on their ability to activate the PI3K/Akt/Nrf2 pathway, other inducers of this pathway can be included among the “source of SODs”. Therefore, the following could be added to the list: (i) the several NRF2 activators such as dimethyl fumarate, bardoxolone methyl, sulforaphane, curcumin, quercetin, and metformin; (ii) the PPARγ activators such as the antidiabetic drugs glitazones, ankaflavin, monascin, and carotenoids; or (iii) the dual Nrf2 and PPARγ activators, genistin, olmesartan, 18β-Glycyrrhetinic acid, and resveratrol, included [47,48,245].

7. Conclusions

The literature data that have been reported herein, covering papers published between 2012 and 2020 on the use of SODs for neurological, cardiovascular, respiratory, gastrointestinal, renal, skin, metabolic and ocular diseases, are indicative of the high efficacy of all the SOD types tested, both natural SOD and SOD mimetics. Although SOD has been an attractive potential therapeutic approach for 50 years, most of the published papers, and even more so in the case of recent works, deal with experimental preclinical studies, and only comparatively few clinical studies are ongoing. Notably, the spread of the pandemic COVID-19 infection, causing the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), further renewed the interest in pharmacological strategies to counteract the oxidative stress response triggered by NOS. Accordingly, J.O.C. Karlsson and coll. (2020) proposed mangafodipir to lower the inflammatory burden in critical SARS-CoV-2 infections [322]. In addition, just in September 2020, Galera Therapeutics, Inc. announced the first randomized, double-blind pilot phase II clinical trial with GC4419 for COVID-19 (ClinicalTrials.gov numbers, NCT04555096, accessed on 19 March 2021). However, none of the tested compounds have been approved to date. Several issues with the testing conditions and the type of compound evaluated have hampered the translation of the evidence for SOD use from the bench to the bedside. These topics can be summarized in three major points.

Firstly, the heterogeneity of the various compounds used to enhance the levels of SOD, from SOD extracts and SOD recombinant forms to SOD mimetics and SOD gene therapy, is an issue, as is the lack of comparative head-to-head studies. This point is strictly correlated with the second, which is the problem of bioavailability and the route of administration for effective doses in humans, and the timing of administration in relation to the dynamics of pathological process. Indeed, the optimal conditions for all the therapeutic approaches have not yet been clearly established. In the absence of comparative studies, even pharmacokinetics and toxicology data are not sufficient for a conclusive consensus on which sources of SOD, doses and administration timings best reflect clinical needs. This is also true for SOD mimetics, which are the most extensively studied type. MnPs are the only compounds for which pharmacokinetics have been clearly defined [25]. However, there is a lack of comparative studies against other sources of SOD and gold-standard comparators—even here.

The third issue is the heterogeneity of the diseases in which SOD strategies have been tested. Indeed, different compounds have been tested for similar applications, but have not been compared. Furthermore, although the same compound has been used under different pathological conditions, reported data cannot still define a specific indication for human use. The spectrum of diseases evaluated is vast, and detrimental contributions by ROS have been comprehensively demonstrated in each. However, oxidative stress can be considered a generic mechanism present in almost all pathological processes, and it is not unique to pathophysiological contexts. Therefore, its role as a drug target may vary according to the disease type and underlying biochemical processes. The way in which SOD affects oxidative stress may be regarded as a composite of direct (scavenger activity) and indirect (stimulating gene transcription of antioxidant pathways) antioxidant effects, as previously discussed.

The role of SOD merits a different type of discussion when considering ALS, in which a mutant overactive SOD1 has been identified, and Down syndrome, in which chromosome 21 trisomy has been associated with the overexpression of SOD1 in patients. These last two diseases remind us that SOD is a hormetic substance; added or over-expressed SOD produces potential beneficial effects in almost all of the conditions tested. However, in some circumstances, the benefits of SOD are either not so clear (i.e., gastrointestinal diseases) or even detrimental (i.e., ALS), in that they can exacerbate cell injury and death [284]. The interpretation of SOD as a hormetic substance draws our attention to another adjunctive issue when defining the therapeutic potential of SODs—the selection of the dose for optimal and tight regulation.

These issues collectively confirm the role of SOD as a supplement, but do not yet allow SOD to be conclusively repositioned as a drug that can be applied in the real world. Further evidence from the ongoing clinical trials is eagerly anticipated.

Acknowledgments

Author Contributions

Conceptualization, A.C.R. and F.D.; formal analysis, A.C.R.; literature search, D.C. and N.C.; writing—original draft preparation, A.C.R.; review and editing, N.B. and F.D. All the authors have approved the submitted version and agree to be personally accountable for their own contributions. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding as it was supported by MIUR-University of Torino “Fondi Ricerca Locale (ex-60%)”.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

References