Antioxidant and Anticancer Properties and Mechanisms of Inorganic Selenium, Oxo-Sulfur, and Oxo-Selenium Compounds

Abstract

Inorganic selenium and oxo-sulfur compounds are widely available in dietary supplements and have been extensively studied for their antioxidant and anticancer properties. Although many in vivo and clinical trials have been conducted using these compounds, their biochemical and chemical mechanisms of efficacy are the focus of much current research. This review discusses the ability of inorganic selenium compounds, such as selenite, and selenate, to prevent damage from reactive oxygen species as well as their ability to promote cell death by reactive oxygen species generation. Oxo-sulfur and selenium compounds, such as allicin, dimethyl sulfone, methionine sulfoxide, and methylselenenic acid also have similar abilities to act as both antioxidants and pro-oxidants, but the mechanisms for these behaviors are distinctly different from those of the inorganic selenium compounds. The antioxidant and pro-oxidant properties of these small-molecule sulfur and selenium compounds are extremely complex and often greatly depend on experimental conditions, which may explain contradictory literature reports of their efficacy.

Introduction

The generation of reactive oxygen species (ROS) has dual functionality in biological systems, with both beneficial and detrimental effects in cells [1, 2]. ROS generation at low or moderate concentrations aids in the defense against infectious agents and functions in several cell signaling pathways [1, 2]. The damaging effects of ROS such as the superoxide anion radical (O ^•−₂ ), hydrogen peroxide (H₂O₂), and hydroxyl radical (^•OH) caused by the overproduction of these species results in oxidative stress, an unavoidable consequence of aerobic cellular respiration [1–6]. ROS damage to lipids, proteins, and DNA [1, 2, 6, 7] is a result of this oxidative stress and leads to several health conditions including aging [1], cancer [1, 8], neurodegenerative diseases such as Parkinson’s and Alzheimer’s [9–12], and cardiovascular diseases such as arteriosclerosis [13–16].

Reactive oxygen species are generated during the reduction of molecular oxygen (O₂) to produce water (H₂O) via metabolic processes catalyzed by cytochrome oxidase in biological systems [5, 17]. The primary ROS formed as a byproduct of this respiratory process is the superoxide anion radical (O ^•−₂ ), generated when molecular oxygen gains an electron from either the mitochondrial electron transport chain (Reaction 1) or as a result of UV-irradiation (Reaction 2) [1, 4, 5, 17]. Further reduction of O ^•−₂ , either directly or through enzyme- or metal-catalyzed

$$ {\text{O}}_{ 2} + {\text{e}}^{ - } \to 2 {\text{O}}_{ 2}^{ \bullet - } $$

(1)

$$ {\text{O}}_{ 2} + hv + {\text{e}}^{ - } \to 2 {\text{O}}_{ 2}^{ \bullet - } $$

(2)

$$ 2 {\text{O}}_{ 2}^{ \bullet - } + 2 {\text{H}}^{ +} \to {\text{H}}_{ 2} {\text{O}}_{ 2} + {\text{O}}_{ 2} $$

(3)

$$ {\text{O}}_{ 2}^{ \bullet - } + {\text{ H}}_{ 2} {\text{O}}_{{ 2 { }}} + {\text{ H}}^{ + } \to ^{\bullet } {\text{OH}} + {\text{O}}_{ 2} + {\text{H}}_{ 2} {\text{O}} $$

(4)

$$ {\text{O}}_{ 2}^{ \bullet - } + \left[ { 2 {\text{Fe}}^{ 2+ } 2 {\text{Fe}}^{ 3+ } 4 {\text{S}}} \right] + 2 {\text{H}}^{ + } \to \left[ {{\text{Fe}}^{ 2+ } 3 {\text{Fe}}^{ 3+ } 4 {\text{S}}} \right] + {\text{H}}_{ 2} {\text{O}}_{ 2} $$

(5)

reactions, results in the formation of secondary ROS such as hydrogen peroxide (H₂O₂; Reaction 3) and the hydroxyl radical (^•OH; Reaction 4) [1, 4, 5, 17]. Hydrogen peroxide is also produced directly by protonation of the superoxide radical anion in solution (Reaction 3), and indirectly upon oxidation of iron-sulfur clusters (Reaction 5) [18, 19].

Compared to other ROS, hydrogen peroxide is a non-radical species with relatively low reactivity [20]. It is one of the more commonly studied ROS, and is produced endogenously by various physiological processes including respiratory burst and oxidative phosphorylation [21]. Calculations to determine the steady-state intracellular concentrations of hydrogen peroxide in unstressed E. coli cells determined a value of ~20 nM, with a high rate of H₂O₂ production ranging from 9 to 22 μM/s [22]. Thus, any imbalance between the rate of H₂O₂ generation and the decomposition may result in significantly increased H₂O₂ concentrations and resultant oxidative stress [22–24].

Imlay and Linn [24] reported bimodal cell killing when E. coli is exposed to H₂O₂. Mode I cell killing occurs at low concentrations of H₂O₂ (1–5 mM) and is faster than Mode II, which occurs at H₂O₂ concentrations greater than 10 mM. Mammalian cells also show the same bimodal killing rates as E. coli upon hydrogen peroxide challenge [25]. Bimodal kinetics are observed for iron-mediated oxidative DNA damage in vitro, where maximal damage occurs at 50 μM H₂O₂ for Mode I conditions, and at H₂O₂ concentrations >10 mM for Mode II conditions [24, 26, 27]. H₂O₂ reacts with redox-active metal ions to generate hydroxyl radical [5, 17, 18, 20, 28]. Hydroxyl radical has an extremely short half-life (~10⁻⁹ s) [29, 30] and reacts quickly with biomolecules near its site of generation, resulting in DNA damage, lipid peroxidation, thiol depletion, and changes in calcium homeostasis [1, 21, 31].

Iron and copper are the most commonly studied redox-active metal ions found in biological systems and are essential in many proteins and enzymes, including ferritin, transferrin, ceruloplasmin, and superoxide dismutase [32]. In E. coli, normal intracellular non-protein-bound (labile) iron concentrations are ~20 μM. However, this concentration increases significantly to 80–320 μM upon disruption of iron homeostasis and oxidative stress [20, 31, 33, 34]. Although the intracellular concentration of non-protein-bound copper was calculated to be approximately 10⁻¹⁸ M in yeast, significant amounts of labile copper are observed in mouse Golgi and mitochondria [35–37]. Studies have also reported extracellular copper concentrations in blood serum and cerebrospinal fluid between 10–25 μM and 0.5–2.5 μM, respectively, whereas copper concentrations in the synaptic cleft are approximately 30 μM [38, 39]. Neural copper concentrations are significantly higher in the locus coeruleus (stress and panic response center) and substantia nigra (dopamine production region) with concentrations of 1.3 and 0.4 mM, respectively [38, 40].

In the reduced state, Fe²⁺ and Cu⁺ are oxidized by H₂O₂ to Fe³⁺ and Cu²⁺, generating hydroxyl radical in the Fenton or Fenton-like reaction (Reaction 6) [1, 21, 28, 41–43].

$$ {\text{Fe}}^{ 2+ } {\text{ or Cu}}^{ + } + {\text{H}}_{ 2} {\text{O}}_{ 2} \to {\text{ Fe}}^{ 3+ } {\text{or Cu}}^{ 2+ } + ^{\bullet} {\text{OH}} + {\text{OH}}^{ - } $$

(6)

This production of ^•OH becomes catalytic in vivo due to the presence of cellular reductants, such as NADH, which reduce Fe³⁺ and Cu²⁺ back to their reduced forms. This iron-mediated generation of ^•OH is the main cause of oxidative DNA damage and cell death in prokaryotes and eukaryotes, including human cells, under oxidative stress conditions [24, 25, 43], and is a root cause of many health conditions such as cancer, aging, and cardiovascular and neurodegenerative diseases [9, 14, 43, 44].

Cellular defenses against the harmful effects of oxidative stress involve both enzymatic and nonenzymatic antioxidant activities [1, 4]. Enzymes such as glutathione peroxidases (GPx), catalases, and superoxide dismutases (SOD) act by directly scavenging ROS or by producing nonenzymatic antioxidants such as glutathione (GSH), thioredoxin, ubiquinone, and menaquinone [1, 4]. Nonenzymatic defenses involve small-molecule antioxidants such as carotenoids, lipoic acid, and vitamins C and E to protect against the damaging effects of oxidative stress [1, 45]. Both vitamins C and E reduce oxidative stress and malformations in the offspring of rats with diabetes [46–48]. Studies have also focused on the ability of various selenium, sulfur, and polyphenol compounds to act as antioxidants by preventing ROS-mediated DNA damage [23, 49–54].

Selenium has been extensively studied for its antioxidant and cancer preventative properties and is an essential trace element in human and animal metabolism [55–58]. It is found in many dietary supplements and multivitamins in forms such as selenite (Na₂SeO₃), selenate (Na₂SeO₄), or selenomethionine (SeMet) [56, 59]. Selenite and selenate are also found in fertilizers, animal feed, infant formulas, and protein shakes [56, 60]. Selenium is incorporated as selenocysteine (SeCys) in selenoproteins P, W, and R, as well as in the active sites of enzymes such as glutathione peroxidases (GPx) and thioredoxin reductases [23, 52, 57, 61–63]. In cells, these selenoproteins have important antioxidant activities and protect the mitochondria, plasma membrane, and DNA from oxidative damage by ROS [60, 64]. For example, GPx is found in the cytosol of cells and exerts its antioxidant activity by reducing intracellular hydrogen peroxide to water, preventing the generation of ROS [62, 63, 65–68]. Although selenoproteins are a significant part of the antioxidant properties of selenium, they have been extensively discussed [62, 69, 70] and are not the focus of this review.

Studies to determine the antioxidant activity of small-molecule selenium- and sulfur-containing compounds have focused mainly on organoselenium and organosulfur compounds since they are more bioavailable and are more readily incorporated into amino acids and proteins when compared to the inorganic forms [62, 71]. Consumption of food products high in selenomethionine (SeMet) results in incorporation of this amino acid into proteins by replacing its sulfur analog, methionine (Met) [72]. SeMet is also more efficiently absorbed and retained than the inorganic sodium selenite and selenate [73]. While the organoselenium compounds have received a significant amount of attention for their role as antioxidants [49, 74–76], several studies indicate that inorganic selenium compounds such as selenite, selenate, selenium dioxide (SeO₂), and sodium selenide (Na₂Se; Fig. 1) also exhibit similar antioxidative properties [23, 51, 55, 57, 58, 77–79]. Fruits, vegetables, and dietary supplements also contain oxo-sulfur compounds (Fig. 2) such as allicin, methylcysteine sulfoxide (MeCysSO), methyl methane thiosulfonate (MMTS), and dimethyl sulfone or methylsulfonyl methane (Me₂SO₂), which are also effective in preventing oxidative damage to cellular components [53, 54, 80–85]. Understanding the effects of oxidation on the antioxidant properties of organosulfur compounds is also important because amino acids such as methionine are susceptible to oxidation by ROS [86–88]. To prevent the disruption of protein function upon methionine oxidation to methionine sulfoxide (MetSO), cells have dedicated methionine reductase enzymes (Msr) to reduce MetSO back to Met [87, 89–92].

Fig. 1

Structures of inorganic and oxo-selenium compounds discussed in this review

Fig. 2

Structures of oxo-sulfur compounds discussed in this review

Research has focused primarily on the ability of organoselenium and organosulfur compounds in their reduced forms to prevent oxidative DNA damage and to treat or prevent diseases caused by oxidative stress. However, it is important to acknowledge the fact that inorganic selenium and oxo-sulfur compounds are abundant in many food products such as dietary supplements, protein shakes, infant formulas, fruits, and vegetables. This review will therefore discuss the role and biochemical mechanisms of inorganic selenium, oxo-selenium, and oxo-sulfur compounds to act as antioxidants and pro-oxidants, both in vivo and in vitro, for the treatment or prevention of ROS-mediated diseases.

Selenium Bioavailability, Related Pathologies, and Biological Effects

Selenium is an important micronutrient for both humans and animals and is obtained through the diet from several sources including cereals, grains, nuts, vegetables, meat, and seafood [62, 93, 94]. The recommended daily allowance (RDA) for selenium ranges from 55 μg/day to an upper limit of 350–400 μg/day, and daily intake comes from dietary supplementation and foods rich in this mineral [56, 95]. Although selenium toxicity has been observed for supplementation greater than 400 μg/day [60, 96], it is important to note that some studies conducted with a selenium intake ranging from 750 to 850 μg/day (~0.01 mg/kg) reported no signs of selenium toxicity in humans [97, 98]. Animal studies reported selenium toxicity within 12 h upon supplementation of 2 mg/kg selenium [99, 100]. These seemingly contradictory observations of selenium toxicity in humans highlight the need for additional studies to establish accurate upper level RDA values for selenium supplementation.

The selenium content of plant and animal products in the diet is important to maintain adequate selenium status and is highly dependent on regionally variable selenium concentrations in soil [48, 56, 62, 63, 66, 68, 98]. In the United States, for example, soil in northern Nebraska and the Dakotas has high selenium levels, but areas such as the Keshan province of China and some parts of Finland, New Zealand, Australia, and North America (northeast, northwest, Midwest, and southeast regions) have low soil selenium levels [66, 68, 98, 101–104]. To increase the selenium content in soils, and therefore increase animal and human consumption of selenium, these countries have implemented the use of fertilizers enriched with selenite or selenate for agricultural crops [105–108]. In the United States, selenite supplementation in animal feed has been shown to improve animal performance and increase the selenium dietary intake for Americans consuming meat products [60].

Pathologies Associated with Selenium Deficiency and Toxicity

Selenium deficiency occurs in regions where the selenium content in soil is low and can result in diseases such as hypothyroidism, weakened immune defenses, and cardiovascular diseases [63, 109, 110]. Keshan disease is a cardiomyopathy endemic to the Keshan province of China. This disease affects young children and women of child-bearing age as a result of low iodine and selenium contents in food products, leading to low blood plasma selenium levels [111–115]. The average intake of selenium for the development of symptoms due to deficiency was 10 μg/day with symptoms such as congestive heart failure, stroke, or sudden death (Fig. 3) [62, 115].

Fig. 3

Pathologies associated with selenium deficiency and toxicity

Also resulting from low selenium and iodine intake is the endemic osteoarthropathy known as Kashin–Beck disease found in several areas of China [113, 114]. Bone and joint deformations in growing children are characteristic of this disease (Fig. 3) [116]. The average serum selenium levels of patients with Kashin–Beck disease is significantly lower (11 ng/mL) than those without these mineral deficiencies (60–105 ng/ml) [117]. Serum thyroxine levels are also much lower in patients with Kashin–Beck disease, resulting in higher incidences of goiter than those unaffected by the disease [118]. In farm animals, selenium deficiency causes a muscular dystrophy known as white muscle disease (Fig. 3) [68, 119]. This disease usually affects growing animals such as lambs and calves between 1 and 3 months old with symptoms including stiffness, inability to move, weakness, tiredness, accelerated breathing, elevated temperatures, difficulty in feeding, and death [68, 119, 120].

Intake of selenium higher than the upper limit range 350–400 μg/day [56, 95] of the RDA is also of major concern to humans and animals since it can result in selenium toxicity or selenosis (Fig. 3) [60, 96]. Acute selenosis is caused by consumption of high levels of selenium in a short period of time. Upon ingestion of 17.2 μg/ml selenium due to incorrect dosage in animal feed, pigs showed signs of acute selenium toxicity, including paralysis, hyperesthesia, anorexia, and tremors [96, 121]. Signs of acute selenosis in buffalo include anorexia, alopecia, mild convulsions, and lowered body temperature [96, 122]. Symptoms of chronic selenosis include hair loss, deformation or cracks on the skin, horns, and hooves of animals, resulting in the sloughing of hooves and staggering [60, 96]. In humans, signs of selenosis include garlic breath, hair and nail loss, thickened and brittle nails, teeth deformation, skin lesions, and lowered hemoglobin levels upon dietary selenium intake of 5 mg/day (Fig. 3) [97].

Although countries have implemented the use of fertilizers containing selenite and selenate to supplement foods grown in selenium-deficient soil, the effects of selenium supplementation vary for each of these inorganic selenium compounds. In an attempt to prevent or reduce the prevalence of selenium deficiency diseases in China, both selenite and selenate were introduced into the soils of rice crops [115]. In unsupplemented soil, selenium content is extremely low, less than 0.06 μg/g. Rice crops grown in selenite- or selenate-enriched soils had significantly increased selenium levels of 0.47 and 0.64 μg/g, respectively, with no adverse effects on the plants [115]. In Chile, ryegrass supplemented with 0.1 μg/g selenite or selenate increases selenium content from 0.07 μg/g for grass grown in untreated soil to 0.28 and 5.72 μg/g, respectively [66]. Soil enrichment at higher levels of selenate (4–10 μg/g), resulted in stunted growth of ryegrass, with selenium content ranging from 150 to 247 μg/g [66]. Surprisingly, selenite-enriched soils (6–10 μg/g) decreased lipid peroxidation in the plants, whereas selenate supplementation of soil at the same levels had the opposite effect. Higher lipid peroxidation levels for plants grown in selenate supplemented soil may account for the observed stunting of plant growth [66].

Because inorganic forms of selenium effectively increase selenium levels in plant crops and prevent selenium deficiency diseases in people that consume them, it is important to understand the effect of selenite and selenate supplementation in crops. Plants more efficiently absorb selenate, as indicated by the higher concentrations of selenium in plants supplemented with selenate as compared to selenite supplementation, but selenite may be safer to use in fertilizers, since there are fewer adverse effects with supplementation at high concentrations [66].

To better treat selenium deficiency and to prevent selenium toxicity, an accurate evaluation of the effects of inorganic selenium compounds in fertilizers is required. Selenite- and selenate-enriched pastures and salt licks are also used to increase selenium concentrations in livestock [68]. It is therefore also important to understand the effects of this supplementation on animals, and further studies are necessary to determine the appropriate levels and forms of inorganic selenium supplementation that are most effective.

In humans, selenium deficiency causes poor immune response by reducing T-cell counts and impairing lymphocyte proliferation and response [63, 123]. Studies have shown that human supplementation of 200 μg/day of sodium selenite over an 8-week period resulted in enhanced T-lymphocyte response [124]. In HIV and AIDS patients, selenium deficiency is associated with decreased immune cell count, higher rates of disease progression, and increased risk of death [125, 126]. Additionally, selenium was found to protect cells from oxidative stress, resulting in slower progression of this disease [127].

A study performed over a period of 5 years on HIV-positive children found that those with low selenium levels died at a younger age than patients with higher selenium status [128]. These experiments were corroborated by another study involving HIV-positive men and women that linked increased death rates with selenium deficiency [129]. Clinical studies performed on male patients with AIDS and AIDS-related complex (ARC) showed that blood selenium levels increased upon supplementation of 400 μg/day Se-enriched yeast from 0.142 to 0.240 μg/ml over a period of 70 days [130, 131]. Similar results were also observed in AIDS patients supplemented with sodium selenite (80 μg/day) [130, 132]. These investigations indicate that both organic and inorganic selenium supplementation is effective for the treatment of patients with immune deficiencies [129].

Numerous studies indicate that selenium also plays an important role in cancer prevention and treatment. Patients with uterine cervical carcinoma were found to have low glutathione peroxidases and selenium levels [133], and, in a random, double-blind cancer prevention trial, the incidence of prostate cancer was reduced by ~50% compared to the placebo group upon selenium supplementation of 200 μg/day as selenium-enriched yeast. Total cancer mortality also decreased by a statistically significant 40% [134]. In a separate trial, a reduction of only 25% in cancer mortality was observed with 200 μg/day selenium supplementation but not with 400 μg/day [135]. In a larger study of selenium and vitamin E supplementation on prostate cancer (SELECT), selenium supplementation was not uniformly protective against the disease [136]. The SELECT trial indicated that selenium, in the form of l-selenomethionine (200 μg/day), was not effective in preventing prostate cancer [137, 138]. This study contradicts similar studies such as the Nutritional Prevention of Cancer (NPC) trial, which showed a reduction in prostate (63%), lung (48%), and colorectal (58%) cancers upon selenium supplementation [134, 137, 139, 140]. Disparities in both the studies have been attributed to the form of selenium used for supplementation, plasma selenium levels of patients, and the type of cancer to be treated [134, 137, 139, 140]. While SELECT trials focused on selenomethionine and prostate cancer, the NPC trials used selenite and selenium-enriched yeast as the selenium source to prevent skin cancer. In addition, the selenium status of patients in SELECT studies was higher (135 ng/ml) than those in the NPC trials (113 ng/ml) [137–139]. The results of these two studies show the importance of understanding the role and mechanism of selenium speciation in preventing or treating cancer. Mechanisms proposed for the observed anticancer properties of selenium include antioxidant activity, regulation of gene expression, cell proliferation control, and angiogenesis inhibition [141, 142].

Antioxidant and Anticancer Activities of Inorganic Selenium Compounds

ROS generation is directly linked to cellular and DNA damage and is the primary cause of many diseases [9, 14, 18, 44, 143]. Antioxidants have been used to prevent or reduce the effects of ROS-mediated DNA and other cellular damage, and selenium has been extensively studied for its antioxidant properties [55, 57, 58]. Inorganic selenium compounds can also act as pro-oxidants to produce DNA damage and cell death, an activity that plays an important role in the treatment of cancer [23, 51, 144–147]. Although sometimes confused in the literature, this distinction between antioxidant (cancer prevention) and pro-oxidant (cancer treatment) behavior is important to make, since the chemical and cellular mechanisms behind each type of activity are distinct. The behavior of these inorganic selenium compounds is complex, and in several studies, both antioxidant and pro-oxidant behavior have been observed for the same selenium compound depending on experimental conditions.

The main inorganic selenium compound used in most cancer treatment studies is sodium selenite; however, a few studies use other forms, such as sodium selenate and selenium dioxide [57, 62, 144–146, 148, 149]. Selenate and selenite (0.1 μg/ml) are effective dietary supplements for the inhibition of tumor cell growth in rodents [146, 148]. These two inorganic selenium compounds also strongly inhibit the growth of mammalian tumor cells at cell cycle phases specific for each compound (Table 1) [145]. Selenite treatment (10 μM) of human lymphocytes resulted in accumulation in the S-phase with irreversible growth inhibition, whereas selenate-treated (250 μM) lymphocytes accumulated in the G₂ phase with reversible inhibitory effects [145]. In a separate study, selenium dioxide (1.5 μM) effectively enhanced lymphocyte progression into the S-phase of the cell cycle in patients with stage IV cancer, restoring immune function and controlling cancer progression [144]. Takahashi et al. [57] showed that both selenite (10 μM) and selenium dioxide (100 μM) induced ~80% apoptosis in human oral squamous carcinoma (HSC-3) cells after treatment for 72 h, whereas selenate had no effect on cell survival (Table 1).

Table 1 The antioxidant or pro-oxidant behavior of inorganic and organic selenium compounds

Ramoutar and Brumaghim [23] have shown that inorganic selenium compounds exert both antioxidant and pro-oxidant activities against iron-mediated oxidative DNA damage (Table 1). Selenite progressively inhibited DNA damage at all concentrations tested (0–5000 μM), with 91% inhibition at the highest concentration under Mode I conditions (50 μM H₂O₂) [23]. Selenate and selenide had no effect on damage under similar conditions, whereas SeO₂ was found to be both a pro-oxidant and an antioxidant, increasing DNA damage by 20% at 50 μM, but inhibiting 100% DNA damage at 5000 μM [23]. Similar studies performed with organoselenium compounds, SeMet (1–1000 μM) showed no antioxidant activity, whereas methyl selenocysteine prevented ~76% iron-mediated DNA damage at very high concentrations (20,000 μM; Table 1) [150].

In contrast, under Mode II conditions (50 mM H₂O₂), Na₂SeO₃ showed pro-oxidant activity at all concentrations tested (0.5–5000 μM), damaging 90% DNA at the highest concentration in the absence of iron. However, SeO₂ was an efficient antioxidant under similar conditions, preventing 81% iron-mediated DNA damage, whereas Na₂SeO₄ and Na₂Se had no effect on such damage (Table 1) [23]. Similar studies were also performed to determine the antioxidant activity of inorganic selenium compounds to prevent copper-mediated DNA damage [151]. In these experiments, both Na₂SeO₃ and SeO₂ were effective antioxidants, preventing 95 and 63% of copper-mediated DNA damage at the highest concentration tested (5000 μM), whereas Na₂SeO₄ and Na₂Se had no effect on such damage. Little activity was observed at more physiologically relevant concentrations (Table 1) [151]. The antioxidant behavior of these inorganic selenium compounds has been attributed to the oxidation state of the selenium atom, rather than the overall charge of the selenium compound [23]. Inorganic selenium compounds in the +4 oxidation state (Na₂SeO₃ and SeO₂) were more effective antioxidants than Na₂SeO₄ and Na₂Se, with selenium oxidation states of +6 and −2, respectively [23]. In a separate study, Hamilton and Wilker [77] found that Na₂SeO₄ (6.2 mM) was effective at inhibiting DNA damage caused by alkylating agents.

High levels of selenite (1 μg/ml) were also shown to increase thioredoxin reductase activity 2-fold in rat kidney, liver, and lung tissues as compared to rats with normal selenite intake (0.1 μg/ml) (Table 1) [152]. In human colon cancer cells supplemented with various dosages of selenite (0.1, 1, and 10 μM), thioredoxin reductase activity increased with increasing selenium concentration, resulting in a 65-fold increase at the highest concentration tested (Table 1) [153].

Although these studies make a strong case for selenium supplementation for the prevention or treatment of cancer, additional studies are required to better compare and elucidate the structural and chemical properties of inorganic selenium compounds that contribute to antioxidant or pro-oxidant behavior. For example, while selenite has been shown to be a more effective antioxidant when compared to selenate in many studies, selenite can also oxidatively damage DNA under conditions of oxidative stress. This pro-oxidant effect is not observed with selenate, suggesting that selenate may be safer for use in human or animal supplementation.

Selenium Speciation and Anticancer Activity

Selenium bioavailability differs for organic and inorganic compounds, and studies have shown that the formulation of the selenium compound, and not the presence of the element itself, is essential for chemopreventative activity [154–156]. It is therefore critical to elucidate the specific selenium compounds that are required for such activity. For example, sodium selenite (5–10 μM) introduced into cell culture media induced DNA single-strand breaks and cell death via necrosis [154, 155, 157, 158]. Organoselenium compounds (10–50 μM), however, caused cell death by apoptosis with no DNA single-strand breaks (Table 2) [154, 155]. Similar results were obtained in a separate study by Lu and co-workers [159] to determine the effect of selenium form on mouse mammary carcinoma cells (Table 2). Although all selenium compounds tested inhibited cell proliferation and induced cell death, selenite and selenide induced both DNA single (51–59%) and double-strand breaks (4.8–14.6%) in a concentration-dependent manner (1-5 μM); no DNA damage was observed for the organic forms, methyselenocyanate (2–7 μM) and methylselenocysteine (20–100 μM) [159].

Table 2 The effect of selenium speciation in anticancer studies

In another study, selenite was found to be more potent than either selenocystamine or selenomethionine in inducing apoptosis in mouse keratinocyte (BALB/cMK2) cells [143]. In this experiment, selenite (10 μg/ml) resulted in 100% apoptosis, whereas selenocystamine resulted in 2.8% apoptosis at the same concentration [143]. Selenocystamine (250 μg/ml) was capable of inducing 100% apoptosis, whereas selenomethionine (5–250 μg/ml) showed no effect on BALB/cMK2 cells (Table 2) [143]. A study to determine whether sodium selenite and methylselenic acid (MSeA) repressed interleukin-6-mediated (IL-6) androgen receptor action in prostate cancer progression indicated that selenite significantly inhibited IL-6 activity in human prostate cancer (LNCaP) cells, but MSeA did not [160].

Although these investigations indicate that the inorganic forms of selenium may be more effective for the prevention or treatment of diseases when compared to the organic forms, further studies are necessary to evaluate the effects of selenium speciation for such purposes. While most studies have focused solely on selenite, the examination of other inorganic selenium compounds such as selenate, selenide, and selenium dioxide, in addition to organoselenium compounds, in antioxidant and anticancer experiments would aid in understanding the effects of selenium speciation within these inorganic and organo-selenium compounds on ROS-induced DNA damage and cell death.

Mechanisms of Antioxidant and Anticancer Activity for Inorganic Selenium Compounds

While the precise mechanisms of cancer prevention or treatment has not been elucidated for inorganic selenium compounds, several reports indicate that the protection against oxidative damage may involve selenoproteins, such as GPx and thioredoxin reductase, and may require supranutritional levels of selenium [55, 152, 153, 161–170]. One proposed mechanism for the effects of cancer treatment by selenium compounds is the direct action of pro-oxidant selenometabolites to generate ROS, resulting in cellular toxicity [57, 149, 171].

This ROS-generation mechanism involves the metabolism of selenite and selenate to generate hydrogen selenide (H₂Se), a by-product of selenium metabolic pathway [149, 172, 173]. High levels of selenide can then react with oxygen to produce ROS, resulting in oxidative damage to cells [149, 172, 173]. It has been suggested that the cytotoxicity of inorganic selenium compounds such as selenite and selenium dioxide is due to the formation of selenotrisulfides (RSSeSR), such as selenoglutathione, upon reaction with disulfide peptides or proteins (Reaction 7) [149, 172, 173]. In more recent studies, this mechanism has also been attributed to the pro-oxidant effect of inorganic selenium compounds in different cell lines [57, 143, 174, 175].

This proposed mechanism for selenite cytotoxicity has been further supported by generation of superoxide upon reduction of selenotrisulfide (GSSeSG) to selenopersulfide anion (GSSe⁻) [149]. The selenopersulfide anion, in turn, is reduced by thiols to generate H₂Se (Reactions 7–9) [149]. Selenide then reacts with oxygen to form elemental selenium (Se⁰) and O ^•−₂ (Reaction 10) [149, 172, 173]. Studies showing that selenite and selenium dioxide, but not selenate, are cytotoxic via this mechanism have been previously reviewed by Spallholz [149].

$$ {\text{SeO}}_{ 3}^{ 2- } + 4 {\text{GSH}} + 2 {\text{H}}^{ + } \to {\text{GSSG}} + {\text{GSSeSG}} + 3 {\text{H}}_{ 2} {\text{O}} $$

(7)

$$ {\text{GSSeSG}} + {\text{GSH}} \to {\text{GSSG}} + {\text{GSSe}}^{ - } + {\text{H}}^{ + } $$

(8)

$$ 2 {\text{GSSe}}^{ - } + 4 {\text{NADPH}} \to 2 {\text{H}}_{ 2} {\text{Se}} + {\text{GSSG}} + 4 {\text{NADP}}^{ + } $$

(9)

$$ {\text{H}}_{ 2} {\text{Se}} + {\text{O}}_{ 2} \to {\text{Se}}^{0} + {\text{O}}_{ 2}^{ \bullet - } + 2 {\text{H}}^{ + } $$

(10)

In contrast, the mechanism for the antioxidant ability of both inorganic and organoselenium compounds in preventing iron- and copper-mediated oxidative DNA damage is through coordination of the metal ion by the selenium compounds [23, 74, 150]. Na₂SeO₃ and SeO₂ (0.5–5000 μM) showed no effect on DNA damage produced by completely coordinated [Fe(EDTA)]²⁻ (400 μM) or [Cu(bipy)₂]⁺ (50 μM), so coordination of Fe²⁺ or Cu⁺ to inorganic selenium compounds is a primary mechanism for both their antioxidant and/or pro-oxidant activities [23, 151]. Similar results were also observed for organoselenium compounds [49]. Although metal coordination to inorganic selenium compounds has not been directly observed in biological systems, iron, copper, mercury, and aluminum react with selenite, selenate or selenium dioxide to form complexes such as Fe₂(H₂O)₄(SeO₃)₂, FeH(SeO₃)₂, Fe(HSeO₃)₃, and Al₂(SeO₃)₃·3H₂O [176, 177], where iron is coordinated through the oxygen atoms of inorganic selenium compounds such as selenite (Fig. 4).

Fig. 4

Coordination of iron to selenite in Fe₂(H₂O)₄(SeO₃)₂ reported by Xiao et al. [177]

Bioavailability and Activity of Oxo-Sulfur and Oxo-Selenium Compounds

Fruits, vegetables, cereal, nuts, and teas have been widely studied for their ability to ameliorate oxidative stress and their potential to prevent or treat cancer, aging, and cardiovascular diseases [54, 62, 93, 94, 178–180]. The antioxidant capabilities of many foods are attributed to their vitamin, polyphenolic, selenium, and sulfur content [53, 54, 62]. Although there are many members of the Allium genus, including onions, leeks, scallions, and chives, garlic has been widely studied for its antioxidant activity [53, 54, 81]. For many centuries, garlic (Allium sativum Liliaceae) has been cultivated and used in food preparation for its distinct flavor and aroma, as well as for its medicinal properties [53, 54, 81]. Throughout the years, this bulb has been used to treat the plague, animal bites, leprosy, and cancer, as well as bacterial, immune, and cardiovascular diseases [53, 54, 81, 181–188].

The characteristic flavor and aroma of garlic are attributed to the volatile organosulfur compounds produced upon tissue damage and enzymatic hydrolysis from non-volatile precursors [81, 189]. The vegetative parts of garlic are odorless and composed of non-volatile sulfur-storage compounds known as S-alk(en)yl-L-cysteine sulfoxides [81, 189]. These compounds (Fig. 2) are stored in the cytosol of undamaged Allium tissues protected from the enzyme alliinase, which is found in the vacuoles [81, 189]. Upon tissue damage, alliinase and S-alk(en)yl-L-cysteine sulfoxides react to generate sulfenic acid that then undergoes condensation to form the volatile thiosulfonate compounds [81, 189]. The S-alk(en)yl-L-cysteine sulfoxides detected in garlic and several other varieties of the Allium genus are S-allyl-L-cysteine sulfoxide (alliin, ACSO), S-methyl-L-cysteine sulfoxide (methiin, MeCysSO, MCSO), S-propyl-L-cysteine sulfoxide (propiin, PCSO), and S-trans-1-propenyl-L-cysteine sulfoxide (isoalliin, TPCSO) [81, 89–192].

The medicinal properties of garlic are primarily attributed to the thiosulfonate compound, allicin produced from allin by alliinase when garlic is crushed (Fig. 5) [193, 194]. While alliin is the main S-alk(en)yl-L-cysteine sulfoxide found in garlic, and is responsible for the volatile odor of cut or crushed garlic, MeCysSO is the most ubiquitous found in onions, chives, leeks, and scallions in various quantities [81].

Fig. 5

Production of allicin from alliin by alliinase

In folk medicine, cauliflower (Brassica oleracea Liliaceae var. botrytis) is also used for its medicinal purposes [80]. The juice extracts from raw cauliflower leaves are expectorants and are used in the treatment of gastric and duodenum ulcers, whereas the stewed leaves have been used as antipyretics or antirheumatics [80]. These medicinal properties have been attributed to S-methyl methane thiosulfonate (MMTS), an oxo-sulfur compound found in cauliflower, broccoli, and cabbage, as well as in the Allium vegetables [80, 195–197].

Dimethyl sulfone or methylsulfonylmethane (Me₂SO₂) is another oxo-sulfur compound found in vegetables including broccoli, peppers, asparagus, and cabbage [82]. It can also be found in trace amounts in fish, meat, unpasteurized milk, beverages, and eggs and has more recently been used as a dietary supplement [83–85]. Currently, Me₂SO₂ is sold in over 30 products in combination with other dietary supplements such as chondroitin sulfate and glucosamine, and in more than 50 different products as a single agent in tablets, capsules, creams, and lotions [198]. Although some physicians suggest a daily dose of only 300 mg, the recommended daily dose of Me₂SO₂ is reported to be between 1000 and 6000 mg when taken as a dietary supplement [82]. Me₂SO₂ is a metabolite of dimethyl sulfoxide (DMSO), a byproduct of algae and phytoplankton decay, and is commercially synthesized by reacting hydrogen peroxide with DMSO to produce Me₂SO₂ and water (Reaction 11) [198, 199]. Studies have shown that ~15% of orally ingested DMSO is recovered as Me₂SO₂ in urine [198–201].

$$ {\text{Me}}_{ 2} {\text{SO}} + {\text{H}}_{ 2} {\text{O}}_{ 2} \to {\text{Me}}_{ 2} {\text{SO}}_{ 2} + {\text{H}}_{ 2} {\text{O}} $$

(11)

Amino acids are major targets for oxidation by reactive oxygen species, and this oxidation can disrupt protein structure and function [86]. Methionine can be oxidized to methionine sulfoxide (MetSO) by ROS and reactive nitrogen species (RNS) such as hydrogen peroxide, hydroxyl radical, and peroxynitrite anion [87, 88, 202]. ROS-mediated oxidation of Met results in mixtures of R- and S-isomers of MetSO [203]. Metal-catalyzed oxidation of methionine occurs through Fenton or Fenton-like reactions when metals such as iron and copper are reduced, subsequently producing hydroxyl radical upon reaction with hydrogen peroxide [87, 204–206].

MetSO can also be formed from methionine oxidized by H₂O₂ alone in the absence of metal ions (Fig. 6) [207, 208], and both peroxynitrous acid (ONOOH) and peroxynitrite also react with methionine to produce MetSO [88, 202]. Methionine oxidation can result in changes in protein hydrophobicity, alterations in protein conformation, and disruption of biological function [91, 204, 209–214]. However, cells contain two enzymes that reduce MetSO to Met, repairing the oxidative damage [87, 89–91]. Methionine sulfoxide reductase A (MsrA) specifically reduces the S-isomer and methionine sulfoxide reductase B (MsrB) is specific for the reduction of the R-isomer [87, 89, 91]. Based on the reversibility of MetSO generation, MetSO formation is proposed to be important in regulating cell functions [87, 91, 204, 213, 215–217]. Interestingly, there have been reports on one biologically relevant oxo-selenium compound formed from the oxidation of selenomethionine (SeMet) by peroxynitrite to methionine selenoxide (MetSeO) [218, 219]. This oxidation is analogous to the peroxynitrite oxidation of methionine [218].

Fig. 6

Oxidation of methionine to methionine sulfoxide by hydrogen peroxide

Antioxidant and Pro-Oxidant Effects of Oxo-Sulfur Compounds in Disease Prevention

Many studies have reported antioxidant effects and disease amelioration by supplementation with aged garlic extracts (AGE) and garlic essential oils [220–222]. Although these products are generally considered to be safe and may be effective in preventing diseases such as cancer, it is difficult to determine the bioactive sulfur component or mixture of components that is responsible for the observed biological effects [220, 221]. Therefore, this review will focus on the ROS damage prevention (antioxidant) or ROS generation (pro-oxidant) abilities of individual oxo-sulfur compounds.

Several studies have shown that oxo-sulfur compounds play a significant role in preventing ROS-mediated cellular damage. For example, allicin, a major component in garlic, acts as a vasodilator, inhibits cholesterol biosynthesis, ameliorates serum lipid in hyperlipidemic rabbits, and lowers intraocular pressure in normal rabbits [193, 223–226]. In a separate study, both allicin and alliin had no effect in reducing lipid peroxidation induced by ferrous sulfate/ascorbic acid in microsomal membranes (Table 3) [227]. Another study to determine the effect of thiosulfonates on platelet aggregation showed that allicin inhibited 74% of such aggregation with an IC₅₀ of 0.27 mM [189]. Hirsh et al. [193] found that allicin (10–40 μM) inhibited cell proliferation in mammary, colon, and endometrial cancer cells with 50% inhibition at 10–25 μM. Alliin, the precursor to allicin, showed no inhibitory effect at all concentrations tested (0–64 μM) [193].

Table 3 Dietary sources and application of various oxo-sulfur compounds

Another oxo-sulfur compound found in Allium vegetables is MeCysSO; however, only a few studies have been performed on this possible antioxidant sulfur derivative, despite the fact that it is a major oxo-sulfur compound in garlic (Table 3). MeCysSO is effective for the treatment of hyperglycemia and hyperlipidemia in diabetic rats [228–230]. In addition, daily oral administration of MeCysSO (200 mg/kg) for 45 days significantly controlled blood glucose and lipids in tissues and serum of diabetic rats [228]. The activities of HMG CoA reductases, liver hexokinase, and glucose-6-phosphate in these animals were close to normal upon MeCysSO treatment, effects similar to the anti-diabetic drugs glibenclamide and insulin [228]. Kumari and Augusti [229] also observed that that MeCysSO was effective in lowering total cholesterol levels in rats.

Similar to allicin, MMTS also has chemopreventative properties [195]. Nakamura et al. [195–197] found that MMTS isolated from cauliflower homogenate showed strong antimutagenic activity against UV-induced mutation in wild-type E. coli (B/r WP2), but not in mutant cell strains lacking excision-repair activities. A separate study by the same group showed that MMTS (10 mg/kg) suppressed the frequency of aflotoxin B₁-induced chromosome aberrations after 2 h [195]. Additionally, Kawamori et al. [231] reported that MMTS (20 and 100 ppm) inhibited 42 and 21% of intestinal neoplasm incidences induced by azoxymethane in rats, respectively.

MMTS also decreased the incidence of phenobarbital-promoted or diethylnitrosamine-initiated hepatocarcinogenesis in rats [232]. In Drosophilia melanogaster and mouse micronuclei, MMTS was found to reduce mitomycin C-induced somatic mutation and recombination via oral administration [233]. Experiments performed with various antioxidants, including Kefir grain extracts, showed that both Kefir extracts (900–21,000 μg/ml) and MMTS (10 μg/ml) stimulated more than 50% thymine dimer repair in UVC-irradiated HMV-1 cells, whereas other antioxidants tested (epigallocatechin and vitamins A, C, E, and K) showed little repair enhancement (≤10–30%) [234].

While the use of Me₂SO₂ in commercially available products (Table 3) is generally considered safe, little data is available to assess the safety and toxicity of this oxo-sulfur compound. Unconfirmed side effects of Me₂SO₂ consumption include headaches, hypertension, gastrointestinal symptoms, insomnia (if taken before bedtime), and increased hepatic enzyme levels [198]. Me₂SO₂ has been used to treat inflammation, parasitic infections, allergies, asthma, cancer, arthritis, and rheumatic pain [198, 199]. Due to its sulfur content, it has also been used for nourishing hair, fingernails, and skin, and in the maintenance of normal connective tissues [198, 235]. In one study to determine the acute (2000 mg/kg/day) and chronic (1500 mg/kg/day) effects on rats, oral administration of Me₂SO₂ for 90 days did not affect body weight, histopathological lesions in tissues and organs, hematological parameters, or mortality [82].

Early pharmacokinetic studies showed that ~64% of [³⁵S]Me₂SO₂ (21 mg/kg) was excreted within 24 h upon intraperitoneal administration to rats [236]. In a more recent study, the distribution of Me₂SO₂ administered orally for 7 days using a ³⁵S radioisotope tracer found that [³⁵S] Me₂SO₂ (470 mg/kg/day) was excreted in urine (~70%) and feces (~10%) [237]. The highest levels of radioactivity were found in hair, blood, and spleen [237]. Further investigations of Me₂SO₂ pharmacokinetics in rats was performed by Magnuson et al. [238]. They also used radiolabled [³⁵S]Me₂SO₂ to determine that Me₂SO₂ is rapidly absorbed, well distributed, and efficiently eliminated [238]. Oral administration of 500 mg/kg in rats showed that the majority of Me₂SO₂ was excreted in urine (~57%), with only 1.6% excreted in feces [238]. After 48 h, Me₂SO₂ was distributed evenly in several tissues, corresponding to blood concentrations of this compound [238].

Two separate studies performed by Cottler-Fox and co-workers [239] and Wever and co-workers [240] found that Me₂SO₂ was present in human plasma and cerebrospinal fluid ranging from 0 to 25 μM. Me₂SO₂ was also detected by magnetic resonance spectroscopy in the brains of patients with memory loss and in normal patients given doses of 1–3 g/day [241]. In this investigation, Me₂SO₂ was found to be distributed equally between white and gray matter of the brain in all patients, ranging from 0.42 to 3.40 mmol/kg, with no adverse neurochemical or clinical effects [241]. Me₂SO₂ was also detected in the brain at a concentration of 2.36 mM in a normal patient after taking Me₂SO₂ at a dosage of 181 mg/kg/day [242].

While the suggested use of Me₂SO₂ is as a dietary supplement for the treatment of arthritis and allergies, only a few studies have reported such activities [199, 243, 244]. A 12-week randomized and controlled study on individuals with knee osteoarthritis was conducted, with each of the 118 patients given Me₂SO₂ (1.5 g/day). One-third reported decrease in pain; joint mobility and improved walking time was also observed [244]. In a second study, 21 patients with osteoarthritis given Me₂SO₂ (3 g) twice daily for a 12-week period showed decreases in pain (25%), stiffness (20%), physical function (31%), and total symptoms (25%) [198]. Interestingly, in this study, patients in the placebo group also showed decreases in pain, stiffness, physical function, and total symptoms (~13, 12, 17, and 14%, respectively) [198]. Although Me₂SO₂ may ameliorate the effects of osteoarthritis, its full effects were not observed in such a short period of time, indicating the need for longer trials [198]. A study to evaluate the efficacy of Me₂SO₂ for reduction of seasonal allergy rhinitis (SAR) symptoms such as headaches, sinus infections, breathing difficulties, and nasal congestion was performed by Barrager et al. [199]. The results of this study indicate that Me₂SO₂ supplementation of 2600 mg/day over a period of 30 days was effective in reducing the symptoms of SAR [199]. Me₂SO₂ was also found to have minimal side effects, suggesting that this oxo-sulfur compound may be therapeutic for the amelioration of SAR-associated symptoms [199].

Antioxidant and Pro-Oxidant Mechanisms of Oxo-Sulfur Compounds

Although numerous studies show that oxo-sulfur compounds can be used in the treatment or prevention of several diseases including cancer and osteoarthritis, the mechanisms of such action have not been elucidated [193–195, 198, 212, 220, 230, 232]. Studies by Rabinkov and co-workers [194] indicate two possible mechanisms for the biological activity of allicin: radical scavenging or its ability to react with thiols. Spin trapping techniques and EPR measurements were used to show that allicin efficiently scavenged ^•OH produced by the Fenton reaction [194]. The ability of allicin to inhibit the functions of thiol-containing proteins such as papain and alcohol dehydrogenase was also suggested as a possible mechanism [194].

The susceptibility of Met to oxidation results in antioxidant activity due to its radical scavenging capabilities, and the ability of Msr to reduce MetSO back to Met [212]. Evidence of such antioxidant activity is observed in studies performed in yeast and Drosophilia, which showed that over-expression of the MsrA gene prevents oxidative stress induced by toxic levels of hydrogen peroxide and paraquat [245, 246]. Studies performed on various bacterial strains and yeast lacking the MsrA gene determined that these strains were more susceptible to paraquat- or H₂O₂-induced oxidative stress [245, 247, 248]. Under normal growth conditions, studies performed on MsrA knockout mice resulted in a 40% decrease of maximal life-span upon exposure to 20% oxygen [249]. In other studies, Met residues on the lipoprotein surface reduced low-density (LDL) and high-density (HDL) lipoprotein-generated peroxides and cholesterolester peroxide to their respective hydroxyl derivatives [91, 250].

The oxidation of methionine to methionine sulfoxide has been implicated in aging, adult respiratory distress syndrome, and emphysema [213, 215]. A study by Costabel and co-workers [251] showed that methionine oxidation by neutrophil-generated ROS results in acute and chronic bronchitis, dependent on the Met/MetSO ratio in lavage fluid in the bronchialveolar. An investigation by Stadtman et al. [211] indicated that surface hydrophobicity of rat liver proteins increased with age over a period of 24 months. This study also indicated that increases in both hydrophobicity and MetSO levels were caused by protein oxidation from ROS [211].

Gradual decreases in the levels of MsrA activity in the liver and the kidney tissues of rats were found to be age-related, whereas in brain tissues, no age-related loss of enzyme activity was observed [252]. MsrA activity in rat kidney and liver tissues decreased from ~296 and ~181 pmol/min/mg, respectively, at 9 months of age to ~158 and ~85 pmol/min/mg, respectively, at 29 months [252]. In addition, the proposed Met/MetSO antioxidant cycle may aid in the prevention of Alzheimer’s disease [87, 253]. A study performed by Markesbery et al. showed that in various brain regions of Alzheimer’s patients, the level of MetSO and protein carbonyls, another measure of oxidative damage, was significantly greater than in the brains of patients without this disease [253].

In an animal study, 67% of Met residues in rat brain calmodulin were oxidized to MetSO in aged rats [254]. Experiments by Wells-Knecht et al. [255] showed that in humans, the methionine sulfoxide content of skin collagen increases from approximately 4%, while young to approximately 12% at 80 years of age. Collectively, these studies indicate the oxidation and reduction processes of methionine, as well as the expression of MsrA, play an important role in the prevention of ROS-mediated diseases. Interestingly, methionine oxidation increases protein hydrophobicity, despite the fact that Met is more hydrophobic than MetSO [211, 256]. Further investigations are therefore needed to better understand effects of the Met/MetSO oxidation cycle at the molecular level, since local changes in protein folding may play a role in hydrophobicity alterations [215].

Similar to DNA inhibition experiments performed with inorganic selenium compounds, the ability of oxo-sulfur compounds to prevent metal-mediated DNA damage has also been investigated (Table 4). MetSO, MeCysSO, MMTS, Me₂SO₂, and methyl phenyl sulfoxide (MePhSO; Fig. 2) showed little inhibition of iron-mediated oxidative DNA damage. While MeCysSO and MMTS inhibited 17 and 20% DNA damage at 1000 μM, respectively, the other oxo-sulfur compounds showed no effect on DNA damage [51]. Likewise, Me₂SO₂ and MePhSO have no effect on copper-mediated DNA damage, whereas MMTS (1000 μM) is a pro-oxidant, producing 35% damaged DNA, and both MetSO and MeCysSO are efficient antioxidants with IC₅₀ values of 18 ± 3 and 8.1 ± 1 μM, respectively [51]. Interestingly, the reduced forms of MeCysSO and MetSO, MeCys and Met, were also effective antioxidants with IC₅₀ values of 8.9 ± 0.02 μM and 11.2 ± 0.02 μM, respectively [49]. Thus, the ability of these sulfur-containing amino acids to prevent copper-mediated DNA damage does not significantly change upon oxidation [51].

Table 4 The antioxidant and pro-oxidant activity of sulfur and oxo-sulfur compounds

Similar to the results obtained with the inorganic selenium compounds, metal coordination is also a mechanism for antioxidant activity of oxo-sulfur compounds [23, 49, 51, 74, 150]. Brumaghim et al. showed that sulfur compounds with the ability to prevent copper-mediated DNA damage have a Cu–S charge transfer band at ~240 nm when combined with Cu⁺, indicative of copper-sulfur coordination. Both MetSO and MeCysSO show similar UV bands with Cu⁺, also indicating copper binding to these compounds [49, 51, 74]. Interestingly, in gel electrophoresis experiments using 2,2′-bipyridine (bipy) to completely coordinate Cu⁺, MetSO and MeCysSO at 1000 μM, inhibited 43 and 88% of [Cu(bipy)₂]⁺-mediated DNA damage, respectively, significantly less than the inhibition observed for uncoordinated Cu⁺ [51]. The results of these studies demonstrate that metal coordination is a primary mechanism for the antioxidant activity of MetSO and MeCysSO, but a second mechanism, such as radical scavenging, may also be responsible for their DNA damage prevention at high concentrations [51]. Under similar conditions, MMTS (5000 μM) is a pro-oxidant, damaging 44% DNA in the presence of [Cu(bipy)₂]⁺/H₂O₂, indicating that copper coordination is not required for its pro-oxidant activities [51]. Taken together, these experiments indicate that the antioxidant or pro-oxidant activities of sulfur compounds are quite complex, and highlight the importance of understanding the conditions and mechanisms for such behavior.

Recently, aryl sulfoxides such as phenyl sulfoxide (PhSO) and MePhSO (Fig. 2) have been found to generate ROS upon irradiation, with implicated use in cancer therapy [257–259]. Interestingly, few studies have reported on the effects of this ROS generation on DNA damage (Table 4). In one such study, Predecki and co-workers [258] showed that photoactivation at 240 nm of PhSO and MePhSO results in 83% DNA damage at 180 and 360 μM, respectively. In these experiments, 2,2,6,6-tetramethylpiperdine-1-oxyl (TEMPO) scavenged carbon-based radicals, indicating that the possible mechanism for photoinduced-DNA damage by aryl sulfoxides is caused by hydrogen abstraction from the DNA backbone to produce a DNA radical [258]. It is assumed that the reaction between the DNA radical and the oxygen is responsible for DNA damage generated by PhSO and MePhSO [258, 259]. Oxygen, in its ground state (³O₂), may also be necessary for the pro-oxidant effects of PhSO, since photo-induced DNA damage was inhibited in the absence of oxygen [258]. These studies clearly indicate that the mechanistic action of oxo-sulfur compounds either as antioxidants or as pro-oxidants is quite complex, and further studies under biologically relevant conditions are essential to better understand the activity of these compounds in biological systems.

Oxo-Selenium Compounds

Many studies have examined the antioxidant and anticancer effects of the oxo-selenium compound, methylselenic acid (MSeA), and this work has been recently reviewed [70, 260]. It is thought that MSeA exerts its observed anticancer properties both through oxidation of protein thiols, similar to selenite, and through formation of MeSe⁻, a species that can generate superoxide upon reacting with O₂ [260]. Thus, MSeA acts as a prooxidant to promote cell dysfunction and death. More recently, MSeA treatment (2.5–5 μM) was found to significantly increase thiol concentration and oxidative stress in the endoplasmic reticulum of human lung carcinoma cells [261]. In addition, MSeA has been found to decrease expression of several kinases involved in cell growth and apoptotic pathways [262], particularly protein kinase C [263], at low micromolar concentrations.

In prostate cancer cell lines (DU145 and LNCaP), Li et al. [264] reported that MSeA induced apoptosis without generating hydroxyl radical or superoxide. Oral administration of MSeA (4 mg Se/kg) significantly inhibited the growth of prostate cancer tumors in xenograft mice (DU145), comparable to effects observed for similar treatment with methylselenocysteine, and more effective than similar selenite treatment. Additionally, MSeA treatment resulted in significantly less genotoxicity and higher inhibition of tumor angiogenesis compared to methylselenocysteine and selenite [265]. Clearly, the observed anticancer properties of MSeA may arise from a combination of several different mechanisms, and more work is required to develop a complete understanding of its biological activity.

The naturally occurring amino acid, selenomethionine (SeMet) is found in many proteins in place of its sulfur analog, methionine [266, 267]. Similar to methionine oxidation to its corresponding sulfoxide, SeMet is also oxidized by peroxynitrite and enzymes such as flavin-containing monooxygenases (FMOs) to methionine selenoxide (MetSeO) [218, 219, 266]. A significant amount of research has investigated Met/MetSO interconversion, but little work has investigated the properties of MetSeO. Oxidation kinetics of SeMet are 10–1000 times faster than oxidation of methionine to produce MetSeO [219]. While the effects of MetSeO in biological systems have not been reported, it is possible that this compound may lead to similar changes in hydrophobicity, conformation, and disruption of biological functions associated with its sulfur analog, MetSO [91, 204, 209–214]. In light of this, Sies and colleagues [218] investigated the reduction of MetSeO to SeMet using glutathione (GSH) as a reductant. This study showed that addition of MetSeO (0.4 mM) to increasing concentrations of GSH resulted in a loss of MetSeO with an increase in SeMet (Reaction 12), suggesting that GSH in low concentrations is effective in the protection against oxidants and that GSH may be responsible for redox cycling of selenoxides [218]. Similar reduction of MetSeO to SeMet was also reported to occur with cysteine, N-acetylcysteine, and ascorbic acid under conditions that did not reduce MetSO to Met [268].

$$ {\text{MetSeO}} + 2 {\text{ GSH}} \to {\text{SeMet}} + {\text{GSSG}} + {\text{H}}_{ 2} {\text{O}} $$

(12)

Apart from its role in protein function, SeMet metabolism by the methionine transsulfuration pathway produces selenocysteine, an essential amino acid for the function of several antioxidant enzymes including glutathione peroxidase and thioredoxin reductase [266, 269]. Interestingly, FMOs also have been shown to oxidize other selenium compounds, including the selenium containing drug ebselen, to their corresponding selenoxides [270]. Thus, further investigations are necessary to determine the effects of oxidation on protein function and activity and the conditions required for antioxidant and pro-oxidant activity of oxidized sulfur and selenium compounds, including selenium-containing drugs.

Conclusions

Investigating the antioxidant and anticancer properties of inorganic selenium and oxo-sulfur compounds in the treatment of diseases such as cancer, aging, and neurodegenerative diseases generated from reactive oxygen species is an active and promising area of research. In the case of inorganic selenium compounds, sodium selenite is the compound of choice for both antioxidant and anticancer studies. While the mechanism for its antioxidant activity is unclear, it has been proposed that the ability of selenocysteine containing enzymes such as GPx and thioredoxin reductases to prevent radical formation is a possible mechanism [57]. The generation of hydrogen selenide from selenometabolites such as selenite to produce toxic ROS is also suggested as a possible mechanism for the pro-oxidant and anticancer properties of inorganic selenium compounds [55, 57].

The antioxidant activity of oxo-sulfur compounds is much less understood than that of the inorganic selenium compounds. Most studies have focused on allicin, an oxo-sulfur compound produced from crushed garlic; however, disparities concerning the antioxidant activity of allicin are attributed to other endogenous components remaining from the extraction process of crude garlic extracts [53, 54]. While MMTS and other oxo-sulfur compounds may be effective in the treatment or prevention of cancer, their mode of action has not been investigated. Similarly, little is known about the antioxidant activity of MeCysSO, MetSeO, and MePhSO; however, the ability of MetSO to prevent ROS-mediated diseases such as aging, emphysema, and Alzheimer’s has been attributed to the cyclic interconversion of MetSO by methionine sulfoxide reductases [213, 215]. Although, MeCysSO and MetSeO are analogous to MetSO, little work has explored the effects of these compounds in biological systems. Since MeCysSO and MetSeO are formed from the oxidation of amino acids essential for protein function, investigating the ability of these compounds to act as antioxidants, as well as their role in disease prevention is of great interest. Therefore, further experiments including both in vitro, and cellular and animal studies are required to investigate such behaviors. In addition, studies are also needed to determine the effects of the sulfur compounds’ structural and chemical properties on their antioxidant activity.

Investigations of the ability of inorganic selenium and oxo-sulfur compounds to inhibit metal-mediated oxidative DNA damage determined that coordination between iron or copper and the antioxidant compound is required for prevention of DNA damage [23, 51]. Additionally, the ability of oxo-sulfur compounds to scavenge radicals was also suggested as a possible mechanism for this activity, particularly at high concentrations [51]. Further investigations to determine the role of metal coordination and radical scavenging mechanisms of antioxidant behavior as well as the conditions required to observe pro-oxidant properties of oxo-sulfur compounds are necessary to fully understand the biological activities of these compounds.

In addition, the role of less studied compounds such as selenate, selenium dioxide, methylcysteine sulfoxide, and methionine selenoxide in the prevention of ROS-mediated disease is necessary to better understand and elucidate a mechanism for the antioxidant and pro-oxidant activities of both inorganic and oxo-sulfur compounds. Although the selenium speciation is an important factor in the biological activities of these compounds [154–156], additional studies using larger numbers of selenium compounds including selenide, selenate, and selenoxides are required to fully understand this effect. Most of the inorganic selenium and oxo-compounds discussed in this review are obtained from food products such as vegetables, fruits, nuts, and dietary supplements. However, these foods and supplements may also contain other antioxidant or bioactive ingredients [1, 46, 47]. It is therefore equally important to understand the effects of other bioactive compounds on the biological activity of inorganic and oxo-sulfur compounds. Additional studies exploring the antioxidant, pro-oxidant, and mechanistic actions of inorganic selenium, oxo-sulfur, and oxo-selenium compounds are essential to fully understand the biological implications of sulfur- and selenium-containing food products and supplements in both humans and animals.