Rationale for the treatment of cancer with sodium selenite

https://doi.org/10.1016/j.mehy.2004.10.012Get rights and content

Summary

Epidemiological studies conducted during several decades of the last century have demonstrated the importance of sufficient nutritional supply of selenium (Se) for human health. More importantly, low blood Se levels were found to be associated with an increased incidence and mortality from various types of cancers. Recently, attention of researchers was drawn to the relationship between free radical generation, known otherwise as oxidative stress, and carcinogenesis. It was therefore thought that antioxidants should be beneficial for prevention and inhibition of different malignancies. However, there appeared to be a paradox, because tumor growth is associated with tissue hypoxia that is accompanied by the formation of reductive rather than oxidative free radicals. Various organic and inorganic Se compounds, generally considered to be antioxidants, produced mixed results when tested in animal models and human subjects. Amongst them, sodium selenite has been shown to be the most effective in an in vitro and in vivo carcinogenesis studies. As recently demonstrated, selenite is not an antioxidant, but possesses oxidizing properties in the presence of specific substrates. Thus selenite is capable of oxidizing polythiols to corresponding disulfides, but does not react with monothiols. Such polythiols associated with cancer membrane-bound proteins appear under the reducing conditions of hypoxic tumor tissue. These thiol groups can, in turn, initiate a disulfide exchange reaction with plasma proteins, predominantly with fibrinogen, to form an insoluble and protease-resistant fibrin-like polymer. As the result, tumor cells become surrounded by a coat which masks specific tumor antigens thus allowing cancer cells to escape immune recognition and elimination by natural killer (NK) cells. Selenite by virtue of oxidizing cell membrane thiols, can prevent the formation of the coat and consequently makes cancer cells vulnerable to the immune surveillance and destruction. In addition, selenite may directly activate NK cells, as well as inhibit angiogenesis without undesirable decrease in the oxidative potential of cellular environment. It is, therefore, postulated that sodium selenite, in view of its relative low toxicity, might become a drug of choice for many types of cancer including leukemia.

Introduction

Although selenium (Se) is an ubiquitous metalloid with properties similar to those of sulfur, it was only relatively recently shown to be an essential element in human nutrition [1]. Human daily intake of Se varies from 60 to 300 μg/day, depending on the food products consumed and the concentration of Se in the soil on which these products were grown. Regional variations in the soil content of Se, ranging from 0.1 μg to 1000 mg/kg, have been well documented. It has been consistently observed that in cancer patients the concentration of blood selenium is lower than that in healthy subjects [2], [3]. Also significant inverse associations were found between Se supply and tumorigenesis in experimental animals [4], [5], [6]. However, potential use of Se as an anticancerogen has received attention only in the past decade [7], [8]. Yet despite strong evidence for the importance of selenium in carcinogenesis, very little is known about the mechanism of its action. Practically no data exists to explain why certain forms of selenium are more active than the other in the inhibition of cancer cell growth. Moreover, known toxicity of elemental selenium confuses the issue when researchers do not specify chemical form of Se compound used in anticancer studies. It is generally accepted that the daily intake of Se should not exceed 600 μg, whether from organic or inorganic Se [9]. This figure is in a sharp contrast with animal studies indicating that reversible toxic symptoms, such loss of hair and nails, occur at doses 1000 times higher than those recommended for humans [10].

Selenium and its compounds are generally considered as antioxidants, particularly that some antioxidant enzymes contain Se in their active centers. Sodium selenite is, however, a unique compound that exhibits oxidative, rather than antioxidative, properties. Since this specific form of selenium has recently been found to be particularly active as an anticancer agent, I felt compelled to offer an explanation, however hypothetical, as to the mechanism of its therapeutic action.

Normal cells and tissues rely on an ample supply of oxygen delivered with an unobstructed flow of blood to be used for metabolic purposes. However, reduction of oxygen to water is associated with generation of the products of one electron addition or subtraction, known as free radicals (FR) that are considered to be responsible for the so called oxidative stress associated with many degenerative diseases including cancer. This conclusion was based on the detection of the oxidation products of nucleic acids, proteins and lipids formed as a consequence of ischemia/reperfusion studies both in vitro and in vivo [11], [12]. Since such conditions are associated with the production of the most biologically active hydroxyl radical (HR), it has been concluded that this specific radical is also an oxidant. Although HRs are generated during hypoxia, their effects are observable only after the period of reoxygenation, which in itself does not cause oxidative damage. The best example is the effect of ionizing radiation on a solution of human serum albumin. In the presence of oxygen, HRs produced by radiation energy led to oxidative degradation of this protein. However, in the absence of oxygen this protein remained not only undegraded, but polymerized in to a huge insoluble aggregate [13] (see Fig. 1).

It is known that a disulfide bond has no inherent stability, but is dependent upon the relative concentrations of appropriate electron donors and acceptors. One of the sensitive redox indices of physiologic fluids is a ratio of cystine (R–S–S–R) to cysteine (R–SH) that in well oxygenated blood is approximately 10:1. Under normal conditions sulhydryl groups of blood proteins also exist in the oxidized state in the form of disulfides that hold together polypeptide chain(s) of proteins in a specific conformation. In case of human serum albumin (HSA) there are 16 such intramolecular disulfide bridges responsible for a globular shape and hydrophilic properties of this most abundant blood protein. When some or all of the disulfides are reduced, the single polypetide chain is unfolded with the exposition of previously buried hydrophobic groups of the side chains of amino acids. Consequently, strong interaction between unfolded chains of HSA molecules results in the formation of large insoluble aggregates. However, limited reduction of disulfides in HSA is of no serious consequence to the tertiary structure of this protein, which in this way function as a scavenger of reducing free radicals. Although it is generally believed that free radicals are responsible for the so called oxidative stress, some of them, notably hydroxyl radical, possess reducing properties particularly when formed under anaerobic conditions. Exposure of plasma to Fenton’s reaction-generated hydroxyl radicals led to the formation of an insoluble complex composed predominantly of fibrinogen [14]. It should be noted that such a complex, due its high hydrophobicity, is resistant to protease-induced degradation, and thus resembles fibrin polymers observed in solid tumors [15].

It was demonstrated over a quarter century ago that cancer tissue is associated with reducing conditions [16]. This revolutionary concept was then confirmed by observations that growth of tumor cells is accompanied by shifting the redox equilibrium towards more reducing environment [17], [18], [19]. The question is, however, what is a cause of a shift from oxidative to reductive conditions in cancer tissues? All evidence indicates that the culprit is hypoxia that has been observed by numerous investigators to be associated with cancer rendering tumors resistant to radiation and chemotherapeutic drugs [20]. More recently, a transcription factor that upregulate a variety of genes when oxygen becomes scarce, was identified in practically all types of cancer [21], [22]. This hypoxia-inducible factor (HIF) is believed to be a result of a metabolic shift from oxidative to anaerobic glycolysis known as the Warburg effect [23]. On the other hand, numerous studies have shown that tumor growth can be controlled by increasing tumor oxygenation [24], [25]. Ozone, a strong oxidant, has also been reported to inhibit cancer growth [26].

It is of interest to note, that virus penetration and infection is facilitated by reductive functions of cell membrane that allow disulfide exchange reaction between its proteins and those of virus envelopes [27]. Disulfide exchange is also a prerequisite for HIV-1 envelope-mediated T-cell fusion during viral entry [28], and in fact many other fusion mechanisms [29]. Such a reaction can also take place between reductively generated sulhydryls of tumor membrane and plasma proteins, according to the following formula:Pm-SH+Pp-SS-PpPm-SS-Ppwhere Pm is a membrane protein, and Pp is a plasma protein.

Fibrinogen is one of the largest molecules of blood (340 kDa) composed of three pairs of polypetide chains linked together by means of numerous disulfide bridges. Reduction of fibrinogen releases individual chains which may then undergo disulfide exchange reaction with protein sulfhydryls on cancer cell membrane. Since fibrinogen after reduction retains its antigenic epitopes, it can be easily identified by immunological methods in cancer tissue. Conversion of fibrinogen into a fibrin clot is initiated by enzyme thrombin generated as a result of blood clotting activation. Cross-linking of such a polymer by means of the action of Factor XIII renders it insoluble in chaotropic solvents. Since a fibrin-like material surrounding solid tumor cells has also been shown to be insoluble in such solvents, it was originally thought to be a result of blood clotting activation [30]. However, there is a substantial difference between fibrin-like material in tumors and that of a fibrin clot, namely that the former is resistant to proteolytic degradation, whereas stabilized fibrin is eventually decomposed by the fibrinolytic enzyme system. In addition, failure of anticoagulation to improve cancer therapy proves that the fibrin deposits in cancer are not formed by the activation of blood coagulation, but by a different mechanism. The formation of a proteolytically resistant “coat” around solid tumor cells was suggested by us to offer a protection against immune recognition and elimination [31]. This concept was based on the fact that the components of such a coat are immunologically indistinguishable from their soluble forms in blood. Consequently, NK cells consider them as “self” and thus spare tumor cells from their attack. The presence of a protective fibrin-like material around tumor cells was also demonstrated by other researchers [32]. Interestingly enough, recent findings indicate that leukemic cells can bind fibrinogen possibly by a disulfide exchange reaction, thus offering a similar protective shield [33].

Similarly to sulfur, Se can exists in various oxidation forms, specifically in association with three oxygen atoms (selenite, Se4+), or four oxygen atoms (selenate, Se6+). Also analogously to sulfur, selenium in organic forms is bivalent (e.g. selenenocysteine or selenomethionine). Metabolism of Se in vivo has been extensively studied showing a complex pathway of a conversion of inorganic into organic forms, as well as to some volatile derivatives. It has been observed that selenite, but not selenate, is readily reduced to elemental form by reducing substances such as ascorbic acid (AA) which becomes oxidized to dehydroascorbic acid. This simple chemical reaction constitutes a direct proof that sodium selenite is an oxidizing agent. Oxidizing properties of selenite can also be demonstrated in vitro by reacting with bifunctional thiol, eg. dithiothreitol (DTT), but not a monothiol such as cysteine or glutathione [34]. Since the product of DDT and other polythiols oxidation is the formation of a corresponding disulfide, this explains why vicinal thiol groups are necessary for the reaction with selenite as shown in the following equation:P[SH]2+Na2SeO3+2H+=PSSP+2NaOH+Se0+2H2O

The appearance of a characteristically reddish elemental selenium is a convenient indicator of a reducing biological milieu. This specific reaction was then used by us to demonstrate the presence of reducing groups at the surface of a lung transplantable tumor, freshly excised from a nude mouse [35]. The presence of polythiols in such a tumor tissue was also confirmed by their characteristic ability to induce disulfide exchange in human serum albumin in vitro. Such a crosslinking was inhibited not only be selenite but by a number of other oxidants including dehydroascorbic acid, natural disulfide ajoene, and chlorine dioxide. Oxidative properties of selenite have also been demonstrated by other researchers in various biological systems [36], [37]. In addition, selenite-induced oxidation has been implicated in death of human prostate cancer cells [38], and was shown to cause apoptosis of human hepatoma cells [39]. Moreover, cytotoxic effect of selenite against HL-60 human leukemia cells was also shown to be associated with oxidation of thiols [40], [41].

It should be emphasized that the oxidant property of selenite is manifested only in the presence of polythiols. This fact has significant consequences for the potential use of selenite in cancer therapy. By contrast to numerous alkylating drugs that indiscriminately block protein thiol groups in cancer and normal cells, selenite reacts only with vicinal thiols and converts them to disulfides. If selenite reacts also with monothiols, particularly with those at the active site of enzymes, it would be a powerful poison.

Section snippets

Conclusions

In this article I argue that, due to prevailing hypoxia, cancer tissue becomes endowed with reducing properties that allow for the disulfide exchange between cell membrane and plasma proteins. Consequently, a proteolytically resistant, fibrin-like coat is formed around tumor cells that is immunologically indistinguishable from the native blood protein. Such a cell membrane protein coat appears as “self” to the natural killer cells, and allows tumor cells to escape immune recognition and

References (45)

  • A. Terada et al.

    Active oxygen species generation and cellular damage by additives of parenteral preparation: selenium and sulhydryl compounds

    Nutrition

    (1999)
  • P.B. Caffrey et al.

    The development of drug resistance of tumor cells in vitro is accompanied by the development of sensitivity to selenite

    Cancer Lett

    (1994)
  • S.A. Rosenberg

    Development of effective immunotherapy for the treatment of patients with cancer

    J Am Coll Surg

    (2004)
  • M.P. Rayman

    The importance of selenium to human health

    Lancet

    (2000)
  • R.A. Passwater

    Selenium as Food and Medicine

    (1980)
  • W.L. Broghamer et al.

    Relationship between serum selenium levels and patients with carcinoma

    Cancer

    (1976)
  • J. Gromadzinska et al.

    Glutathione peroxidase activity, selenium and lipid peroxides levels in blood of cancer children

    Ann Clin Res

    (1988)
  • N.V. Gopee et al.

    Sodium selenite-induced apoptosis in maurine B-lymphoma cells is associated with inhibition of protein kinase C, nuclear factor kappa-B, and inhibitor of apoptosis protein

    Toxicol Sci

    (2004)
  • E. Rudolf et al.

    Combined effect of sodium selenite and camphotecin on cervical carcinoma cells

    Neoplasma

    (2004)
  • F.G. Combs et al.

    Chemopreventive agents: Selenium

    Pharmacol Ther

    (1998)
  • K. Bahr et al.

    Selenium supplementation by selenium yeast and sodium selenite: analysis of the selenium status as well as risk deficiency and intoxication

    J Lab Med

    (1999)
  • P.D. Whanger

    Selenium in the treatment of heavy metal poisoning and chemical carcinogenesis

    J Trace Elem Electrolytes Health Dis

    (1992)
  • Cited by (0)

    In memory of Laszlo G. Egyud.

    View full text