Organic selenium compounds as potential chemotherapeutic agents for improved cancer treatment

ABSTRACT Selenium(Se)‐containing compounds have attracted a growing interest as anticancer agents over recent decades, with mounting reports demonstrating their high efficacy and selectivity against cancer cells. Typically, Se compounds exert their cytotoxic effects by acting as pro‐oxidants that alter cellular redox homeostasis. However, the precise intracellular targets, signalling pathways affected and mechanisms of cell death engaged following treatment vary with the chemical properties of the selenocompound and its metabolites, as well as the cancer model that is used. Naturally occurring organic Se compounds, besides encompassing a significant antitumor activity with an apparent ability to prevent metastasis, also seem to have fewer side effects and less systemic effects as reported for many inorganic Se compounds. On this basis, many novel organoselenium compounds have also been synthesized and examined as potential chemotherapeutic agents. This review aims to summarize the most well studied natural and synthetic organoselenium compounds and provide the most recent developments in our understanding of the molecular mechanisms that underlie their potential anticancer effects. Graphical abstract Figure. No Caption available. HighlightsSelenium compounds have gained substantial interest as anticancer agents.The review covers naturally occurring and synthetic organoselenium compounds.Their chemical properties and main mechanisms of action are discussed herein.


Abstract
Selenium(Se)-containing compounds have attracted a growing interest as anticancer agents over recent decades, with mounting reports demonstrating their high efficacy and selectivity against cancer cells. Typically, Se compounds exert their cytotoxic effects by acting as pro-oxidants that alter cellular redox homeostasis. However, the precise intracellular targets, signaling pathways affected and mechanisms of cell death engaged following treatment vary with the chemical properties of the selenocompound and its metabolites, as well as the cancer model that is used. Naturally occurring organic Se compounds, besides encompassing a significant antitumor activity with an apparent ability to prevent metastasis, also seem to have fewer side effects and less systemic effects as reported for many inorganic Se compounds. On this basis, many novel organoselenium compounds have also been synthesized and examined as potential chemotherapeutic agents. This review aims to summarize the most well studied natural and synthetic organoselenium compounds and provide the most recent developments in our understanding of the molecular mechanisms that underlie their potential anticancer effects.

Graphical abstract
Keywords: Organic selenium compounds Cancer therapy

Introduction:
The primary focus of selenium (Se) and cancer research over last four decades has been dedicated to the chemo-preventive effects of Se, wherein Se supplementation directly or indirectly exerts anti-oxidant functions thereby increasing the cellular defense against oxidative stress (2-5). However, during recent years, Se containing compounds have gained substantial interests as anti-cancer therapeutics, due to their pro-oxidant properties at higher levels, with evidences demonstrating their high efficacy and selectivity (1,6).
Chemical derivatives of Se include inorganic compounds like selenite and organic compounds like selenomethylselenocysteine and selenomethione. Currently, a plethora of research is dedicated towards understanding the anti-tumor potential of these compounds (7). Generally, the inorganic Se compounds exert higher genotoxic stress, which in turn, may account for a lower therapeutic window, higher systemic toxicity and an increased risk of metastatic burden. Contrarily, organic Se compounds retain significant anti-tumor activity along with increased ability to prevent metastasis, have fewer side effects and lower systemic effects (8,9). Organic Se compounds comprise a vast group of chemically diverse nucleophilic molecules. In order to further improve the selectivity, specificity and efficacy and, to lower the toxicity, several novel organic Se compounds have been synthesized. Taken together, due to their aforementioned characteristics, the use of organoselenium compounds has become a promising candidate in cancer therapeutics.
Rationale: Selenium as cancer therapeutic agent All aggressive neoplastic cells exhibit a common phenotype of uncontrolled growth that might potentially metastasize leading to lethal malignancy. In order to support this relentless cell division and proliferation, cancer cells undergo aberrant metabolic adaptation that promotes their survival. Although the cellular transformation mechanisms during carcinogenesis in different cell types can occur via various pathways, the metabolic requirements of resulting cancer cells are generally very similar (10).
Equally important to the energy adaptation, acquired by the cancer cells, is their prominent feature to counteract the increased oxidized environment, which in turn is a result of an aberrant metabolism. In order to evade the ROS induced death, the cancer cells up-regulate their antioxidant scavenging molecules like glutathione peroxidase (GPx), glutathione-Stransferase (GST), glutaredoxin (Grx), thioredoxin (Trx), superoxide dismutase (SOD) and catalase (CAT) (10, 11). Several of these antioxidants rely on the reducing power of NADPH, which is also often found to be induced in cancer cells and is a key metabolite of the pentose phosphate pathway, to maintain their activity to scavenge ROS and repair ROS-induced damage. This leaves the cancer cells with very little or no room to further adapt to any additional induction of ROS, and hence, even a slight additional ROS induction, would lead to oxidative stress dependent cell death (10, 12). Contrarily, normal healthy cells generally possess relatively low steady state levels of ROS and reducing equivalents. In case of any anomaly, the elevated ROS levels in normal non-cancerous cells can therefore be counteracted via induction of the antioxidant systems (13,14). Hence, inducing ROS or targeting the strained antioxidant defense system to reach a threshold that is incompatible with cell viability may selectively kill the cancer cells without affecting the normal cells (12,15,16). Consequently, inducing oxidative stress in cancer cells using redox modulators provides an interesting therapeutic window and has been established as a mean of successful anti-cancer therapy (17).
Se compounds have been well documented to act as redox modulators and have shown higher selectivity and sensitivity in malignant cells. (1,6,17,18). Hence, with respect to induction of oxidative stress using pro-oxidants as a cancer therapeutic strategy, Se compounds seem to be promising candidates. The pro-oxidant properties of Se compounds, responsible for their cytotoxicity and anti-cancer activity, can be broadly attributed to ROS generation, oxidation of protein thiols, direct or indirect DNA binding as detailed below

Specificity of Se compounds towards cancer cells
Apart from ROS generation, the specificity of Se compounds is conferred by their selective uptake, localization and accumulation in cancer cells. The use of 75 Se-sodium selenite and 75 Se-selenomethionine (SeMet) as radiotrackers in diagnosis of tumors in humans' dates back to the 1960's. It was shown that 75 Se displayed high accuracy in localizing intracranial tumors as well as thoracic and abdominal neoplasms (48-50). However, the uptake mechanism varies between different Se compounds, and is not clearly elucidated. Due to their high reactivity, deducing the exact transport mechanism for each Se compound remains a challenge. One of the important determining factors of Se uptake is the intra-as well as extracellular redox status. It was well demonstrated that the addition of reduced extracellular thiols facilitated selenite uptake in the keratinocyte model-HaCaT cells. The authors showed that reduction of selenite with dithiothreitol (DTT) to, presumably, selenide resulted in a significant stimulation of Se uptake and that this uptake was mediated by a membrane associated ATP-dependent transporter (51). Later studies, in a variety of cell lines, confirmed that a strong reductive microenvironment, achieved by production of extracellular thiols by cells is required for an efficient selenite uptake (52). It was demonstrated that the overexpression of the cysteine/glutamate antiporter xCT in cancer cells generated a more reducing extracellular microenvironment, leading to the reduction of selenite to selenide, thus facilitating Se uptake. Moreover, addition of an xCT inhibitor (monosodium glutamate) and a thiol scavenger (DTNB) decreased the cytotoxicity of selenodiglutathione and selenocystine in non-small cell lung carcinoma, suggesting that reduction of these compounds is required for their uptake (52). Additionally, the selective uptake of Se by cancer cells has been attributed to their aberrant metabolism, which requires an increased amount of Se, generally being rate limiting for its incorporation into central selenoproteins such as GPx and thioredoxinreductase (TrxR), which are key antioxidants enzymes in upholding the redox homeostasis (53).

Naturally Occurring Organoselenium Compounds
In recent decades, the anticancer properties of organoselenium compounds have been thoroughly investigated. Naturally occurring organoselenium compounds comprise a large group of chemically diverse molecules, which are primarily acquired through the diet and, in general, can be classified as selenides/diselenides, selenoamino acid derivatives, seleninic acids or metabolites of inorganic Se (Figure 2). While the specific metabolic pathway engaged differs from one chemical species to another, most naturally occurring organoselenium compounds typically undergo enzymatic transformation into intermediary methylselenol (CH3Se -), prior to a demethylation reaction into biologically available selenide (HSe -). Our understanding of the metabolism and active forms, as well as the anticancer activity of several promising natural organoselenium compounds are discussed below.

Se-methylselenocysteine (MSC)
Se-methylselenocysteine (MSC, 1) is a monomethylated seleno-aminoacid derivative that is converted into CH3Se -by selenocysteine Se-conjugated β-lyases (54). Given the high volatility and reactivity of CH3Se -, it has not been possible to directly identify and measure the metabolism of MSC into CH3Se -in cell or animal models; however, β-lyase activity has been indirectly monitored in rat liver supernatant using labelled Se-containing selenoamino acids (55). In the same study, CH3Se -was shown to be demethylated via a hydrolysis reaction into HSe -. The in vitro cytotoxicity of MSC has been determined in the micromolar range for various human cancer cell lines including colon, breast, lung and oral squamous cells (56, 57). Additionally, MSC treatment has also been shown to reduce vascular endothelial growth factor (VEGF) expression in vitro (56) . Moreover, investigations into the mechanisms of cell death associated with MSC treatment have supported a role for caspasedependent apoptosis, while the involvement and extent of mitochondrial signaling in this process still remains unclear (58-60). Another important consideration for the use of MSC is that its efficacy is completely dependent upon the production of the active metabolite, CH3Se -, by β-lyase activity, which varies across cells, tissues and organs (61). This could partly explain, why the promising findings in in vivo studies, discussed below, are not as apparent in in vitro studies.
The in vivo anticancer potential of MSC has been investigated primarily in combination therapy with traditional chemotherapy drugs. In particular, MSC has gained attention for its capacity to modulate a wide range of cellular and physiological processes that underpin metastasis, as well as for its potential to act synergistically with a wide range of anticancer compounds. In mice bearing FaDu human head and neck squamous cell carcinoma xenografts, treatment with MSC resulted in a series of promising antiangiogenic effects, ultimately resulting in tumor growth inhibition and enhanced doxorubicin delivery associated with decreased microvessel density, increased vascular maturation, improved vessel functionality and reduced vascular permeability (62). Importantly, these outcomes were more pronounced in a histologically identifiable subset of tumors that displayed fewer or no hypoxic regions and were uniform in structure. Similarly, in a mouse xenograft model of MCF-7 breast cancer, a synergistic effect of MSC and tamoxifen was observed on tumor growth inhibition (63). Likewise, in mouse xenograft models of colon carcinoma and squamous cell carcinoma of the head and neck, a synergistic effect of MSC and irinotecan was observed; moreover, this chemotherapeutic synergy effects was evident in both sensitive (HCT-8 and FaDu) and drug-resistant (HT-29 and A253) xenografts (64). Perhaps even more promisingly, not only has MSC been shown to synergistically enhance the efficacy of chemotherapeutic drugs, it has also displayed a capacity to protect against their toxicity in animal models. Specifically, these effects have been observed in combination therapy of MSC with the aforementioned tamoxifen and irinotecan, as well as with cisplatin and oxaliplatin (63-66).

Selenocystine (SeCys)
Selenocystine (SeCys, 2) is a diselenide oxidation product of the amino acid selenocysteine (Sec). It is highly redox active, being efficiently reduced to Sec by both low-molecular weight thiols and disulfide reductases. In turn, Sec, given its low pKa value, is highly nucleophilic, rendering it highly unstable and reactive. The capacity of SeCys to modulate redox homeostasis has made it an interesting candidate as a pro-oxidant for cancer therapy.
The cytotoxic effects of SeCys have been determined in the low to medium micromolar range across lung, breast, cervical and liver cancer cells, as well as in melanoma cells (67-71). Moreover, across a broad panel of cancer cells, SeCys has been repeatedly shown to act by increasing ROS production and, subsequently, inducing DNA damage, leading to mitochondrial-mediated apoptosis associated with p53 phosphorylation. These effects have also been observed by combining SeCys with 5-FU treatment, additionally highlighting a role for ERK and AKT inactivation (72). Perhaps unsurprisingly, SeCys has been shown to potentiate the cytotoxicity of auranofin, a potent TrxR inhibitor, and TrxR knockdown primarily by enhancing intracellular ROS accumulation (70, 73, 74). Hence, SeCys can enhance oxidative stress associated with the inhibition or consumption of intracellular antioxidants.
In addition to the accumulation of intracellular ROS, other pathways are affected by SeCys treatment in cancer cells. For example, apoptotic cell death via the extrinsic/death-receptor pathway has been observed in melanoma cells following treatment with SeCys (67). More recently, SeCys treatment in cervical cancer cells has been associated with two different types of cell death, depending on the stage of the cell cycle during treatment. Both apoptotic cell death and a paraptotic-like cell death associated with ER stress and induction of the unfolded protein response (UPR) were observed (71). Interestingly, in triple-negative breast cancer cells, apoptotic cell death has been linked more specifically with S-phase cell cycle arrest (75). Hence, the mechanism of cell death associated with SeCys appears to vary across cell types.

Selenomethionine (SeMet)
Selenomethionine (SeMet, 3) is a naturally occurring amino acid that contains a sulfur (S)to-Se modification. In humans, SeMet is usually incorporated into proteins non-specifically in place of methionine (Met). Although the anticancer mechanisms of action of SeMet have been explored, there is limited work examining its anticancer potential. SeMet has displayed notable cytotoxicity toward several cancer cells, including lung, breast, colorectal, prostate and melanoma, among others (56, 77-79). While these cytotoxic effects have been achieved at higher concentrations (medium to high micromolar range) than those observed for commonly studied redox-active Se compounds, a strong selectivity toward cancer cells over normal cells has still been realized (78). However, the redox potential of SeMet in cancer therapy has been brought into question since, in rat liver lysates, neither γ-lyase cleavage of SeMet to CH3Se − nor its subsequent demethylation to HSewas detected (55).
Nevertheless, several other pathways have been implicated in SeMet-induced apoptosis, including the activation of caspases and p53, ER stress, HDAC inhibition, altered expression of Bcl-xl, Bax, Bad and Bim, decreased cyclooxygenase-2 expression and decreased glutathione peroxidase activity (56, 77, 80, 81). Moreover, SeMet used in combination with ionizing radiation enhanced treatment selectivity toward NCI-H460 and H1299 lung cancer cells, while having little effect on WI-38 human diploid lung fibroblasts (82). Importantly, in a mouse xenograft model of colorectal carcinoma (SW480), low systemic toxicity and high tumor selectivity has been reported (81).

Selenodiglutathione/g-glutamyl-selenomethylselenocysteine (SDG)
Selenodiglutathione (SDG, 4) is the primary metabolite produced during the reduction of selenite by glutathione. In this reaction, rather than the formation of oxidized glutathione (GSSG), a Se metabolite conjugated with two glutathione moieties, SDG (GS-Se-SG), is produced. Like inorganic Se forms such as selenate and selenite, SDG is reduced to HSeby GSH and by activity of the Grx and Trx systems (83, 84). Importantly, SDG has chemical properties that are very different from those of GSSG, which is not a substrate for mammalian TrxR. Moreover, the reduction of SDG by Trx is greatly altered compared to that of GSSG. SDG has exhibited cytotoxicity typically in the low micromolar range toward a panel of different cancer cells, including breast, ovarian, cervical, lymphoma, promyelotic leukemia and oral squamous (71, 85-90). In general, it has been determined as either an equally potent or more potent anticancer compound than selenite and has displayed selectively higher activity toward cancer cells than non-transformed cells. Given the similarities between the metabolism of SDG and selenite, it has long been thought that the adverse side effects associated with selenite treatment would also be associated with SDG treatment. However, more recently it has been demonstrated that despite both SDG and selenite being metabolized to HSe -, their mechanisms of action are notably different.
Specifically, SDG was shown to glutathionylate free protein thiols, which accounted for the different intracellular targets, expression patterns and, ultimately, mechanism of cell death (apoptosis-like) associated with SDG treatment compared to selenite (necroptosis-like cell death) treatment (71).

Synthetic Organoselenium Compounds
Although the first report on the synthesis of an organoselenium compound, diethyl selenide, was back in 1836, the development of synthetic organoselenium compounds as antitumor agents is a relatively new and expanding field of research. In fact, until '90s, the chemistry of organoselenium compounds was scarcely developed in comparison with that of organosulfur compounds, mainly because of the high instability and toxicity of many Secontaining compounds. In the last three decades, on the contrary, advances in the field of synthesis and reactivity of organoselenium as well as the discovery of the fundamental role of Se and selenoproteins in cancer, has tremendously motivated an intense development of new organic Se-containing therapeutics. Below, the major and most widely studied are discussed, based on their chemical and anticancer properties.

Methylseleninic acid
Methylseleninic acid (MSA; CH3SeO2H), is an oxoacid Se-containing compound which has been studied since 2000 as a chemo-preventive and chemotherapeutic agents. In vitro, it was shown to be effective against a great variety of tumors, including human lung (91), prostate (92-95), breast (47, 96) and pancreatic (21), by acting as a pro-oxidant. In fact, in cells, similarly to selenite, it is readily metabolized to CH3Se − , which in turn generates superoxide upon reacting with O2, thus leading to cell dysfunction and death.
Mechanistically, MSA has been found to induce apoptotic cancer cell death in different cell lines by the activation of multiple caspases (caspase-3, -7, -8 and -9), ER stress, UPR induction, cytochrome c release and PARP cleavage (93-95, 97). In particular, MSA has been shown to induce apoptosis in both p53 wild-type, p53-mutant (94), and in p53-null cells Additionally, MSA treatment resulted in significantly less genotoxicity and higher inhibition of tumor angiogenesis compared to MSC and selenite (103) and compared to selenomethionine (104). In combination therapy, MSA was effective in enhancing paclitaxel efficacy in the treatment of triple-negative breast cancer (105) and etoposide/paclitaxel activity in prostate cancers (106).
Recently, Cai and coworkers highlighted the potential of MSA as antiangiogenic therapeutic.
In particular, they have proved that MSA inhibits angiogenesis by down-regulating the expression, and the clustering of integrin β3, which in turn, leads to the inhibition of the phosphorylation of AKT, IκBα, NFκB (107).

Selenides
Many different classes of synthetic selenides have been evaluated for their antitumor potential so far ( Figure 3). Plano and coworkers synthesized and tested a series of methylimidoselenocarbamates (5) and demonstrated their efficacy as cytotoxic agents (at micromolar doses) against human prostate, colon and breast cancer cells. From a mechanist point of view, they showed that these compounds act as multi-kinase inhibitors, being effective in blocking PI3K/AKT/mTOR and ERK1/2, and thus leading to autophagic and apoptotic cancer cell death. In addition, a quinolone-based imidoselenocarbamate derivative was promisingly found to be effective in vivo, in a prostate cancer xenograft model (108)(109)(110).
The same research group also developed a series of quinazoline and pyrido[2,3-d] pyrimidine Se compounds (6). When tested on a panel of several human cancer cell lines (i.e. lymphocytic leukemia cells as well as colon, lung and breast cancer cells), these compounds displayed different activity profiles, depending on the position of the Se moiety (111). Another interesting class of selenides, which deserves attention, is that of quininebased seleniocompounds (7). These compounds, designed by Jacob and co-workers, containing both the Se and quinone redox centers with the aim to efficiently strengthening the ROS-generating capacity of the molecule, were endowed with a noteworthy and selective antitumor activity by acting as cancer cell specific redox catalysts. Upon testing them against cells derived from the most common solid tumors (such as colorectal, lung and breast cancer cells) as well as leukemia cells, these compounds were shown to be highly cytotoxic; however the human non-cancer cells were less affected (112)(113)(114). The same authors reported similar results obtained with peptidomimetic compounds containing Se (8) (115). Mechanistically, the biological activity of these compounds was consistent with a cancer specific redox modulation and a subsequent induction of apoptotic cell death.
On these bases, with the aim of developing other novel hybrid compounds with two redox centers, some Se-containing β-lapachone derivatives (9) were recently developed by Vieira and coworkers. When evaluated against several cancer cell lines (leukemia, human colon carcinoma, prostate, human metastatic prostate, ovarian, central nervous system and breast), these compounds showed IC50 values in the low and sub micromolar range whereas they were barely effective against non-cancer cells (116).  (118). Se containing AZT derivatives have been evaluated for their potential toxicity and shown low toxicity against immune cells (119,120).

Diselenides
In addition to selenides, symmetrical diselenides ( Figure 4) were highly studied during the past years (121). Posser and coauthors showed for the first time that diphenyldiselenide (11) possess a significant cytotoxicity potential against neuroblastoma cells, and is highly effective in inducing cancer cells death through ERK1/2 mediated apoptosis (122).
Subsequently, Nedel and coworkers demonstrated that substituted diaryldiselenides (3-(trifluoromethyl)-diphenyldiselenide and 4-methoxydiphenyl diselenide) have the potential to induce apoptosis in human colon cancer cells by inducing a cell-cycle arrest and activating caspase-dependent or independent pathways (123).
Recently, De Freitas and Rocha showed that diphenyldiselenide can act as a substrate for TrxR, thus laying down the basis for the development of novel diselenides as potential anticancer agents (124). In fact, different series of diselenides containing bioactive pharmacophores (e.g., naphthalene, cyclic imides) or pharmacologically relevant heterocycles (12) (e.g., thiazolidinone, tetrazole, pyrazole and thiazolopyrimidine) were developed, some of which were shown to exhibit cytotoxicity at sub-micromolar concentrations towards different cancer cell types (125)(126)(127)(128)(129). From a mechanistic point of view, these selenocompounds induced a down regulation of Bcl-2 and Ki-67 expression level and activated caspase-8, thus leading to apoptotic cancer cell death (130).

Selenocyantes
Extensive studies have shown that selenocyanate incorporation into pharmacologically relevant scaffold allows the obtainment of molecular entities with improved pharmacological potentials as compared to the original precursor compounds.
Naturally occurring and synthetic selenocyanates ( Figure 5) have proven their effectiveness in the prevention and treatment of a variety of cancers, both in vitro and in vivo (131,132).
In particular, phenylalkyl selenocyanates (13) were shown to be effective in vitro towards melanoma, glioblastoma, sarcoma, prostate, breast, and colon cancer cell lines. Structureactivity relationships (SARs) evidenced that an increase in compound lipophilicity by increasing the alkyl chain length was consistent with an increase in the antitumor potential in vitro. In vivo, the most promising phenylalkyl selenocyanate was able to significantly decrease tumor growth in a melanoma xenograft model, without inducing any systemic toxicity sign or by affecting blood parameters indicative of liver, kidney or cardiac related toxicity. Mechanistically, it was assessed that these compounds were able to redox cycle in the presence of GSH forming nitric oxide, superoxide and other ROS, thus affecting cancer cell redox state (133). In addition, their effects on specific phase II detoxifying and antioxidant enzymes (glutathione-S-transferase, superoxide dismutase and catalase) and, in particular, on selenoenzymes TrxR and GPx have also been reported (133-135). This, in turn, was associated in cells to a caspase-dependent apoptotic cell death through the induction of p53, Bax and suppression of Bcl-2 (136).
This class of compounds was also proven to be extremely promising in terms of combination treatments. The diphenylmethylselenocyanate and a naphthalimide based selenocyanate compound 2-[5-selenocyanato-pentyl]-7-amino benzo[de]isoquinoline-1,3dione (14) were found extremely effective not only in significantly potentiating the therapeutic efficacy of classical chemotherapeutic drugs, such as cisplatin and cyclophosphamide, but also were very promising in reducing the chemotherapy-induced adverse effects and toxicity of these drugs (136-138).
The synthetic compound 1,4-phenylenebis(methylene)selenocyanate (p-XSC, 15) has also been proven to be effective in vitro, at low micromolar concentrations, against androgen responsive and androgen unresponsive human prostate cancers by inducing cell cycle arrest in the G1 phase and activating JNK, p38, ERKs 1&2 and Akt cancer cell signaling (139-141), and by inhibiting the m-TOR downstream effectors (142).
As for selenides, some multi-target selenocyanate compounds were recently developed. By combining an indoleheterocycle microtubule inhibitor with a selenocyanate moiety, Krishnegowda and coworkers developed a series of selenocyanate derivatives (16) endowed with a very promising in vitro antitumor activity (in the low micromolar range) against colon, breast and lung cancers and against melanoma. The conjugation of the indoleheterocycle and the selenocyanate moieties in the same molecular entity yield to selenoisatin-analogs was able to induce cancer cell death through a mechanism encompassing both microtubule and Akt inhibition (143).
In addition, selenocyanate containing triterpenes based on homobetulin scaffold (17)  Among sugar-based selenoesters, both xylitolselenious ester (xylitol-Se, 19) and sucrose selenious ester (sucrose-Se, 20) have been synthesized. While barely characterized, the xylitol-Se compound has shown selective cytostatic effects on human hepatoma cells, in a dose-dependent manner, by acting as GSH depleting agents (148,149). Interestingly, this cytotostatic effect was not detected in normal human hepatic cells. Similarly, the Sucrose-Se was selectivity inhibited the proliferation of several cancer cell lines in a dose-dependent manner without affecting the proliferation of non-cancer human liver cells. Additionally, in an acute toxicity test in mice, the sucrose-Se showed a very high median lethal dose compared to sodium selenite (290.0 and 13.1 ppm, respectively), thus laying down the basis for a further preclinical development of this type of selenocompounds (150).

Ethaselen
Ethaselen (1,2-[bis(1,2-benzisoselenazolone-3(2H)-ketone)]ethane, Figure 7, 21), also known as BBSKE, has been synthesized and extensively studied by Zeng and co-workers in a number of cell lines (151). They have found that it targets and inhibits thioredoxin reductase (TrxR), a central antioxidant and redox regulatory enzyme, which has been proposed as a promising target for anticancer drugs (152). Ethaselen, via its TrxR inhibitory effect, has also been shown to inhibit the proliferation of tumor cells through S phase arrest (153). More detailed studies revealed that TrxR inhibition by ethaselen occurs through targeting of the C-termina,l but not the N-terminal active site of the enzyme, and that it does not inhibit thioredoxin or glutathione reductase. The same study showed that downstream effects of ethaselen caused oxidation of Trx and increased the levels of ROS (154).
Concomitantly, in vivo experiments have also shown promising results with ethaselen, with tumor inhibitory effects in the range of 40-80% in prostate, tongue and liver cancer models (155)(156)(157). The metabolism and secretion of ethaselen has further been suggested to occur via oxidation, methylation and glucoronidation (158). Ethaselen has also been tested in vivo in several combination therapies. One study used ethaselen in combination with cisplatin

Ebselen
Ebselen (Figure 7, 23), a structurally related compound to ethaselen, has mainly been considered as a glutathione peroxidase mimic and hence, also, as a strong antioxidant. In comparison to ethaselen, ebselen is an efficient substrate for TrxR, and the reaction is known to be stimulated by Trx, leading to a rapid oxidation of Trx (167). The antioxidative properties of ebselen has been widely studied, but a few studies have also suggested that ebselen might possess antiproliferative and anticancer properties via ROS production (168). In addition, ebselen has been proposed to alter the mitochondrial function, including Bax activation (168). These anticancer properties have also been suggested to be via the inhibition of Quiescin sulfhydryl oxidase 1, an enzyme that promotes the growth and invasion of tumor cells and alters the extracellular matrix composition. In the same study, the authors described that a daily oral treatment with ebselen resulted in a 58% reduction in tumor growth in mice bearing human pancreatic tumor xenografts (169). By combining ebselen with gamma radiation, a major antitumor efficiency was accomplished, by both inducing apoptosis and inhibiting cancer progression, as well as by modulating the response of pro-and anti-inflamatory cytokines (170).

Selenium-containing 5-membered rings
Se-containing 5-membered ring compounds (selenophenes, selenazoles and selenadiazoles, see Figure 8) have been prepared and evaluated for their pharmacological potential as anticancer agents.

Selenophene
Among compounds endowed with antitumor potential, one of the most promising derivatives is selenophene, D-501036, 2,5-bis(5-hydroxymethyl-2-selenienyl)-3hydroxymethyl-N-methylpyrrole (25), which was identified as a potent antineoplastic agent with a broad spectrum of activity against human cancer cell lines. It elicited IC50 values in the low micromolar range, being highly selective against cancer cells compared to normal cells and highly effective against multi drug resistant (MDR) cancer cells.
Furthermore, D-501036 exhibited strong antitumor activity in vivo, in a mouse xenograft model of human renal carcinoma. Concerning its mechanism of action, it was found to induce cell death through a p-53 and mitochondrial mediated apoptotic pathway (174)(175)(176).
Selenopheno quinolinones and coumarins were also shown to promote cancer cell death.

Selenadiazoles
Selenadiazoles have also attracted attention in the past years owing to their potential as anticancer agents. Initially, a series of 4-methyl-1,2,3-selenadiazole-5-carboxylic acid  (188,189). Later, they also developed some 1,4-osxaselenins (31) and demonstrated their inhibitory effect on the proliferation of human ovarian and cervical cancer cells by inducing apoptosis (190).

Selenium-nonsteroidal anti-inflammatory drug (Se-NSAID).
A few years back, a study performed by Bi et al. presented evidence for a synergistic effect between a nonsteroidal anti-inflammatory drug (NSAID), sulindac, and Se (191). In this study, a non-toxic dose of sulindac was used in combination with Se, which significantly inhibited intestinal tumorigenesis, determined by reduced incidences of tumors (52%) and tumor multiplicities (80%). The results were linked to a significant induction of the expression of p27 and p53 as well as JNK1 phosphorylation. Recently, a Se-containing selenocoxib derivative (Figure 9, 32) was developed and the authors highlighted the ability of this compound to decrease melanoma cell growth by arresting cells in the G0-G1 phase of the cell cycle, promoting apoptosis and inhibiting cellular proliferation (192).
Mechanistically, the Se-containing selenocoxib derivative retains COX-2 inhibitory activity and PI3K/Akt inhibitory activity. Similarly, Plano and coworkers later incorporated Se into NSAIDs, generating Se-NSAID hybrids (33) (193). The most effective was a SeCN-aspirin analogue, which was shown to be extremely selectivity toward cancer cells and was found to cause cell cycle arrest in G1 and G2/M phases and induce apoptosis by activating caspase 3/7 and PARP cleavage in colorectal cancer cells.

Se containing histone deacetylase inhibitors
Histone deacetylases (HDACs) are involved in the epigenetic regulation of gene expression by the remodeling of chromatin through their deacetylase activity. Inhibition of HDACs has emerged as a target for specific epigenetic changes associated with cancer and other diseases. The use of HDAC Inhibitors is therefore believed to be an attractive therapeutic strategy against cancer. Se containing derivatives of the FDA approved HDAC inhibitor suberoylanilide hydroxamic acid (SAHA, 34) have been synthesized and studied in melanoma and lung cancer cell lines. Both of the reported compounds, bis(5phenylcarbamoylpentyl) diselenide and 5-phenylcarbamoylpentyl selenocyanide were able to inhibit HDAC activity and were also found to possess PI3 kinase pathway inhibitory properties. More importantly, both were significantly more effective in inducing cytotoxicity towards different cancer cell lines compared to SAHA (194)(195)(196).
HDAC inhibition has also been reported for MSC and SeMet, which are transaminase substrates of glutamine transaminase K (GTK) and L-amino acid oxidase (197). GTK and Lamino acid oxidase are able to convert MSC to the corresponding alpha-keto acid, βmethylselenopyruvate (MSP), while L-amino acid oxidase converts selenomethionine to its corresponding alpha-keto acid, alpha-keto-gamma-methylselenobutyrate (KMSB). These metabolites structurally resemble the known HDAC inhibitor, butyrate, and both MSP and KMSB have been shown to inhibited HDAC activity (43).

Selenium containing nanoparticles in the treatment of cancer
Nanoparticles (NPs), both metallic and non-metallic (polymers, oxides and semiconductors), are highly emerging and substantial alternatives for oncotherapy due to their excellent bioavailability, biological activity and low toxicity (71). Se containing nanoparticles (SeNPs), have recently gained interest as potential novel therapeutic agents, due to their high biocompatibility, stability and pronounced selectivity (119,182). During the last decade, a surplus of SeNPs have been synthesized using distinct and novel strategies for therapeutic purposes. Primarily the colloidal SeNPs can be synthesized using different precursor materials, like sodium selenite, sodium dioxide and selenicious acid, using physical, chemical or biological methods (74). Chemical synthesis involves reduction of ions, which is then followed by physical synthesis where bulk material is converted to smaller units, microwave synthesis, laser ablation and gamma irradiation (74). The

RAs/Raf/MEK/ERK pathways (84). In another study, apoptosis inducing aptamer -NAS-24
conjugated to SeNPs loaded with epirubicin significantly reduced viability in MCF-7 and C26 cells as well as reduced tumor growth in cancer-bearing mice as compared to EPI treatment alone (199). Mechanistic studies have shown that gold NPs (AuNPs) with Se ligands (AuNP/Se) significantly induced and enhanced the production of ROS leading to cellular apoptosis, which was initially a limitation for only AuNPs (200). However, the exact mechanism of action and/or synergism that Se imparts to other canonical drugs is yet to be understood and needs to be studied further, and novel more sophisticated compounds are warranted.

Conclusion
Many different classes of naturally and synthetic organoselenium compounds have been explored as antiproliferative agents, and the field is constantly emerging, with several compounds demonstrating pronounced cytotoxic activity against cancer cells compared to non-transformed ones. The current knowledge of their dose efficacy range, target cancers and mode of cell death are summarized in Table 1. Coherently, in vivo experiments clearly confirm their scarce systemic toxicity and enhanced tolerability, both when used as monotherapy and, strikingly also when administered in combination with conventional chemotherapeutic agents. In combination with other treatments, Se compounds not only increase the therapeutic potential, but they also seem to protect against unwanted side effects. With this said, more profound in vivo experiments are required for many of these compounds to elucidate their specific anticancer properties. However, what seems to be clear is that organoselenium compounds are more efficient, selective, and well-tolerated than the corresponding non-chalcogenated molecules, thus illustrating the importance of further exploiting the potential of novel orgnoselenium compounds and SeNP for the development of novel cancer-specific anticancer agents.