Foods contain various amounts and chemical forms of Se, most of which is covalently bound to carbon in organic molecules including as selenomethionine (SeMet), selenocysteine (SeCys) and Se-methylselenocysteine (SeMSC) [31,32]. These forms can each be metabolized to selenide (H₂Se), which serves as the obligate intermediate in selenoprotein synthesis (Figure 1) [33,34,35,36,37]. SeMet, unlike other forms of Se, can substitute for methionine (Met) in protein synthesis for which reason it is incorporated non-specifically into proteins [38,39]. Selenide can also be metabolized by methylation and sugar-derivation to produce a variety of excreted products (Figure 1) [36]. These include selenosugars, which are the major urinary metabolites in humans, and methylselenol (CH₃SeH), which is regarded as the major anticarcinogenic Se metabolite [40].

The Se-species that are likely to be present in cells (H₂Se, CH₃SeH and SeMet) may be biologically active, executing effects by several means: producing ROS by redox cycling of H₂Se and CH₃SeH; modification of protein-thiols by CH₃SeH; and mimicking Met in protein synthesis by SeMet (Figure 2). It has been suggested that these activities are underlying the cellular effect of high doses of Se [41,42], which includes the regulation of cell cycle and apoptosis. Although many of these biochemical pathways have been characterized in several different cell types, these molecular events may also occur in bone cells.

Only 25 selenoprotein genes have been identified in the human genome (24 in the mouse) [3,46]. These include the GPXs, TRRs, iodothyronine deiodinases (DIs), selenprotein P (SEPP1) and several less well characterized selenoproteins [3,46]. Their synthesis involves the co-translational biosynthesis of SeCys by a unique process in which the UGA codon is recoded to specify SeCys incorporation instead of translational stop [35]. Selenoprotein biosynthesis is dependent on dietary Se intake, genotype and inflammatory tone [3,35,46,47]. At limiting levels, this is thought to result in a hierarchal pattern of selenoprotein expression correlating with the relative significance of selenoproteins in cellular function [3,46]. The nutritional roles of Se are thought to be discharged by these selenoproteins most of which appear to exhibit oxidoreductase activities [3,46]. Accordingly, these selenoproteins have been shown to protect cells from ROS, reactive nitrogen species and oxidative damage [47].

Inadequate Se intake can result in increased levels of ROS/oxidative stress, particularly in individuals of low status with respect to other antioxidants (e.g., vitamins E and C) [48]. All of the enzymatically characterized seleoproteins catalyze not only redox reactions at sulfhydryl groups/disulfides, but also C–I bonds catalyzed by deiodinases or S–O bonds catalysed by MetSulfoxide Reductase B. [47,48]. An increase in oxidative stress, achieved experimentally either by elevating intracellular ROS or adding exogenous H₂O₂ at micromolar concentrations, has been shown to inhibit growth in a wide variety of mammalian cells [49]. Many studies have established that ROS-related oxidative stress induces cell cycle arrest, senescence, apoptosis and/or necrotic cell death [49,50].

ROS have been shown to be important signaling molecules at submicomolar levels, and a biphasic effect has been demonstrated on cellular proliferation with ROS—especially hydrogen peroxide and superoxide—in which low levels (usually submicromolar concentrations) induce growth but higher concentrations (usually >10–30 μM) induce apoptosis or necrosis [49,50]. For example, cell surface receptors produce ROS upon activation; this includes receptors for epidermal growth factor, platelet-derived growth factor, insulin-like growth factor, vascular endothelial growth factor and various cytokines [47,51,52]. At higher intracellular levels, ROS are damaging to organelles, particularly mitochondria [53], which may result in energy depletion, accumulation of cytotoxic mediators and cell death [53]. Thus, compromised oxidative defense may predispose tissues to damage by environmental stress factors. The maintenance of intracellular redox homeostasis is dependent on a complex web of antioxidant factors including both low molecular weight molecules (e.g., glutathione and thioredoxin) and protein antioxidants (e.g., certain selenoproteins) [53].

Under normal physiological conditions, most cells appear to require Se for normal growth because of their need of selenoproteins. For example, 50 nmol/L Se, as either selenite or SeMet, is required for the growth of human fibroblasts, lymphocytes and Chinese hamster ovary cells [54,55,56]. Selenium plays a critical role in cell cycle progression; its omission results in G2 cell cycle arrest [56]. Studies have shown that treatment with either selenite or SeMet can up-regulate the expression of several cell cycle-related genes: c-Myc, cyclin C, proliferating cell nuclear antigen, cyclin-dependent kinase (cdk)1, cdk2, cdk4, cyclin B and cyclin D2 mRNA [56]. In addition, Se increased total cellular phosphorylated proteins [56]. These observations are consistent with the finding that the up-regulation of cdk1, cdk2, cdk4, cyclin B and cyclin D2 led to the promotion of cell cycle progression, particularly G2/M transition and/or the reduction of apoptosis, in vivo and in vitro [42,57,58]. A few cases have been reported in which certain cells can survive in the absence of Se [59]. This includes most hepatocellular carcinoma cell lines [59]. We found [60] that Se-deprivation did not affect cell cycle progress and apoptosis in human colon Caco-2 cells even though GPX activity was limited to 8% of that of controls. In that model, Se treatment increased the expression of humoral defense and tumor suppressor-related genes while decreasing the expression of pro-inflammatory genes [60].

Some, but not all, studies have found immune cell proliferation to be sensitive to Se-deprivation [11,56]. For example, when derived immune T and B cells (e.g., Jurkat and HL60 cells) were cultured in a Se-deficient, serum-free medium, GPX and TRR activities were low and cells did not survive [11,56]. That Se-deprivation increased ROS production in these cells was indicated by the fact that treatment with a lipid soluble radical-scavenging antioxidant (vitamin E) restored survival without affecting selenoenzyme expression [11]. Selenium-deficient lymphocytes are less able to proliferate in response to a mitogen, but the response can be improved by supplementing with Se [61].

Adequate Se status has been shown to be essential for an optimum immune response, affecting both the innate and acquired immunity [62]. Inadequate Se intake has been associated with low titers of IgG and IgM [63]. Redox tone is known to play a role in modulating the activation of T cells to effectors [61], and the absence of selenoproteins in T cells led to decreased pools of mature T cells and defective T cell activation [64].

Only a few Studies have addressed mechanisms underlying the relationship of Se status and bone cell proliferation. However, it is likely that the cellular mechanisms described in other cell types may also occur in bone cells. Indeed, selenoproteins are expressed in osteoblastic cells, and Se supplement restores the antioxidative capacity and prevents cell damage in bone cells (e.g., BMSCs) [12,13,14,18]. For example, TRR has been found to be the responsive gene of 1-α,25(OH)₂-vitamin D₃, the hormone that stimulates bone cell growth and differentiation [65].

Normal bone remodeling appears to depend on the controlled function of ROS. Remodeling is a lifelong process involving the removal of old and damaged bone matrix by osteoclasts and replacement with new bone formed by osteoblasts [66]. Osteoclasts are bone-specific multi-nucleated cells generated from hematopoietic monocyte precursor cells [67]; this differentiation is regulated through both receptor-activation of nuclear factor κB ligand (RANKL) and macrophage-colony stimulating factor (M-CSF) [66,67]. Bone loss, regardless of cause, reflects increased function of osteoclasts relative to that of osteoblasts [66,67]. The latter bone cells generate superoxide by NADPH oxidase [68], and intracellular ROS is known to stimulate the differentiation and formation of mature osteoclasts [69]. That Se and other antioxidants may have a role in this process is suggested by the finding that Se-treatment suppressed RANKL-induced gene expression and phosphor-IκBα activtion [70,71,72] to reduce RANKL-mediated ROS generation and inhibit the signaling of osteoclast differentiation.

Adequate Se intake appears to play an essential role in osteoclast/osteoblast cell proliferation and differentiation by the regulation of ROS status. Signaling by extracellular signal-regulated kinases ERK1/2 is known to mediate the inhibitory effect of H₂O₂ on osteoblastic differentiation in rabbit BMSCs and MC3T3-E1 preosteoblastic cells [15,73].

Recent findings indicate that Se-treatment can protect BMSCs against H₂O₂-induced inhibition of osteoblastic differentiation by inhibiting oxidative stress and ERK activation [74]. Treatment with nanomolar amounts of Se, which increased GSH concentraitions and GPX1 expression, also enhanced the expression of type I collagen and alkaline phosphatase, and the deposition of calcium in rat marrow stromal cells (MSCs) [74]. These observations suggest that, at the cellular level, Se was capable of enhancing osteoblastic differentiation of MSCs by reducing basal oxidative stress [74]. In contrast, when selenoprotein expression was decreased due to an inadequate Se intake, ROS levels and phosphorylation status remained elevated and contributed to pathologically exacerbated signaling and enhanced osteoclast activity [13]. Because osteoblasts express several selenoproteins (e.g., GPX, SelP) at high levels it would follow that adequate Se intake is necessary to support this system of osteoblast antioxidative defense that may be relevant for the protection against ROS produced by osteoclasts during bone remodeling [13].

Cells treated with high levels of Se typically show alterations in proliferation and differentiation. While these effects have been found to be strongest in cancer cells (and would appear to contribute to the anticarcinogenic activity of Se), they also occur in noncancerous cells [75]. Cells treated with high levels of selenite typically show arrested at the S/G2-M phases with an increase in cdk2 kinase activity and DNA damage-inducible gadd gene [76,77,78,79,80]. In contrast, the metabolite methylselenol (CH₃SeH) and its metabolic precursors have been found to induce caspase-mediated apoptosis in those cells [81,82]. Methylselenol has also been shown to activate the caspase cascade and apoptosis [76,77,83]. Submicromolar levels of CH₃SeH have been found to cause cell cycle arrest, leading to increases in cells in the G1 and G2 phases with concomitant reductions of cells in the S-phase [75,76,77,84]. A metabolic precursor of CH₃SeH, methylselenocyanate (SeMSC), has been shown to exert moderate anti-proliferative effects through G1 cell cycle arrest; whereas, selenite rapidly blocked DNA synthesis and arrested cells in the S phase [76,77,84]. Cells exposed to methylselenocysteine (CH₃SeCys) were arrested at the G1 phase of the cell cycle as there was a decrease in the cdk2 kinase activity accompanied by a decrease in cyclin E-cdk2 content [78]. In addition, CH₃SeH exposure of noncancerous NCM460 colon cells inhibited cell growth and led to increases in cells in both the G1 and G2 phases with a concomitant reduction of cells in S-phase and an induction of apoptosis [75]. These responses were associated with reduced phosphorylation of ERK1/2, p38 mitogen-activated protein kinase (MAPK) and cellular myelocytomatosis oncogene (c-Myc) and up-regulation of phosphorylation of sarcoma and focal adhesion kinase survival signals [75]. In general, the doses of Se of these in vitro findings may be higher than that of the in vivo findings (human studies) [75,85,86,87]. However, the underlying biochemical mechanisms are likely to be the same.

High-dose Se treatment appears to enhance the capacity of lymphocytes to respond to mitogen or alloantigen stimulation, to proliferate, and to differentiate into cytotoxic effector cells. Humans receiving a supranutritional Se dose (200 μg/day) showed increased potential for cytotoxic T lymphocyte (CTL)-driven tumor lysis, mitogen-induced lymphocyte proliferation, and mixed lymphocyte reaction proliferation compared with a placebo group [85]. In both animals and humans, CD4⁺ T cells play a crucial in initiating immune responses, and antigen-specific CD4⁺ T cell proliferation and activation have been found to be increased in response to treatment with supranutritional amounts of Se [86]. Also, Se appears to abrogate the age-related deficiency of lymphocytes to respond to stimulation by proliferation and differentiation into cytotoxic effector cells [85]. However, the molecular mechanisms mediating these effects remain largely unexplored.

Because bone contains appreciable amounts of Se [87], it likely that Se plays a role in bone health. The hematopoietic system has been implicated as a primary target of Se at high doses [88], and osteoclasts are derived from hematopoietic progenitor cells of the monocyte-macrophage lineage [43]. Accordingly, selenite-treatment of osteoclast-like cells (e.g., RAW 264.7 cells) induced apoptosis as revealed by morphological changes, internucleosomal DNA fragmentation, activation of caspase-3, and generated the superoxide anion [28]. These findings indicate that selenite can induce apoptosis through the mitochondrial pathway in mature osteoclasts.

Osteoclasts are known to be activated by inflammatory cytokines released at low levels by osteoblasts [30]. That Se can alleviate the NF-κB dependent regulation of the inflammatory response suggests that it may have a role in mediating the osteoblast-ostoclast crosstalk [44,45]. Recent data have shown that supplementation of osteoblasts with MSeA reduced the activation of NF-κB, leading to decrease in interleukin (IL)-6, monocyte chemoattractant protein-1 (MCP-1), cyclooxygenase-2 (COX-2) and inducible nitric oxide synthase (iNOS) [30]. Thus, it is likely that Se, at least at high doses, can prevent bone resorption through the inactivation of osteoclasts. Consistent with this observation, a recent study demonstrated that Se status was inversely related to bone turnover and positively correlated with BMD in healthy euthyroid postmenopausal women independent of thyroid status [8]. However, the other study indicated that elevated Se intake negatively affected bone mass measurements in postmenopausal women over the age of 51 but only if calcium intake was less than 800 mg/day [89]. Therefore, more future human studies are needed to determine the beneficial effect of Se on bone health.