Interactions between alpha-tocopherol, polyunsaturated fatty acids, and lipoxygenases during embryogenesis

Abstract

α-Tocopherol is a lipid-soluble antioxidant that is specifically required for reproduction and embryogenesis. However, since its discovery, α-tocopherol’s specific biologic functions, other than as an antioxidant, and the mechanism(s) mediating its requirement for embryogenesis, remain unknown. As an antioxidant, α-tocopherol protects polyunsaturated fatty acids (PUFAs) from lipid peroxidation. α-Tocopherol is likely required during embryonic development to protect PUFAs that are crucial to development, specifically arachidonic (ARA) and docosahexaenoic (DHA) acids. Additionally, ARA and DHA are metabolized to bioactive lipid mediators via lipoxygenase enzymes and α-tocopherol may directly protect, or it may mediate the production and/or actions of these lipid mediators. In this review, we discuss how α-tocopherol 1) prevents the nonspecific, radical-mediated peroxidation of PUFAs, 2) functions within a greater antioxidant network to modulate the production and/or function of lipid mediators derived from 12- and 12/15-lipoxygenase and 3) modulates 5-lipoxygenase activity. The application and implication of such interactions with be discussed in the context α-tocopherol requirements during embryogenesis.

α-Tocopherol and Lipid Peroxidation

α-Tocopherol, a lipid-soluble antioxidant, is one of the eight vitamin E forms synthesized by plants [1], and is the only form that meets human vitamin E requirements [2]. α-Tocopherol scavenges peroxyl radicals during the propagation of lipid peroxidation (FIGURE 1), and is termed a chain-breaking antioxidant because it prevents the chain reaction of lipid peroxidation, but it does not prevent the formation of the initial lipid peroxyl radical [3]. α-Tocopherol is particularly enriched in neuronal tissue, especially the brain, where it is tenaciously retained during inadequate vitamin E intake even after the peripheral tissues become α-tocopherol-depleted [4]. Overt vitamin E deficiency occurs rarely in humans, but does occur in patients with fat malabsorption syndromes or genetic defects in the hepatic α-tocopherol transfer protein (α-TTP) [5] and in severe malnutrition [6, 7]. Humans with vitamin E deficiency present initially with a mild sensory neuropathy, which leads to a progressive, peripheral neuropathy caused by a dying back of large-caliber, sensory neurons, that advances to a spinocerebellar ataxia and ultimately death [5].

Figure 1. Lipid peroxidation and the antioxidant network.

During the propagation stage of lipid peroxidation, a carbon-centered radical (R·) abstracts an allylic hydrogen from a neighboring, unsaturated fatty acid. Molecular oxygen reacts with the fatty acid radical, generating a peroxyl radical, which can be reduced by α-tocopherol, creating a hydroperoxy fatty acid and the α-tocopheryl radical. The hydroperoxy fatty acids are converted to hydroxy fatty acids via phospholipid hydroperoxide glutathione peroxidase (GPx4), using two glutathiones (GSH) as the reducing agents and creating oxidized gluthatione disulfide (GSSG). To replenish the α-tocopherol, the α-tocopheryl radical is reduced by ascorbate. The oxidized ascorbyl radical is subsequently reduced back to ascorbate via GSH.

α-Tocopherol protects cellular membranes from lipid peroxidation in association with a larger antioxidant network (FIGURE 1). Once α-tocopherol reduces lipid peroxyl radicals to lipid hydroperoxides, the selenium-dependent enzyme, phospholipid hydroperoxide glutathione peroxidase (GPx4), converts the hydroperoxides to the less toxic lipid hydroxides at the expense of glutathione. Ascorbate (vitamin C) reduces the α-tocopherol radical, regenerating active α-tocopherol [8]. Subsequently, ascorbate is regenerated at the expense of glutathione. Maintenance of this antioxidant network is crucial to protect cellular membranes against radical-mediated degradation [9]. For example, vitamin E disappears faster from plasma in individuals who smoke, but vitamin C supplementation corrects the rapid α-tocopherol disappearance [10, 11]. In adult zebrafish, chronic vitamin E deficiency causes a secondary depletion of vitamin C, and concomitantly, a severe degeneration of skeletal muscle [12].

α-Tocopherol protects polyunsaturated fatty acids (PUFAs), notably arachidonic acid (ARA, 20:4 ω-6) and docosahexaenoic acid (DHA, 22:6 ω-3). Indeed, human α-tocopherol requirements increase in parallel with dietary PUFA consumption or with an increasing index of fatty acid unsaturation [13]. α-Tocopherol is postulated to co-localize with PUFA-enriched phospholipid domains of the cell membrane [14]. Moreover, a peroxyl radical generated from the PUFA will localize to the air-water interface, where the hydroxyl group of α-tocopherol intercepts the radical and reduces it [15]. Studies in animals fed experimental diets have demonstrated the importance of α-tocopherol in protecting PUFAs. For example, when zebrafish are fed diets that require them to synthesize ARA and DHA from their respective precursors, the vitamin E deficient fish have decreased percentages of total ω-6 and ω-3 PUFAs compared with those fed a vitamin E-sufficient diets [16], suggesting that vitamin E protects these long chain PUFAs. Similarly, feeding fish oil to pregnant rats decreased fetal brain α-tocopherol concentrations [17].

Is vitamin E deficiency a significant cause of spontaneous embryonic death?

In 1922, α-tocopherol was discovered because vitamin E deficient, pregnant rats fed rancid fat failed to carry their offspring to term [18]. α-Tocopherol and the α-tocopherol transfer protein (α-TTP) have critical roles in embryonic development [19]. α-TTP is expressed in the human yolk sac [20]; therefore, we used zebrafish (Danio rerio) embryos, which abundantly express α-TTP by 48 hour post-fertilization (hpf) and up-regulate α-TTP in response to oxidative stress [21]. Remarkably, adult α-tocopherol deficient zebrafish could spawn and produce viable eggs, but within days the embryos suffered developmental impairment and increased mortality [22]. Similar findings have been reported for α-TTP knockout mice [23]. Importantly, we found that the zebrafish embryonic brain accumulates α-tocopherol and expresses α-TTP. α-TTP knockdown caused head and eye malformations prior to 15 hpf [19]. Intriguingly, the α-tocopherol requirement for neurologic development in zebrafish coincides with increased synthesis of highly peroxidizable lipids by the zebrafish embryo, evidenced by increased Elovl4 [24] and Elovl5 [25] expression in the head/brain region.

Our knowledge of how vitamin E is delivered to the brain is very limited. α-Tocopherol is transported in the circulation by all lipoproteins and is delivered by pathways that deliver other lipids to cells [26]. α-Tocopherol is readily exchanged between HDL and apoB-containing lipoproteins [27]. HDL, however, is likely more important in brain development, because the central nervous system (CNS) does not contain apoB, but rather large apoE particles serve this function [28]. Scavenger receptor class B, type I (SR-BI) facilitates selective uptake of HDL-associated α-tocopherol by the blood brain barrier in vitro [29]. In vivo the cerebrospinal fluid (CSF) only contains HDL [30] and CSF-HDL contain α-tocopherol [31].

The importance of both HDL and apoB-containing lipoproteins is illustrated in patients with the autosomal recessive disorder, abetalipoproteinemia. These patients have extraordinarily low circulating lipids because they have only high density lipoproteins (HDL) and no lipoproteins containing apolipoprotein B (apoB; e.g. chylomicrons, very low density (VLDL), or low density (LDL)) [32]. Vitamin E deficiency in humans was first described in abetalipoproteinemic patients [33]. When these patients were supplemented with large vitamin E doses (150 mg/kg body weight), they did not experience the neurologic defects seen in unsupplemented patients; moreover, they were able to bear normal children [34, 35].

Abetalipoproteinemia is distinct from homozygous hypobetalipoproteinemia [36]. The latter is caused by a defect in the apoB gene [36], while abetalipoproteinemia is caused by a defect in the gene for the microsomal triglyceride transfer protein (MTP) [37]. MTP lipidates apoB and is critical in humans for normal secretion of chylomicrons and VLDL [38]. Importantly, both cause the virtual absence of circulating apoB-containing lipoproteins, which then results in vitamin E deficiency. Knockdown of either apoB [39, 40] or MTP [41] is embryonically lethal in mice.

Importantly, mouse apoB or MTP knockout embryos had neural tube defects, which were proposed to result from insufficient α-tocopherol [39–41], while mice expressing either apoB100 or apoB48 had normal embryos and did not have vitamin E deficiency [42]. Similarly, α-TTP knockout in mice was also embryonically lethal, but the pregnancy could be rescued by feeding the mother large amounts of vitamin E or synthetic antioxidant (presumably salvaging the minor amounts of remaining tissue α-tocopherol fed) [23]. Thus, there is a close relationship between α-tocopherol and normal embryonic development.

Human vitamin E deficiency caused by inadequate diets is largely assumed to be non-existent in the US. The USDA’s 2010 Dietary Guidelines (http://www.cnpp.usda.gov/dietaryguidelines.htm) gave short shrift to the observation that 96% of women do not consume sufficient α-tocopherol to meet the estimated average requirement (EAR) [43]. The expectation from the nutrition community is that requirements are too high and that the amount of α-tocopherol needed by humans to prevent deficiency is very low, thus vitamin E deficiency as a result of inadequate intake in humans is unlikely. However, our studies using the embryonic zebrafish model shows that the embryo with inadequate α-tocopherol undergoes severe developmental defects [22], and the α-TTP knockdown causes even more severe defects, especially impaired head/brain and eye formation [22]. Likely, a vitamin E continuum exists, where sufficient α-tocopherol at critical time points allows the embryo to progress to the next developmental stage.

Our discoveries in zebrafish embryos also suggest that α-tocopherol is critical in humans for embryonic development prior to when a woman knows she is pregnant (e.g. hours post fertilization (hpf) for zebrafish [44], and days for rats [45] or humans [46–48]). This early requirement for vitamin E is analogous to the situation in humans where there is an early requirement for folic acid to prevent neural tube defects. Surprisingly, folic acid supplements were not as effective in preventing neural tube defects as were the combination of folic acid and multivitamin, as a review of five human trials showed [49]. Specifically, the Hungarian trial to evaluate neural tube defects contained 15 mg vitamin E and various other vitamins [50]. Hypothetically, some of the neural tube defects observed in humans are due to inadequate α-tocopherol status. Studies in China suggest, based on maternal and cord blood vitamin E levels, that higher infant vitamin E status at birth improves cognitive function assessed at age 2 y [51], again emphasizing the importance of good vitamin E status during pregnancy for brain development.

PUFAs and Lipid mediators derived from PUFAs

Since its discovery more than 90 years ago [18], α-tocopherol’s specific in vivo biologic functions, and the mechanism(s) mediating its requirement, remain unknown. The unclear requirement of α-tocopherol for reproduction exemplifies our present gap in knowledge. It is possible that α-tocopherol is required during embryonic development specifically to protect ARA and DHA, or mediate the production and/or actions of bioactive lipid mediators derived from these PUFAs.

ARA and DHA are required for proper embryonic development. DHA is highly enriched in the central nervous system (CNS), comprising upwards of 50% of CNS PUFA [52]. DHA deficiency during pregnancy distinctly and adversely affects neurodevelopment. For example, DHA deficiency inhibits fetal neurogenesis [53, 54] and synaptogenesis [53], alters the synaptic proteome [55], serotoninergic neurotransmission [56], dopaminergic regulatory protein composition [57], neuronal phospholipid composition [58] and signaling [59], and impairs neuronal migration [60]. Adverse developmental outcomes caused by DHA inadequacy persist even after repletion with DHA [61, 62], demonstrating long-lasting effects of embryonic DHA deficiency regardless of later restitution with an adequate diet.

ARA is the most abundant 3–6 neuronal fatty acid throughout gestation and postnatal development [63]. An appropriate balance between DHA and ARA is required during neonatal development, as infant formula supplemented with DHA, but lacking ARA, impaired infant growth [64]. Indeed, higher infant ARA concentrations are positively correlated with infant birth weight and length [65]. Conversely, ARA inadequacy is associated with delayed postnatal development and reduced growth [17, 66].

Production of bioactive lipid mediators derived from ARA and DHA can occur through enzymatic peroxidation or non-enzymatic radical-mediated peroxidation. The three known enzymatic pathways that act upon PUFAs include the cyclooxygenase (COX; two major isoforms: COX-1 and COX-2), lipoxygenase (LOX; four major isoforms: 5-LOX, platelet-type 12-LOX (12-LOX), leukocyte-type 12-LOX (12/15-LOX), and 15-LOX [67]), and cytochrome P450 (CYP450) pathways. The oxidation of ARA gives rise to hydroxyeicosatetraenoic acids (HETEs) and the eicosanoids, a class of lipids, which encompasses the prostaglandins, prostacyclins, thromboxanes, leuokotrienes, lipoxins, and isoprostanes (FIGURE 2). Similarly, the oxidation of DHA gives rise to the docosanoids, which include the resolvins (D-series), neuroprotectins, and maresins, as well as intermediary monohydroxy lipids termed HDHAs. The synthesis of docosanoids requires the complex coordination of multiple enzymes, including LOXs and COXs [68].

Figure 2.

Lipid mediators derived from arachidonic acid.

Arachidonic acid (ARA) is cleaved from membrane phospholipids by phospholipase A2. The lipoxygenase (LOX) pathways convert ARA to 5-hydroperoxyeicosatetraenoic acid (5-HpETE) via 5-LOX, 12-hydroperoxyeicosatetraenoic acid (12-HpETE) via 12-LOX, or 15-hydroperoxyeicosatetraenoic acid (15-HpETE) via 15-LOX (not pictured; LOXs can also oxidize membrane ARA without prior release by phospholipase A2 [110]). 12/15-LOX can generate 12- or 15-HpETE (not pictured). The HpETEs are reduced by a peroxidase to produce 5-HETE and 12-HETE. 5-LOX converts the peroxide of 5-HpETE to an epoxide, generating leukotriene A4 (LTA4). LTA4 is the precursor for leukotriene B4 (LTB4), lipoxins, and cysteinyl leukotrienes (CysLTs). The cyclooxygenase (COX) enzymes, which are either inducible (COX-2) or constitutively expressed (COX-1), convert ARA to prostaglandin G2 (PGG2), which is a precursor for other prostaglandins, prostacyclins, and thromboxanes. Non-enzymatic, radical-mediated peroxidation of ARA yields the isoprostane class of signaling molecules (8-F₂-isoprostane pictured). Finally, cytochrome P450s (CYP450) convert ARA to 20-HETE or the epoxyeicosatrienoic acids (-EETs; 8,9-EET pictured).

An emerging body of evidence indicates that oxidation of ARA (and potentially DHA) is needed during embryogenesis. For example, knockdown of COX-1 in zebrafish embryos led to gastrulation arrest [69], while inhibition of COX-1 after completion of gastrulation caused defective vascular tube formation [70]. In mice, COX-2 regulates ovulation and embryonic implantation [71], however no effect of COX-2 knockdown was noted in zebrafish embryos [69, 70]. In zebrafish embryos, inhibition of five lipoxygenase-activating protein (FLAP), a membrane protein required for 5-LOX function, resulted in pericardial edema and reduced intersegmental vasculature and vessel/axial blood flow [72]. Neuronal growth-cone collapse requires functional 12/15-LOX [73, 74], as does hematopoietic stem cell function [75] and epidermal barrier formation [76]. Knockdown of 12/15-LOX in zebrafish embryos diminished the production of ARA and DHA lipid mediators, altered PUFA profiles, and caused abnormal brain, eye, and tail development by 24 hpf [76, 77]. Interestingly, studies in 12/15-LOX knockout mice are somewhat different than the findings generated in zebrafish. In one study, deletion of 12/15-LOX altered bone mass in developing mice [78]; however, in another study 12/15-LOX deficient mice had no gross morphologic defects and were born at the expected Mendelian ratios [79], suggesting that 12/15-LOX is not needed for mouse embryonic development. Similarly, deletion of platelet 12-LOX in mice resulted in expected Mendelian ratios in offspring [80], also suggesting that platelet 12-LOX is also not needed for mouse embryonic development. However, platelet 12-LOX knockout 6–10 wk old mice have abnormal epidermal permeability barrier function [81].

α-Tocopherol modulates the actions of lipid mediators

Programmed cell death is a key function throughout embryonic development, regulating the formation and remodeling of complex multicellular tissues, but it must be closely regulated. In numerous experiments, PUFAs, LOX, and the antioxidant network have been implicated in the induction and control of apoptosis. Mechanisms from such studies could be applied to the processes occurring during embryogenesis and lend insight into the requirement of α-tocopherol for reproduction. For example, glutathione depletion has long-been recognized to induce neuronal cell death. Specifically, glutathione depletion leads to an increase in 12-LOX activity, which is a requisite step for neuronal apoptosis induced by glutathione depletion [82–84]. Furthermore, neuronal apoptosis following knockdown of GPx4 requires 12/15-LOX activity and induces the apoptosis-inducing factor (AIF) translocation from the inner mitochondrial membrane to the nucleus [85]. Other mechanisms by which LOX activity and its products may induce apoptosis include altering membrane fluidity and permeability, increasing calcium influx into the cell, disruption of mitochondrial integrity, and release of cytochrome C [86]. Notably, α-tocopherol entirely prevents 12/15-LOX mediated cell death [85]. In neuroblastoma cells, 15-LOX converts DHA to a series of hydroperoxy DHAs, including 4-HpDHA, 7-HpDHA, 14-HpDHA and 17-HpDHA [87]. Exposure to either DHA or 17-HpDHA, but not the monohydroxy DHA product 17-HDHA, potentiates apoptosis in neuroblastoma cells [87, 88]. Notably, co-treatment of neuroblastoma cells with α-tocopherol prevents apoptosis induced by DHA [88] or 17-HpDHA [87]. Similarly, ARA depletes intracellular glutathione and induces apoptosis in cultured cortical neurons, which is attenuated by LOX inhibitors or the α-Tocopherol analog, trolox [89], suggesting that ARA-derived 12-LOX products are responsible for apoptosis. Indeed, 12-HETE, the major ARA derived 12-LOX product, mimics ARA-induced apoptosis [89]. Given that neuronal tissue is highly enriched with DHA, it is notable that DHA hydroperoxides are more cytotoxic in neuroblastoma cells than are hydroperoxides derived from either linoleic acid or ARA [90]. This evidence illustrates important interactions between α-tocopherol, ARA and DHA, LOX, and hydroperoxy fatty acids, suggesting that these interactions mediate neuronal function. The implications of these findings are that regulation of lipid mediators that control apoptosis are essential for embryonic development and thus may explain the requirement for α-tocopherol to protect these key regulators. Notably, when the α-tocopherol transfer protein was knocked down in zebrafish embryos, the brain failed to form appropriately, suggesting that apoptosis was dysregulated [91].

12/15-LOX, 5-LOX and α-tocopherol

Vitamin E must be administered to the mother on post-fertilization days 5 to 9 to prevent fetal resorption in vitamin E-deficient rodents [92, 93]. Interestingly, this is the same critical period when the 12/15-LOX pathway appears to mediate implantation [94] and when GPx4-knockout mice embryos are resorbed [95]. Importantly, in the studies discussed above, 5-LOX and COX inhibitors did not prevent neuronal cell death induced by ARA or DHA in vitro [82, 83, 85, 89]; only 12/15-LOX inhibitors and α-tocopherol rescued the cells from death. These findings suggest that neuronal α-tocopherol is required for protection against excessive 12/15-LOX activity and PUFA peroxidation. Additionally, given the involvement of the larger antioxidant network (i.e. GPx4 and glutathione), the above studies suggest that this protection is mediated through an oxidative stress mechanism.

In contrast to the role of α-tocopherol mediating 12/15-LOX-dependent apoptosis in neuronal cells, the interactions of α-tocopherol and 5-LOX activity appears to be important in inflammation, but the mechanisms are unclear. Using 5-LOX isolated from potato tubers, α-tocopherol inhibited the enzyme with an IC50 of 5 micromolar and was apparently irreversibly bound to the enzyme [96]; however, these were studies using plant material. More recently, the inhibitory effect of α-tocopherol on human 5-LOX activity has been confirmed [97]. Additionally, α-tocopherol supplementation was found to suppress 5-LOX activity in peripheral blood mononuclear cells from end-stage renal disease patients undergoing hemodialysis [98–100]. Leukotriene B(4) (LTB(4)) is one of the products of the 5-LOX pathway, and has been demonstrated to increase tumor necrosis factor-alpha (TNF-alpha) and interleukin-1 beta (IL-1 beta) protein levels in fibroblasts [101]. α-Tocopherol inhibits LTB(4) production by human blood neutrophils or differentiated HL-60 cells, as does δ;-tocopherol and 13′-carboxy-δ;-Tocopherol [97]. The various vitamin E forms did not inhibit 5-LOX directly, but rather impaired intracellular calcium increase and influx, as well as the 5-LOX translocation from cytosol to the nucleus, a key event for 5-LOX activation [97]. α-Tocopherol also reduced the release of TNF-alpha and IL-1 beta from human activated monocytes [102, 103], an effect that was abrogated by addition of LTB(4). DHA also inhibits the production of IL-1 beta and TNF-alpha by monocytes, via inhibition of 5-LOX activity and LTB(4) synthesis [104], likely through substrate competition with ARA for 5-LOX [105], or by decreasing the availability of ARA. Thus, it is clear that α-tocopherol inhibits 5-LOX, however the mechanisms remain unclear. 5-LOX may be directly inhibited by α-tocopherol, other forms of vitamin E or their metabolites. It is also equally likely that α-tocopherol inhibits the production of various oxidized lipid products and it is the oxidized lipids that stimulate 5-LOX activity.

α-Tocopherol-deficient zebrafish embryos have increased levels of nonspecific (i.e. malondialdehde) lipid peroxidation products as compared with α-tocopherol sufficient embryos, confirming the role of α-tocopherol in radical-mediated lipid peroxidation. Interestingly, α-tocopherol deficient embryos have reduced mRNA expression of 5-LOX (Miller GM et al, unpublished data), suggesting this pathway is important and that α-tocopherol deficiency during embryogenesis may affect the transcription and/or mRNA processing of 5-LOX in addition to its activity. However, specific oxidized lipids produced via 5-LOX need to be investigated in α-tocopherol deficient embryos to determine if this pathway mediates the requirement of α-tocopherol for embryogenesis.

Implications and Future Directions

A complex interaction takes place between α-tocopherol, PUFAs, and lipid mediators. Apparently, α-tocopherol 1) prevents the nonspecific, radical-mediated peroxidation of PUFAs, 2) functions within a greater antioxidant network to modulate the production and/or function of lipid mediators derived from 12-LOX and 12/15-LOX, and 3) from 5-LOX. All of these mechanisms likely mediate the requirement of α-tocopherol for embryogenesis. The above evidence of the interaction between α-tocopherol, PUFAs, and lipid mediators warrants a new perspective in vitamin E research. Viewing oxidized PUFAs as only an outcome of α-tocopherol deficiency is too simplistic; oxidation of PUFAs gives rise to lipid mediators with potent physiological roles that are not simply non-specific by-products of lipid peroxidation. Furthermore, current guidelines recommend increasing ω-3 consumption during pregnancy and as a preventative/treatment measure for other diseases (ex: cardiovascular health [106]). The recommendations to increase PUFA intake are occurring concurrent with reports regarding the negative effects of α-tocopherol on health outcomes (ex: prostate [107] and bone health [108]). It has been estimated by the USDA that 96% of women do not meet the dietary recommended intake for vitamin E [109]. It is concerning given the interaction between α-tocopherol, PUFAs, and lipid mediators, and the possible applications to embryonic development, that PUFA intake in humans may increase while α-tocopherol intake may decrease. Nutritional recommendations must take into account nutrient-nutrient interactions and whether or not supplementation with one nutrient during critical life stages, such as pregnancy and embryogenesis, may alter the requirements for another nutrient.

HIGHLIGHTS.

We discuss how α-tocopherol:

1. prevents the nonspecific, radical-mediated peroxidation of PUFAs,

2. functions within a greater antioxidant network to modulate the production and/or function of lipid mediators derived from 12- and 12/15-lipoxygenase,

3. modulates 5-lipoxygenase activity, within the context of embryogenesis.