Alpha-Tocopherol Transfer Protein (α-TTP): Insights from Alpha-Tocopherol Transfer Protein Knockout Mice

α-T transfer protein

α-TTP is a 32 kDa cytosolic lipid binding protein known as a member of the cis-retinal binding motif sequence (CRAL-TRIO) protein family (Panagabko et al., 2003). α-TTP was first described in 1975 by Catignani et al. (1975) as a hepatic α-T binding protein. α-TTP is expressed primarily in liver, but has been detected in various tissues including lung, spleen, kidney, brain, adrenals, uterus (pregnant rodent) and placenta (Copp et al., 1999; Gohil et al., 2004; Hosomi et al., 1998; Jishage et al., 2001; Kaempf-Rotzoll et al., 2003; Kaempf-Rotzoll et al., 2002).

All forms of dietary vitamin E are absorbed and transported via chylomicrons through lymph into the circulation (Arita et al., 1995). Unlike non-specific absorption of vitamin E, α-TTP mediates the selective transport of α-T in its RRR- or 2R-forms into nascent very low density lipoproteins (VLDL) and is responsible for secretion of α-T from the liver into blood to maintain plasma α-T concentrations (Arita et al., 1995).

Liver α-TTP has been isolated and its cDNA sequences reported from a variety of species. The human gene has been localized to the 8q13.1-13.3 region of chromosome 8 (Arita et al., 1995; Doerflinger et al., 1995). α-TTP has been crystallized and the α-T binding pocket identified (Meier et al., 2003; Min et al., 2003). α-TTP has a hinge and a cover that closes entrapping α-T in the binding pocket that causes α-T to fold such that the 2-position is critical for its fit (Meier et al., 2003; Min et al., 2003).

α-TTP has ligand specificity and relative affinities towards different tocopherols (α-> β-> γ-> δ-T) in vitro (Hosomi et al., 1997). Tocopherol binding affinity and transfer activity of α-TTP and ligand selectivity discriminating physiological factor for α-T retention were also demonstrated in vitro models (Morley et al., 2004; Panagabko et al., 2003; Panagabko et al., 2002). However, binding affinity of α-TTP to α-T and roles of α-TTP beyond incorporation of α-T to very low density lipoproteins (VLDL) were not demonstrated in various tissues from in vivo models.

A recent study has shown that high concentration of hepatic α-tocopherol caused by high α-T injection increased hepatic α-T concentrations but didn't induce either hepatic α-TTP mRNA and protein levels in rats (Mustacich et al., 2006). The result suggests that hepatic α-T levels doesn't regulate α-TTP gene expression.

To elucidate the importance of vitamin E in vivo, α-TTP gene deletion leading to vitamin E deficiency was required due to difficulty of dietary depletion of vitamin E in humans and animals (Cuddihy et al., 2004). α-TTP^-/- mice were generated by targeted disruption of α-TTP (Jishage et al., 2001; Terasawa et al., 2000).

Infertility in females with α-T deficiency

Vitamin E was first recognized as a "reproductive factor" in 1922 by Evans and Bishop (Evans and Bishop, 1922). Impaired fertility, such as resorption of the fetus, was reported in female α-TTP^-/- mice (Jishage et al., 2001; Terasawa et al., 2000) while no signs of male infertility were detected (Terasawa et al., 2000). In pregnant α-TTP^-/- mice, placental failure was effectively abrogated by dietary α-T (567 mg/kg diet) or a synthetic antioxidant (BO-653) (Jishage et al., 2001). Female rats with vitamin E deficiency demonstrated structural changes of ovary and uterus and abnormal estrus cycle (Das and Chowdhury, 1999). α-T levels increased in the placental membrane as gestation progressed during feto-placental development in normal humans and rodents (Jishage et al., 2001; Qanungo et al., 1999; Sen & Mukherjea, 1998). Moreover, α-TTP was not expressed in the placenta but in the uterus to supply α-T for placenta and fetus during the development of placental labyrinthine tropoblasts in mice (Jishage et al., 2001). These results suggest that α-TTP is a critical fertility factor to provide enough α-T for the maintenance of a regular estrous cycle and protection of placenta and fetus during pregnancy. It should be noted that mice and humans have very different placental structures and that α-TTP has been detected in human placenta (Muller-Schmehl et al., 2004).

Neurological disorders in α-T deficiency

Vitamin E deficiency is associated with neurological abnormalities in humans and animals suggesting α-T's importance in the central nervous systems. The α-TTP mRNA expression was demonstrated in cerebral cortex, especially Purkinje cells and in cerebellum (Hosomi et al., 1998) possibly to protect the brain from oxidative stress by providing α-T (Meier et al., 2003). AVED patients with genetic defects of α-TTP had very low or undetectable plasma α-T levels and progressive spinocerebellar dysfunctions (Hoshino et al., 1999; Hosomi et al., 1998; Ouahchi et al., 1995; Yokota et al., 1997; Yokota et al., 2000).

Although specific discrimination of α-T by α-TTP was shown in in vitro models (Hosomi et al., 1997; Panagabko et al., 2003; Panagabko et al., 2002) some human α-TTP mutants showed α-T transfer activity similar to that of wild type protein suggesting that α-TTP may have additional physiological functions in vivo (Morley et al., 2004). Therefore, we should consider the importance of an in vivo animal model to investigate effects of α-T in AVED and other possible genetic disorders in inbreeding mice as well.

Gene profiles related to neurodegeneration

α-TTP^-/- mice showed many features of neurological diseases as seen in humans with AVED. Specifically, the mice with ataxia had very low plasma and brain α-T levels (Yokota et al., 2001). Gene profile studies in α-TTP^-/- cortex identified putative genes affected by genetic deletion of α-TTP and may account for neurological disorders (Gohil et al., 2004; Gohil et al., 2003). These studies showed down-regulation in α-TTP^-/- cortex of genes related to neurogenesis and increased expression of genes related to neurodegeneration. The authors suggested that some of the affected genes may be regulated by retinoic acid receptor related orphan receptor α. In female α-TTP^-/- cerebral cortex, selective genes, such as myelin associated oligodendrocyte specific protein and proteolipid protein, associated with synaptogenesis and myelination were repressed. Induction of neurodegenerative genes, such as synulein-α, repression of synulein-β and induction of mitochondrial proapoptotic factor, cytochrome c, were also shown in α-TTP^-/- cerebral cortex (Gohil et al., 2004; Gohil et al., 2003). Although apparent neurological signs were not reported in 3-4 months old mice (Gohil et al., 2003; Jishage et al., 2001; Terasawa et al., 2000; Yokota et al., 2001), neurological behaviors such as balance, coordination, strength, exploratory behavior and anxiety were affected in α-TTP^-/- mice fed a α-T deficient or a normal diet up to at age of 6 months (Gohil et al., 2003; Jishage et al., 2001; Terasawa et al., 2000; Yokota et al., 2001). Learning and memory skills were significantly decreased in old female α-TTP^-/- mice at age of 12 months old (Gohil et al., 2004). However, by the age of 18 months α-TTP^-/- mice fed normal α-T diets developed abnormalities gradually while those fed deficient diets had obvious ataxia (Yokota et al., 2001). These differences in response along with the slow progression of abnormalities might be due to different dietary α-T concentrations and/or the duration of feeding, as well as different genetic backgrounds, ages and genders of mice (Table 1). The results suggest that the combination of genetic α-TTP defect (or deficiency) and dietary α-T deficiency would exacerbate neurodegeneration caused by prolonged α-T deficiency. The neurological abnormalities were prevented or attenuated when high dietary α-T was supplemented to α-TTP^-/- mice, as well as to patients with AVED (Mariotti et al., 2004; Schuelke et al., 2000; Yokota et al., 2001; Yokota et al., 1997). Recent studies in phospholipid transfer protein (PLTP) deficient mice having brain α-T concentrations at higher levels than in apparently normal α-TTP -/-mice demonstrated vitamin E deficiency-related anxiety but did not show any evidence of neurological disorders, including neuromotor coordination (Desrumaux et al., 2005). It is likely that there will be tissue specific responses that are also dependent on α-T concentrations in various brain region.

Table 1
Relationship of plasma (µmol/l) and tissue α-T (nmol/g) and dietary α-T in α-TTP -/- mice

• Click for larger image

Gene profiles related to vasculogenesis, signal transduction pathways and cell proliferation

The gene profile studies also suggested that deletion of α-TTP leading to α-T deficiency causes tissue specific change of genes associated with vasculogenesis, signal transduction pathways and cell proliferation (Gohil et al., 2003; Vasu et al., 2007). In α-TTP^-/- liver, changes in genes associated with neovascularization including quaking, NMDA receptor regulated gene 1 and hyalurinidase I were found. In addition, cell proliferation pathways such as growth factor binding proteins, mitogen activated protein kinase kinase 3 and apoptosis inhibitor 6 were activated in α-TTP^-/- liver (Gohil et al., 2003). Moreover, genes related to signal transduction such as casein kinase 1, doublecortin and calcium/calmodulin-dependent protein kinase-like 1 were downregulated in heart gene profile study of α-TTP^-/- mice (Vasu et al., 2007).

Taken together, the gene profile studies demonstrated that α-TTP^-/- mice are a useful model to find target genes related to neurological disorders or neurodegeneration which could be prevented or attenuated by dietary vitamin E treatment and vasculogenesis, signal transduction pathways and cell proliferation. Further gene profile studies identifying gender differences in different feeding regimen (time and dose) at different age and verifying reversibility of the neurological symptoms are recommended for future application of dietary α-T in patients with AVED.

Oxidative stress, inflammatory and immune response related markers

Vitamin E is a widely accepted antioxidant and a potent anti-inflammatory agent. α-TTP mRNA level was decreased in rat liver exposed to hypoxia (Ban et al., 2002). α-TTP deletion increased levels of lipid peroxidation measured by thiobarbituric acid reactive substances (TBARS) and 4-hydroxy-2-nonenal (HNE) in brain and by malondialdehydes in plasma (Terasawa et al., 2000; Yokota et al., 2001). α-TTP deletion with vitamin E deficiency increased another lipid peroxidation biomarkers demonstrated by 8-iso-prostaglandin F_2α (8-iso-PGF_2α) and hydroxyoctadecadienoic acid (tHODE) in retina and retinal pigment epithelium (RPE) fraction (Tanito et al., 2007). Moreover, F₂-isoprostanes levels in aortic lesions were increased in α-TTP^-/-ApoE^-/- mice, which are atherosclerosis susceptible and vitamin E deficient, when compared to that of α-TTP^+/+ApoE^-/- mice (Terasawa et al., 2000). α-TTP^-/- mice challenged with lipopolysaccharide affected the inflammatory markers including heme oxygenase -1, inducible nitric oxide synthetase expression and nitric oxides (NOx) release in lung and liver (Schock et al., 2004).

Additionally, protein oxidation (protein carbonyl) in lung lavage fluid, lung and liver and protein kinase C activity were increased in polymorphonuclear cells of α-TTP^-/- mice. However, levels of vitamin C and urate in lung tissues remained unchanged in α-TTP^-/- mice (unpublished data).

As a supportive evidence, genes regulated by antioxidant response elements such as metallothionein II and glutathione-s-transferase π were selectively induced in α-TTP^-/- liver. However, classical antioxidant enzymes such as catalase, gluthathione peroxidase, and superoxide dismutase were not induced in α-TTP^-/- liver, lung and heart (Gohil et al., 2003; Oommen et al., 2007; Vasu et al., 2007). It was similar patterns shown in lung tissues of α-TTP^+/+ mice fed an vitamin E deficient diet (Gohil et al., 2004) as shown in the lung tissues of α-TTP^-/- mice (Gohil et al., 2007). Interestingly, genes involved in detoxification of xenobiotics (cytochrome p450 1a2 and 2d9) were induced in α-TTP^-/- liver and herat (Gohil et al., 2003; Vasu et al., 2007).

Vit E has been known as a regulator of immune functions and vit E deficiency modulates immune responses (Beharka et al., 1997). The immunoglobulin genes including immunoglobulin kappa chain variable 8, immunoglobulin joining chain, immunoglobulin heavy chain, complement factor D were down-regulated (Vasu et al., 2007).

These results suggest that α-TTP may play a role in modulation of inflammatory-immune responses and antioxidant systems. Moreover, α-TTP^-/- mice are mildly susceptible to oxidative stress and might use other indirect antioxidant defense systems instead of classical antioxidant defense mechanism. Therefore, α-TTP^-/- mice are a practical model to discriminate antioxidant and anti-inflammatory functions α-T and to identify biomarkers of oxidative stress and inflammation in various disease states.

The relationship between dietary α-T and plasma or tissue α-T

α-TTP is the primary determinant of plasma α-T levels (Gohil et al., 2003; Jishage et al., 2001; Leonard et al., 2002; Schock et al., 2004; Terasawa et al., 2000; Yokota et al., 2001). Plasma and tissue α-T levels were very low in α-TTP^-/- mice fed α-T deficient or normal diets but with vitamin E supplementation concentrations were comparable to those of wild-type mice fed normal diets (Table 1). During chylomicron catabolism vitamin E is transported to other lipoproteins bypassing the sorting function of α-TTP in the liver (Traber et al., 1990). If vitamin E intakes are sufficiently high, then α-TTP is not necessary to maintain plasma α-T concentrations.

Liver α-T levels in α-TTP^-/- mice were 34-47% of those of wild-type mice when fed normal diets (30-150 mg/kg) (Table 1). Brain α-T concentrations in α-TTP^-/- mice were less than 10% of those of wild-type mice fed normal diets (Traber et al., 2005). In long term (38 weeks) vitamin E deficient wild-type mice, the brain retained approximately 50% of its α-T while liver was approximately 95% deficient (Cuddihy et al., 2004; Traber et al., 2005). Interestingly, liver α-T in α-TTP^-/- mice on normal diets was relatively well-maintained compared to wild-type mice with dietary vitamin E deficiency. This might be due to non-specific transport of α-T from gut to liver. Our group reported that α-T modulates cytochrome p 450 system especially, cyp3a, which is responsible for xenobiotic metabolism and is regulated by pregnane X receptor (Goodwin et al., 2002; Traber et al., 2005). This regulatory role of α-T in metabolizing toxic substances may be the reason that the liver maintains relatively high α-T levels compared to other organs. The results also suggest that α-TTP^-/- mice are more susceptible to neurological disorders because brain α-T levels are lower in α-TTP^-/- mice compared to dietary vitamin E deficient wild-type mice.

Lung, spleen, kidney and heart α-T levels in α-TTP^-/- mice were very low (only 10-15%) compared with those of wild-type mice fed normal diets (Leonard et al., 2002; Traber et al., 2005). In chow (vitamin E, ~99 mg/kg)-fed mice for 27 weeks, α-T levels of α-TTP^-/- liver, adipose tissue, adrenal gland and aorta were 15-35% of those of wild type mice (Terasawa et al., 2000). α-T levels were undetectable in the plasma (<0.46 µmol/l) and tissues (<0.2 nmol/g) of α-TTP^-/- mice when fed α-T deficient diet for ~43 weeks. After α-T supplementation (600 mg/kg diet), α-T levels in α-TTP^-/- mice increased to 70-90% of wild-type levels in the plasma (~3.5 µmol/l) and heart (~21 nmol/g) but were slightly increased in the cerebrum cortex (6.8%), cerebellum (19.6%), and spinal cord (23.8%) as compared to those of wild type mice fed normal diets (Yokota et al., 2001). Another study showed selective transport of RRR- or 2R-α-T from 1:1 mixture of natural and synthetic α-T by α-TTP and relatively higher plasma and tissue α-T levels compared to those of other studies in both α-TTP^-/- and wild-type mice; these latter diets were high in polyunsaturated fat (Leonard et al., 2002).

Our group also demonstrated tissue specific gene expression profiles in 5 different tissues including liver, lung, brain, adrenal and spleen of α-TTP^-/- mice (Gohil et al., 2004). The number of genes which are sensitive to α-TTP in tissues was not correlated with levels of α-T in the same tissues. Additionally, the levels of α-TTP expressed in tissues are not directly correlated with the levels of α-T (unpublished data). Each tissue might specifically preserve α-T for its own functions i.e. spleen and thymus for immune functions, plasma and liver for maintenance of homeostasis and lung and skin for protection from environmental insults such as ozone. In the plasma, the distribution of α-T to tissues is related to plasma levels of low density lipoprotein (LDL) and high density lipoprotein (HDL) which deliver α-T to tissues (Traber et al., 1992). Thus, α-T levels in plasma are dependent on plasma concentrations of these lipoproteins (Mardones et al., 2002). Scavenger receptor class B1 type (SR-B1) binds HDL, then mediates selective lipid uptake from HDL (Trigatti et al., 2000). SR-B1 knockout mice showed significant decreases in α-T levels in various tissues including liver, spleen and kidney. SR-B1 may be responsible for tissue α-T levels by regulating α-T influx into cells (Mardones et al., 2002). These results suggest that the levels of lipoproteins and SR-B1 maybe responsible for regulation of plasma and tissue α-T concentrations.

Variability of α-T levels shown in the same tissues treated with normal diets maybe due to the dietary form (natural or synthetic) of α-T, the amount of dietary fat in diets and/or the duration of dietary treatments. With α-T supplementation plasma and tissue levels of α-T in α-TTP^-/- mice on the diets containing different levels of α-T showed considerable variability suggesting that these differences are likely to be tissue specific.

Since there are still limitations to evaluate the relationship between dietary α-T levels and tissue α-T concentrations, and the roles of α-TTP in different tissues further studies in this animal model are recommended.