**4. Mechanisms of the Intestinal Absorption**

The selective absorption for carotenoids in humans cannot be explained by the simple diffusion mechanism alone. On the other hand, recent studies have suggested that the carotenoid uptake is partly mediated by facilitated diffusion [46–53]. For example, the ratio of the uptake mediated by scavenger receptor class B type 1 (SR-B1) to the total uptake of carotenoids in Caco-2 cells was as follows: 

50% for Ά-carotene; 20% for Ά-cryptoxanthin and 7% for lutein/zeaxanthin [53]. The efficiency of 

Ά-carotene absorption was remarkably reduced in SR-B1 knockout mice [54]. The physiological relevance of SR-B1 as an mediator of intestinal uptake for provitamin A carotenoids was indicated 

by the report that retinoic acid and the intestinal transcription factor ISX regulated expressions of 

both SR-B1 and Ά-carotene-15,15ȝ-oxygenase (BCO1), an enzyme responsible for vitamin A production [55]. The facilitated diffusion may cause the selective absorption of carotenoids in humans. However, even if SR-B1 does not mediate intestinal uptake of the highly polar epoxy xanthophylls, they can pass across membranes via the simple diffusion pathway. Thus, these absorption mechanisms could not account for the strict selectivity that was observed in humans. The strict selective absorption might occur if most parts of the highly polar epoxy xanthophylls taken up by intestinal epithelial cells were excreted back into intestinal lumen. 

The ATP-binding cassette (ABC) transporters such as ABCG5 and ABCG8 are well known to mediate the excretion of dietary phytosterols [56,57]. Although phytosterols such as Ά-sitosterol and campesterol are ingested from vegetables, grains, and cooking oils, the serum concentrations of the phytosterols are much lower than that of cholesterol in mammals [56,57]. Interestingly, ABCG5 polymorphism was suggested to be associated with the lutein bioavailability from egg in human subjects [58]. ABCG5 may excrete lutein and highly polar epoxy xanthophylls to intestinal lumen. 

Multi-drug resistance 1 (MDR1, ABCB1) is well known as a major efflux pump for lipid-soluble compounds. As the affinity of substrates for MDR1 has been suggested to be related to their 

polarity [59], the highly polar xanthophylls may be excreted by MDR1. Carotenoids were evaluated for a substrate of MDR1 expressed in certain cancer cells. Neoxanthin and violaxanthin, compared with other carotenoids tested, showed higher affinity for transfected-human MDR1 in mouse lymphoma L1210 cells [60], but similar results were not found in several human breast and colon cancer cell lines [61,62]. Further study is required to confirm the involvement of MDR1 in the excretion of carotenoids in intestinal cells. Thus, the selectivity in the intestinal absorption of carotenoids in humans is likely to be caused by these proteins that mediate uptake and excretion (Figure 2). The specificity of these proteins would cause the differences in the intestinal absorption of carotenoids among animal species. 

## **5. Metabolism of Xanthophylls in Mammals**

It is necessary to explore the metabolism of carotenoids after intestinal absorption in order to elucidate the mechanism of their biological activities, and to achieve safe and effective applications to human subjects. Although Ά-carotene is known to be metabolized to vitamin A through action of BCO1, little is known about the metabolic transformation of non provitamin A xanthophylls in mammals. 

Recently, we obtained evidence that the oxidative transformation of fucoxanthin and lutein to 

keto-carotenoids occurred in mammals. Fucoxanthinol and amarouciaxanthin A were found in the plasma and liver of mice fed with fucoxanthin, whereas fucoxanthin itself was not detected [32,33]. Fucoxanthinol was hydrolyzed from fucoxanthin in the intestinal tract, circulated in the body, and 

then oxidatively converted into amarouciaxanthin A (Figure 3). The conversion of fucoxanthinol into amarouciaxanthin A was also found to occur in human hepatoma HepG2 cells. Moreover, we found for the first time that the oxidative conversion was mediated in mouse liver microsomal fractions and required NAD+ as a cofactor, demonstrating the metabolic conversion of the 3-hydroxyl end group in xanthophylls at the level of enzyme reaction in animals [33]. 

Several proposed metabolites of lutein, as shown in Figure 4, were previously known to be present in such human tissues as plasma, milk, liver, and retina [18,63– 66]. Moreover, we found a remarkable accumulation of metabolites in mice fed with lutein [67]. 3ȝ-Hydroxy-<sup>Ή</sup>,<sup>Ή</sup>-caroten-3-one and lutein were the predominant carotenoids in the plasma, liver, kidney, and adipose, accompanied by 

<sup>Ή</sup>,<sup>Ή</sup>-carotene-3,3<sup>ȝ</sup>-dione, indicating that mice actively convert lutein to ketocarotenoids by oxidizing the secondary hydroxyl group. However, 3-hydroxy-Ά,Ήcaroten-3ȝ-one (3ȝ-oxolutein), the major metabolite of lutein in human plasma [67] and the retina [64], was not detected in the tissues of the mice. 

**Figure 4.** Chemical structures of lutein and its metabolites. 

These metabolites would be formed by the same enzyme that mediated the conversion of fucoxanthinol to amarouciaxanthin A. The combined level of the lutein metabolites in the liver of the mice was 72.4% of the total (intact lutein and the metabolites) [67]. This indicates that quantification of the metabolites is necessary to estimate the lutein bioavailability. Moreover, intact lutein and the metabolites may differ in their biological activities. Differences among lutein and its metabolites as antioxidants and blue light filters deserved further study. 

Similar to the case of lutein in mice, the oxidative metabolism of the other xanthophylls was reported to occur in human subjects. After the ingestion of paprika juice containing capsanthin as a major xanthophyll, capsanthon in addition to capsanthin was found in the plasma [68]. Capsanthon may be formed from capsanthin by the oxidation of the 3ȝ-hydroxyl group to the 3ȝ-keto group. After an oral dose of 4,4ȝ-dimethoxy-Ά-carotene in peanut oil, both 4-keto-Ά-carotene and canthaxanthin were found in the plasma [69]. These studies certainly indicate that humans have potential metabolic activity for the oxidation of secondary hydroxyl groups in various xanthophylls. 

In human tissues, other metabolites of lutein were detected. 3ȝ-Epilutein might be formed by a back reduction of 3ȝ-oxolutein that was produced from lutein [64]. *meso*-Zeaxanthin, which is detected in the retina only, might be formed by double bond migration from lutein [64]. The dehydration products of lutein, 3-hydroxy-3<sup>ȝ</sup>,4<sup>ȝ</sup>didehydro-Ά,·-carotene and 3-hydroxy-2<sup>ȝ</sup>,3<sup>ȝ</sup>-didehydro-Ά,<sup>Ή</sup>-carotene [19] were thought to be formed non-enzymatically under acidic conditions of stomach [70,71].

Recent studies have indicated the cleavage reaction of xanthophylls occurred in mammals. BCO1 catalyzes the central cleavage of provitamin A carotenoids, while Ά-carotene 9<sup>ȝ</sup>,10<sup>ȝ</sup>-oxygenase (BCO2) expressed *in vitro* can cleave a double bond at C-9ȝ and C-10ȝ of Ά-carotene, lycopene and 

xanthopylls [72–74]. Nonsense mutation of BCO2 was found to be associated with a yellow fat phenotype in sheep, in which xanthophylls were accumulated in adipose tissues [75]. The BCO2 knockout mice fed with lutein remarkably accumulated lutein metabolites, compared with the 

wild-type mice [76]. BCO2 might reduce the accumulation of xanthophylls by converting to smaller molecules, although the cleavage products and their further  metabolites have not been detected in animal tissues yet. Thus, in addition to oxidation of secondary hydroxyl group in xanthophylls, the cleavage reaction of carbon skeleton by BCO2 would be also a major metabolic transformation of xanthophylls in mammals. 
