Mechanisms of Xanthophyll Uptake in Retinal Pigment

MECHANISMS OF XANTHOPHYLL UPTAKE IN RETINAL PIGMENT

EPITHELIAL CELLS

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Sara Elizabeth Thomas

The Ohio State University Nutrition Graduate Program

The Ohio State University

2016

Dissertation Committee:

Earl H. Harrison, Ph.D., Advisor

Martha A. Belury, Ph.D.

Joshua A. Bomser, Ph.D.

Richard S. Bruno, Ph.D.

Copyrighted by

Sara Elizabeth Thomas

2016

Abstract

Lutein and zeaxanthin are dietary carotenoids that selectively accumulate in the

macula of the primate retina. Meso-zeaxanthin is a non-dietary carotenoid that concentrates in the fovea of the macula and is thought to form from lutein. Macular carotenoids may play a role in light absorption, protection from oxidative stress and inflammation, and perhaps influence visual performance. Strong evidence indicates that the macular carotenoids protect against the development of age-related macular degeneration (AMD), the leading cause of blindness in adults over 50 years of age in the

United States. Understanding the mechanisms of selective accumulation of carotenoids and the influences on accumulation in retinal tissues may help in developing treatments for the prevention of AMD. To reach the retina, macular carotenoids must cross the highly selective retinal pigment epithelium (RPE). To better understand carotenoid accumulation in the retina, we studied the uptake of xanthophyll delivered by lipoproteins and human serum under different conditions using differentiated ARPE-19 cells, an RPE cell model. We also looked for products of xanthophyll metabolism after cell uptake to determine if any metabolism occurs in the RPE.

Prior to investigating carotenoid accumulation in the RPE, we developed an in vitro model using ARPE-19 cells to deliver human serum and lipoproteins enriched with

xanthophylls and determined that the highest cell uptake of lutein and best time to

perform future experiments occurred between 6-10 weeks of cell differentiation. We

measured lipoprotein mobility and separation of fetal bovine, calf, bovine, horse, and

human serum using agarose gel and decided to use human serum for separation and

delivery of lipoproteins since it is biologically relevant to the human ARPE-19 cell line.

Human lipoproteins were clearly separated using the agarose gel electrophoresis and the

iodixanol gradient. We confirmed that the distribution of carotenoids among human

lipoproteins was similar using iodixanol to separate lipoproteins as when using other

methods. We compared methods of lipoprotein enrichment and subsequent cell uptake of

xanthophylls and determined that lutein cell uptake was similar regardless of the method

of lipoprotein enrichment or the presence of a different unenriched lipoprotein. We

decided to use the method of first separating lipoproteins from human serum and then

enriching the isolated lipoprotein due to the simplicity of reaching specific xanthophyll

concentrations using this method.

After developing a ARPE-19 cell model to deliver xanthophyll-enriched human

serum and lipoproteins, our first objective was to determine uptake of β-carotene, lutein, zeaxanthin, and meso-zeaxanthin in human serum or lipoproteins. Previous studies show that xanthophylls associate mostly with HDL in human serum while β-carotene associates mostly with LDL. HDL binds to cells via a scavenger receptor class B1 (SRB1)-

iii

dependent mechanism while LDL binds via the LDL receptor (LDLR). Therefore, we

hypothesized that xanthophylls are delivered via HDL to the RPE in an SRB1-dependent

manner. Carotenoid-enriched whole human serum and lipoproteins were added to ARPE-

19 cells. A chemical inhibitor, BLT-1, and a biological inhibitor, serum amyloid albumin

(SAA), were used to block binding of HDL to the SRB1 receptor. Carotenoids were extracted and measured using HPLC analysis. For lutein and β-carotene, LDL delivery

resulted in the highest rate and extent of uptake. HDL was more effective in delivering

meso-zeaxanthin and zeaxanthin. Using whole serum to deliver lutein or zeaxanthin in

the presence of increasing amounts of β-carotene, lutein, or zeaxanthin, we found that β- carotene decreased lutein uptake. Zeaxanthin uptake was unaffected by either lutein or β- carotene. Compared to the control treatment, zeaxanthin delivery via HDL was reduced by 49% and 52% by SAA and BLT, respectively. Our results suggest a selective HDL- mediated uptake of meso-zeaxanthin and zeaxanthin via SRB1 and potentially a LDL- mediated uptake mechanism for lutein via LDLR. Finally, we carefully looked for formation of meso-zeaxanthin from lutein and xanthophyll metabolism to apocarotenoids using HPLC and found no evidence of the formation of either indicating that xanthophylls are not metabolized in the RPE. Overall, the results provide a plausible mechanism for the selective uptake of zeaxanthin over lutein in the retina.

Acknowledgments

I would like to thank my advisor, Dr. Earl Harrison, who has supported me through the research process with untmost kindness and patience. I have learned a great deal from his mentoring and thank him for his advice and many poster, abstract, and paper revisions. I would also like to thank the other members of my committee, Dr.

Martha Belury, Dr. Josh Bomser, Dr. Rich Bruno, and Dr. Steve Schwartz, for their advice and time invested in ensuring progress in my project.

I would also like to thank all the members of my lab who have been there to support my research both inside and outside the lab. In particular, I would like to thank

Vanessa Reed who significantly set the ground work for a successful Ph.D., while training me during my Master’s program and beginning my Ph.D. work in cell culture techniques, Shiva Raghuvanshi for her time in helping me develop an agarose gel electrophoresis protocol for lipoproteins, Dr. Carlo dela Seña for his patience in helping me with HPLC, and Jian Sun for her help in cell culture. I would also like to thank Yan

Yuan for her friendship and support during the Ph.D. process and for performing Western

Blot analysis on my cells.

I would like to thank the organizations who provided financial support during my education and research experience. These include: Gladys Mason Scholarship, Joseph &

Nina Mae MATTUS Scholarship, Education and Human Ecology Dissertation

Fellowship Award, Education and Human Ecology Travel Award, OSU Interdisciplinary

Ph.D. Program in Nutrition Travel Award, Marguerite E. Ellis Scholarship, my advisor,

Earl Harrison, and the College of Education and Human Ecology.

Finally, I would like to thank my parents, Jeanie Lisak and Bill Thomas, for their encouragement and support.

Vita

1996...... B.S. Zoology, The Ohio State University

2006 - 2008 ...... M.S. Nutrition, The Ohio State University

2008 - 2009 ...... Dietetic Intern, The Ohio State University

2009 - 2013 ...... Consumer Relations Dietitian, Abbott

Nutrition

2013 – 2016 ...... Graduate Research Associate/Graduate

Teaching Associate, The Ohio State

University

2014 - 2015 ...... Gladys Mason Scholarship, The Ohio State

University

2014 - 2015 ...... Joseph Nina Mae MATTUS Scholarship,

The Ohio State University

2015 - 2016 ...... Gladys Mason Scholarship, The Ohio State

University

2015 - 2016 ...... EHE Dissertation Research Fellowship, The

Ohio State University

vii

2016...... Marguerite E. Ellis Scholarship, The Ohio

State University

Publications

DiSilvestro, Robert A., Thomas, Sara E., Harrison, Earl, and Epitropoulos, Alice. 2015.

“A pilot comparison of phospholipidated lutein to conventional lutein for effects on plasma lutein concentrations in adult people”, Nutrition Journal, 14:104 - 108.

Thomas, Sara E. and Harrison, Earl H. 2016. “Mechanisms of Selective Delivery of

Xanthophylls to Retinal Pigment Epithelial Cells by Human Lipoproteins.” Journal of

Lipid Research, 57 (10): 1865 - 1878.

Fields of Study

Major Field: Ohio State University Nutrition Program

viii

Table of Contents

Abstract ...... ii

Acknowledgments...... v

Vita ...... vii

List of Figures ...... xii

Chapter 1: Introduction ...... 1

1.1 Introduction ...... 2

1.2 Aims ...... 4

Chapter 2: Literature Review ...... 7

2.1 Carotenoids...... 8

2.2 Macular Xanthophylls ...... 8

2.3 Digestion and Absorption of Carotenoids in Humans ...... 9

2.4 Delivery to the Retina ...... 11

2.5 Anatomy of the Eye...... 14

2.6 Xanthophylls and Other Nutrients in the Retina ...... 19 ix

2.7 Measuring Macular Pigment and Retinal Function ...... 22

2.8 Xanthophyll Uptake in the RPE and Retinal Distribution ...... 23

2.9 Carotenoid Cleavage Enzymes...... 26

2.10 Xanthophyll Function in the Retina ...... 26

2.11 Age-Related Macular Degeneration (AMD) ...... 27

Chapter 3: Lipoprotein Separation and Xanthophyll Delivery to ARPE-19 Cells -

Methods Development ...... 30

3.1 Introduction ...... 31

3.3 Materials and Methods ...... 33

3.4 Results and Discussion ...... 37

3.5 Acknowledgements ...... 44

3.6 Figures and Tables ...... 45

Chapter 4: Mechanisms of Selective Delivery of Dietary Xanthophylls to Retinal Pigment

Epithelial (ARPE-19) Cells from Human Lipoproteins* ...... 54

4.1 Abstract ...... 55

4.2 Introduction ...... 56

4.3 Materials and Methods ...... 60

4.4 Results ...... 66

4.5 Discussion ...... 72

4.6 Acknowledgements ...... 81

4.7 Figures ...... 82

Chapter 5: Epilogue ...... 99

List of References ...... 106

List of Figures

Figure 2.1 Distribution of carotenoids among lipoproteins ...... 13

Figure 2.2 Anatomy of the human eye……………………………...... 15

Figure 2.3 Distribution of rods and cones in the retina...... 17

Figure 2.4 Macular pigment and ratio of lutein/zeaxanthin in relation to distance from

foveal center ...... 21

Figure 2.5 Cross-section of primate retina showing macular pigment and it’s absorption

of blue light ...... 22

Figure 2.6 Potential pathway for xanthophyll uptake, transport, and accumulation in the

human retina...... 25

Figure 3.1 Lutein uptake as a function ot weeks of cell differentiation ...... 45

Figure 3.2 Lutein recovery from cells, medium, and total recovery at 2, 4, 6, 8, and 10 weeks of cell differentiation...... 46

Figure 3.3 Agarose gel of whole human serum and lipoproteins stained with Sudan Black

...... 47

Figure 3.4 Agarose gel of whole horse serum and lipoproteins stained with Sudan

Black… ...... 48

xii

Figure 3.5 Iodixanol fractionation of lipoproteins ...... 49

Figure 3.6 Agarose gel confirmation of isolated human lipoproteins...... 50

Figure 3.7 Carotenoid distribution among lipoproteins ...... 51

Figure 3.8 Lipoprotein delivery of lutein and cell uptake using different methods ...... 52

Figure 4.1 Macular xanthophylls ...... 82

Figure 4.2 Agarose gel confirmation of lipoproteins…………………………………….83

Figure 4.3 Carotenoid distribution among lipoproteins………………………………….84

Figure 4.4 ARPE-19 carotenoid uptake: delivery in whole serum………………………85

Figure 4.5 Kinetics of carotenoid uptake from LDL and HDL………………………….86

Figure 4.6 Concentration-dependent carotenoid uptake from LDL or HDL…………….88

Figure 4.7 Interactions of carotenoids during cell uptake……………………………...... 90

Figure 4.8 Impact of excess lipoprotein on carotenoid uptake……………………….….91

Figure 4.9 Effect of inhibition of SR-B1 by BLT-1 and SAA on zeaxanthin or lutein uptake…………………...... 93

Figure 4.10 Xanthophylls are not extensively metabolized to apocarotenoids by ARPE-19

cells………………………………………………………………………………………95

Figure 4.11 Lutein is not converted to meso-zeaxanthin by ARPE-19 cells…………….97

xiii

Chapter 1: Introduction

1.1 Introduction

Age related macular degeneration (AMD) is an incurable disease among adults 50

years of age and older and is the leading cause of vision loss in this population (1). A

2014 meta-analysis predicts that 196 million people will have AMD by 2020 increasing

to 288 million by 2040 and another 7.3 million are at risk for vision loss from the disease

(2). Age related macular degeneration occurs due to deterioration of the macula located

in the retina of the eye. Breakdown of the macula impacts central vision and the ability

to see fine detail. Because no cure is available, there is interest in the mechanisms of the

disease as well as its prevention. Lutein and zeaxanthin are naturally occurring

carotenoids in the diet that have become of interest since these carotenoids preferentially

accumulate in the retina of the eye. Meso-zeaxanthin is a non-dietary carotenoid that also accumulates in the macula. These macular xanthophylls may provide protection to the eye and help in the prevention of AMD by filtering blue light and acting as antioxidants.

However, their mechanisms of action in the eye are largely unknown. Elucidating some of these mechanisms is the first step in determining possible prevention and/or treatment of AMD through the diet, supplements, or medications.

Entry of nutrients from the blood into the eye requires crossing over the highly selective retinal pigment epithelium (RPE) from the choroid. Lutein and zeaxanthin preferentially accumulate in the retina of the eye with the majority of these carotenoids located in the macula (3). Higher intakes of these nutrients are related to a lower risk of developing AMD (4). To cross the RPE and gain entry to the retina, lutein, zeaxanthin, and other nutrients must be transported into the RPE cell. A previous study performed by

2 our lab provides evidence that zeaxanthin transport into the RPE is at least partially mediated by SR-B1 and zeaxanthin is preferentially taken up compared to β-carotene.

Studies have shown that lutein and zeaxanthin associate more with HDL than with LDL in the human plasma (5) and HDL may be an important carrier of nutrients to the retina

(6). Furthermore, SR-B1 has been shown to bind HDL with high affinity. The objective of this study is to determine mechanisms of xanthophyll uptake into the RPE cell by evaluating transport, uptake, competition with other carotenoids, and possible xanthophyll metabolites formed in the RPE. This has important long-term implications in formulating dietary supplements and/or making diet recommendations for the prevention and treatment of AMD.

1.2 Aims

Aim 1: Determine if lutein and zeaxanthin are more efficiently delivered to ARPE-

19 cells via HDL or LDL. Studies have shown that lutein and zeaxanthin associate more with HDL than LDL in human serum (5). Some studies suggest that HDL may be an important carrier of nutrients to the retina (6). Evidence has shown that transport of xanthophylls might be partially facilitated by SR-B1, which is involved in cholesterol trafficking in the enterocyte, in Caco-2 and ARPE-19 cells. Furthermore, SR-B1 has been shown to bind HDL with high affinity. Our working hypothesis of this aim is that xanthophylls are more efficiently delivered to ARPE-19 cells by HDL than by LDL.

Aim 2: Examine the impact of other carotenoids on xanthophyll uptake and accumulation into ARPE-19 cells. In the intestinal cell model, Caco-2 cells, xanthophyll uptake is inhibited when other nutrients such as β-carotene or vitamin E are present and vice versa (7-9). In ARPE-19 cells, zeaxanthin is preferentially accumulated in the cell compared to β-carotene (10). Competition among xanthophylls, and carotene for entry into the enterocyte may be due to competition for lipid transporters involved in uptake of carotenoids. Our working hypothesis for this aim is that xanthophyll uptake into ARPE-19 cells is impacted when other carotenoids are present due to competition/interaction for lipid transporters.

Aim 3: Measure uptake of zeaxanthin delivered by lipoproteins in ARPE-19 cells when lipid receptors such as SR-B1 are inhibited. Increasing evidence supports that xanthophylls are selectively taken up into the enterocyte through a facilitative process as opposed to diffusion into the cell. Furthermore, xanthophylls are selectively accumulated in the retina indicating that a transporter is responsible for this selective uptake. Our lab has shown that zeaxanthin is preferentially taken up by ARPE-19 cells by an SR-B1 dependent mechanism. Our working hypothesis for this aim is that zeaxanthin uptake via

HDL is inhibited when SR-B1 is inhibited.

Aim 4: Account for all lutein/zeaxanthin/meso-zeaxanthin and measure any metabolites formed once introduced to RPE cells. The ratio of lutein found in human blood is 5 times that of zeaxanthin, which is similar to amounts found in the American diet. Likewise, the overall ratio of lutein to zeaxanthin in the whole retina is 2:1.

However, zeaxanthin preferentially accumulates in the macular region of the retina in a ratio of less than 1:2 lutein to zeaxanthin. About 50% of the zeaxanthin within the macula is present as the meso-isomer of zeaxanthin, a metabolite of lutein, indicating that lutein may be converted to meso-zeaxanthin in the macula region. Lutein and zeaxanthin are delivered via plasma lipoproteins and must first cross over the retinal pigment epithelium from the blood supply in the choroid before being transferred to the retina.

Since lutein is possibly converted to meso-zeaxanthin upon transfer into the macula, some of this conversion might take place in the RPE cells. In addition, it is unclear whether BCO2 (beta-carotene oxygenase 2) is present and/or active in the macula of the

retina. Thus, it is worthwhile to determine if xanthophyll products of BCO2 activity are

present after uptake in the RPE. Our working hypothesis is that meso-zeaxanthin, a

metabolite of lutein, is formed in ARPE-19 cells.

Chapter 2 provides background and a review of the literature related to AMD, macular carotenoids, transport and delivery of carotenoids to the retina, anatomy of the retina, and possible functions of xanthophylls in the retina. This review provides a context for the research aims and experiments. Chapter 3 includes the preliminary research performed to develop a protocol for the separation, enrichment, and delivery of carotenoids to ARPE-

19 cells and investigate the work proposed in the research aims. Chapter 4 was published in the Journal of Lipid Research and contains the bulk of the results for research aims 1-

4. Finally, Chapter 5 is the epilogue and provides a summary of the findings and proposed future research on the topic.

Chapter 2: Literature Review

2.1 Carotenoids

Carotenoids are a class of hydrophobic polyisoprenoid hydrocarbons (carotenes)

and their oxygenated derivatives (xanthophylls) (11) present as pigments that are

responsible for yellow, orange, red, purple colors of fruits, vegetables, and many animals.

These eight isoprenoid units form a long chain of 40 carbon atoms and 13 conjugated

bonds (12) and various carotenoids are formed by hydrogenation, dehydrogenation,

cyclization, and/or oxidation. This base structure is modified by cyclization of the ring

structure and addition of oxygen groups, which leads to their individual colors and

antioxidant functions (13). The conjugated double bonds allow isomerization to cis and

trans forms. These different chemical structures among carotenoids result in differences

in absorption, metabolism, and biological activities. Carotenoids are synthesized by

plants and microorganisms, but not animals, which must obtain them from diet. Common

carotenoids in the diet include β-carotene, lycopene, lutein, β-cryptoxanthin, and zeaxanthin. More than 600 carotenoids have been identified in nature, but only 40 are present in the human diet mostly from fruits and vegetables (13).

2.2 Macular Xanthophylls

While carotenes are composed of only carbon and hydrogen, xanthophylls include hydroxyl groups making them slightly more hydrophilic than the other carotenoids.

Common xanthophylls in the human diet include lutein, zeaxanthin, and β-cryptoxanthin.

(14). The nonprovitamin A carotenoids include lutein and zeaxanthin which are unable to form vitamin A. Lutein and zeaxanthin differ in structure due to the position of

double bonds and methyl groups on the carbon chain (3). Sources of xanthophylls in the

human diet include kale, spinach, eggs, and broccoli (9).

The oxygen containing xanthophylls are found in plants and the human retina

which are both highly exposed to light. Three xanthophylls preferentially accumulate in

the macula of the retina and include dietary lutein and zeaxanthin and non-dietary meso-

zeaxanthin, a stereoisomer of zeaxanthin. Zeaxanthin is present in 3 stereoisomer forms

(15). (3R,3’R) –zeaxanthin is the form most referred to as zeaxanthin and is naturally

present in most diets. (3R,3’S)-zeaxanthin (meso-zeaxanthin) is found in the macula of the retina along with very small amounts of (3S,’3S) – zeaxanthin. The dietary xanthophylls selectively accumulate in the primate eye and brain compromising 80-90% of the total carotenoids and exclusively accumulate in the neural retina and lens (16).

2.3 Digestion and Absorption of Carotenoids in Humans

Humans are unable to synthesize carotenoids and must obtain them from the diet.

For carotenoids to reach tissues such as the retina after consumption, they must first undergo digestion. Carotenoids are processed during digestion similarly to other dietary lipophilic compounds. Thus, dietary lipids including fat-soluble, nonpolar vitamins must be released from the food matrix, emulsified in the lipid phase of chyme, and solubilized in mixed micelles. Much research has focused on digestion since many factors can promote or inhibit carotenoid absorption and thus distribution in important tissues such as the eye.

Bioaccesibility is a measure of incorporation from the food matrix into the mixed micelle for uptake into the intestinal cell. Factors that affect carotenoid bioaccessibility

9 include: type of food matrix (3), foods eaten with the carotenoid (17), and competition between carotenoids and other nutrients (8,18,19). Carotenoids are poorly soluble in digestive fluid so they must first be released from the food matrix. The food matrix in fruits and vegetables can interfere with the release of carotenoids due to the solid structure of the plants cell walls. Processing and heat treatment can help with the breakdown and make them more accessible. Lutein in plants is available in the esterified form while lutein in plasma is present in the free form indicating that esterified lutein is enzymatically cleaved during digestion (3). There is some debate whether supplements containing the free form of lutein are more easily absorbed than esterified lutein. One recent study showed that blood levels were 20% higher with supplementation of free lutein than when supplemented with esterified lutein (20). Another studied compared a supplement containing conventional lutein ester and one containing lutein plus phospholipid, similar to the phospholipid composition of an egg (21). The lutein and phospholipid supplement resulted in higher serum lutein concentrations than the conventional lutein ester supplement. It’s hypothesized that esterified lutein must go through the process of removing the fatty acids before it can be absorbed thus delaying absorption.

Other foods and nutrients consumed along with carotenoids can also affect bioaccessibility. Carotenoids are hydrophobic, so including fat in a meal facilitates incorporation into the mixed micelle (22). There is also some evidence that certain types of fats facilitate better absorption. For lutein and zeaxanthin, it was found that greater absorption occurs with more saturated fatty acids in butter and palm oil then with olive

and fish oils (23). This is also true in the previous study mentioned above with the lutein in phospholipid supplement resulting in higher serum lutein levels than conventional lutein ester (21). Studies have also shown competition among carotenoids and other nutrients such as vitamin E (24). Although lutein did not interact with β-carotene for absorption, β-carotene and lycopene did interact (9).

Once incorporated into the mixed micelle, carotenoids are ready to cross into the enterocyte. The classical view assumes that intestinal absorption of carotenoids occurs by passive diffusion. However, variability of absorption in individuals observed during post-prandial studies (25) as well as observed competition between absorption of carotenoids (18,26) and other compounds indicate the possibility of a protein-mediated mechanism. Lipid transporters, specifically the cholesterol transporters, SR-B1 and CD-

36 have been shown to facilitate transport of β-carotene (27). Another study using Caco-

2 cells identified SR-B1 as playing a role in intestinal uptake of lutein (7). It is also possible that ABC transporters on the apical side of the epithelial cell may efflux lutein back into the lumen.

2.4 Delivery to the Retina

Once within the enterocyte, xanthophylls are packaged into chylomicrons and released into the lymphatic system where they enter the blood stream. Next, they are delivered to the liver and packaged with lipoproteins and released into the blood. Lutein, in its unesterified, free form, and zeaxanthin are transported via lipoproteins to other tissues such as adipose, retina, and brain. When comparing distribution of carotenoids in the human serum (Figure 2.1), lutein was to shown to be mostly associated with high-

11 density lipoprotein (HDL) while lycopene and β-carotene are most associated with low- density lipoprotein (LDL) (5). Some studies suggest that HDL might be an important carrier of nutrients to the retina (6). In one study, a group of individuals classified as cholesterol absorbers had a higher serum HDL cholesterol and a 229% higher increase in plasma lutein after drinking a lutein-enriched egg yolk drink (28). Wisconsin hypoalpha mutant (WHAM) chickens display a ˃90% reduction in plasma HDL making it a relevant model to study HDL deficiency (29). WHAM chickens fed a high lutein diet showed an increase in tissues except for the retina indicating HDL as an important carrier (30).

Thus, investigating transport of carotenoids to tissues may provide clues to its uptake into these tissues especially where preferential uptake occurs as in the retina.

Figure 2.1: Distribution of carotenoids among lipoproteins.

Endogenously (A) and after laboratory enrichment (B) (5).

2.5 Anatomy of the Eye

Figure 2.2 shows a simplistic image of the human eye. The orbit is a bony cavity in the skull that supports the eyeball, muscle, nerves, and blood vessels (31). The orbit is pear shaped and formed by bones. The eyeball contains an outer layer white in color called the sclera, which is covered by a thin mucous membrane called the conjunctiva.

The cornea is the first place light enters and helps focus the light on the retina located in the back of the eye. After light passes through the cornea, it travels through the pupil, which is the black dot in the center of the iris. The pupil controls the amount of light entering the eye by expanding and contracting. The lens is located behind the cornea and its purpose is to focus light onto the retina by changing shape. The lens becomes thicker to focus on objects close by and thinner to view objects farther away.

Figure 2.2: Anatomy of the human eye (16)

The retina, located at the back of the eye, is a multi-layered sensory tissue containing millions of photoreceptors to capture light and convert them to neural impulses that travel along the optic nerve to the brain (31). Once in the brain, these neural impulses are converted into images that the brain recognizes. The retina originates embryologically from the central nervous system. In primates, the macula is located in the center of the retina temporal to the optic nerve and is about 4.5-6 mm in diameter.

The macula is most sensitive to light and responsible for perception of fine detail. The 15 two main photoreceptor types located within the retina are rods and cones. The presence and number of rods and cones varies between animal species based on the type of environment in which they live. Rods dominate most animal retinas. For example, mice and rats are nocturnal animals, so their retinas are mostly rods with only 3-5% containing cones. Primate retinas contain a cone rich fovea at the center of the macula, which allows for daylight vision and the ability to discriminate color and fine detail.

The primate macula and fovea (located at the very center of the macula) contain mostly cones, which are responsible for detailed central and color vision (Figure 2.3).

The rods are responsible for night and peripheral vision and are much more sensitive to light than cones but unlike cones, they do not process color. The rods are mostly located in the peripheral areas of the retina while cones are more concentrated as you move centrally toward the macula. The fovea is the very central part of the retina and is the region with the greatest visual acuity (32). The fovea contains the highest density of cone receptors (199,000/mm2) along with Muller cells and is a rod-free zone.

Figure 2.3 Distribution of rods and cones in the retina (33)

The remaining part of the retina is the peripheral retina located outside the temporal retinal arteries (32). The major cell types of the retina include RPE cells, photoreceptor cells, interneurons, ganglion cells, and glial cells. The retinal pigment

17 epithelium (RPE) is a layer of pigmented cells located between the blood rich choroid and the photoreceptors toward the back of the retina (31). The adult retina contains approximately 3.5 million RPE cells, which are 14 um in the central retina, and 60 um in the peripheral retina in diameter. RPE cells are highest in density within the fovea compared to its density in the peripheral retina. Within the central retina, RPE cells are tightly packed to form a single layer of cuboidal epithelium. The RPE cells form tight junctions of the outer blood-retina barrier to prevent the free flow of molecules from the choriocapillaris and the photoreceptors. The basal and apical surfaces of RPE cells are infolded to increase the surface area and facilitate active transport of nutrients between the choriocapillaris and the photoreceptors.

Light must pass through the entire retina before reaching the sensory photoreceptors. The location of the photoreceptors positions them close to the RPE, supplying the photoreceptors with retinaldehyde and other nutrients while performing phagocytosis of the outer segments of the photoreceptor for regeneration (32). Among one of the most biologically active tissues in the body, the RPE is very important in maintaining the function and health of the retina and photoreceptors. The RPE serves as the outer retinal epithelial barrier and carries essential nutrients from the choroid to the photoreceptors. A healthy RPE is essential for nutrient delivery from the choriocapillaris to the photoreceptors. The RPE also removes metabolic waste generated by the photoreceptors provides additional protection from photooxidation.

The lipid enriched photoreceptor outer segments are subject to intense light in an oxygen rich environment leading to the production of free radicals and damage (32). To

be rid of accumulated peroxidation damage, photoreceptors regularly shed their outer

segments. To maintain photoreceptor excitability, the RPE cells are involved in

phagocytosis of the photoreceptor outer segments. The outermost tips of the

photoreceptor outer segments contain the most free radicals and photo-damaged proteins

and lipids and undergo shedding and renewal to maintain photoreceptor function (34).

The outer segments of the photoreceptors are phagocytosed by the RPE and the outer

segments are digested within the RPE cell. The outer segments of the photoreceptors are

rich in docosahexaenoic acid and retinal, which are recycled within the RPE and delivered back to the photoreceptors. The RPE converts all trans retinyl ester to 11-cis-

retinal and delivers it to the photoreceptors to commence the visual cycle. Breakdown of

the RPE cells or failure of the RPE cells to achieve nutrient exchange and turnover of the

photoreceptor cells can lead to diseases of the eye such as Age-Related Macular

Degeneration (AMD).

2.6 Xanthophylls and Other Nutrients in the Retina

As mentioned previously, the RPE takes up nutrients such as carotenoids, glucose,

retinol, and fatty acids from the blood through active or facilitated transport and delivers

these to the photoreceptors. The RPE accumulates lutein, zeaxanthin, ascorbate, α-

tocopherol, and β-carotene (34), which are thought to function as antioxidants and absorb

blue light. Research has focused mostly on the xanthophylls, lutein and zeaxanthin, since

they are concentrated in the retina, especially in the fovea of the macula. Lutein and

zeaxanthin represent 80% of carotenoids in the retina while β-carotene is present in only

trace amounts (35). Xanthophylls accumulate in the macula of the retina, which is aptly

19 named macula lutea due to the yellow macular pigment. Figure 2.4 summarizes the concentration of macular pigment and the lutein:zeaxanthin ratio as you move away from the center of the fovea in the macula. The overall ratio of lutein to zeaxanthin in the whole retina is 2:1 while the ratio in the macular region is less than 1:2. About 50% of the zeaxanthin within the macula is present as a meso-isomer of zeaxanthin called meso- zeaxanthin. Meso-zeaxanthin is a metabolite of lutein indicating that lutein may convert to meso-zeaxanthin within the macula. Figure 2.5 shows a cross-section of the primate retina where the yellow color indicates the macular pigment made up of lutein, zeaxanthin and meso-zeaxanthin. Interestingly, lutein and zeaxanthin show a linear relationship between their regional relationship and the regional relationship of rods and cones suggesting that lutein is associated with rods and zeaxanthin with cones (36).

Figure 2.4 Macular pigment and ratio of lutein/zeaxanthin in relation to distance from foveal center (16)

Figure 2.5 Cross-section of primate retina showing macular pigment and it’s absorption of blue light (16)

2.7 Measuring Macular Pigment and Retinal Function

Mice and non-primates do not contain a macula making the mass study of macular pigment and retinal function a challenge with the lack of easily obtainable animal models. Non-human primates contain a macula similar to humans making them among the best test subjects for invasive methods. Studies in monkeys have shown similar oxidative products of xanthophylls as those found in humans indicating similar pathways of these products (37,38). However, the high cost of using these animals limits their use (39). Alternatively, recent studies have used quail to study carotenoids and

AMD since the quail retina is similar to the macula and disease progression occurs quickly (40). Noninvasive assessment of macular pigment can be performed using optical density (MPOD, macular pigment optical density) which measures a unit of optical density as equivalent to 0.025 ng macular pigment in a 1 mm2 retinal tissue area 22

(41). This method of measuring macular pigment has also been correlated not only to

lutein and zeaxanthin amounts in the peripheral retina (42) but also to amounts in the

brain (43) making it an easily used biomarker of lutein and zeaxanthin status in the retina and brain.

2.8 Xanthophyll Uptake in the RPE and Retinal Distribution

Macular carotenoids selectively accumulate within the retina suggesting a protein-

mediated mechanism of uptake. Previous research in our lab found that xanthophylls,

specifically zeaxanthin, are preferentially taken up into ARPE-19 cells from micelles in a

SR-B1-dependent mechanism compared to β-carotene(10). Interestingly, SR-B1 is an

HDL receptor. Since xanthophylls are mostly carried within HDL, it makes sense that

SR-B1 would be responsible for xanthophyll uptake in the RPE. Other potential lipid transporters in the RPE include CD-36 and LDLR. However, the previous study found that knock-down of CD-36 did not impact xanthophyll uptake in ARPE-19 cells.

The stability of xanthophylls in the macula despite constant assault by free radicals and UV light and the specific localization of lutein, zeaxanthin, and mesozeaxanthin in the retina indicate the presence of transport and binding proteins specific to these xanthophylls to distribute and stabilize them in the retina. Lutein and zeaxanthin bind saturably and specifically to membrane proteins within the human retina

(44). These potential binding proteins within the retina have been identified as steroidogenic acute regulatory domain protein 3 (StARD3) for lutein (45) and glutathione

S-transferase P1 (GSTP1) for zeaxanthin and meso-zeaxanthin (46). When bound to

GSTP1, zeaxanthin and meso-zeaxanthin were more resistant to degradation (47) which

23 may explain the relative stability of these xanthophylls in the retina. Figure 2.6 summarizes these events for the possible uptake and distribution of macular carotenoids in the retina.

Figure 2.6 Potential pathway for xanthophyll uptake, transport, and accumulation in the human retina (33).

2.9 Carotenoid Cleavage Enzymes

BCO1 and BCO2 are carotenoid cleavage enzymes responsible for the cleavage

of carotenes (48) and xanthophylls (49) respectively. While BCO2 can cleave both

carotenes and xanthophylls, BCO1 requires its substrates to have at least one non-

substituted β-ionone ring and therefore cannot cleave xanthophylls. BCO2 cleaves the

asymmetric lutein molecule to yield 3- hydroxy-β-apo-10’-carotenal and 3-hydroxy-α- apo-10’-carotenal (50). The cleavage of the symmetric zeaxanthin molecule by BCO2 yields 3-hydroxy-β-apo-10’-carotenal. BCO1 and BCO2 have been immunolocalized to

the human retina and RPE (51). Thus, researchers are curious about the accumulation of

xanthophylls in the retina in the presence of BCO2. One study suggests that BCO2 is

inactive in the human retina leading to the accumulation of lutein and zeaxanthin (52)

while another proposes that BCO2 is active but localized to the inner mitochondrial

membrane away from lutein and zeaxanthin (53). Thus, it is unclear what role if any that

BCO1 or BCO2 plays in macular carotenoid accumulation.

2.10 Xanthophyll Function in the Retina

Xanthophylls are antioxidants preferentially accumulated within the retina. While

it is not known exactly how they function within the retina there is evidence that

xanthophylls filter blue light protecting the eye from photoxidation, act as antioxidant to

protect from free radicals created, and possibly interact in signaling pathways. Both

lutein and zeaxanthin absorb blue light protecting the retina from the photoxidation to

which it is routinely exposed. In quail, long-term xanthophyll supplementation results in

increased xanthophyll in the retina with subsequent protection from light induced

26 photoreceptor death (54). Rhesus monkeys supplemented with lutein and zeaxanthin following a deficiency resulted in protection of the fovea from blue light damage (55).

Lutein and zeaxanthin may act as quenchers of singlet oxygen and reactive oxygen species protecting the retina from oxidative stress and inflammation (56). One study showed that along with the known actions of DHA in neuroprotection of photoreceptors, xanthophylls promote differentiation of photoreceptors (57). Their results suggest that xanthophylls may activate cell-signaling pathways to support cell development outside of their antioxidant capacity. However, their experiments did not elucidate a mechanism for this action.

2.11 Age-Related Macular Degeneration (AMD)

Age-related macular degeneration is an eye condition most common in adults 50 years an older, which leads to central vision loss (1). The disease is characterized by deterioration of the macula, which is responsible for central vision and fine detail. There are two types of AMD: Wet AMD and Dry AMD. Wet AMD occurs due to abnormal blood vessel growth below the retina leading to leakage and blurred vision. Dry AMD, the most common form, occurs due to aging and thinning tissues, breakdown of the RPE, and formation of drusen. Environmental factors, such as smoking and exposure to chronic oxidative stress can, as well as genetic susceptibility and aging all factor into development of AMD.

The RPE is essential in maintaining the health of the retina by providing nutrients from the choroid to the photoreceptors, disposing of metabolic waste, maintaining turnover of the photoreceptor outer segments, and protecting the retina from

photooxidation (34). Photoreceptor outer segment phagocystosis may contribute to the development of AMD. Over time, the RPE is exposed to high amounts of oxidized lipids and proteins from phagocytosis of the photoreceptor outer segments. In addition, the

RPE is exposed to high amounts of oxygen radicals since it is exposed to light and oxygen. RPE cells start to decrease in number later in life leading to slower turnover of the photoreceptor outer segments. Exposure to these combined stresses leads to atrophy of the RPE cells. AMD begins in the macula, which contains the highest turnover of photoreceptor segments as well as exposure to light. Because there is no cure for AMD, scientists are very interested in learning about mechanisms of the disease as well as how nutrients present in the eye may protect against the disease.

Lutein and zeaxanthin are found throughout the retina and selectively accumulate

in the fovea of the macula. They likely function as filters of blue light while also acting

as an antioxidant to prevent DNA damage in the RPE. A higher amount of xanthophylls

in the RPE may protect from lipofuscion accumulated during phagocytosis of the

photoreceptor outer segments. The first Age-Related Eye Disease Study (AREDS)

conducted by the National Eye Institute released results of their study which showed that

adults showing early stages of AMD had a 25% lowered risk after >5 years when given

high doses of vitamin E, zinc, vitamin C, and β-carotene. As a result of this study, they

observed that individuals consuming the highest amount of carotenoids that included

lutein and zeaxanthin from foods had a 43% lower risk of developing AMD (58). This

led to a second clinical trial, AREDS 2, designed to research the effects of lutein, zeaxanthin, and DHA/EPA on progression to advanced AMD in those at risk. Results of

the study show daily supplementation with lutein/zeaxanthin and the original AREDS

formulation (high doses of vitamin C and E, beta-carotene, zinc and copper) showed no

significant overall effect on progression to advanced AMD (59). However, a potential

interaction was found when people were supplemented with beta-carotene along with

lutein and zeaxanthin compared with those supplemented with these two xanthophylls

and no beta carotene. When beta carotene was replaced with lutein and zeaxanthin in the

formulation, participants had an 18% reduced risk of developing AMD than those

supplemented with beta carotene in the formulation and no zeaxanthin and lutein.

Competition between carotenoids in the body may have resulted in β-carotene masking

the effects of lutein and zeaxanthin. Further research is needed to clarify whether

supplementation of lutein and zeaxanthin in the AREDs formulation without beta-

carotene might slow the progression of AMD.

Initial results of the AREDs 1 study support high intakes of lutein and zeaxanthin in the diet to a lowered risk for AMD progression. However, AREDs 2 study results did not see an effect, but found a potential interaction of lutein and zeaxanthin when taken along with beta-carotene. Elucidating the mechanisms of lutein and zeaxanthin uptake and metabolism in the RPE along with other nutrient interactions is an important step in determining its action within the cell and for the design of future supplements and diet recommendations.

Chapter 3: Lipoprotein Separation and Xanthophyll Delivery to ARPE-19 Cells -

Methods Development

3.1 Introduction

Previous research in our lab has shown that lutein associates more with HDL and

less with LDL within human plasma (5,60) than does beta-carotene. We hypothesized that lutein and zeaxanthin delivered via HDL to the ARPE-19 cell line will result in

greater uptake than when delivered via LDL. Because our eventual goal is to compare

delivery of xanthophylls to ARPE-19 cells via LDL and HDL, we evaluated which

animal serum (fetal bovine, calf, adult bovine, horse, and human) provided the best

separation of lipoprotein classes using a simple ultracentrifugation technique using an iodixanol gradient. The previous classic technique for lipoprotein fractionation consisted of adding potassium bromide to form density gradients to separate lipoprotein fractions

(61). These methods required long hours of centrifugation (24-78 hours) and overnight dialysis to remove salts prior to most other applications. In addition, the high salt gradient can elute apolipoproteins and interfere with the hydration status of lipoproteins

(62). Iodixanol is a nonionic, endotoxin tested medium capable of creating self- generated, continuous gradients for human lipoproteins in approximately 3 hours (63,64).

In addition, dialysis to remove salts is not required prior to agarose gel electrophoresis

and other methods.

We evaluated different animal serums, including fetal bovine, calf, adult bovine,

horse, and human serums, to determine which would provide the best vehicle for

lipoprotein delivery of these carotenoids. We developed simple techniques to separate

LDL and HDL from the serums, enrich the serums with lutein, and quantify lutein within these lipoprotein fractions. In addition, we measured the rate of lutein uptake delivered via fetal bovine serum to the ARPE-19 cell line at different weeks of differentiation to determine the best time to study xanthophyll delivery for future experiments. Finally, we established an enrichment and delivery method of LDL, HDL, and whole serum to ARPE-19 cells for our future studies

3.3 Materials and Methods

Chemical and Supplies

Unless otherwise stated, supplies and chemicals were purchased from Fisher

Scientific (Pittsburgh, PA or Sigma Alderich (St. Louis, MO). Fetal bovine, calf, adult bovine, and horse serums were received as samples from Sigma Alderich. Human serum

(type AB, sterile) was purchased from Valley Biomedical (Winchester, VA).

Cell Culture

Human retinal pigment epithelial cells in the form of ARPE-19 cells (ATCC®-

CRL-2302™) were purchased from the American Type Culture Collection (Rockville,

MD). ARPE-19 cells were maintained in Ham's F12 Media: DMEM (1:1) (Dulbecco's modified Eagle's medium) with 10% fetal bovine serum (FBS) (Gibco, Life

Technologies, Inc.) as monolayers at 37°C with 5% CO2 in T-75 flasks. Cultures of

ARPE-19 cells were seeded at 1.5 X 105 cells/cm2 with Ham’s F12 Media: DMEM (1:1) with 10% FBS for differentiation experiments on 6-well flat bottom plates. For cell differentiation experiments, cells were plated and allowed to become confluent after 7 days when the first week of differentiation started. The cell medium was changed 2-3 times per week.

Lipoprotein Fractionation

LDL and HDL lipoproteins were isolated from fetal bovine, calf, adult bovine, horse (Sigma Aldrich, St. Louis, MO), and human serum (type AB, sterile, Valley

Biomedical, Winchester, VA) using a method previously developed (64,65). Briefly,

33 chylomicrons were first removed by centrifugation at 100,000 g for 10 minutes in a

Beckman Coulter Optima L-90K Ultracentrifuge. Isolation of VLDL, LDL, and HDL from chylomicron-free plasma was performed by using a continuous iodixanol gradient consisting of a either 12% (wt:vol) iodixanol-plasma mixture or a combination of 12%

(wt:vol) iodixanol-plasma mixture, a 20% iodixanol-saline solution, a 6% iodixanol- saline solution and saline. The 12% iodixanol in serum was prepared by combining a

60% iodixanol solution with human serum and the placed in a 4.7 ml polypropylene centrifuge tube. The rest of the resulting solution was either overlayed with PBS or overlayered on top of 20% iodixanol with a final 6% iodixanol solution on the top. The tube was capped and centrifuged in a Beckman Coulter Optima™ TLX Ultracentrifuge at

350,000 g for 2.5 hours at 16°C. Lipoprotein gradients were removed by tube puncture using a syringe, volumes recorded, and collected into separate vials. Lipoprotein fractions were confirmed using agarose gel electrophoresis and staining with Sudan

Black. Protein amounts in human serum and lipoproteins were measured using the

Modified Lowry Method (Thermo Scientific Pierce Modified Lowry Method kit).

Collected lipoproteins were used immediately following isolation.

Agarose Gel Confirmation

Agarose gel confirmation was used for confirming lipoproteins in collected fractions using the method described above. Freshly prepared lipoprotein fractions and whole serum were evaluated using agarose gel electrophoresis on precast 1% agarose gel (mini agarose gel, BioRad Laboratories, Philadelphia, PA). Before loading samples onto the gel, the agarose gel was pre-run at 100V for 10 minutes in 1X barbital buffer to allow the

gel to equilibrate in running buffer. Samples were mixed in loading buffer (40%

sucrose/0.25% bromophenol blue) in a ratio of 3:1 and 12 ul of the resulting mixture was

loaded onto the gel. Electrophoresis (BioRad Mini-Sub Cell GT and BioRad power

supply, BioRad Laboratories, Philadelphia, PA) was performed at 100 V for 1 hour or

until the dye reached the edge of the gel. After electrophoresis, the gels were fixed for 15

minutes in a 5% glacial acetic acid solution in 75% ethanol. After fixation, the gel was

dried overnight to facilitate staining of the gel. For staining, the gel was immersed in a

working solution of Sudan Black B stock (200 mg Sudan Black B in 100 ml 60%

ethanol) mixed with 60% ethanol (1:1) for 1 hour. The stained gel was then de-stained in

60% ethanol for 1-2 hours. Lipoprotein bands were visualized using an Odyssey Licor imaging system (Licor Biotechnology, Lincoln, NE) at 700 nm.

Carotenoid Enrichment of Serum or Lipoproteins

Whole serum or lipoproteins isolated by centrifugation were enriched using a procedure previously reported (5). Carotenoids were added to serum or lipoproteins at the indicated concentrations dissolved in ethanol (zeaxanthin and lutein) or tetrahydrofuran (β-carotene) (<2% final volume). The solution was vortexed gently, sonicated in a water bath sonicator, and incubated under nitrogen at 4°C in the dark on a mixer overnight for 24 hours. Aliquots of the carotenoid-enriched serum or lipoprotein solution were removed for extraction and HPLC analysis against standards to confirm the initial concentration added to cells prior to experiments. Carotenoid-enriched serum or lipoproteins were added to cell medium to reach the designated carotenoid concentration for that experiment.

Carotenoid Extraction

Extraction of cells and medium was performed as previously described with slight modifications (66,67). For human serum or lipoproteins, a volume (1 volume) was treated with 2-propanol-dichloromethane (2:1, v/v) (3 volumes) and vortexed. All samples were kept on ice under yellow light to prevent oxidation during extraction. The mixture was centrifuged for 3 minutes at 2,000 RPM. The top layer was removed and dried under nitrogen. The resulting residue was re-suspended in 200 ul of mobile phase, filtered through a 0.22 µm pore sized filter and run on HPLC. For carotenoid extraction from cells, cell medium was removed and cells were rinsed with 2 mls of ice-cold PBS followed by the addition of 1 ml of 2-propanol-dichloromethane (2:1) for 30 minutes.

This was performed 3 times at room temperature. Extracts were then collected, dried under nitrogen, re-suspended in mobile phase, filtered through a 0.22 µM pore sized filter and run on HPLC.

HPLC Analysis

Lutein, zeaxanthin, and β-carotene were analyzed using an Agilent Technologies

1200 Series Diode Array and Multiple Wavelength Detector HPLC system (Santa Clara,

CA) using a method previously described (68,69). A column C30 Type Carotenoid, 4.6

X 250 mm, 3 μm (YMC, Inc., Milford, MA) was used with methanol-methyl-tert-butyl- ether (90:10, v/v) at a flow rate of 0.9 ml/min as mobile phase. Carotenoids are monitored at 450 nm and quantified using external standard curves established for each carotenoid tested.

3.4 Results and Discussion

Delivery of lutein as a function of time and cellular differentiation

Previous research in our lab determined that lutein and zeaxanthin delivered by

detergent micelles to ARPE-19 cells had maximum uptakes between weeks 6-10 of

differentiation. Prior to evaluation of xanthophyll enrichment of animal serum, we

determined maximal uptake of lutein ARPE-19 cell via fetal bovine serum as a function of time and cellular differentiation. To start the experiment, ARPE-19 cells were grown to achieve 1.0 X 105 cells per well on flat bottomed 6-well plates. Once cells were plated, they were allowed to differentiate for 7 days after which the experiment began.

The experiment was performed over a 10-week period with 4 plates assigned at weeks 2,

4, 6, 8, and 10. For each week point, a plate was reserved for the following time-points:

15 minutes, 2 hours, 4 hours, 7 hours, and 24 hours. An exception to this set-up occurred

during week 4 when only 3 plates were available and assigned as 2, 7, and 24 hours.

Results of this experiment are in agreement with a previous experiment (10) showing that

the highest lutein uptake in the ARPE-19 cell line occurs between 6-10 weeks of

differentiation (Figures 3.1 and 3.2) coinciding with the 5-week time point in which these

cells develop RPE characteristics (70). However, in contrast to the previous study, we

found very high uptake for lutein in all weeks with a maximum uptake of 94% of the

initial 1 µM amount added after 24 hours during week 6 and a minimum of around 30%

at weeks 2 and 4 after 24 hours. Total recovery of lutein from what was initially added

was high for all weeks. Total lutein recovery (from cells and medium) declined over time

except during weeks 6 and 8 when most of the initial amount of lutein was recovered in

the cells making the total lutein recovery higher (Figure 3.2). Previous research has also shown that expression of RLBP1 and RPE65 mRNA, specific markers highly expressed in differentiated RPE cells under similar conditions, increased significantly between 6-8 weeks (71). Thus cells for experiments were used when they were differentiated for greater than 6 weeks.

Evaluation of Whole Animal Serum

Prior to using the iodixanol method for lipoprotein separation, we evaluated electrophoretic mobility of whole fetal bovine, calf, adult bovine, horse, and human serum using agarose gel electrophoresis. Agarose gel electrophoresis is a common technique where a small amount of serum is run on an agarose gel, lipoproteins are separated by size and are visualized after staining with Sudan Black. Research indicates that fetal bovine serum does not contain VLDL and contains about equal amounts of LDL

(43.8 mg/100 ml) and HDL (34.5 mg/100 ml) (72). Calf serum contains a very small amount of VLDL and a larger amount of HDL (167.4 mg/100 ml) than LDL (30.7 mg/100 mls). An adult bovine’s serum also contains VLDL and a much higher amount of HDL (365 mg/100 mls) compared to LDL (73.4 mg/100 mls). Similarly, horse serum contains a higher proportion of HDL than LDL. Human serum, on the other hand, contains a higher proportion of LDL than HDL. Agarose gel electrophoresis of fetal bovine, calf, horse, and human resulted in a variable separation of lipoprotein classes.

Due to the similarity in density of bovine lipoprotein classes, adult bovine lipoprotein agarose gel bands were much closer together than other animal lipoprotein bands (data not shown). This is in agreement with other studies indicating that bovine lipoproteins

contain a light and heavy class of HDL (73-75). Light HDL seems to overlap in density to LDL in bovine serum making it difficult to separate the two lipoproteins on agarose gel. Based on our results, we decided against using fetal and calf serum due to the low amount of lipoproteins in general within serum. We narrowed our focus for iodixanol separation of lipoproteins to bovine and horse serum due to the high amount of HDL and human serum since it is more biologically relevant to the human ARPE-19 cells.

Lipoprotein Separation

The iodixanol method of lipoprotein separation has been implemented in humans, but not other animals except for the adult bovine (73). Our goal was to implement the

centrifugation method in adult bovine, horse, and human to find the best fit serum for

future xanthophyll enrichment of isolated lipoproteins. Iodixanol can be used several

ways to generate gradients that separate lipoprotein classes. We experimented with

mixing serum with 12% (w:v) iodixanol which has been shown to separate human

VLDL, LDL, and HDL (64). In addition, we experimented using equal volumes of 6% and 12% iodixanol gradients which separates LDL subclasses more readily and has been shown in adult bovine to separate HDL from VLDL, but not LDL from HDL (74). In agreement with other results, we were able to separate human lipoprotein fractions

VLDL, LDL, and HDL (Figure 3.3). Similarly, horse lipoproteins were separated using

this method (Figure 3.4). Using other ultracentrifugation techniques, horse lipoproteins

were found to separate similarly to human lipoproteins (75,76). In agreement with other

studies, bovine lipoprotein fractions were not as well separated (results not shown) due to

the complex lipoprotein classes with overlapping light LDL and HDL (73,74). Although

horse serum lipoproteins were well-separated, the ratio of HDL to LDL is higher than

human serum. Figure 3.5 shows a centrifuge tube after separation of lipoproteins in

human serum using the iodixanol method. VLDL floats to the top of the tube followed

by LDL, HDL, and serum proteins on the bottom. Based on these results, we decided to

use human serum for future lipoprotein separation since it is easiest to separate using the

iodixanol gradient technique and it the most biologically relevant for delivery of

carotenoids to human ARPE-19 cells.

Carotenoid Distributions Among Human Lipoproteins

After confirming that the iodixanol method was successful in separating human

lipoprotein classes, we determined the carotenoid-lipoprotein distribution in human serum. We evaluated endogenous amounts in each lipoprotein fraction for β-carotene,

lutein, and zeaxanthin. After centrifugation of human serum and separation and removal

of lipoprotein fractions, the fractions were analyzed on agarose gel with Sudan black

staining. For this experiment, five fractions were collected into separate vials from the

tube starting at the top via tube puncture with a syringe. An aliquot of each fraction was

removed for agarose gel electrophoresis confirmation of lipoproteins and for HPLC

analysis of each carotenoid. As shown in figure 3.6, VLDL was mostly present in

fraction 1 (F1), LDL was present in fractions 2 and 3 (F2 and F3) while HDL was mostly

present in fractions 4 and 5 (F4 and F5). The mixed fraction represents whole serum

which contains all lipoproteins. After removal of lipoprotein fractions, carotenoids (β-

carotene, lutein, and zeaxanthin) were extracted from each fraction as described in the

methods section and analyzed using HPLC. Each carotenoid was quantified and

40 compared to the total amount of that carotenoid present in whole serum (Figure 3.7). β- carotene mostly associated with the LDL fraction (64 ± 0.4%) followed by HDL (25 ±

2%) and VLDL (10 ± 1%). Lutein and zeaxanthin mostly associated with HDL (54 ± 9% and 51 ± 14%) followed by LDL (36 ± 4% and 40 ± 10%) and VLDL (10 ± 5% and 8 ±

3%). These data are agreement with other studies showing similar carotenoid distributions among lipoproteins (5,60,77).

Comparing Cell Uptake with Different Lipoprotein Enrichment Methods

Our next goal was to compare ARPE-19 cell uptake when whole serum was enriched with lutein and then lipoproteins separated and added to cells or when lipoproteins were first isolated, then enriched with lutein, and added to cells. We then compared these results with a third experiment where we determined if once a particular lipoprotein was enriched with lutein if the presence of the other unenriched lipoprotein would change the cell uptake. In the first experiment, LDL and HDL were separated using the iodixanol gradient described in the “Materials and Methods” section. Equal volumes of LDL, HDL, and whole serum were enriched with 0.8 µM lutein (final concentration when added to cells and cell medium) as described previously and lutein cell uptake was measured at 0, 4, 6 and 24 hours. This was compared to the second method where whole human serum was first enriched with lutein overnight as described previously and then the lipoproteins were separated using the iodixanol method. Once

LDL and HDL were isolated, we used an aliquot to measure the lutein concentration using HPLC. Next, we calculated the amount of water to add to each lipoprotein to achieve the same concentration of lutein so that the total volume and lutein concentration

41 were equal when added to cells. We achieved a final concentration of 0.2 µM and measured cell uptake at 0, 4, 6, and 24 hours. For the third experiment, we enriched whole human serum with lutein, separated the lipoproteins using iodixanol, and measured the lutein concentration. Like the second experiment, we added water to LDL, HDL, and whole serum so that their lutein concentrations were equal. We also reserved an aliquot of the diluted lipoprotein and serum for protein analysis using Modified Lowry method.

Next, unenriched LDL was added to enriched HDL, unenriched HDL was added to enriched LDL, and unenriched whole serum was added to enriched whole serum. The mixtures added to the cells all had the same volume (10% total volume added to cells), protein concentration (~4.5 mg protein), and lutein concentration (0.4 µM), with the exception of enriched/unenriched whole serum which had a protein concentration of 6.7 mg. Finally, we measured cell uptake of lutein at 0, 6, and 24 hours.

Figure 3.8 shows the results of the previous three experiments summarized as cell uptake as a percentage of the initial amount of lutein added to cells. Regardless of the delivery method or presence of another lipoprotein, after 24 hours, HDL resulted in the lowest cell uptake of lutein (Experiment 1 = 10%, Experiment 2 = 5%, and Experiment 3

= 10%). LDL delivery of lutein was highest in Experiment 1 (22%), Experiment 2

(20%), and Experiment 3 (25%) after 24 hours. Our results indicate that lutein cell uptake delivered by LDL, HDL, or whole serum is consistent regardless of delivery method or the presence of the other unenriched lipoprotein.

The purpose of our preliminary study was to determine the best conditions to complete our specific aims. We confirmed that the best time to deliver xanthophylls to

ARPE-19 cells is between 6-10 weeks of differentiation when xanthophyll uptake is

highest. Using iodixanol to separate lipoproteins from various sera, we determined that

human serum is best separated into lipoproteins using this method and is the most

biologically relevant vehicle to deliver xanthophylls to ARPE-19 cells. We also determined lipoprotein electrophoretic mobility and carotenoid distribution among lipoproteins. Since there was no noticeable difference in enrichment method of lipoproteins, we chose to use the post-enrichment of lipoproteins after centrifugation to ensure consistency in lutein concentration when delivering carotenoids to cells. With this information, we could better clarify the mechanisms of uptake in these cells when xanthophylls are delivered by lipoproteins.

3.5 Acknowledgements

We would like to thank Shiva Raghuvanshi for her assistance in developing protocols for agarose gel confirmation of lipoproteins and in the iodixanol separation of lipoproteins.

We would also like to thank Dr. Eunice Mah and Dr. Richard Bruno for their assistance in developing the iodixanol gradient for lipoprotein separation.

3.6 Figures and Tables

100

60 Week 2 % Lutein Recovered (from 50 Week 4 initial) 40 Week 6 Week 8 30 Week 10 20

0 0 5 10 15 20 25 30 Time (hours)

Figure 3.1 Lutein uptake as a function of weeks of cell differentiation.

Each time point per week = average of 4 cell wells.

Week 2

100 % lutein recovered in cells 50 from initial total 0 % lutein recovered in % Lutein 0.25 2 8 24 medium from initial Time (hours) % Total Recovery (Lutein Week 4 from initial)

100 50 % lutein recovered in cells from initial total

% Lutein 0 2 7 24 % lutein recovered in medium from initial Time (hours) Week 6

100 50 % lutein recovered in cells from initial total 0 % Lutein % lutein recovered in 0.25 2 7 24 medium from initial Time (hours)

Week 8 100 50 % lutein recovered in cells from initial total 0 % Lutein % lutein recovered in 0.25 2 7 24 medium from initial Time (hours)

Week100 10 50 % lutein recovered in cells from initial total 0 % Lutein % lutein recovered in 0.25 2 7 24 medium from initial Time (hours)

3.2 Lutein recovery from cells, medium, and total recovery at 2, 4, 6, 8, and 10 weeks of cell differentiation. N = average of 4 wells/week/time point. 46

HDL

VLDL

LDL

5 4 3 2 1

Figure 3.3 Agarose gel of whole human serum and lipoproteins stained with Sudan

Black.

Lane 1: Whole human serum; Lane 2: VLDL; Lane 3: LDL; Lane 4: LDL; Lane 5:

HDL.

HDL

VLDL

LDL

5 4 3 2 1

Figure 3.4 Agarose gel of whole horse serum and lipoproteins stained with Sudan

Black.

Lane 1: Whole horse serum; Lane 2: VLDL; Lane 3: LDL; Lane 4: LDL; Lane 5: HDL.

VLDL

LDL

HDL

Figure 3.5 Iodixanol fractionation of lipoproteins

HDL

VLDL

LDL

F1 F2 F3 F4 F5 Mixed Fraction

Figure 3.6 Agarose gel confirmation of isolated human lipoproteins.

F1-F5 represents fractions removed after tube puncture starting at the top of the tube and

ending at the bottom. The dominant lipoproteins are: Fl – VLDL, F2 – LDL, F3 – LDL,

F4 – HDL, F5 – HDL. The mixed fraction represents all of the lipoproteins in whole

serum.

100

30 % of carotenoid recovered

0 β-Carotene Lutein Zeaxanthin

VLDL LDL HDL

Figure 3.7 Carotenoid distribution among lipoproteins

Lipoprotein fractions from human serum were separated and endogenous levels of β- carotene, lutein, and zeaxanthin were measured in each lipoprotein fraction. Carotenoid amounts in each lipoprotein fraction are listed as a percentage of the total amount recovered in all lipoprotein fractions. Total recovery from lipoprotein fractions from the initial amount measured in whole serum was as follows: 110 ± 26 % β-carotene, 107 ±

30% lutein, and 113 ± 34% zeaxanthin. Data represent means ± SD of triplicate separations of lipoprotein fractions.

Experiment 1: Lipoproteins Enriched Post-Centrifugation 25 20 15 LDL 10 HDL

% Cell Uptake 5 Whole Serum 0 0 5 10 15 20 25 30 Time (hours)

Experiment 2: Lipoproteins Enriched Pre-centrifugation 25

15 LDL

10 HDL Whole Serum % Cell Uptake 5

0 0 5 10 15 20 25 30

Experiment 3: Enriched and Unenriched Lipoproteins Mixed 25 20 Enriched LDL + 15 Unenriched HDL

% Cell Uptake 10 Enriched HDL + 5 Unenriched LDL 0 Enriched Serum + 0 5 10 15 20 25 30 Unenriched Serum Time (hours)

Figure 3.8 Lipoprotein delivery of lutein and cell uptake using different methods

In Experiment 1, lipoproteins were separately enriched and then the same volume and

Continued 52

Figure 3.8 Lipoprotein delivery of lutein and cell uptake using different methods continued: concentration was added to cells. In Experiment 2, whole serum was enriched with lutein, lipoproteins separated, and diluted to the same volume and concentration and added to cells. In Experiment 3, whole serum was enriched with lutein, lipoproteins separated, diluted to the same concentration and mixed with unenriched lipoprotein so that volume, lutein concentration, and protein concentration was the same when added to cells. N = 3-4 cell wells/treatment.

Chapter 4: Mechanisms of Selective Delivery of Dietary Xanthophylls to Retinal Pigment

Epithelial (ARPE-19) Cells from Human Lipoproteins*

*Sara E. Thomas and Earl H. Harrison. J.Lipid Res. 2016. 57: 1865- 1878.

4.1 Abstract

The xanthophylls, lutein and zeaxanthin, are dietary carotenoids that selectively

accumulate in the macula of the eye providing protection against age-related macular

degeneration (AMD). To reach the macula, carotenoids cross the retinal pigment

epithelium (RPE). Xanthophylls and β-carotene mostly associate with HDL and LDL,

respectively. HDL binds to cells via a scavenger receptor class B1 (SR-B1)-dependent

mechanism while LDL binds via the LDL receptor (LDLR). Using an in-vitro, human

RPE cell model (ARPE-19), we studied the mechanisms of carotenoid uptake into the

RPE by evaluating kinetics of cell uptake when delivered in serum or isolated LDL or

HDL. For lutein and β-carotene, LDL delivery resulted in the highest rates and extents of uptake. In contrast, HDL was more effective in delivering zeaxanthin and meso- zeaxanthin leading to the highest rates and extents of uptake of all four carotenoids.

Inhibitors of SR-B1 suppressed zeaxanthin delivery via HDL. Results show a selective

HDL-mediated uptake of zeaxanthin and meso-zeaxanthin via SRB1 and a LDL-mediated

uptake of lutein. This demonstrates a plausible mechanism for the selective accumulation

of zeaxanthin > lutein and xanthophylls over β-carotene in the retina. We found no evidence of xanthophyll metabolism to apocarotenoids or lutein conversion to meso-

zeaxanthin.

4.2 Introduction

Age related macular degeneration (AMD) is an incurable disease in adults 55

years of age and older and is the leading cause of vision loss in this population (1). A

2014 meta-analysis predicts that 196 million people will have AMD by 2020 increasing

to 288 million by 2040 (2). AMD occurs due to deterioration of the macula located in the

retina of the eye impacting central vision and the ability to see fine detail. The

xanthophylls, lutein and zeaxanthin (Figure 4.1A and 4.1B), are dietary carotenoids of

interest since they accumulate in the retina of the eye and may provide protection from

AMD. Common xanthophylls in the human diet include lutein, zeaxanthin, and β-

cryptoxanthin (14) and sources include corn, kale, spinach, eggs, and broccoli (9). Meso-

zeaxanthin (Figure 4.1C), a stereoisomer of zeaxanthin, is present in the macula of the

eye but is not a common dietary component.

Of the 700 carotenoids found in nature, only about 25 are found in the diet and

human serum (78). The top 6 carotenoids in the human plasma include α-carotene, β-

carotene, β-cryptoxanthin, lutein, zeaxanthin, and lycopene (79). Although the

concentration of carotenoids in the serum vary widely among individuals, the typical

order from highest to lowest is β-carotene > lutein > zeaxanthin (80,81). Xanthophylls comprise about 20% of the carotenoids in the human plasma with a lutein: zeaxanthin ratio between 2:1 and 4:1 (79,82,83). Xanthophylls accumulate in the macula of the retina, imparting the yellow macular pigment color named macula lutea. The functions of xanthophylls in the macula are not fully understood, but they are thought to filter light preventing damage to the macula and might provide protection as antioxidants. Unlike

carotenoid concentrations in the blood, lutein and zeaxanthin represent 80% of

carotenoids in the retina while β-carotene is present in only trace amounts (35). In the

peripheral macula, lutein dominates over zeaxanthin with a ratio of between 2:1 and 3:1

(82,84). Moving closer to the central macula from the peripheral macula, the ratio

changes to predominantly zeaxanthin with a 1:2 ratio of lutein to zeaxanthin. About 50%

of the zeaxanthin within the central macula is present as a stereo-isomer of zeaxanthin called meso-zeaxanthin. It is unclear what mechanisms are responsible for the selective accumulation of xanthophylls and particularly zeaxanthin in the retina.

Meso-zeaxanthin is not found in detectable amounts in the blood, liver, or in the typical human diet to account for the concentrations found in the macula (85). It is hypothesized that lutein is converted to meso-zeaxanthin either enzymatically or induced by light (33,86). Thus, primates given a xanthophyll-free diet from birth followed by a lutein supplement showed the presence of meso-zeaxanthin in the retina while those provided no xanthophylls or with just zeaxanthin alone did not have meso-zeaxanthin in the retina (87). However, no studies have been able to show how this conversion occurs in the macula. Even so, the discovery of meso-zeaxanthin as an additional macular pigment xanthophyll has prompted supplement companies to promote products containing lutein, zeaxanthin, and meso-zeaxanthin for the prevention of AMD. When supplemented, meso-zeaxanthin is found in the human serum (88), leading researchers to investigate whether its supplementation could increase macular pigment and thus improve the outcome of retinal diseases like AMD. Human studies measuring serum concentration and macular pigment optical density have shown that taking a supplement

predominantly composed of meso-zeaxanthin and small amounts of lutein and zeaxanthin

results in the presence of all three xanthophylls in human serum and an increased macular

optical pigment density compared to those un-supplemented (89). Since meso- zeaxanthin is found in the blood and may increase macular pigment, it is worth investigating the mechanisms of its retinal uptake.

The RPE is similar to the blood-brain barrier for the retina in that it serves as a cellular and metabolic interface between the retina and the blood supply from the choroid

(90). RPE cells display polarity with a basolateral side of tight junctions creating a barrier to the choriocapillaris and an apical side where villi extend and perform phagocytosis on photoreceptors of the retina. This unique position of the RPE allows it to provide nutrients to the photoreceptor cells while also eliminating waste from the retina. Differentiated ARPE-19 cells are often used as a model for the study of retinal metabolism since they show structural and functional properties similar to the human

RPE (70). ARPE-19 cells express lipoprotein receptors necessary for studying lipoprotein delivery of carotenoids including SR-B1, SR-B2, LDLR, CD-36, and ABCA1

(10,91-96). Based on analysis of the expression of lipoprotein transporters and receptors in the retina and two different RPE-cell lines, including ARPE-19 cells, it was proposed that circulating LDL and HDL enter the basalateral side of the RPE via LDLR and the

SR-B1 (96).

Dietary carotenoids are incorporated into lipoproteins for distribution to various tissues in the human body. Previous work in our laboratory (5,60) demonstrated that xanthophylls associated mostly with HDL, while β-cryptoxanthin, lycopene, and β-

carotene predominantly associated with LDL in human serum. We also showed that

zeaxanthin delivered by detergent micelles to ARPE-19 cells is preferentially taken up via an SR-B1-dependent mechanism compared to β-carotene (97). The current work was designed to extend those studies and to investigate in detail the uptake and metabolism of

lutein, zeaxanthin, meso-zeaxanthin, and β-carotene delivered in their physiologically

relevant transport vehicles. Using differentiated ARPE-19 cells showing structural and

functional properties similar to human RPE cells (70), we used human serum and the

lipoproteins LDL and HDL. We evaluated the kinetics of uptake, the possible

interactions of the carotenoids, and the effects of the presence of other lipoproteins and

specific inhibition of SR-B1. We demonstrate that xanthophylls, lutein and zeaxanthin,

transported in both HDL and LDL show quite different uptake kinetics. Next, we show

that meso-zeaxanthin is very similar to zeaxanthin in its uptake. Our experimental

evidence provides strong support for HDL-dependent selective uptake of zeaxanthin and meso-zeaxanthin via SRB1 in contrast to an LDL-mediated uptake of lutein and β-

carotene. Finally, we found no evidence of xanthophyll conversion to meso-zeaxanthin

or lutein conversion to apocarotenoids providing evidence that RPE cells do not

extensively metabolize xanthophylls.

4.3 Materials and Methods

Materials

All-trans- lutein and all-trans-zeaxanthin (≥98% assay (UV) purity) were

purchased from Indofine Chemical Company, Inc. (Hillsborough, NJ). All-trans-β- carotene and solvents used for HPLC were purchased from Sigma Aldrich (Saint Louis,

MO). Meso-zeaxanthin was a gift from DSM Nutritional Products (Heerlen,

Netherlands). Recombinant human serum amyloid A (human SAA1 α except for the presence of an N-terminal methionine and substitution of asparagine for aspartic acid at position 60 and arginine for histidine at position 71) was purchased from PeproTech

(Rockyhill, NJ). Human and bovine serum albumins were purchased from Thermo

Scientific. Block lipid transport 1 (BLT-1) (≥98% assay HPLC purity) and dimethyl sulfoxide (DMSO) were purchased from Sigma Alderich.

Cell Culture

Human retinal pigment epithelial cells, ARPE-19 cells (ATCC®-CRL-2302™), were purchased from the American Type Culture Collection (Rockville, MD). ARPE-19 cells were maintained in Ham's F12 Media: DMEM (1:1) (Dulbecco's modified Eagle's medium) with 10% fetal bovine serum (FBS) (Gibco, Life Technologies, Inc.) as monolayers at 37°C with 5% CO2 in T-75 flasks. Cultures of ARPE-19 cells were seeded

at 1.5 X 105 cells/cm2 with Ham’s F12 Media: DMEM (1:1) with 10% FBS for

experiments on 6-well flat bottom plates. Cells were plated and allowed to become

confluent after 7 days. The cell medium was changed 2-3 times per week. Cells were

used after differentiation for 6-8 weeks. Pilot experiments confirmed that maximal

xanthophyll cell uptake occurs at this point of differentiation (results not shown).

Isolation of Lipoproteins

VLDL, LDL and HDL were isolated from human serum (type AB, sterile, Valley

Biomedical, Winchester, VA) using a method previously developed (64). Briefly, chylomicrons were first removed by centrifugation at 100,000 g for 10 minutes in a

Beckman Coulter Optima L-90K Ultracentrifuge. Human serum was then mixed with

OptiPrep™ (Sigma Alderich) (4:1 v/v, 12% iodixanol final concentration) and 3.5 mls was transferred to an OptiSeal™ tube (Beckman Coulter). The remaining tube was filled with Phosphate-Buffered Saline (PBS) (Gibco, Life Technologies). The tube was capped and centrifuged in a Beckman Coulter Optima™ TLX Ultracentrifuge at 350,000 g for

2.5 hours at 16°C. Lipoprotein fractions were removed by tube puncture using a syringe.

The syringe was inserted into the tube just below the lipoprotein band starting with

VLDL at the top followed by LDL and then HDL at the bottom. The volumes were recorded and collected into separate vials. Lipoprotein fractions were confirmed using agarose gel electrophoresis and staining with Sudan Black. Protein amounts in human serum and lipoproteins were measured using the Modified Lowry Method (Thermo

Scientific Pierce Modified Lowry Method kit). Collected lipoproteins were used immediately following isolation.

Carotenoid Enrichment of Human Serum and Lipoproteins

Whole human serum or lipoproteins isolated by centrifugation were enriched with carotenoids using a procedure previously reported (5). This method was previously

shown to successfully enrich the lipoprotein with the intended carotenoid without

influencing lipoprotein integrity or redistributing carotenoids among lipoproteins in

whole serum when incubated in vitro (60). Carotenoids were added to human serum or

lipoproteins dissolved in ethanol (zeaxanthin, meso-zeaxanthin, and lutein) or tetrahydrofuran (β-carotene) (<2% final volume) so that when diluted in serum-free medium they would meet the desired concentrations. The solution was mixed and incubated under nitrogen at 4°C in the dark on a mixer overnight for 24 hours. Aliquots of the carotenoid-enriched human serum or lipoproteins were removed for analysis to confirm the initial concentration of carotenoids added to cells prior to experiments.

Addition of Carotenoids to Differentiated ARPE-19 Cells

After 6-8 weeks of cell differentiation, cell medium was removed and cells were rinsed 3 times with PBS. For cell uptake experiments, carotenoid-enriched human serum or isolated lipoproteins were added to serum-free cell medium (10% final volume unless otherwise described) to meet the desired carotenoid concentration and added to cells. For

BLT-1 and SAA experiments, HDL protein was first measured using Lowry Assay and carotenoids were added to 10 ug HDL protein/ml to reach the desired final concentration in serum-free medium.

Carotenoid Extraction

Extraction of cells and medium was performed as previously described with slight

modifications (66,67). For human serum or lipoproteins, one volume was treated with 3

volumes 2-propanol-dichloromethane (2:1, v/v) and vortexed. All samples were kept on

ice under yellow light to prevent oxidation during extraction. The mixture was

centrifuged for 3 minutes at 2,000 RPM. The top layer was removed and dried under

nitrogen. The resulting residue was re-suspended in 200 ul of mobile phase, filtered through a 0.22 µm pore sized filter and injected on to the HPLC. For carotenoid extraction from cells, cell medium was removed and cells were rinsed with 2 mls of ice-

cold PBS followed by the addition of 1 ml of 2-propanol-dichloromethane (2:1) for 30

minutes. This was performed 3 times at room temperature. Extracts were then collected,

dried under nitrogen, re-suspended in mobile phase, filtered through a 0.22 µM pore

sized filter and injected on to the HPLC.

Method 1: HPLC Analysis of Carotenoids

Lutein, zeaxanthin, and β-carotene were analyzed using an Agilent Technologies

1200 Series Diode Array and Multiple Wavelength Detector HPLC system (Santa Clara,

CA) using a method previously described (69,97). A column C30 Type Carotenoid, 4.6

X 250 mm, 3 μm (YMC, Inc., Milford, MA) was used with methanol: methyl-tertiary-

butyl-ether (MTBE) (90:10, v/v) at a flow rate of 0.9 ml/min as mobile phase. When

only β-carotene was being measured, the same column was used but with a gradient of

75:25, v/v methanol: MTBE and a flow rate of 1.4 ml/min. Carotenoids are monitored at

450 nm and quantified using external standard curves established for each carotenoid

tested.

Method 2: Chiral Column Separation of Xanthophylls

Separation and identification of meso-zeaxanthin from zeaxanthin was analyzed using the same HPLC system described above except using a chiral column (ChiralPak

AD, 25 cm length X 4.6 mm ID, Chiral Technologies, Exton, PA) and a method previously described but with slight modifications (85). A three-step gradient was used starting with a mobile phase consisting of 94.5% hexanes and 5.5% 2-propanol for 40 minutes. From 40-50 minutes, 2-propanol was linearly increased to 15% while hexane was reduced to 85 percent. From 50-55 minutes the gradient of hexane and 2-propanol was changed to a 50:50 mixture and maintained for 15 minutes (70 minutes into the total run-time). At 70 minutes, the gradient was re-equilibrated to the initial gradient of 94.5% hexanes and 5% 2-propanol from 70-80 minutes. The flow rate during the run was 0.7 mls/minute and monitored at 453 nm.

Inhibition of SR-B1

BLT-1 is a chemical inhibitor of lipid transport via the SR-B1 pathway (98).

Differentiated ARPE-19 cells were pretreated with equal volumes (0.1% final volume) of either DMSO or BLT-1 dissolved in DMSO (10 µM) for 1 hour in serum-free cell medium. This concentration of BLT-1 has been shown to be effective in inhibition of lipid transport from HDL to SR-B1 (7,99). After removal of cell medium, the cells received the previous treatment in addition to 0.1 µM of zeaxanthin- or lutein- (final concentration when added to cell medium) enriched HDL or LDL (~10 ug protein/ml

final volume), respectively, in serum-free cell medium for 3 hours at 37°C. After the incubation, zeaxanthin or lutein was extracted from cells using the method previously prescribed. Lipid-free serum amyloid A (SAA) is an inhibitor of SR-B1-dependent binding and selective cholesterol uptake from HDL (100). SAA or BSA (control) were added to differentiated cells at a concentration of 10 ug/ml along with 0.1 µM of zeaxanthin- or lutein-enriched HDL or LDL (~10 ug/ml), respectively, in serum free medium for 3 hours at 37°C. The amounts of SAA and HDL added in this experiment have previously shown a significant decrease in SR-B1-specific HDL binding. After incubation, zeaxanthin or lutein was extracted from cells as described previously above.

Statistical Analysis

Statistical analyses were conducted using R Data Analysis Software. Values are listed as means ± the standard deviation (SD). Results were analyzed using a mixed- effects ANOVA model. Where appropriate, this was followed by a multiple comparison of means using Tukey contrasts. P values <0.05 were considered significant.

4.4 Results

Lipoprotein Separation and Carotenoid Distribution

After centrifugation of human serum and separation and removal of lipoprotein

fractions, the fractions were analyzed on agarose gel with Sudan black staining. Figure

4.2 shows the presence of only LDL and HDL staining in lanes 1 and 2, respectively, and

the presence of all lipoproteins in whole serum in lane 3. After removal of lipoprotein

fractions, carotenoids (β-carotene, lutein, and zeaxanthin) were extracted as described in

the methods section and analyzed using HPLC. Each carotenoid was quantified and

compared to the total amount of that carotenoid present in whole serum (Figure 4.3). β-

carotene mostly associated with the LDL fraction (64 ± 0.4%) followed by HDL (25 ±

2%) and VLDL (10 ± 1%). Lutein and zeaxanthin mostly associated with HDL (54 ± 9%

and 51 ± 14%) followed by LDL (36 ± 4% and 40 ± 10%) and VLDL (10 ± 5% and 8 ±

3%). These data are agreement with other studies showing similar carotenoid

distributions among lipoproteins (5,60,77).

Carotenoid Uptake from Whole Serum and Isolated Lipoproteins

We first studied the uptake of β-carotene, lutein, meso-zeaxanthin, and zeaxanthin from human serum enriched separately with each of the carotenoids (Figure 4.4).

Carotenoid uptake is expressed as a percentage of the initial amount added to cells.

Zeaxanthin uptake was highest in ARPE-19 cells (69 ± 20%) followed by lutein (20 ±

3%) and β-carotene (3 ± 0.01%) after 24 hours. Meso-zeaxanthin uptake was very

similar to zeaxanthin uptake (62 ± 2%). After establishing the kinetics of carotenoid

uptake in whole serum, LDL and HDL were isolated and enriched with carotenoids to

compare their delivery to cells by the isolated lipoproteins (Figure 4.5). Strikingly, HDL

delivered zeaxanthin (4.5A) and meso-zeaxanthin (4.5B) most efficiently at all 3 time points with a greater than 50% cell uptake at 24 hours and a statistically higher uptake than from LDL at all time points. In marked contrast and surprisingly, lutein (4.5C) was more efficiently delivered from LDL than HDL similar to β-carotene.

We next studied the concentration dependence of the initial rate of cell uptake of lipoprotein-delivered carotenoids. After separation and enrichment of lipoproteins with

1, 10, 20, 30, and 40 µM of zeaxanthin, meso-zeaxanthin, or lutein, we measured cell delivery after 3 hours. As shown in figure 4.6, the initial rate of uptake of all three compounds was linear over the concentration range used. The results confirm that HDL more efficiently delivers zeaxanthin (4.6A) and meso-zeaxanthin (4.6B) while lutein

(4.6C) is more efficiently delivered by LDL.

Influence of Other Carotenoids on Cellular Uptake

We next asked whether the presence of increasing amounts of another carotenoid would inhibit uptake of either zeaxanthin or lutein. Aliquots of human serum (5% (v/v) in serum-free medium) were enriched with 1 µM of zeaxanthin or lutein (final concentration after mixing with cell medium) and added to cells along with increasing amounts of 0, 1, 3, or 5 µM of β-carotene-, lutein-, or zeaxanthin- enriched whole serum

(5% (v/v) in serum-free medium). The amount of whole serum added to cells was equal in all treatments (i.e. 10% v/v) while only the amount or presence/absence of the carotenoid added varied. Cells were incubated with the appropriate carotenoid treatment for 3 hours before cells were extracted. Compared to the control treatment, the amount of

zeaxanthin taken up remained unchanged when increasing amounts of β-carotene and

lutein were present (Figure 4.7A) (P > 0.05). A small but significant increase (P < 0.05)

of 9% of lutein taken up occurred in the presence of 5 μM of zeaxanthin (Figure 4.7B)

likely reflecting the presence of a small amount of lutein in the added zeaxanthin. More

strikingly, the presence of increasing amounts of β-carotene resulted in an 8% (P < 0.05) and 41% (P < 0.001) reduction in delivery of lutein to cells at 3 μM and 5 μM of β- carotene compared to baseline, respectively (Figure 4.7B). In summary, zeaxanthin uptake to cells remained unchanged with increasing amounts of β-carotene and lutein, while lutein cell uptake decreased markedly with increasing amounts of β-carotene.

Impact of Excess Lipoprotein on Carotenoid Uptake

Since we found that LDL most efficiently delivers lutein while HDL most

efficiently delivers zeaxanthin to ARPE-19 cells, the goal of the next experiments was to determine if increasing amounts of LDL or HDL devoid of carotenoid would affect the cell delivery of 1 µM lutein-LDL or zeaxanthin-HDL respectively. Excess unenriched

LDL resulted in an overall decline in lutein-enriched LDL cell delivery (Figure 4.8A).

The addition of 1X, 3X, 5X, and 10X unenriched LDL resulted in a 5, 10, 15, and 27% decline in lutein-enriched LDL cell delivery compared to the control treatment. Adding increasing amounts of unenriched HDL had a marked impact on zeaxanthin-enriched cell

delivery resulting in a 5, 2, 28, and 87% decline in zeaxanthin-enriched HDL cell

delivery compared to the control treatment (Figure 4.8B). It should be pointed out that

these effects are not merely due to the “dilution” effect since we have previously shown

that xanthophylls do not exchange among lipoproteins when incubated in vitro (60).

Effect of Blocking SR-B1 on the Uptake of Zeaxanthin from HDL or Lutein from

LDL

A chemical inhibiter, BLT-1, and SAA, a protein ligand of SR-B1 were used to

block HDL-dependent uptake of zeaxanthin into ARPE-19 cells (Figure 4.9). After pre-

incubation with 10 uM BLT-1 in serum-free medium for 1 hour followed by co-

incubation with 10 uM of BLT-1 and 0.1 uM zeaxanthin-enriched HDL cell medium for

3 hours, BLT-1 decreased zeaxanthin uptake by 52% compared to the control. Similarly, after exposing cells to 10 ug/ml of SAA and 0.1 uM of zeaxanthin-enriched HDL cell medium for 3 hours, zeaxanthin cell delivery was decreased by 49% compared to the control treatment. Conversely, we repeated the same experiments using lutein-enriched

LDL (Figure 4.9) and saw no significant difference compared to the control treatment when BLT-1 and SAA were used to block SR-B1. These results suggest that zeaxanthin entry into ARPE-19 cells via HDL is largely SRB1-dependent while lutein entry via LDL is not.

Carotenoids are not Extensively Metabolized by ARPE-19 cells

β-carotene 9’-10’-oxygenase (BCO2) cleaves non pro-vitamin A carotenoids, lutein and zeaxanthin, to create 3-hydroxy-apo-10’-carotenals (50). BCO2 cleaves the asymmetric lutein molecule to yield 3-hydroxy-β-apo-10’-carotenal and 3-hydroxy-α- apo-10’-carotenal. The cleavage of the symmetric zeaxanthin molecule by BCO2 yields

3-hydroxy-β-apo-10’-carotenal. Previous research in our lab using the same method 1 described in “Materials and Methods” to detect 3-hydroxy-apo-10’-carotenals (50), shows that these would appear in our chromatograms with retention times between 4-5

minutes. To determine if ARPE-19 cells formed these 3-hydroxy-apo-10’-carotenals, we looked carefully at the uptake and recovery of xanthophylls.

HPLC chromatograms (Method 1) of lutein, zeaxanthin, and meso-zeaxanthin standards are shown in Figure 4.10A and 4.10B. Differentiated ARPE-19 cells incubated with lutein-enriched human serum for 24 hours had a HPLC profile of 93 ± 2% all-trans- lutein, 4 ± 1% all-trans-zeaxanthin, and 3 ± 1% of unidentified peaks (Figure 4.10C).

Cell medium after 24 hours of incubation with lutein-enriched human serum included 91

± 4% all-trans-lutein, 5 ± 1% all-trans-zeaxanthin, and 4 ± 3% unidentified peaks. The

HPLC profile of the initial lutein-enriched human serum prior to cell addition contained

95 ± 1% all-trans-lutein, and 5 ± 1% all-trans-zeaxanthin. Zeaxanthin-enriched human serum prior to cell addition contained 96 ± 2% all-trans-zeaxanthin and 4 ± 2% all-trans- lutein. After 24 hours of incubation with zeaxanthin-enriched human serum, differentiated ARPE-19 cells contained 97 ± 1% all-trans-zeaxanthin, 1 ± 0.5% all-trans- lutein, and 2 ± 1% unidentified peaks (Figure 4.10D). Cell medium from the 24-hour incubation contained 92 ± 3% all-trans-zeaxanthin, 4 ± 1% all-trans-lutein, and 4 ± 2% other peaks. Lutein and zeaxanthin are highly conserved after 24 hours in cells and medium with unidentified peaks occurring more in cell culture medium than in cells. In no case were there any detectable metabolites at the retention times of the 3-hydroxy- apo-10’-carotenals (4.10C-4.10E). Rather, unidentified peaks are likely isomers of the xanthophylls identified in previous studies using the same HPLC conditions (69,97).

HPLC chromatogram profiles are similar to a study using micelles to deliver lutein and zeaxanthin to differentiated ARPE-19 cells (97). Thus, xanthophylls are not metabolized

to 3-hydroxy-apo-10’-carotenals in ARPE-19 cells. Furthermore, western blot analysis of

these cells did not detect any BCO2 protein (data not shown). The antibody did detect

BCO2 protein in other human cell lines.

It has been hypothesized that meso-zeaxanthin is formed from lutein in the macula

of the retina. To determine if meso-zeaxanthin is formed from lutein in the RPE, we incubated ARPE-19 cells for 24 hours with 10 µM lutein-enriched human serum and measured xanthophylls using the chiral column method 2. The chiral column is able to separate meso-zeaxanthin, zeaxanthin, and lutein as shown in the standard mixture in

Figure 4.11A. After 24 hours of receiving lutein-enriched human serum, the cell extract contained 95 ± 0.2% all-trans lutein, 3 ± 0.1% all-trans zeaxanthin, and 2 ± 0.1% other peaks. Cell medium contained 94 ± 0.1% all-trans lutein, 1 ± 0.1% all-trans zeaxanthin, and 5 ± 1% other peaks. Meso-zeaxanthin was not detected in cell extracts (Figure

4.11B) or cell medium (not shown) after 24 hours. Similarly, we incubated cells with 10

µM of zeaxanthin- or meso-zeaxanthin-enriched whole serum for 24 hours and measured the uptake of xanthophylls using chiral method 2 to look for any conversion to the other xanthophylls. After incubation with zeaxanthin, cell extracts contained 97 ± 0.2% all trans zeaxanthin, 1 ± 0.1% all trans lutein, and 2 ± 0.1% other peaks (Figure 11C) while cell medium contained 93 ± 1% all trans zeaxanthin, 0.4 ± 0.2% all trans lutein, and 7 ±

1% other peaks. Incubation with meso-zeaxanthin resulted in cell extracts containing 97 ±

0.5% meso-zeaxanthin, 1 ± 0.01 all trans zeaxanthin, and 1 ± 0.7% other peaks (Figure

4.11D) and cell medium containing 94 ± 1% meso-zeaxanthin, 0.5 ± 1% all trans zeaxanthin, and 5 ± 2% other peaks.

4.5 Discussion

Xanthophylls preferentially accumulate in the macula of the eye despite higher

levels of other carotenoids in the blood, indicating a selective mechanism of delivery in

the retina. Human studies have shown a correlation between increased serum

xanthophyll concentrations and increased macular pigment density (101). Xanthophylls

are transported mostly in HDL while carotenes such as β-carotene are transported mostly in LDL. The RPE is a highly selective transfer point from the blood vessels of the choroid where lipoprotein-bound xanthophylls must cross to reach the photoreceptors of the macular retina. Using differentiated ARPE-19 cells as a model for the RPE, we previously demonstrated that zeaxanthin delivered by detergent micelles is preferentially taken up compared to β-carotene via SR-B1 (97). We now show that despite both lutein and zeaxanthin being mostly associated with HDL in serum, zeaxanthin is much more efficiently delivered to ARPE-19 cells via HDL while lutein is more efficiently delivered via LDL. Thus, there is a selective uptake of zeaxanthin over lutein even though both are transported in HDL. Furthermore, our results show that zeaxanthin delivered by HDL occurs in an SR-B1 dependent process while lutein cell uptake from LDL may involve the LDL receptor. Given the similarities to zeaxanthin in structure and cell delivery, meso-zeaxanthin is also likely taken up from HDL in a SR-B1-dependent process in

ARPE-19 cells.

SR-B1 is a transmembrane glycoprotein homodimer embedded in the plasma membrane containing two N- and C-terminal transmembrane domains and a central extracellular domain clustered in caveolae-like domains (102). SR-B1 mediates selective

lipid uptake of HDL cholesterol to the liver and other steroidogenic tissues through

facilitated diffusion (103). The exact mechanism of selective lipid uptake of cholesterol

from HDL via SR-B1 is not well understood, but differs from the LDL receptor pathway

where LDL binds to the LDL receptor forming a clathrin-coated pit and the entire LDL

particle is taken up into the cell. Instead, SR-B1-mediated selective cholesterol uptake

occurs in a two-step process where the HDL particle binds to the extra-cellular domain

where selective delivery of cholesterol into the cell occurs without internalization of the

HDL particle (104). It was hypothesized that a lipophilic channel forms between the

HDL particle and the plasma membrane localized SR-B1 where selective uptake of

cholesterol esters occurs (105). This is supported by a recent publication identifying the

crystal structure of the extracellular domain of LIMP-2 (lysosome membrane protein 2)

and using this as a model for SR-B1 (106). LIMP-2 belongs to the CD36 superfamily of scavenger-receptor proteins along with SR-B1 and CD-36. In this model, they show a large, predominantly hydrophobic cavity running through the molecule where cholesterol esters are delivered from bound HDL through the plasma membrane.

Several studies report SR-B1 as the transfer protein for carotenoids in intestinal cell delivery (7,26,99,107) as well as the RPE (97). Genetic evidence links variants in several genes including SR-B1 to AMD risk in the Carotenoids in Age-Related Macular

Degeneration Study (CAREDS) and these variants were related to levels of lutein and zeaxanthin in human serum and the macula (104). However, given the role of SR-B1 in intestinal absorption of carotenoids and its possible role in other cell types, it is difficult to determine whether these effects are due to variants in SR-B1 at the intestinal level or

SR-B1 related uptake in the retina. Using ARPE-19 cells as a model for the RPE, we

were able to directly compare delivery of xanthopylls in LDL or HDL to ARPE-19 cells.

After 24 hours, zeaxanthin delivery in HDL is highest (66%) while lutein delivery from

HDL is only 13% and is higher when lutein is delivered in LDL (33%). We hypothesized

that HDL-dependent uptake of zeaxanthin occurs via SR-B1 while LDL-dependent

uptake of lutein may occur via the LDL receptor. Indeed, ARPE-19 cells express SR-B1

and LDLR and it was proposed that lipoproteins likely enter the RPE through these

lipoprotein receptors (96). Adding increasing amounts of unenriched LDL to lutein-

enriched LDL creates competition for entry into the cell and decreased lutein uptake by

up to 27 percent. Likewise, adding increasing amounts of unenriched HDL decreased

zeaxanthin-enriched HDL uptake by 87 percent supporting competition via SR-B1, an

HDL-specific receptor.

To further support SR-B1 as the protein transporter for zeaxanthin, we used BLT-

1, an HDL-specific inhibitor of SR-B1 selective lipid uptake (98). It has been effectively used to inhibit carotenoid uptake by SR-B1 (7,99,108) in intestinal cells and it does not inhibit other protein transporters (109). Interestingly, it’s been found that BLT-1 inhibits cholesterol uptake by SR-B1 in a Cys384-dependent manner (110). BLT-1 attaches to

Cys384 which is located in the lumen of the SR-B1 tunnel and blocks cholesterol transport. This is evidenced by the fact that when Cys384 is converted to serine, the effect of BLT-1 on cholesterol transport is lost. Thus it seems that Cys384 contributes to the ability of SR-B1 to mediate selective uptake which is inhibited by BLT-1. It is interesting to note that CD36 does not have the same Cys384 equivalent to SR-B1’s, yet

can bind to HDL but is unable to mediate efficient lipid uptake. In our study, BLT-1

inhibited HDL-delivered zeaxanthin cell uptake into ARPE-19 cells by 52 percent while no significant effect was seen with LDL-delivered lutein. Thus, it seems that zeaxanthin may be transported from HDL through the SR-B1 tunnel by a mechanism similar to cholesterol esters. This is consistent with results of another study showing that an SR-B1 antibody reduced zeaxanthin uptake in ARPE-19 cells by 58% while an antibody for CD-

36 had no effect (97). Lipid-free SAA, an acute inflammatory protein, was used as an inhibitor of SR-B1 due to its ability to compete with HDL for binding to SR-B1.

Zeaxanthin uptake was significantly decreased by 49% when lipid-free SAA was added further supporting SR-B1 as the zeaxanthin transport protein. Lipid-free SAA had no significant effect on LDL-delivered lutein.

Surprisingly, our results demonstrated that LDL was more efficient in delivering lutein to ARPE-19 cells, despite its greater extent of association with HDL in serum. One explanation could be that SR-B1 contains a zeaxanthin-specific transfer factor within the

SR-B1 tunnel that discriminates the chemical structure of zeaxanthin, so not only does it recognize HDL for binding and cholesterol transfer, but zeaxanthin as well. A similar function for SR-B1 was characterized in Drosophila Malannogaster where disruption of the NinaD gene codes for an SR-B1 protein that when disrupted causes blindness (111).

These authors further identified 2 isoforms of the NinaD gene located in different subcellular compartments where NinaD-1 acts on the plasma membrane to specifically mediate the uptake and transfer of zeaxanthin from micelles (112). Furthermore, NinaD-

1 uptake was preferential for zeaxanthin compared to β-carotene. The function of NinaD-

11, which is localized to intracellular membranes, is yet to be determined. Another

example is carotenoid delivery by lipophorin, a lipid transporter similar to HDL in

humans that occurs in the silkworm, Bombx mori (113). Two genes encode for a high-

density lipoprotein receptor-2 (Cameo2) and scavenger receptor class B member 15

(SCRB15) demonstrating selective affinity for xanthophylls and carotenes respectively.

Similar to xanthophylls and HDL, carotenoids are carried by lipophorin and occur as a

complex mixture in the lipoprotein yet Cameo2 and SCRB15 demonstrate selective

delivery of specific carotenoids. Likewise, while HDL may transport both lutein and

zeaxanthin, cell uptake to the RPE may occur selectively and specifically by SR-B1 for zeaxanthin. Although not typical in the human diet, meso-zeaxanthin, being a stereoisomer to zeaxanthin, is similar in structure and when consumed in the diet, may be recognized by SR-B1 and taken up into the RPE like zeaxanthin.

Our results show a 41% decline in lutein cell uptake in the presence of 5-fold molar excess of β-carotene when delivered in whole serum. This is indicative of competition occurring between lutein and β-carotene during uptake by the RPE.

Although lutein is found mostly in HDL, LDL delivers a higher amount to the RPE. β- carotene mainly associates with LDL in the blood; therefore, any excess β-carotene may somehow inhibit LDL-delivered lutein cell uptake. If translated to retinal uptake, a large amount of β-carotene present in the blood compared to lutein could reduce the amount taken up by the RPE to the retina thus reducing macular pigment and increasing the risk of AMD. This was suggested in the Age-Related Eye Disease Study 2 (AREDS2) randomized clinical trial (59). Oral supplementation of the original AREDs formulation

(vitamins C and E, β-carotene, and zinc) reduced the risk of progression to advanced

AMD. Adding lutein and zeaxanthin to the supplement did not further reduce the risk of

progression to advanced AMD. However, lutein serum levels were lower in those

receiving the original AREDs formulation including lutein and zeaxanthin along with β-

carotene compared to those with the lutein and zeaxanthin added and without β-carotene.

Post-hoc analyses showed that receiving the AREDs supplement without β-carotene but including lutein and zeaxanthin was even more impressive in reducing progression to advanced AMD compared to those receiving the original AREDs supplement with β-

carotene. Although further studies are needed, the results suggest an inhibitory effect on

the uptake of lutein by β-carotene. Together, these results suggest that the AREDs

formulation containing lutein and zeaxanthin without β-carotene may maximize the

delivery of xanthophylls to the retina.

Meso-zeaxanthin is present as approximately half the zeaxanthin found in the

central macula of the retina, but is not present in the peripheral regions of the macula

where lutein dominates. Meso-zeaxanthin is naturally found in shrimp carapace, fish

skin, and turtle fat which are not common foods consumed in the human diet (114-116).

As such, meso-zeaxanthin is not found in detectable amounts in human blood. Meso- zeaxanthin in the macula is likely converted from lutein as predicted by chemistry (117).

Support of this was found when monkeys fed a lutein-only diet accumulated meso- zeaxanthin while those fed zeaxanthin only did not (87). Similar results were found in leghorn chicks fed a xanthophyll-free diet compared to a control diet containing lutein and zeaxanthin (118). Furthermore, meso-zeaxanthin was produced in the retina of

leghorn chickens in a developmentally regulated manner (86). However, the exact

mechanism for this conversion in the retina is unknown. We were unable to detect any

conversion to meso-zeaxanthin when ARPE-19 cells were incubated for 24 hours with 10

μM lutein or zeaxanthin. Our results suggest that if meso-zeaxanthin is converted from lutein in the macula of the eye, this conversion does not occur robustly in the RPE or at least in our cell model. It’s possible that lutein and zeaxanthin are transported across the

RPE to the retinal cells where conversion of lutein to meso-zeaxanthin may occur.

Binding proteins, StARD3 (steroidogenic acute regulatory domain) and GSTP1

(glutathione S-transferase), were identified as having a high affinity for lutein and zeaxanthin in the central macula, respectively (46,117). GSTP1 was found to also have a high affinity for not only zeaxanthin, but meso-zeaxanthin, so it’s possible that these proteins are involved in the process of converting lutein to meso-zeaxanthin within the macula.

Immunohistochemical analysis has detected the presence of BCO2 in the human

RPE (119). However, the presence of BCO2 in the RPE does not explain why xanthophylls seemingly accumulate in the retina without BCO2 conversion to its metabolites. One explanation is that unlike mouse and chicken BCO2 enzymes, human

BCO2 enzyme is inactive in the retina (52) which explains why only primates accumulate xanthophylls. An amino acid insertion was discovered near the substrate binding tunnel of human BCO2 that when inserted into the mouse BCO2 enzyme leads to its inactivation. Additionally, zeaxanthin accumulated in the retinas of BCO2 knockout mice. However, another group finding a similar result of accumulation of zeaxanthin in

the retinas of BCO2 knockout mice offers an alternative explanation. They propose that

human retinal BCO2 resides in a different cellular compartment of the retina than the

xanthophylls (53,120). We were unable to detect any 3-hydroxy-apo-10’-carotenals in cell extracts after incubation with lutein or zeaxanthin nor did we detect the presence of

BCO2 protein.

In summary our results suggest that zeaxanthin is taken up by RPE cells via a mechanism of selective uptake from HDL similar to that of cholesterol ester by way of transfer through the lipophilic channel of SR-B1. Meso-zeaxanthin likely follows a similar pathway given its similar structure and kinetics of uptake compared to zeaxanthin.

We provide strong evidence that lutein is transported into the RPE by a different mechanism, perhaps involving the LDL receptor. This would also explain the inhibitory interaction between lutein and β-carotene, as both are taken up into the RPE mainly from

LDL. Finally, xanthophylls are transported intact into the RPE and we did not detect any measurable conversion to meso-zeaxanthin or to other metabolites. Thus the RPE acts as a transporter of xanthophylls from the choroid to the retina where conversion of lutein to meso-zeaxanthin may occur in the macula mediated by other binding proteins and/or enzymes.

These results suggest that lutein and zeaxanthin are selectively taken up by different mechanisms and that this leads to markedly preferential uptake of zeaxanthin.

The exact molecular mechanism by which zeaxanthin and meso-zeaxanthin are taken up preferentially from HDL that also contains lutein remains to be resolved. Lutein and β- carotene appear to interact for entry into the retina via LDL possibly due to competition

for the LDLR. Thus nutritional treatment of AMD should focus on supplementation of

xanthophylls without β-carotene. Future studies are needed to confirm that lutein uptake

occurs via LDLR in the RPE and to explore this mechanism in detail. The preferential

uptake of zeaxanthin compared to lutein and the possible conversion of lutein to meso-

zeaxanthin in the macula of the eye may indicate that focus should be placed on

consuming higher amounts zeaxanthin. For example, the AREDS2 supplements 10 mg

of lutein and 2 mg of zeaxanthin (59). Given the preferential uptake of zeaxanthin compared to lutein in the RPE and its lower amount in the blood compared to lutein, it remains to be seen if supplementing a higher amount of zeaxanthin compared to lutein would be more beneficial for lowering the risk of AMD.

4.6 Acknowledgements

The authors would like to thank Carlo dela Seña for assistance with HPLC identification of BCO2 cleavage products of lutein and zeaxanthin, Shiva Raghuvanshi for help with lipoprotein separation and identification, and Yan Yuan for western blot analysis of

BCO2 protein in ARPE-19 cells. The work was supported by NIH grant RO1-HL049879.

4.7 Figures

Figures

Figure 4.1 Macular xanthophylls

Structure of xanthophylls which selectively accumulate in the macula of the eye: Lutein

(A), zeaxanthin (B), and meso-zeaxanthin (C). Zeaxanthin and meso-zeaxanthin are

stereo-isomers while lutein differs in the placement of a double bond.

Figure 4.2 Agarose gel confirmation of lipoproteins

After ultracentrifugation, lipoprotein fractions were confirmed using agarose gel confirmation and staining with Sudan Black. Lane 1 contains whole serum and stains for

VLDL, LDL, and HDL. Lanes 2, 3, and 4 indicate a single band for HDL, LDL, and

VLDL fractions, respectively.

100

30 % of carotenoid recovered

0 β-Carotene Lutein Zeaxanthin

VLDL LDL HDL

Figure 4.3 Carotenoid distribution among lipoproteins

30% lutein, and 113 ± 34% zeaxanthin. Data represent means ± SD of triplicate separations of lipoprotein fractions.

Figure 4.4 ARPE-19 carotenoid uptake: delivery in whole serum

Kinetics of ARPE-19 cell uptake of 1 µM of β-carotene, lutein, zeaxanthin, and meso- zeaxanthin delivered in whole serum. Cells were grown to confluency in 6-well plates

(3-4 wells/treatment) and allowed to differentiate for 6 weeks. Carotenoid amounts are represented as a percentage of the amount of that carotenoid measured in whole serum prior to addition to cells. Carotenoids were measured using HPLC. After 24 hours, we recovered 92 ± 10% β-carotene, 92 ± 14% lutein, 103 ± 30% zeaxanthin, and 89 ± 3% meso-zeaxanthin in cells and medium. Data are means ± SD of 2-3 independent experiments.

Figure 4.5 Kinetics of carotenoid uptake from LDL and HDL

Continued

Figure 4.5 Kinetics of carotenoid uptake from LDL and HDL Continued:

ARPE-cell uptake of 1 µM of (A) zeaxanthin, (B) meso-zeaxanthin, (C) lutein and (D) β- carotene delivered in either LDL or HDL was measured at 0, 4, 6, and 24 hours. Cells were grown to confluency in 6-well plates (3-4 wells/treatment) and allowed to differentiate for 6 weeks. LDL and HDL were separated and enriched with 1 µM of the indicated carotenoid, added to cells, and uptake was measured at the indicated time points. Total zeaxanthin, meso-zeaxanthin, lutein, and β-carotene recovered from cells and medium delivered by: LDL (82 ± 4%, and 72 ± 9%, 98 ± 19%, and 88 ± 11%) and

HDL (103 ± 16% 86 ± 5%, 97 ± 8%, and 92 ± 4%) respectively. Data are means ± SD of two or more independent experiments. * P < 0.05, LDL vs. HDL at the time indicated.

Figure 4.6 Concentration-dependent carotenoid uptake from LDL or HDL

Continued

Figure 4.6 Concentration-dependent carotenoid uptake from LDL or HDL

Continued:

Different concentrations of (A) zeaxanthin, (B) meso-zeaxanthin, or (C) lutein were delivered in LDL or HDL and the amount of xanthophyll taken up by the cells was measured after 3 hours. Cells were grown to confluency in 6-well plates (2 wells/treatment) and allowed to differentiate for 6 weeks. LDL and HDL were separated and enriched with 1, 5, 10, 20, 30, and 40 µM of the indicated carotenoid, added to cells, and cell uptake of xanthophylls was measured by HPLC after 3 hours. Data are means ±

SD of 3 independent experiments.

Figure 4.7 Interactions of carotenoids during cell uptake

Impact of increasing concentrations (1, 3, 5 µM) of lutein or β-carotene on cell uptake of

1 µM zeaxanthin (A) and impact of increasing concentrations (1, 3, 5 µM) of zeaxanthin or β-carotene on cell uptake of 1 µM lutein (B). Xanthophyll cell uptake is represented as a percentage of the control without the added carotenoid ((1 µM zeaxanthin (A) or 1

µM lutein (B) only)). Whole serum was separately enriched with the given amounts of carotenoids and added at the same time. Total amount of whole serum added to each treatment was the same. Cells were analyzed for carotenoid content after 4 hours using

HPLC. Data are means ± SD of 3 independent experiments. * P < 0.05 compared with the control treatment

Figure 4.8 Impact of excess lipoprotein on carotenoid uptake

Continued

Figure 4.8 Impact of excess lipoprotein on carotenoid uptake Continued:

ARPE-19 cells received 50 ul of lutein-enriched LDL (A) or zeaxanthin-enriched HDL

(B) along with the following treatments: 0X (control), 1X, 3X, 5X, or 10X of un-enriched

LDL (LDL-lutein treatment) (A) or HDL (HDL-zeaxanthin treatment) (B). Cellular carotenoid content was measured after 3 hours and compared to the control (lutein- enriched LDL without excess LDL or zeaxanthin-enriched HDL without excess HDL).

Data are means ± SD >3 independent experiments. * P < 0.05 compared with the control treatment.

110

100

80 *

70 * 60

30 Xanthophyll (% Uptake Control) Xanthophyll

0 Zeaxanthin Lutein

BLT-1 SAA

Figure 4.9 Effect of inhibition of SR-B1 by BLT-1 and SAA on zeaxanthin or lutein uptake

ARPE-19 cells were pre-incubated for 1 hour with 10 µM of BLT-1 or control (DMSO).

After pre-incubation with BLT-1, medium was removed from cells and 0.1 µM of zeaxanthin-enriched HDL or lutein-enriched LDL was added to ARPE-19 cells along with 10 µM of BLT-1 or control. After 3 hours, cells were analyzed for zeaxanthin or lutein content using HPLC. SAA or control (bovine serum albumin) was added to ARPE-

19 cells at a concentration of 10 ug/ml at the same time as 0.1 µM of zeaxanthin-enriched

Continued

Figure 4.9 Effect of inhibition of SR-B1 by BLT-1 and SAA on zeaxanthin or lutein uptake Continued:

HDL or lutein-enriched LDL. After 3 hours of incubation, zeaxanthin or lutein content was measured in the cells using HPLC. The amount of zeaxanthin or lutein in the cells treated with SAA was compared with the control. Data are means ± 3 independent experiments. * P < 0.05 compared with the control treatment

Figure 4.10 Xanthophylls are not extensively metabolized to apocarotenoids by

ARPE-19 cells Continued

Figure 4.10 Xanthophylls are not extensively metabolized to apocarotenoids by

ARPE-19 cells Continued:

HPLC chromatograms of a mixture of lutein and zeaxanthin standards (A), meso-

zeaxanthin standard (B), 24-hour cell extract of lutein-treated cells (C), 24-hour cell extract of zeaxanthin-treated cells (D), and 24-hour cell extract of meso-zeaxanthin- treated cells (E). Under these conditions, 3-hydroxy-apo-10’-carotenals, the products of

BCO2 catalytic conversion of lutein and zeaxanthin, would be seen between retention times 4 and 5 minutes. A column C30 Type Carotenoid, 4.6 X 250 mm, 3 μm (YMC,

Inc., Milford, MA) was used with methanol-MTBE (90:10, v/v) at a flow rate of 0.9 ml/min as mobile phase. mAU = absorbance units, min = minutes.

Figure 4.11 Lutein is not converted to meso-zeaxanthin by ARPE-19 cells

HPLC chromatogram (A) showing the separation of meso-zeaxanthin, zeaxanthin, and lutein in a mixture of standards. HPLC chromatogram of cell extracts after 24 hours of incubation with 10 µM of lutein (B), zeaxanthin (C), or meso-zeaxanthin (D) in whole

Continued

Figure 4.11 Lutein is not converted to meso-zeaxanthin by ARPE-19 cells Continued: serum. A Chiralpak AD, 25 cm length X 4.6 mm ID column was used with a 3-step gradient as described in the “Materials and Methods” section. mAU = absorbance units, min = minutes

Chapter 5: Epilogue

AMD is an incurable disease among adults 55 years of age and older and is the

leading cause of vision loss in this population (1). Age related macular degeneration

occurs due to deterioration of the macula located in the retina of the eye (34).

Breakdown of the macula impacts central vision and the ability to see fine detail.

Xanthophylls, such as lutein, zeaxanthin, and meso-zeaxanthin, preferentially accumulate

in the macula of the retina in the eye, so supplementation of these nutrients has been the

focus of potential treatments for this incurable disease (3). Entry of nutrients from the

blood into the eye requires crossing over the highly selective RPE from the choroid. By

elucidating the mechanisms of this selective uptake, we can develop better treatments for

prevention of AMD. A previous study performed by our lab provides evidence that

zeaxanthin transport into the RPE is at least partially mediated by SR-B1 and zeaxanthin is preferentially taken up compared to β-carotene by this receptor (10). This previous study delivered 1 μM carotenoids using micelles and chylomicrons secreted from Caco-2

cells and showed an uptake of 1.6% for beta-carotene, 2.5% for lutein, and 3.2% for

zeaxanthin. The overall aim of this work was to study the mechanisms of xanthophyll

uptake when delivered by lipoproteins and to determine if any metabolism of

xanthophylls occurs in the RPE. The aim of Chapter 3 was to further develop a

biologically relevant delivery system to deliver carotenoids to the RPE using human

serum or lipoproteins so that we could further study xanthophyll cell uptake.

While developing a model for delivery of xanthophylls to ARPE-19 cells, we delivered

lutein enriched in fetal bovine serum and found that as much as 90% of the lutein

delivered was taken up by these cells. We established that the time of differentiation of

100

ARPE-19 cells between 6-8 weeks is the highest rate of cell uptake in agreement with other studies on ARPE-19 cell differentiation (10,70,71). We characterized different animal serum and human serum using agarose gel electrophoresis and determined that separation of human serum into lipoproteins was the best delivery method to ARPE-19 cells. Using an iodixanol gradient to separate human lipoproteins, we found that β- carotene resides mostly in LDL (64%), while lutein and zeaxanthin associate mostly with

HDL (54% and 51%, respectively) which is in agreement with other studies using different lipoprotein separation methods (5,60,77). This method of separating lipoproteins is fast, non-toxic to cells and does not interfere with the hydration status of the lipoprotein (62-64). We compared delivery of lutein to ARPE-19 cells using different human serum and lipoprotein enrichment techniques and determined that regardless of enriching before or after lipoprotein separation, lutein uptake into the cell was similar.

Furthermore, once a lipoprotein was enriched with lutein, the presence of another unenriched lipoprotein did not affect lutein uptake. Thus, once a lipoprotein is enriched with lutein, lutein does not appear to switch lipoproteins. Similar results were discovered in another study in our lab where unlike triglycerides and cholesterol esters which are known to transfer among lipoproteins via cholesterol ester transfer protein, carotenoids did not transfer among lipoproteins (60). In summary, in Chapter 3, we developed a solid model for xanthophyll enrichment, lipoprotein separation, and xanthophyll delivery in human serum and lipoproteins to ARPE-19 cells to study the mechanisms of xanthophyll uptake and metabolism.

101

Chapter 4 contains the bulk of the four research aims outlined in Chapter 1. In

our first aim, we hypothesized that since lutein and zeaxanthin are transported in the

blood mainly via HDL they must have similar uptake to the RPE, perhaps via SR-B1, an

HDL specific receptor. Since the previous study in our lab focused on zeaxanthin (10)

uptake into ARPE-19 cells, we initially narrowed our experiments to lutein cell uptake.

Surprisingly and unexpectedly, we found that lutein was more efficiently delivered to the

RPE via LDL, not HDL. Zeaxanthin was much more efficiently taken up by HDL as

hypothesized. One explanation for this difference could be the presence of a zeaxanthin-

specific transfer factor within the hydrophobic SR-B1 tunnel that discriminates the

structure of zeaxanthin from other carotenoids. One such example of this, as mentioned

in Chapter 4, occurs in the silkworm, Bombx mori (113). In the silkworm, carotenoids

are carried by lipophorin as a complex mixture, yet Cameo2 and SCRB15 demonstrate

selective delivery of specific carotenoids. To further investigate our first aim, we found that cell uptake of LDL-delivered lutein and HDL-delivered zeaxanthin decreased in the presence of increasing amounts of unenriched LDL and HDL, respectively. We summarized that although lutein and zeaxanthin are carried in the blood mainly by HDL, they differ in their mechanisms of delivery perhaps with lutein delivered by LDL to

LDLR and zeaxanthin delivered by HDL to SR-B1. This was further characterized in aim 3.

For aim 3, we found that only HDL-delivered zeaxanthin was significantly reduced when SR-B1 was blocked by BLT-1 or SAA. LDL-delivered lutein on the other hand was not significantly reduced with either treatment. Thus, this provides further

102

support that while zeaxanthin is delivered to SR-B1 via HDL, lutein may be delivered by

LDL. Future studies are needed to determine if blocking the LDLR would reduce lutein uptake into the RPE when delivered by LDL.

We investigated the three carotenoids supplemented in the AREDs trials, β-carotene, lutein, and zeaxanthin, and determined if the presence of the other would inhibit its uptake into the RPE as part of aim 2. We found that the presence of β-carotene significantly reduced lutein cell uptake. Zeaxanthin was unaffected by the presence of either β-carotene or lutein. Interestingly, the AREDs 2 study found that removing β- carotene from the AREDs supplement and replacing it with lutein and zeaxanthin was even more beneficial in reducing the progression to advanced AMD (59). Perhaps, given the similar uptake of β-carotene and lutein by LDL, there is an interaction at the LDLR of the RPE where β-carotene interferes with the cell uptake of lutein. On the other hand, the selective uptake of HDL-delivered zeaxanthin by SR-B1 is unaffected by the presence of

β-carotene or lutein.

Finally, we investigated whether ARPE-19 cells convert lutein to meso- zeaxanthin or form any BCO2 generated metabolites of lutein or zeaxanthin as part of our final aim. We carefully measured lutein, zeaxanthin, and meso-zeaxanthin uptake and found no evidence of any conversion to meso-zeaxanthin nor did we detect any 3- hydroxy-apo-10’-carotenals in cell extracts. Furthermore, we did not detect any BCO2 protein in ARPE-19 cells when measured using Western Blot analysis. This is in conflict with one study that found a small presence of meso-zeaxanthin in extracts from human

RPE-choroid samples (121). However, this amount was very small in comparison to the

103 amount of lutein and zeaxanthin extracted and may be the result of incomplete separation of the RPE-choroid from the rest of the retina. These results are not surprising since lutein and zeaxanthin appear to be highly conserved and stable in the retina despite exposure to UV light and oxidants. Also, most meso-zeaxanthin is found in the fovea and central region of the macula and tends to increase as you move away from the RPE (122).

Lutein is likely converted to meso-zeaxanthin after it exits from the RPE further into the retina by other binding proteins (46,123).

There are several limitations, but also benefits for using ARPE-19 cells as a model for xanthophyll uptake and metabolism. Clearly, we are limited in the study using a cell line since we cannot take into consideration the effect of the whole organ system on

RPE dynamics. Studying an animal or human with a macula would provide more real world results. However, the study of humans and primates, which are the only ones with a true macula, is both expensive and difficult in terms of invasive techniques. Using an

ARPE-19 cell line as a model for the RPE allows us to directly study the uptake dynamics of xanthophylls and carefully track any metabolism or xanthophyll products created by the cell. Our results suggest a preferential uptake of zeaxanthin by the RPE and by a mechanism different from lutein. Whether the LDLR is involved in lutein uptake delivered by LDL remains to be studied. Given the reduced uptake of lutein in the presence of β-carotene, supplementation of lutein and zeaxanthin without β-carotene should be the focus of AMD treatment and prevention. The preferential uptake of zeaxanthin compared to lutein in the RPE and its lower amount in the blood compared to lutein may indicate a higher need for zeaxanthin in the RPE. It remains to be seen if

104

supplementing a higher amount of zeaxanthin compared to lutein would be more

beneficial for lowering the risk of AMD. Finally, we did not detect meso-zeaxanthin in

the RPE indicating that future studies on lutein conversion to meso-zeaxanthin should focus on binding proteins located in the macula and fovea region. We did not detect 3- hydroxy-apo-10’-carotenals or BCO2 protein in the RPE which is consistent with the

conservation of xanthophylls in the RPE. The presence and/or activity of BCO2 in the

retina is controversial at this point and is an interesting avenue for future research

(52,53,120).

105

List of References

(1) Zarbin, M., P. Rosenfeld. 2010. Pathway-based therapies for age-related macular degeneration: an integrated survey of emerging treatment alternatives. Retina 30: 1350-67.

(2) Wong, W.L., X. Su, X. Li, C.M. Cheung, R. Klein, C.Y. Cheng, T.Y. Wong. 2014. Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis. Lancet. 2: 106-16.

(3) Kijlstra, A, Y. Tian, E. Kelly, T. Berendschot. 2012. Lutein: more than just a filter for blue light. Prog. Retin. Eye Res. 31: 303-15.

(4) Moeller, S., N. Parekh, L. Tinker, C. Ritenbaugh, B. Blodi, R. Wallace, J. Mares, CAREDS Research Study Group. 2006. Associations between intermediate age- related macular degeneration and lutein and zeaxanthin in the Carotenoids in Age- related Eye Disease Study (CAREDS): ancillary study of the Women's Health Initiative. Arch. Ophthalmol. 124: 1151-62.

(5) Romanchik, J. E., E.H. Harrison, D.W. Morel. 1997. Addition of lutein, lycopene, or β-carotene to LDL or serum in vitro: Effects on carotenoid distribution, LDL composition, and LDL oxidation. J. Nutri. Biochem. 8: 681-688.

(6) Johnson, E., H. Chung, S. Caldarella, D. Snodderly. 2008. The influence of supplemental lutein and docosahexaenoic acid on serum, lipoproteins, and macular pigmentation. Am. J. Clin. Nutr. 87: 1521-9.

(7) Reboul, E., L. Abou, C. Mikail, O. Ghiringhelli, M. André, H. Portugal, D. Jourdheuil-Rahmani, M.J. Amiot, D. Lairon, P. Borel. 2005. Lutein transport by Caco-2 TC-7 cells occurs partly by a facilitated process involving the scavenger receptor class B type I (SR-BI). Biochem. J. 387: 455-61.

(8) Reboul, E., S. Thap, E. Perrot, M. Amiot, D. Lairon, P. Borel. 2007. Effect of the main dietary antioxidants (carotenoids, gamma-tocopherol, polyphenols, and vitamin C) on alpha-tocopherol absorption. Eur. J. Clin. Nutr. 61: 1167-73.

106

(9) Reboul, E., P. Borel. 2011. Proteins involved in uptake, intracellular transport and basolateral secretion of fat-soluble vitamins and carotenoids by mammalian enterocytes. Prog. Lipid Res. 50: 388-402.

(10) During, A., S. Doraiswamy, E.H. Harrison. 2008. Xanthophylls are preferentially taken up compared with beta-carotene by retinal cells via a SRBI-dependent mechanism. J. Lipid Res. 49: 1715-24.

(11) Pfander, H. 1992. Carotenoids: An Overview. Meth. Enzymol. 213: 3 - 13.

(12) During, A., E.H. Harrison. 2004. Intestinal absorption and metabolism of carotenoids: insights from cell culture. Arch. Biochem. Biophys. 430: 77-88.

(13) Rao, A., L. Rao, L.G. 2007. Carotenoids and human health. Pharmacological Research 55: 3.

(14) Kotake-Nara, E., A. Nagao. 2011. Absorption and Metabolism of Xanthophylls. Mar. Drugs. 9: 1024-1037.

(15) Sajilata, M., R. Singhal, M. Kamat. 2008. The Carotenoid Pigment Zeaxanthin—A Review. Comprehensive Reviews in Food Science and Food Safety 7: 29-49.

(16) Mares, J. 2016. Lutein and Zeaxanthin Isomers in Eye Health and Disease. Annu. Rev. Nutr. 36: 571-602.

(17) Poulaert, M., P. Borel, B. Caporiccio, Z. Gunata, C. Dhuique-Mayer. 2012. Grapefruit juices impair the bioaccessibility of β-carotene from orange-fleshed sweet potato but not its intestinal uptake by Caco-2 cells. J. Agric. Food Chem. 60: 685-91.

(18) Kostic, D., W. White, J. Olson. 1995. Intestinal absorption, serum clearance, and interactions between lutein and beta-carotene when administered to human adults in separate or combined oral doses. Am. J. Clin. Nutr. 62: 604-10.

(19) Tyssandier, V., N. Cardinault, C. Caris-Veyrat, M. Amiot, P. Grolier, C. Bouteloup, V. Azais-Braesco, P. Borel. 2002. Vegetable-borne lutein, lycopene, and beta- carotene compete for incorporation into chylomicrons, with no adverse effect on the medium-term (3-wk) plasma status of carotenoids in humans. Am. J. Clin. Nutr. 75: 526-34.

(20) Norkus, E., K. Norkus, T. Dharmarajan, J. Schierle, W. Schalch. 2010. Serum lutein response is greater from free lutein than from esterified lutein during 4 weeks of supplementation in healthy adults. J. Am. Coll. Nutr. 29: 575-85.

107

(21) DiSilvestro, R., S. Thomas, E. Harrison, A. Epitropoulos. 2015. A pilot comparison of phospolipidated lutein to conventional lutein for effects on plasma lutein concentrations in adult people. Nutrition Journal 14: 104 - 108.

(22) Reboul, E., M. Richelle, E. Perrot, C. Desmoulins-Malezet, V. Pirisi, P. Borel. 2006. Bioaccessibility of carotenoids and vitamin E from their main dietary sources. J. Agric. Food Chem. 54: 8749-55.

(23) Gleize, B., F. Tourniaire, L. Depezay, R. Bott, M. Nowicki, L. Albino, D. Lairon, E. Kesse-Guyot, P. Galan, S. Hercberg, P. Borel. 2012. Effect of type of TAG fatty acids on lutein and zeaxanthin bioavailability. Br. J. Nutr. 1-10.

(24) Reboul, E., S. Thap, F. Tourniaire, M. André, C. Juhel, S. Morange, M. Amiot, D. Lairon, P. Borel. 2007. Differential effect of dietary antioxidant classes (carotenoids, polyphenols, vitamins C and E) on lutein absorption. Br. J. Nutr. 97: 440-6.

(25) Borel, P., N. Mekki, Y. Boirie, A. Partier, M. Alexandre-Gouabau, P. Grolier, B. Beaufrere, H. Portugal, D. Lairon, V. Azais-Braesco. 1998. Comparison of the Postprandial Plasma Vitamin A Response in Young and Old Adults. J. gerontol. A Biol. Sci. Med. Sci. 53: B133.

(26) During, A., M. Hussain, D. Morel, E. Harrison. 2002. Carotenoid uptake and secretion by CaCo-2 cells: beta-carotene isomer selectivity and carotenoid interactions. J. Lipid Res. 43: 1086-95.

(27) van Bennekum, A., M. Werder, S. Thuahnai, C. Han, P. Duong, D. Williams, P. Wettstein, G. Schulthess, M. Phillips, H. Hauser. 2005. Class B scavenger receptor- mediated intestinal absorption of dietary beta-carotene and cholesterol. Biochemistry 44: 4517-25.

(28) van der Made, S., E. Kelly, T. Berendschot, A. Kijlstra, D. Lütjohann, J. Plat. 2014. Consuming a buttermilk drink containing lutein-enriched egg yolk daily for 1 year increased plasma lutein but did not affect serum lipid or lipoprotein concentrations in adults with early signs of age-related macular degeneration. J. Nutr. 144: 1370-7.

(29) Attie, A., Y. Hamon, A. Brooks-Wilson, M. Gray-Keller, M. MacDonald, V. Rigot, A. Tebon, L. Zhang, J. Mulligan, R. Singaraja, J. Bitgood, M. Cook, J. Kastelein, G. Chimini, M. Hayden. 2002. Identification and functional analysis of a naturally occurring E89K mutation in the ABCA1 gene of the WHAM chicken. J. Lipid Res. 43: 1610-7.

(30) Connor, W., P. Duell, R. Kean, Y. Wang. 2007. The prime role of HDL to transport lutein into the retina: evidence from HDL-deficient WHAM chicks having a mutant ABCA1 transporter. Invest. Ophthalmol. Vis. Sci. 48: 4226-31. 108

(31) Kolb, H. 2003. How the Retina Works. Am. Sci. 91: 1.

(32) Reynolds, J. and S. Olitsky. Pediatric Retina. New York: Springer Verlag 2011. Print..

(33) Bernstein, P., B. Li, P. Vachali, A. Gorusupudi, R. Shyam, B. Henriksen, J. Nolan. 2016. Lutein, zeaxanthin, and meso-zeaxanthin: The basic and clinical science underlying carotenoid-based nutritional interventions against ocular disease. Prog. Ret. Eye Res. 50: 34-66.

(34) Strauss, O. 2005. The retinal pigment epithelium in visual function. Physiol. Rev. 85: 845-81.

(35) Schmitz, H., C. Poor, E. Gugger, J. Erdman. 1993. Analysis of carotenoids in human and animal tissues. Meth. Enzymol. 214: 102-16.

(36) Sommerburg, O., W. Siems, J. Hurst, J. Lewis, D. Kliger, F. van Kuijk. 1999. Lutein and zeaxanthin are associated with photoreceptors in the human retina. Curr. Eye Res. 19: 491-5.

(37) Handelman, G., D. Snodderly, A. Adler, M. Russett. 1992. Measurement of Carotenoids in Human and Monkey Retinas. Meth. Enzymol. 213: 220-30.

(38) Snodderly, D., G. Handelman, A. Adler. 1991. Distribution of individual macular pigment carotenoids in central retina of macaque and squirrel monkeys. Invest. Ophthalmol. Vis. Sci. 32: 268-79.

(39) Lee, C., A. Boileau, T. Boileau, A. Williams, K. Swanson, K. Heintz, J. Erdman. 1999. Review of Animal Models in Carotenoid Research. J. Nutr. 129 (12): 2271- 2277.

(40) Toyoda, Y., L. Thomson, A. Langner, N. Craft, K. Garnett, C. Nichols, K. Cheng, C. Dorey. 2002. Effect of dietary zeaxanthin on tissue distribution of zeaxanthin and lutein in quail. Invest. Ophthalmol. Vis. Sci. 43: 1210-21.

(41) Bone, R., J. Landrum, L. Fernandez, S. Tarsis. 1988. Analysis of the macular pigment by HPLC: retinal distribution and age study. Invest. Ophthalmol. Vis. Sci. 29: 843-9.

(42) Bone, R., J. Landrum, Z. Dixon, Y. Chen, C. Llerena. 2000. Lutein and Zeaxanthin in the Eyes, Serum and Diet of Human Subjects. Exp. Eye Res. 71: 239-245.

109

(43) Vishwanathan, R., W. Schalch, E. Johnson. 2015. Macular pigment carotenoids in the retina and occipital cortex are related in humans. Nutritional Neuroscience 19: 95-101.

(44) Yemelyanov, A., N. Katz, P. Bernstein. 2001. Ligand-binding Characterization of Xanthophyll Carotenoids to Solubilized Membrane Proteins Derived from Human Retina. Exp. Eye Res. 72: 381-392.

(45) Li, B., P. Vachali, J. Frederick, P. Bernstein. 2011. Identification of StARD3 as a lutein-binding protein in the macula of the primate retina. Biochemistry 50: 2541-9.

(46) Bhosale, P., A. Larson, J. Frederick, K. Southwick, C. Thulin, P. Bernstein. 2004. Identification and Characterization of a Pi Isoform of Glutathione S-Transferase (GSTP1) as a Zeaxanthin-binding Protein in the Macula of the Human Eye. J. Biol. Chem. 279 (47): 49447-54.

(47) Bhosale, P., P. Bernstein. 2005. Synergistic effects of zeaxanthin and its binding protein in the prevention of lipid membrane oxidation. Molecular Basis of Disease 1740 (2): 116-121.

(48) dela Seña, C., S. Narayanasamy, K. Riedl, R. Curley, S. Schwartz, E. Harrison. 2013. Substrate specificity of purified recombinant human β-carotene 15,15'- oxygenase (BCO1). J. Biol. Chem. 288: 37094-103.

(49) Mein, J., G. Dolnikowski, H. Ernst, R. Russell, X. Wang. 2011. Enzymatic formation of apo-carotenoids from the xanthophyll carotenoids lutein, zeaxanthin and β-cryptoxanthin by ferret carotene-9',10'-monooxygenase. Arch. Biochem. Biophys. 506: 109-21.

(50) dela Seña, C., J. Sun, S. Narayanasamy, K. Riedl, Y. Yuan, R. Curley, S. Schwartz, E. Harrison. 2016. Substrate specificity of purified recombinant chicken β-carotene 9,10-oxygenase (BCO2). J.Biol.Chem. 288: 37094-103.

(51) Lindqvist, A., Y. He, S. Andersson. 2005. Cell type-specific expression of beta- carotene 9',10'-monooxygenase in human tissues. J. Histochem. Cytochem. 53: 1403-12.

(52) Li,B., P. Vachali, A. Gorusupudi, Z. Shen, H. Sharifzadeh, B. Besch, K. Nelson, M. Horvath, J. Frederick, W. Baehr, P. Bernstein. 2014. Inactivity of human β,β- carotene-9',10'-dioxygenase (BCO2) underlies retinal accumulation of the human macular carotenoid pigment. Proc. Natl. Acad. Sci. 111: 10173-8.

110

(53) Babino, D., G. Palczewski, M. Widjaja-Adhi, P. Kiser, M. Golczak, J. von Lintig. 2015. Characterization of the Role of β-Carotene 9,10-Dioxygenase in Macular Pigment Metabolism. J. Biol. Chem. 290: 24844-57.

(54) Thomson, L., Y. Toyoda, A. Langner, F. Delori, K. Garnett, N. Craft, C. Nichols, K. Cheng, C. Dorey. 2002. Elevated retinal zeaxanthin and prevention of light-induced photoreceptor cell death in quail. Invest. Ophthalmol. Vis. Sci. 43: 3538-49.

(55) Barker, F., D. Snodderly, E. Johnson, W. Schalch, W. Koepcke, J. Gerss, M. Neuringer. 2011. Nutritional manipulation of primate retinas, V: effects of lutein, zeaxanthin, and n-3 fatty acids on retinal sensitivity to blue-light-induced damage. Invest. Ophthalmol. Vis. Sci. 52: 3934-42.

(56) Bian, Q., S. Gao, J. Zhou, J. Qin, A. Taylor, E. Johnson, G. Tang, J. Sparrow, D. Gierhart, F. Shang. 2012. Lutein and zeaxanthin supplementation reduces photooxidative damage and modulates the expression of inflammation-related genes in retinal pigment epithelial cells. Free Radical Biology and Medicine 53: 1298- 1307.

(57) Chucair, A., N. Rotstein, J. Sangiovanni, A. During, E. Chew, L. Politi. 2007. Lutein and zeaxanthin protect photoreceptors from apoptosis induced by oxidative stress: relation with docosahexaenoic acid. Invest. Ophthalmol. Vis. Sci. 48: 5168-77.

(58) Seddon, J., U. Ajani. 1994. Dietary carotenoids, vitamins A, C, and E, and advanced age-related macular degeneration. J. Amer. Med. Assoc. 272 (18): 1413-20.

(59) Age-Related Eye Disease Study 2 Research Group. 2013. Lutein + zeaxanthin and omega-3 fatty acids for age-related macular degeneration: the Age-Related Eye Disease Study 2 (AREDS2) randomized clinical trial. J. Amer. Med. Assoc. 309: 2005-15.

(60) Romanchik, J., D. Morel, E. Harrison. 1995. Distributions of carotenoids and alpha- tocopherol among lipoproteins do not change when human plasma is incubated in vitro. J. Nutr. 125: 2610-7.

(61) Hatch, F. 1968. Practical methods for plasma lipoprotein analysis. Adv. Lipid Res. 6: 1-68.

(62) Timasheff, S. 1993. The control of protein stability and association by weak interactions with water: how do solvents affect these processes? Annu. Rev. Biophys. Biomol. Struct. 22: 67-97.

111

(63) Billington, D., E. Maxwell, J. Graham, P. Newland. 2007. Large-scale preparation of human low- and high-density lipoproteins by density gradient centrifugation using iodixanol. Anal. Biochem. 367: 137-9.

(64) Graham, J., J. Higgins, T. Gillott, T. Taylor, J. Wilkinson, T. Ford, D. Billington. 1996. A novel method for the rapid separation of plasma lipoproteins using self- generating gradients of iodixanol. Atherosclerosis 124: 125-35.

(65) Davies, I., J. Graham, B. Griffin. 2003. Rapid separation of LDL subclasses by iodixanol gradient ultracentrifugation. Clin. Chem. 49: 1865-72.

(66) Barua, A. 2001. Improved normal-phase and reversed-phase gradient high- performance liquid chromatography procedures for the analysis of retinoids and carotenoids in human serum, plant and animal tissues. J. Chrom.. A 936: 71-82.

(67) Barua, A., Olson, J. 1998. Reversed-phase gradient high-performance liquid chromatographic procedure for simultaneous analysis of very polar to nonpolar retinoids, carotenoids and tocopherols in animal and plant samples. J. Chrom. B 707: 69-79.

(68) During, A., E. Harrison. 2005. An in vitro model to study the intestinal absorption of carotenoids. Food Research International 38: 1001-1008.

(69) Aman, R., J. Biehl, R. Carle, J. Conrad, U. Beifuss, A. Schieber. 2005. Application of HPLC coupled with DAD, APcI-MS and NMR to the analysis of lutein and zeaxanthin stereoisomers in thermally processed vegetables. Food Chemistry 92: 753-763.

(70) Dunn, K., A. Aotaki-Keen, F. Putkey, L. Hjelmeland. 1996. ARPE-19, A Human Retinal Pigment Epithelial Cell Line with Differentiated Properties. Exp. Eye Res. 62: 155-170.

(71) Lornejad-Schäfer, M., C. Lambert, D. Breithaupt, H. Biesalski, J. Frank. 2007. Solubility, uptake and biocompatibility of lutein and zeaxanthin delivered to cultured human retinal pigment epithelial cells in tween40 micelles. Eur J Nutr European 46: 79-86.

(72) Forte, T., J. Bell-Quint, F. Cheng. 1981. Lipoproteins of fetal and newborn calves and adult steer: a study of developmental changes. Lipids 16: 240-5.

(73) Gardner, R., N. Ogden, P. Cripps, D. Billington. 2003. Separation of bovine plasma lipoproteins by a rapid ultracentrifugation method. J. Comp. Pathol. 128: 15-23.

112

(74) Grummer, R., C. Davis, H. Hegarty. 1983. Comparison of ultracentrifugation and gel filtration for the isolation of bovine lipoproteins. Lipids 18: 795-802.

(75) Hollanders, B., A. Mougin, F. N'Diaye, E. Hentz, X. Aude, A. Girard. 1986. Comparison of the lipoprotein profiles obtained from rat, bovine, horse, dog, rabbit and pig serum by a new two-step ultracentrifugal gradient procedure. Comp. Biochem. 84: 83-9.

(76) Papadopulo, I., F. de La Farge, J. Braun, P. Valdiguié, A. Rico. 1987. Analysis for lipoproteins in horse serum. Clin. Chem. 33 (6): 1081.

(77) Krinsky, N., D. Cornwell, J. Oncley. 1958. The transport of vitamin A and carotenoids in human plasma. Archives of Biochemistry and Biophysics 73: 233-246.

(78) Namitha, K., P. Neggi. 2010. Chemistry and biotechnology of carotenoids. Crit. Rev. Food Sci. Nutr. 50: 728-60.

(79) Brevik, A., L. Andersen, A. Karlsen, K. Trygg, R. Blomhoff, C. Drevon. 2004. Six carotenoids in plasma used to assess recommended intake of fruits and vegetables in a controlled feeding study. Eur. J. Clin. Nutr. 58: 1166-73.

(80) Chung, H., A. Ferreira, S. Epstein, S. Paiva, C. Castaneda-Sceppa, E. Johnson. 2009. Site-specific concentrations of carotenoids in adipose tissue: relations with dietary and serum carotenoid concentrations in healthy adults. Am. J. Clin. Nutr. 90 (3): 533-9.

(81) Peng, Y., Y. Peng, Y. Lin, T. Moon, D. Roe, C. Ritenbaugh. 1995. Concentrations and plasma‐tissue‐diet relationships of carotenoids, retinoids, and tocopherols in humans. Nutrition and Cancer 23: 233-246.

(82) Bone R., J. Landrum, L. Friedes, C. Gomez, M. Kilburn, E. Menendez, I. Vidal, W. WANG. 1997. Distribution of Lutein and Zeaxanthin Stereoisomers in the Human Retina. Exp. Eye Res 64: 211-218.

(83) Thomas, J., D. Duewer, I. Mugenya, K. Phinney, L. Sander, K. Sharpless, L. Sniegoski, S. Tai, M. Welch, J. Yen. 2012. Preparation and value assignment of standard reference material 968e fat-soluble vitamins, carotenoids, and cholesterol in human serum. Anal. Bioanal. Chem. 402: 749-762.

(84) Bone, R., J. Landrum, G. Hime, A. Cains. 1993. Stereochemistry of the Human Macular Carotenoids. Invest. Ophthalmol. Visual Sci. 34 (6): 2033-40.

113

(85) Khachik, F., F. de Moura, D. Zhao, C. Aebischer, P. Bernstein. 2002. Transformations of selected carotenoids in plasma, liver, and ocular tissues of humans and in nonprimate animal models. Invest. Ophthalmol. Vis. Sci. 43: 3383-92.

(86) Gorusupudi. A., R. Shyam, B. Li, P. Vachali, Y. Subhani, K. Nelson, P. Bernstein. 2016. Developmentally Regulated Production of meso-Zeaxanthin in Chicken Retinal Pigment Epithelium/Choroid and Retina. Invest. Ophthalmol. Vis. Sci. 57: 1853-61.

(87) Johnson E., M. Neuringer, R. Russell, W. Schalch, D. Snodderly. 2005. Nutritional manipulation of primate retinas, III: Effects of lutein or zeaxanthin supplementation on adipose tissue and retina of xanthophyll-free monkeys. Invest. Ophthalmol. Vis. Sci. 46: 692-702.

(88) Thurnham, D., A. Trémel, A. Howard. 2008. A supplementation study in human subjects with a combination of meso-zeaxanthin, (3R,3'R)-zeaxanthin and (3R,3'R,6'R)-lutein. Br. J. Nutr. 100: 1307-14.

(89) Bone, R., J. Landrum, Y. Cao, A. Howard, F. Alvarez-Calderon. 2007. Macular pigment response to a supplement containing meso-zeaxanthin, lutein and zeaxanthin. Nutr Metab Lond Nutrition & Metabolism 4: 1-8.

(90) Fliesler, S., L. Bretillon. 2010. The ins and outs of cholesterol in the vertebrate retina. J. Lipid Res. 51: 3399-413.

(91) Akanuma, S., A. Yamamoto, S. Okayasu, M. Tachikawa, K. Hosoya. 2009. High- density lipoprotein-associated alpha-tocopherol uptake by human retinal pigment epithelial cells (ARPE-19 Cells): the irrelevance of scavenger receptor class B, type I. Biol. Pharm. Bull. 32: 1131-4.

(92) Dunn, K., A. Marmorstein, V. Bonilha, E. Rodriguez-Boulan, F. Giordano, L. Hjelmeland. 1998. Use of the ARPE-19 cell line as a model of RPE polarity: basolateral secretion of FGF5. Invest. Ophthalmol. Vis. Sci. 39: 2744-9.

(93) Gordiyenko, N., M. Campos, J. Lee, R. Fariss, J. Sztein, I. Rodriguez. 2004. RPE cells internalize low-density lipoprotein (LDL) and oxidized LDL (oxLDL) in large quantities in vitro and in vivo. Invest. Ophthalmol. Vis. Sci. 45: 2822-9.

(94) Finnemann, S., R. Silverstein. 2001. Differential roles of CD36 and alphavbeta5 integrin in photoreceptor phagocytosis by the retinal pigment epithelium. J. Exp. Med. 194: 1289-98.

114

(95) Tserentsoodol, N., N. Gordiyenko, I. Pascual, J. Lee, S. Fliesler, I. Rodriguez. 2006. Intraretinal lipid transport is dependent on high density lipoprotein-like particles and class B scavenger receptors. Molecular Vision 12: 1319-33.

(96) Tserentsoodol, N., J. Sztein, M. Campos, N. Gordiyenko, R. Fariss, J. Lee, S. Fliesler, I. Rodriguez. 2006. Uptake of cholesterol by the retina occurs primarily via a low density lipoprotein receptor-mediated process. Molecular Vision 12: 1306-18.

(97) During, A., S. Doraiswamy, E. Harrison EH. Xanthophylls are preferentially taken up compared with beta-carotene by retinal cells via a SRBI-dependent mechanism. J. Lipid Res. 49: 1715-24.

(98) Nieland, T., M. Penman, L. Dori, M. Krieger, T. Kirchhausen. 2002. Discovery of chemical inhibitors of the selective transfer of lipids mediated by the HDL receptor SR-BI. Proc. Natl. Acad. Sci.. 99: 15422-7.

(99) Moussa, M., J. Landrier, E. Reboul, O. Ghiringhelli, C. Coméra, X. Collet, K. Fröhlich, V. Böhm, P. Borel. 2008. Lycopene absorption in human intestinal cells and in mice involves scavenger receptor class B type I but not Niemann-Pick C1-like 1. J. Nutr. 138: 1432-6.

(100) Cai, L., M. de Beer, F. de Beer, D. van der Westhuyzen. 2005. Serum amyloid A is a ligand for scavenger receptor class B type I and inhibits high density lipoprotein binding and selective lipid uptake. J. Biol. Chem. 280: 2954-61.

(101) Bone, R., J. Landrum. 2010. Dose-dependent response of serum lutein and macular pigment optical density to supplementation with lutein esters. Archives of Biochemistry and Biophysics 504: 50-55.

(102) Babitt, J., B. Trigatti, A. Rigotti, E. Smart, R. Anderson, S. Xu, M. Krieger. 1997. Murine SR-BI, a high density lipoprotein receptor that mediates selective lipid uptake, is N-glycosylated and fatty acylated and colocalizes with plasma membrane caveolae. J. Biol. Chem. 272: 13242-9.

(103) Phillips, M. 2014. Molecular mechanisms of cellular cholesterol efflux. J. Biol. Chem. 289: 24020-9.

(104) Meyers, K., J. Mares, Z. Liu, C. Karki, B. Blodi, R. Igo, B. Truitt, S. Iyengar, A. Millen, M. Klein, E. Johnson, C. Engelman, K. Gehrs, L. Tinker, R. Wallace, J. Robinson, E. LeBlanc, G. Sarto, P. Bernstein, J. SanGiovanni. 2013. Genetic evidence for role of carotenoids in age-related macular degeneration in the carotenoids in age-related eye disease study (CAREDS). Invest.Ophthalmol.Vis.Sci. 55: 587-599.

115

(105) Rodrigueza, W., S. Thuahnai, R. Temel, S. Lund-Katz, M. Phillips, D. Williams. 1999. Mechanism of scavenger receptor class B type I-mediated selective uptake of cholesteryl esters from high density lipoprotein to adrenal cells. J. Biol. Chem. 274: 20344-50.

(106) Neculai, D., M. Schwake, M. Ravichandran, F. Zunke, R. Collins, J. Peters, M. Neculai, J. Plumb, P. Loppnau, J. Pizarro, A. Seitova, W. Trimble, P. Saftig, S. Grinstein, S., S. Dhe-Paganon. 2013. Structure of LIMP-2 provides functional insights with implications for SR-BI and CD36. Nature 504: 172-6.

(107) During, A., H. Dawson, E. Harrison. 2005. Carotenoid transport is decreased and expression of the lipid transporters SR-BI, NPC1L1, and ABCA1 is downregulated in Caco-2 cells treated with ezetimibe. J. Nutr. 135: 2305-12.

(108) Reboul, E., A. Klein, F. Bietrix, B. Gleize, C. Malezet-Desmoulins, M. Schneider, A. Margotat, L. Lagrost, X. Collet, P. Borel. 2006. Scavenger receptor class B type I (SR-BI) is involved in vitamin E transport across the enterocyte. J. Biol. Chem. 281: 4739-45.

(109) Nieland, T., A. Chroni, M. Fitzgerald, Z. Maliga, V. Zannis, T. Kirchhausen, M. Krieger. 2004. Cross-inhibition of SR-BI- and ABCA1-mediated cholesterol transport by the small molecules BLT-4 and glyburide. J. Lipid Res. 45: 1256-65.

(110) Yu, M., K. Romer, T. Nieland, S. Xu, V. Saenz-Vash, M. Penman, A. Yesilaltay, S. Carr, M. Krieger. 2011. Exoplasmic cysteine Cys384 of the HDL receptor SR-BI is critical for its sensitivity to a small-molecule inhibitor and normal lipid transport activity. Proc. Natl. Acad. Sci. 108: 12243-8.

(111) Kiefer, C., E. Sumser, M. Wernet, J. Von Lintig. 2002. A class B scavenger receptor mediates the cellular uptake of carotenoids in Drosophila. Proc. Natl. Acad. Sci. 99: 10581-6.

(112) Voolstra, O., C. Kiefer, M. Hoehne, R. Welsch, K. Vogt, J. von Lintig. 2006. The Drosophila class B scavenger receptor NinaD-I is a cell surface receptor mediating carotenoid transport for visual chromophore synthesis. Biochemistry 45: 13429- 13437.

(113) Tsuchida, K., T. Sakudo. 2015. Recent progress in molecular genetic studies on the carotenoid transport system using cocoon-color mutants of the silkworm. Arch. Biochem. Biophys. 572: 151-7.

(114) Maoka, T., A. Arai, M. Shimizu, T. Matsuno. 1986. The first isolation of enantiomeric and meso-zeaxanthin in nature. Comparative Biochemistry 83: 121-4.

116

(115) Nolan, J., K. Meagher, S. Kashani, S. Beatty. 2013. What is meso-zeaxanthin, and where does it come from? Eye 27: 899-905.

(116) Thurnham, D., J. Nolan, A. Howard, S. Beatty. 2015. Macular response to supplementation with differing xanthophyll formulations in subjects with and without age-related macular degeneration. Graefe's Archive for Clinical and Experimental Ophthalmology 253: 1231-43.

(117) Li, B., P. Vachali, J. Frederick, P. Bernstein. 2011. Identification of StARD3 as a lutein-binding protein in the macula of the primate retina. Biochemistry 50: 2541-9.

(118) Wang, Y., S. Connor, W. Wang, E. Johnson, W. Connor. 2007. The selective retention of lutein, meso-zeaxanthin and zeaxanthin in the retina of chicks fed a xanthophyll-free diet. Exp. Eye Res. 84: 591-8.

(119) Lindqvist, A., Y. He, S. Andersson. 2005. Cell type-specific expression of beta- carotene 9',10'-monooxygenase in human tissues. J. Histochem. Cytochem. 53: 1403-12.

(120) Palczewski, G., J. Amengual, C. Hoppel, J. von Lintig. 2014. Evidence for compartmentalization of mammalian carotenoid metabolism. FASEB 28: 4457-69.

(121) Khachik, F., F. de Moura, D. Zhao, C. Aebischer, P. Bernstein. 2002. Transformations of selected carotenoids in plasma, liver, and ocular tissues of humans and in nonprimate animal models. Invest. Ophthalmol. Vis. Sci. 43: 3383-92.

(122) Bone, R., J. Landrum, Friedes, L., C. Gomez, M. Kilburn, E. Menendez, I. Vidal, W. Wang. 1997. Distribution of Lutein and Zeaxanthin Stereoisomers in the Human Retina. Experimental Eye Research 64: 211-218.

(123) Li, B., P. Vachali, P. Bernstein. 2010. Human ocular carotenoid-binding proteins. Photochemical & Photobiological Sciences 9: 1418-1425.

117