Identification of Additional Retinoic Acid Response

The Pennsylvania State University

The Graduate School

The Department of Nutritional Sciences

IDENTIFICATION OF ADDITIONAL RETINOIC ACID RESPONSE

ELEMENTS INVOLVED IN THE INDUCIBILITY OF CYP26A1

IN LIVER

A Thesis in

Nutrition

Yao Zhang

Submitted in Partial Fulfillment of the Requirements for the Degree of

Master of Science

December 2008

The thesis of Yao Zhang has been reviewed and approved* by the following:

A. Catharine Ross Professor and Occupant of Dorothy Foehr Huck Chair in Nutrition Thesis Advisor

Gary J. Fosmire Associate Professor of Nutrition

Okhee Han Assistant Professor of Nutrition

Sharon M. Nickols-Richardson Associate Professor of Nutrition Chair of Graduate Program

* Signatures are on file in the Graduate School

ii ABSTRACT

All-trans retinoic acid (ATRA) is considered as the major active metabolite of vitamin A. ATRA regulates many physiological processes, such as embryonic development, cellular proliferation and differentiation, testicular function, and immune function. Clinically, retinoids have been demonstrated to be useful for cancer treatment and chemoprevention. Oral ATRA have been effective in the treatment of the patients with acute promyelocytic leukemia (APL).

ATRA concentration in tissues is tightly controlled in a spatial and temporal manner, accomplished by a balance of synthesizing and metabolizing enzymes. Several retinaldehyde dehydrogenases have been characterized for the irreversible conversion of retinaldehyde to ATRA. Metabolism of ATRA is mediated by cytochrome P450 activities, and is known to be induced by RA itself in many tissues.

CYP26A1 (P450RA1) is a cytochrome P450 enzyme, and specifically catalyzes the conversion of ATRA to polar metabolites including 4-OH-RA, 4-oxo-RA, and 18-

OH-RA. Moreover, CYP26A1 is remarkable for its strong inducibility by ATRA in vivo.

In the liver of intact ATRA-treated rats, CYP26A1 mRNA increased nearly 2000-fold within 10 hours and then declined rapidly. Loudig et al. have previously reported on one retinoic acid response element (RARE, referred as RARE1), proximal to the transcription start site, which cooperated with guanine-guanine-rich element (GGRE) in the induction of CYP26A1 promoter activation in F9 and P19 (murine embryonal or teratocarcinoma cells).

To better understand the mechanism of the strong regulation of CYP26A1 gene expression in the liver, we have measured endogenous levels of CYP26A1 mRNA in

iii ATRA-treated human liver cells (HepG2 cells), and further characterized the CYP26A1 promoter region extending 2.2 kb upstream from the translational start site.

Whereas HepG2 cells expressed very little endogenous CYP26A1 mRNA prior to retinoid treatment, they responded rapidly and dose-dependently to 1-1000 nM RA, and therefore closely resembled intact liver. Examination of the putative promoter of

CYP26A1 extending 2.2 kb upstream from the translation state site revealed three additional potential regulatory elements: two 5-bp-spaced direct repeat (DR-5) RAREs

(RARE2, RARE3) and a nearby half-site (RARE4). Sequence analysis shows that all three RAREs and the half site are highly conserved among mouse, rat and human. Based on the human CYP26A1 nucleotide sequence, we prepared a series of luciferase reporter constructs that contained either nucleotide deletions or mutations within RARE1, 2, 3, and the half site, RARE4, of the full-length construct. Relative luciferase activity for the full-length construct was dose-dependently regulated by RA in HepG2 cells, and correlated with the level of endogenously expressed CYP26A1 mRNA. Comparative studies revealed that RA increased CYP26A1 promoter activity significantly more in

HepG2 cells as compared to HEK293T cells, and that all three of the RAREs and the half site synergistically regulated RA-mediated CYP26A1 expression in HepG2 cells.

Cotransfection of each of the individual retinoic acid receptors (RAR α, β, γ) significantly increased relative luciferase activity. Interestingly, in electrophoretic mobility shift assays, similar patterns of nuclear factor complex binding to RARE2 and RARE3 and to the RARE4 half site sequence in conjunction with proximal DNA were observed. RAR α was likely to play a major role in CYP26A1 expression in HepG2 cells.

iv Together, these experiments demonstrate that additional RA-responsive elements upstream of the CYP26A1 coding sequence are both active and necessary for the strong inducibility of CYP26A1 expression in HepG2 liver cells.

TABLE OF CONTENTS

LIST OF ABBREVIATIONS ...... vii ACKNOWLEDGEMENTS ...... ix

CHAPTER 1 INTRODUCTION AND HYPOTHESIS ...... 1 CHAPTER 2 BACKGROUND AND SIGNIFICANCE ...... 4 Absorption, Intestinal Metabolism of Dietary Vitamin A ...... 4 Retinol Esterification and Storage ...... 8 Plasma retinol transport...... 10 Retinoic Acid...... 11 Retinoic acid biosynthesis ...... 12 Retinoic acid catabolism...... 14 Distribution and function of CYP26 family members ...... 17 Molecular mechanism for the regulation of CYP26A1 gene expression...... 22 Retinoic acid signaling: retinoic acid receptors and retinoid X receptors...... 23 Conclusion...... 27 CHAPTER 3 ADDITIONAL RARES ARE INVOLVED IN RA-INDUCED CYP26A1 EXPRESSION...... 28 Experimental Procedures ...... 29 Results...... 33 CHAPTER 4 DISCUSSION ...... 47 CYP26A1 is essential for RA homeostasis in normal and clinical conditions...... 47 RAREs in the promoter region of CYP26A1 are highly conserved...... 54 All the three RAREs and the half site synergistically regulated RA-mediated CYP26A1 expression...... 55 CYP26A1 expression and regulation exhibit cell specificity...... 57 Future Directions...... 60 REFERENCES...... 62

LIST OF ABBREVIATIONS

ADH: alcohol dehydrogenase

AF: transactivation function domain

APL: acute promyelocytic leukemia

ARAT: acyl-coenzyme A:retinol acyltransferase

ATRA: all-trans retinoic acid

AUC: area under the concentration-time curve

CRABP: cellular retinoic acid-binding protein

CRBP: cellular retinol-binding protein

CYP: cytochrome P450

DBD: DNA binding domain

DMEM: Dulbecco's modified eagle medium

DR5: a direct repeat of hexanucleotides separated by 5 spacer nucleotides

EMSA: electrophoretic mobility shift assays

FBS: fetal bovine serum

FL: full-length construct

GGRE: guanine-guanine-rich element

GADPH: glyceraldehyde-3-phosphate dehydrogenase

HAT: histone acetyltransferase

HDAC: histone deacetylase

HPLC: high performance liquid chromatography

LBD: ligand biding domain

vii

LRAT: lecithin:retinol acyltransferase

NCoR1: nuclear receptor corepressor 1

NPE: nuclear protein extract

PPAR: peroxisome proliferator-activated receptor

RA: all-trans -Retinoic acid

RALDH: retinaldehyde dehydrogenase

RAR: retinoic acid receptor

RARE: retinoic acid response element

RBP: plasma retinol binding protein

RE: retinyl ester

REH: retinyl ester hydrolase

RDH: retinol dehydrogenase

RT-PCR: real-time polymerase chain reaction

RXR: retinoid X receptor

SMRT: silencing mediator for retinoid and thyroid hormone receptor

SP: stable protein

STRA6: stimulated by retinoic acid 6

TF: transcription factor

TTR: transthyretin

TR: thyroid hormone receptor

VDR: vitamin D receptor

VDRE: vitamin D responsive element

viii

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude and appreciation to my advisor, Dr.

Ross, who has been extraordinary in helping me grow to be a future scientist. Dr. Ross has constantly nourished me with her broad knowledge and creative ideas in all fields of nutrition, biochemistry, molecular biology, and cell biology. The time I spent in her lab working with her has been very great and I have learned from her how to design experiments, interpret results, and most importantly to think logically, creatively, and critically. I would like to thank her for all of her time, guidance, and wisdom. I would like to thank Reza and Qiuyan for all of their help. They gave me a lot of very helpful advice allowed me to be successful in lab. To my colleagues in lab, Lili, Kat, Amanda,

Libo, Yong, Nan, and Maddie, for their help and friendship which made working in lab productive and happy.

I would like to thank all other committee members: Dr. Gary J. Fosmire, and Dr.

Okhee Han. They have provided, with kindness, their insight and suggestions, which are precious to me.

I would like to thank my family for their love, care and support. To my husband

Song, his love, friendship, and support have helped me reach my goals. His constant encouragement and support made all of my accomplishments possible. To my son

Alexander, his arrival brightened my life and supported me to finish my study. To my parents, they are always the best listeners whenever I am happy or blue, and they are always on my side.

Finally, I would like to thank everyone that has helped me, especially staff and my friends in the department of Nutritional Sciences.

CHAPTER 1 INTRODUCTION AND HYPOTHESIS

All-trans retinoic acid is the active metabolite of vitamin A that mediates the majority of vitamin A-dependent functions, such as embryonic development, cellular proliferation and differentiation, testicular function, and immune function. RA induces gene expression through binding the nuclear receptors, retinoic acid receptors (RARs) and retinoid X receptors (RXRs) [3, 7]. RARs and RXRs bind to each other forming functional heterodimers, to regulate transcription. A specific DNA sequence has been characterized as the binding site for RAR/RXR heterodimers; this sequence has been named as retinoic acid response element (RARE) [3, 8]. Over 500 genes have been reported to be regulated by RA, and RAREs have been found to be localized in many of these genes including RAR β, CRBPI, CRABPII and members of the Hox families [9].

RA concentration in tissues is tightly controlled in a spatial and temporal manner, accomplished by a balance of synthesizing and metabolizing enzymes. Several retinaldehyde dehydrogenases have been characterized for the irreversible conversion of retinaldehyde to RA [10-13]. Metabolism of RA is mediated by cytochrome P450 activities [14], and is known to be induced by RA itself in many tissues [15, 16].

CYP26A1 (P450RA1) is a cytochrome P450 enzyme, which specifically catalyzes the conversion of RA to polar metabolites including 4-OH-RA, 4-oxo-RA, and 18-OH-RA

[17]. Moreover, CYP26A1 is remarkable for its strong inducibility by RA in many tissues

[6, 17, 18]. The level of CYP26A1 mRNA in the liver changes rapidly and dramatically in response to treatment with RA [6]. The liver plays a major role in the uptake of RA

from plasma and its rapid oxidation to polar metabolites [19]. Understanding the mechanism of RA-induced CYP26A1 in the liver may be especially important for understanding retinoid homeostasis.

We are interested in investigating transcription factors that interact with the

CYP26A1 promoter region and which are involved in the RA-induced expression of

CYP26A1. The proximal region of mouse CYP26A1 promoter has been studied previously by Louding et al [20], and a RARE and a G-rich element necessary for RA induction and regulation were found in that region. In that study, the proximal region of mouse CYP26A1 promoter (-238bp to +18 bp) was fused with luciferase reporter gene, transfected into F9, COS-1 or Hela cell lines. Addition of 1 M RA resulted in an about

2-fold increase of transcriptional activity. However, in the liver of intact RA-treated rats,

CYP26A1 mRNA increased very rapidly, 10-fold above control after 3 hours, and reached a peak of nearly 2000-fold within 10 hours [6]. According to these observations, we hypothesize that additional retinoic acid response elements could be involved in the strong regulation of CYP26A1 expression in liver by ATRA. Therefore, four specific aims have been designed to address our hypothesis.

First, to determine whether there are additional RARE consensus sequences embedded in the putative promoter region of gene CYP26A1. This would be tested by analyzing DNA sequences of 2.2kb upstream of the translation start site of gene

CYP26A1 and by comparing sequences among different species, including human, mouse and rat since conserved RAREs are most likely the real RAR binding elements and involved in the regulation of CYP26A1 expression.

Second, to determine which fragments of the upstream region of gene CYP26A1 are involved in the RA-induced CYP26A1 expression in liver. This would be tested by examining the effects of RA on endogenous CYP26A1 mRNA and the activities of reporter gene linked with upstream regions of translation start site of gene CYP26A1 by using HEK293T cells and HepG2 cells as cell models.

Third , to determine whether the additional RAREs contribute to the RA inducibility of CYP26A1 expression. This would be tested by examining the effects of

RA on the activities of reporter gene linked with putative promoter with selectively mutated RAREs.

Forth, to determine which RAR subtype(s) are involved in the RA inducibility of

CYP26A1 expression. This would be tested by examining RA effects on activities of reporter gene linked with the putative promoter of CYP26A1 gene with overexpression of

RAR α, RAR β, or RAR γ, and RXR α, and by examining RARs binding affinities with

RAREs in vitro by EMSA assay.

CHAPTER 2 BACKGROUND AND SIGNIFICANCE

Since the discovery of vitamin A in 1910’s, research on vitamin A has made steady progress [21]. Early studies demonstrated that vitamin A plays an important role in the maintenance of epithelial tissue and that vitamin A deficiency causes pathologic changes in the morphology of cell, particularly the keratinization of epithelial cells [22].

Later, the chemistry of vitamin A was delineated and its organic synthesis was developed

[23, 24]. Further studies elucidated the role of vitamin A in vision [25]. And then later, great progress was made in understanding the transport and metabolism of vitamin A

[21]. More recent and ongoing studies have focused on identifying the molecular mechanisms by which vitamin A affects development, reproduction, cell proliferation and differentiation, immunity, and more.

Absorption, Intestinal Metabolism of Dietary Vitamin A

Vitamin A (retinoid structures shown in Figure 1) is present in the diet as retinyl esters and as provitamin A carotenoids, of which β-carotene is the most active form.

Retinyl ester is hydrolyzed to retinol by enzymes prior to uptake in the small intestine.

Dietary β-carotene usually undergoes central oxidative cleavage and is absorbed as retinal, which later is reduced to retinol by retinol dehydrogenases. Intestinal retinol is esterified within the enterocyte and packaged into chylomicrons along with a portion of

β-carotene in carotenoid-absorbing species, and with triglycerides, cholesterol, phospholipids, and other fat-soluble vitamins [26, 27].

Chylomicrons deliver vitamin A to many extrahepatic tissues, including bone marrow, blood cells, spleen, adipose tissue, muscle, lungs, and kidneys [28]. The majority of chylomicron remnant containing retinyl esters and β-carotene are cleared into hepatocytes. In the liver, retinyl esters are rapidly hydrolyzed to unesterified retinol which can be re-esterified and stored in the stellate cells, or retinol can become associated with retinol-binding protein (RBP) and transported to other tissues or converted into the active metabolite RA. A model of retinoid metabolic pathway is illustrated in the Figure

2. Further detailed pathways will be discussed below.

Figure 1. Metabolites and derivatives of retinol (vitamin A) and synthetic retinoid (Am580)

RDH ADH REH RALDH CYP26 Polar RE Retinol Retinal RA metabolites LRAT ARAT

Secretion and RAR/RXR transport or oxidize to polar metabolites RA-induced genes

differentiation, proliferation, apoptosis, etc.

Figure 2. Retinoid metabolic pathway. ADH, alcohol dehydrogenase; ARAT, acyl-coA: retinol acyltransferase; CYP26, enzymes of cytochrome P450 family; LRAT, lecithin:retinol acyltransferase; RA, retinoic acid; RALDH, retinaldehyde dehydrogenase; RE, retinyl ester; REH, retinyl ester hydrolase; RDH, retinol dehydrogenase; RAR/RXR, nuclear retinoic acid and retinoid X receptors.

Retinol Esterification and Storage

The major storage organ of vitamin A is the liver. Under conditions of vitamin A adequacy, most mammals, including humans, store more than 90% of their total body vitamin A as retinyl esters in liver stellate cells. Two different microsomal retinol esterifying enzymes have been identified: lecithin:retinol acyltransferase (LRAT) and acyl-coenzyme A:retinol acyltransferase (ARAT).

LRAT is a 25 kD microsomal enzyme, and is present in many tissues but most abundant in the liver, small intestine, testis, adrenal gland, and eyes [5]. Retinyl ester formation is catalyzed mainly by LRAT. LRAT activity and mRNA expression in the liver are highly regulated by vitamin A status. LRAT activity was almost undetectable in the liver of vitamin A deficient rats [29]. However, LRAT mRNA and activity were induced rapidly after treatment with retinol, as well as with RA, and RAR-selective retinoids [29-31]. However, LRAT in the small intestine and the testis was not reduced during vitamin A deficiency [5]. Therefore, the small intestine is capable of esterifying retinol immediately after vitamin A is delivered; however, in the liver, LRAT is sensitive to vitamin A, which helps regulate the balance between retinol storage, and retinol oxidation and retinol secretion into plasma throughout the vitamin A homeostasis.

Retinol bound to cellular retinol binding protein (CRBP) is the preferred substrate for LRAT. In vivo studies have shown that CRBP enhanced the affinity of retinol for

LRAT [32, 33]. CRBP belongs to the fatty acid-binding or cellular retinoid-binding protein gene family. Four CRBPs (CRBP-I, CRBP-II, CRBP-III, and CRBP-IV) have been identified. CRBPs make retinoids water soluble, protect retinoids from degradation, protect membranes from accumulating retinoids, and escort retinoids to enzymes that

metabolize them [2, 27, 34]. CRBP-I and CRBP-II possess a central role in retinol trafficking and homeostasis. CRBP-I is widely distributed in retinol-metabolizing tissues such as liver, kidney, lung, and testes [35, 36]. CRBP-II is abundant in the absorptive cells of the small intestine and its ligands are both all-trans retinol and retinal [33, 37].

CRBP-II mRNA levels were significantly increased in the small intestine of vitamin A- deficient rats, implying its importance in the absorption of vitamin A [38]. Mice lacking

CRBP-I or CRBP-II developed normally, and grew healthy as long as they were fed a vitamin A-sufficient diet [39, 40]. However, the storage of retinyl ester in the liver was reduced in knock- out mice. When mice were challenged with a vitamin A-deficient diet, vitamin A storage in CRBP-I or CRBP-II null mice was depleted much faster than in wild type mice. Thus, these CRBPs are essential to the conservation of vitamin A.

The sequences of CRBP-III and CRBP-IV are ~50%-60% homologous to CRBP-I and CRBP-II. CRBP-III is present mainly in liver, kidney, mammary tissue, and heart, and it binds all-trans retinol and several other retinoids , as well as fatty acids [41, 42].

CRBP-IV has a high affinity for retinol, but the way of binding is different from other

CRBPs. CRBP-IV is relatively widely expressed, but most abundant in kidney, heart, and colon [43].

ARAT is a microsomal protein that has been shown to esterify retinol in several tissues, including mammary tissue, intestine, and liver. Intestinal ARAT is physiologically important to vitamin A homeostasis [44]. Large dietary doses of vitamin

A significantly increased ARAT levels in intestinal cells. Retinol complexed to CRBP is the preferred substrate for LRAT [33]. In contrast, uncomplexed retinol in membranes may be esterified by ARAT [32]. ARAT may serve to esterify excess retinol when large

doses of vitamin A are ingested and CRBP II becomes saturated. Furthermore, ARAT activity also depends upon the availability of fatty acyl-CoA and the overall lipid metabolism in hepatic stellate cells [45].

Plasma retinol transport

Even though retinyl esters are found in a variety of tissues such as lung and intestine, liver storage represents the major endogenous source of vitamin A for peripheral tissues [46]. As retinol is required by peripheral tissues, retinyl esters within stellate cells are hydrolyzed by retinyl ester hydrolases (REHs), and retinol is transferred back to hepatocytes. Retinol is secreted from liver in association with plasma retinol- binding protein (RBP) and circulates to peripheral tissues as the retinol-RBP- transthyretin(TTR) complex [47]. The generation of the RBP- TTR complex can reduce the glomerular filtration and renal catabolism of RBP and help stabilize the interaction of retinol with RBP [28].

The concentration of plasma retinol is normally very stable and tightly controlled within a narrow range, averaging ~2-2.2 M for adults [48]. Plasma retinol falls when vitamin A status becomes marginal, and continues to decline when vitamin A is deficient.

In addition to vitamin A status, many other factors affect the transport of retinol. For example, protein and calorie malnutrition, aging, inflammation, renal disease, and diabetes alter plasma retinol concentration [26]. Therefore, understanding the reason why an individual has low plasma retinol is very important for appropriate clinical management.

The mechanisms responsible for retinol uptake into target tissue are not quite clear. Some evidence supports receptor-dependent uptake, whereas other evidences were consistent with the dissociation of RBP and retinol followed by diffusion of retinol into plasma membrane [34, 49, 50]. However, more recently, Kawaguchi et al identified a protein, called stimulated by retinoic acid 6 (STRA6), as the membrane receptor for RBP, which mediated cellular uptake of vitamin A [48].

Retinoic Acid

RA is considered to be the major active metabolite of vitamin A. RA regulates many physiological processes, such as embryonic development, cellular proliferation and differentiation, testicular function, and immune function. Retinoids have been used as cancer therapeutic and chemopreventive agents. They exert anti-proliferative, differentiation-inducing, pro-apoptotic, and other biological effects [51]. Oral all-trans

RA has been demonstrated to induce complete remission in a high proportion of patients with acute promyelocytic leukemia (APL) [52].

The absorption of RA from the intestine is efficient in rats, and nearly two-thirds of the newly digested RA is distributed throughout the body [53]. However, overall, the dietary contribution of RA to total-body RA pools is minimal. The endogenous conversion of retinol to RA represents the primary mechanism for generating adequate amounts of RA.

RA circulates in plasma bound to albumin [53]. Under normal physiological and dietary conditions, RA is present in the circulation at very low levels. The fasting plasma level of RA is 4-14 nmol/L in humans and 7.3-9 nmol/L in rats [54, 55]. The sources of

plasma RA are still not well understood [28]. However, RA uptake and accumulation from plasma were tissue- and cell type-specific, with most of the tissues examined taking up significant amounts of the retinoid [56]. For instance, the liver and brain derived more than 75% of their RA from plasma, whereas the testes derived less than 1% from the circulation. Therefore, plasma delivery of RA to tissues appears to be an important process that is necessary for maintaining tissue RA pools. On the other hand, RA need be produced by individual tissues or cells as they need it, which provides an explanation for why dietary RA will not substitute for dietary retinol in the maintenance of spermatogenesis in rats [56].

Within cells, RA binds to cellular retinoic acid binding proteins (CRABP-I,

CRABP-II). Both CRABPs binds to all-trans RA. CRABP-I is expressed at low level in many tissues [57], and it has been implied in the oxidation of RA [2]. However, CRABP-

II has a more limited distribution range but is inducible by RA [57], and CRABP-II may facilitate the distribution of RA within cells and as a cotranscription factor in the nucleus

[58, 59].

The concentration of RA is tightly controlled in a spatial and temporal manner.

This is accomplished by a balance of the expression of RA synthesizing and metabolizing enzymes [26, 60]. Each of these pathways will be discussed in further detail below.

Retinoic acid biosynthesis

The oxidation of retinol to retinal can be catalyzed by several different enzymes.

This is a reversible reaction. Among the enzymes, some are members of the alcohol dehydrogenase gene family (ADH), others of the short-chain dehydrogenase or reductase

gene family (RDH) [2, 26]. They are all capable of metabolizing both free and CRBP- bound retinol. Studies of ADH knock-out mice have highlighted the existence and importance of redundancy in RA biosynthetic enzymes [61, 62].

The oxidation of the C-15 terminal group of retinal to retinoic acid is a physiologically irreversible process. Four retinal dehydrogenase isozymes (RALDHs:

RALDH1, RALDH2, RALDH3 and RALDH4) have been characterized [10-13], and each is able to synthesize RA from free or CRBP-bound retinal. Most cells and tissues appear to possess binding proteins and enzymes necessary for RA synthesis, enabling them to be dependent only on circulating retinol levels to meet their RA needs. All four of the RALDH genes are expressed in mouse liver, whereas RALDH2 is thought to be a key enzyme in the localized production of RA, especially during development [63, 64]. 9- cis-retinoic acid is formed by RALDH4 [12]. Vitamin A status also affects the mRNA levels of RALDH1and RALDH2 tissue-dependently [2].

The retinal concentrations outside of the eye are undetectable, less than pmol/g of tissue [2]. And RALDHs have high K 1/2 and V max values. They function in their linear rate versus product range under physiological conditions, which allows available retinal to be completely converted to RA. Thus, the availability of retinal is a limiting factor for synthezing RA; in other words, the oxidation of retinol to retinal is the presumed rate- limiting step in RA biosynthesis. However, on the other hand, the expression of the enzyme RALDH is also a control factor for the synthesis of RA. For example, during digit formation in the embryo, RA concentration is high within the interdigital region, associated with RALDH2 expression pattern [65, 66].

Retinoic acid catabolism

RA has a short elimination half life in vivo and in cultured cells, in minutes to several hours [2, 67]. In general, the generation of both oxidized derivatives (phase I metabolism) and glucuronidated conjugates of retinoic acid (phase II metabolism) maintains optimal cell and tissue levels of RA (Figure 3).

Phase I Phase II Oxidation Conjugation

Figure 3. Retinoic acid catabolic pathway [1]: oxidized derivatives (phase I metabolism) and glucuronidated conjugates (phase II metabolism)

RA can form two distinct conjugates. First, a molecule of glucuronic acid can be conjugated to the carboxyl end of the molecule to form retinoyl β-glucuronide. Members of the UDP-glucuronyltransferase gene family have been identified to catalyze conjugated reactions [1]. Second, glucuronyl transferase activity in liver was also capable of conjugating several metabolites of RA, including 4-hydroxy- and 4-oxo-RA, to form the 4-O-β-glucuronide of 4-hydroxy-RA [1]. Glucuronidation of retinoic acid convert these lipophilic molecules into water-soluble forms, then the glucuronides are excreted in bile and eliminate by the fecal route.

Glucuronide formation may also provide a reclaimable pool of RA during vitamin

A need. All-trans-retinoyl β-glucuronide fed to vitamin A-deficient rats, but not to vitamin A-sufficient rats, was converted to RA as the major metabolite [68, 69].

RA is also subject to oxidation of the ring and side chain, as well as chain- shortening reactions to produce polar molecules for excretion. At least two reactions initiate pathways of metabolism from physiological amounts of RA. One catalyzes 4- hydroxylation to produce 4-OH-RA, and then generate 4-oxo-RA; the other catalyzes 18- hydroxylation to produce 18-OH-RA [70]. Other than the glucuronide, the major steady- state metabolites are 4-derivatized ones, although the metabolite concentrations vary in different tissues [70, 71]. A member of enzymes have been reported for these reactions, but one cytochrome P450 gene family, CYP26, appears to play a prominent role and highly inducible by RA.

Distribution and function of CYP26 family members

To date, the CYP26 gene family comprises four members: CYP26A1, CYP26B1,

CYP26C1 and CYP26D1. CYP26A1, also known as P450RAI-1, can oxidize all-trans

RA to polar metabolites, including 18-OH-RA, 4-OH-RA, 4-oxo-RA and polar, water- soluble metabolites [17]. The CYP26A1 enzymatic activity shows relatively strong selectivity for all-trans RA, not 9-cis or 13-cis RA [72]. CYP26A1 was first cloned from zebrafish [73], and subsequently, homologs have been isolated from human, mouse, chick and Xenopus [74-77]. The CYP26A1 gene, located on chromosomes 10, 19 and 1 in human, mouse and rat respectively, is also highly homologous among human, rodent, and lower vertebrates. The CYP26A1 gene spans approximately 4.4 kb and has a similar exon-intron organization in each of the species studied. The deduced protein sequences for mouse and human are 93% identical, and similar to that for zebrafish, indicating strong conservation of the gene during vertebrate evolution [60].

Similar to CYP26A1, CYP26B1 is expressed in several species, including zebrafish, mouse, and human [78, 79]. The amino acid sequence of CYP26B1 is 42% identical to CYP26A1, and CYP26B1 is able to rapidly catabolized all-trans RA with efficiency and specificity comparable to that of CYP26A1 [78]. In addition to CYP26A1 and CYP26B1, CYP26C1 [80] and CYP26D1 [81] were identified and characterized.

Both CYP26C1 and CYP26D1 share significant amino acid identity with both each other and with CYP26A1 and CYP26B1. Unlike CYP26A1 and CYP26B1, CYP26C1 is able to metabolize both all- trans - and 9-cic -RA with equal efficiency to more polar metabolites [80]. CYP26D1 can metabolize all-trans RA, 9-cis RA, and 13-cis RA [81].

The various CYP26 genes are expressed in patterns that are specific with respect to location and timing during embryonic development and are essential for establishing the body patterns of the embryo. CYP26A1 is expressed throughout the murine developmental process and can be detected as early as embryonic day 8.5 in mouse embryos [82]. Genetic ablation of CYP26A1 is embryonic lethal and results in many of the defects observed when embryos are exposed to teratogenic levels of RA [83]. Murine

CYP26B1 is expressed in the developing hindbrain, forebrain, spinal cord, lung, kidney, heart, eye, and limb bud and its expression is essential for normal development [75].

However, differential expression of CYP26A1 and CYP26B1 were shown during murine organogenesis [75]. In addition, CYP26C1 is expressed in the developing hindbrain, inner ear, and branchial arch of mice [80], while CYP26D1 is expressed during zebrafish development [81, 84]. Thus, the members of the CYP26 family regulate RA levels during development, thus protecting the developing tissues from overexposure to RA.

In adults, the members of CYP26 family also show different tissue-specific expression patterns. CYP26A1 shows low level expression in most of the tissues, but it is present at the relatively high level in the liver, duodenum, colon, placenta, and in some regions of the brain including olfactory bulb, temporal cortex, and hippocampus [78, 82,

85]. CYP26A1 levels are highly induced by RA in a number of human cell lines, and it has been demonstrated that several tissues exhibit induced expression of CYP26A1 transcript following exposure to RA, implying an important role of CYP26A1 in the RA homeostasis. CYP26B1, in the adult, is broadly expressed at detectable levels in most tissues, but is predominantly expressed in brain tissues, notably the pons and cerebellum, in which CYP26A1 is absent in any of the corresponding brain tissues, suggesting that

CYP26B1 is required for protecting this tissue from exposure to RA and for normal brain functioning. To date, the exact roles that CYP26C1 and CYP26D1 play in RA metabolism have not been clearly delineated yet. Although the full-length CYP26C1 cDNA was derived from RNA isolated from human adrenal tissue, attempts to characterize CYP26C1 expression did not convincingly show the presence of CYP26C1 transcript in any of the other RNA samples isolated from adult human tissues [80]. The identification and characterization of CYP26D1 in mammalian species has not been reported. Thus, CYP26A1 and CYP26B1 are the predominant forms in adult tissues, and most likely to catalyze RA metabolism. In addition to the CYP26 family, several other cytochrome P450 isoforms have been identified that are capable of converting RA into 4- hydroxy- and 4-oxo-RA, including CYP2C8, CYP3A4, CYP2C9, and CYP1A1 have been shown to catalyze the 4-hydroxylation of all-trans-RA metabolism in human liver microsomes [86].

RA is known to induce its own metabolism in vivo and in cultured cells, and multiple tissues and cell types convert RA into polar metabolites [2, 27, 87] . CYP26A1 expression is rapidly and strongly induced by RA in many cell lines, such as nonsmall cell lung carcinoma LC-T, breast adenocarcinoma MCF7, acute promyelocytic leukemia- derived NB4, and hepatocarcinoma HepG2, but not all cell types [17]. CYP26A1 expression is regulated by diet and RA in the liver and some extrahepatic tissues [6], including lung, small intestines, and testis, but not in brain [82]; however, the response of

CYP26A1 to RA is most dramatic in the liver. In vitamin A-deficient rats, little, if any,

CYP26A1 mRNA was detected in the lung, small intestine, and testis, but treatment with

RA resulted in a strong induction, equaling 9-, 6- and 2.5-fold, respectively [6]. In

comparison, CYP26A1 mRNA levels were virtually undetectable in VA-deficient rats, while a single dose of RA resulted in a significant increase in CYP26A1 mRNA levels by as early as 3 hours post-treatment, which reached a peak of nearly 2000-fold within 10 hours. Moreover, CYP26A1 mRNA increased linearly in response to increasing exogenous RA doses [6]. In addition, CYP26A1 mRNA expression increased progressively with dietary VA intake and CYP26A1 mRNA was directly correlated with liver retinol levels: vitamin A deficient < vitamin A marginal < control < vitamin A supplemented << RA treated [87]. CYP26B1 is also regulated by high-dose exogenous

RA treatment in the liver [88], but the increase of CYP26B1 was not as dramatic as for

CYP26A1. Taken together, regulation of RA oxidation involves local control in peripheral tissues as well as strong, potentially high capacity, central metabolism in the liver. Moreover, CYP26A1 in the liver may be considered as a “central enzyme”, sensing and catabolizing RA to polar metabolites (Figure 4).

+ all-trans RA

CYP26A1 CYP26A1 RXR RAR mRNA protein

RAREs CYP26A1 gene

4 -OH RA 4 -oxo RA 18-OH RA

Polar Derivatives

Figure 4 . Autoregulatory feedback loop to control RA levels. High physiologic levels of RA can induce CYP26A1 expression through transcriptional regulation of the CYP26A1 gene, consequently increasing CYP26A1 protein. The increased CYP26A1 enzyme will act to normalize RA levels. Once RA levels are back to normal, CYP26A1 expression falls off.

Molecular mechanism for the regulation of CYP26A1 gene expression

The molecular basis of the rapid increase in CYP26A1 mRNA by RA was first studied by Loudig et al.[20], who characterized the putative promoter region of the

CYP26A1 gene, which is well conserved among human, mouse, and zebrafish, and contains a classical retinoic acid response element (RARE), located approximately 100 bp upstream of the translation start site, and guanine-guanine-rich element (GGRE) upstream of the RARE. This RARE, RARE1, is composed of two direct hexanucleotide repeats (agttca), spaced by 5 base-pairs (DR5). In vitro , RARE1 could bind RAR γ/RXR α heterodimers, as shown by retardation in electrophoretic mobility shift assays (EMSA) of a DR5-containing probe, representing R1, that was incubated with in vitro -translated receptor proteins and nuclear extracts from RA-treated embryonic stem cell lines, F9 and

P19. In F9 cells transfected with a 256 bp fragment of wild-type or mutated mouse

CYP26A1 promoter fused with a luciferase reporter gene, the addition of 1 µM RA increased transcriptional activity 2-fold, while mutation of the RARE resulted in a nearly complete loss of inducibility, while cotransfection of RAR γ and RXR α increased basal activity 2-fold and RA-induced activity nearly 4-fold. Similar results were found in COS-

1 cells (derived from African green monkey kidney) and Hela cells (derived from human cervical cancer cells). This report also showed that TAAT motifs flanking the RARE positively influence the RARE, as mutation of these motifs greatly decreased basal and

RA-inducible promoter activity. By DNase I footprinting analysis both the RARE and

GGRE sequences were hypersensitive to RA treatment, indicating GGRE might be involved in the RA inducibility. The GGRE, which interacted with Sp1/Sp3 in vitro , was

important for supporting RARE activity, as its elimination reduced both basal and RA- inducible promoter activity.

Retinoic acid signaling: retinoic acid receptors and retinoid X receptors

It is now known that RA can influence gene expression and protein production in many ways, but in terms of molecular mechanisms, a single predominant, classical pathway has been characterized through two families of nuclear hormone receptors composed of three subtypes each: retinoic acid receptors (RARs: RAR α, β, γ) and retinoid X receptors (RXRs: RXR α, β, γ) [3, 7]. For each isotype of receptors, there are at least two isoforms generated by differential promoter usage and alternative splicing that differ only in their N-terminal region. The RAR family is activated by all-trans RA and its 9-cis isomer, whereas the RXR family is activated exclusively by 9-cis RA.

Artificial retinoids have also been synthesized to bind certain RAR or RXR specifically, such as Am580 [89], which activates RAR α only and is not as easily catabolized as natural retinoids.

The retinoid receptor gene family belongs to the superfamily of steroid hormone receptors. As most nuclear hormone receptors, each protein contains several domains that are similarly organized and well conserved within the RAR and RXR subfamilies [3, 90].

The amino terminus of RAR or RXR contains a ligand-independent transcriptional transactivation function domain (AF-1) domain, which is variable among different receptors. This domain contains several consensus phosphorylation sites, and the phosphorylation of the AF-1 domain helps the ligand-dependent recruitment of

coactivators and chromatin modifiers. The DNA-binding domain (DBD) in the center of the protein contains two zinc-finger motifs. The Carboxyl terminus harbors the ligand- binding domain (LBD). It is functionally complex as it contains the ligand-binding pocket, the main dimerization domain and the ligand-dependent transactivation function

(AF-2). The DBD and LBD of all subtypes are highly conserved.

RARs and RXRs commonly work together, in the form of heterodimers, binding to DNA and regulating gene expression (Figure 5). The binding site for RAR/RXR heterodimers is a specific DNA sequence known as the retinoic acid response element

(RARE) [3, 8]. RAREs consist of a direct repeat of a core hexameric sequence, PuG

(G/T) TCA, separated by 1, 2, or 5 base pairs (DR1, DR2, or DR5, respectively). RXR homodimers, on the other hand prefer DR-1, although they bind to DR-1 less efficiently than do RAR/RXR heterodimers. In the absence of ligand, RAR/RXR heterodimers are bound to response elements located in the promoter of target genes, and recruit corepressors complexes, such as nuclear receptor corepressor 1 (NCoR1), and silencing mediator for retinoid and thyroid hormone receptor (SMRT), which recruit histone deacetylase (HDAC), resulting in chromatin condensation and gene silencing. Upon ligand binding, RAR and RXR change conformation, which favors the dissociation of corepressors and the recruitment of coactivators complexes that display histone acetyltransferase (HAT), methyltransferase, and ATP-dependent remodeling activities.

This results in decompacting of chromatin, making it accessible to transcriptional machinery to initiate transcription [3, 7, 8].

A Corepressors complexes

AF-2 AF-2 Histone deacetylation RXR/RAR Repression of transcription AF-1 AF-1 heterodimer

DBD DBD RARE

ligand

B Coactivators complexes

AF-2 AF-2 AF-1 AF-1 Histone acetylation/ methylation/phosphorylation DBD DBD RARE

C Mediator and transcription machinery AF-2 AF-2

AF-1 AF-1 Activation of transcription DBD DBD RARE

Figure 5 . Mechanism of retinoid receptor action [3, 4]. RAR and RXR form heterodimers that bind within the regulatory region of target genes through RARE. (A) In the absence of ligand, the heterodimer are associated with transcriptional corepressors which then recruit HDACs, resulting in chromatin condensation and gene silencing. (B) Binding of ligand induces the release of the HDAC complex and results in the recruitment of coactivators which lead to decompact repressive chromatin. (C) Subsequently, mediator complexes and transcription machinery are recruited, resulting in transcription initiation.

Recently, Balmer and Blomhoff [9] classified over 500 genes reported to be regulated by RA into categories of genes that are either directly regulated through classical RAR/RXR-mediated transcriptional activation or are indirectly regulated by RA through nonconventional pathways. Twenty-seven of these genes had been reported to be unquestionably controlled through the classical pathway. For the other genes, regulation may occur through indirect, or unknown, mechanisms that are still important physiologically. CYP26A1 qualifies as a directly regulated target gene.

RA is known to autoregulate its own homeostasis. The concept at the level of gene transcription was proposed based on the findings of an RARE within the promoter of genes encoding several retinoid-related proteins, such as RAR β2, RAR α2, CRABP-II,

CRBP-I, CRBP-II, and CYP26A1. In addition to transcriptional regulation, it is likely that other autoregulatory mechanisms contribute to retinoid homeostasis, such as retinoid-dependent proteolytic degradation of retinoid receptors, and regulation of retinoid receptor stability by phosphorylation [3].

RARs and RXRs are expressed early in embryonic development, a time at which

RA helps to specify body pattern formation, and they are also expressed in a wide variety of adult cell types, in amounts and proportions that differ among tissues [91, 92]. In addition, because RXRs functions as a partner for several families of nuclear receptors other than RARs, such as thyroid hormone receptor (TR), vitamin D receptor (VDR), peroxisome proliferator-activated receptor (PPAR), and so on, RXRs and its ligand have the potential to influence a wider variety of genes than are regulated by RARs.

Conclusion

The liver plays a major role in the uptake of RA from plasma and its rapid oxidation to polar metabolites [19]. The level of CYP26A1 mRNA in the liver changes rapidly and dramatically in response to treatment with RA, while in the steady-state the level of CYP26A1 mRNA increases in proportion to dietary vitamin A intake [6]. Thus, understanding the regulation of CYP26A1 gene expression in the liver may be especially important for adult retinoid homeostasis. Investigating the mechanism of CYP26A1 response to RA and the role of nuclear retinoid receptor (RAR/RXR) could be a promising start for understanding retinoid homeostasis in greater detail.

CHAPTER 3 ADDITIONAL RARES ARE INVOLVED IN RA-INDUCED CYP26A1 EXPRESSION

Our inspection of the upstream region of the CYP26A1 promoter suggested there could be additional retinoid response elements besides RARE1, which we tentatively designated RARE2, and RARE3, as well as a hexanucleotide half-site, RARE4. The current studies were designed to test the ability of these elements, alone and in combination, to activate expression of CYP26A1 promoter-luciferase constructs in cell lines representative of human embryonic kidney and adult liver cells, and the binding interactions of the RAREs with the retinoid receptors RAR α, β, and γ.

Experimental Procedures

Cell Culture and Reagents : HEK293T cells (human embryonic kidney, transformed) and HepG2 cells (human hepatocellular carcinoma) were maintained in

Dulbecco's modified eagle medium (DMEM) supplemented with 10% fetal bovine serum

O (FBS) and 0.5% penicillin-streptomycin at 37 C in a 5% CO 2-air incubator. In most experiments, the cells were plated at approximately 70% confluency. All-trans RA

(prepare in ethanol) was obtained from PreproTech Inc. (Rocky Hill, NJ). Am580 (RAR α agonist [93]) was provided by Dr. H. Kagechika (Tokyo, Japan).

DNA Plasmid: Promoter constructs were based on the human CYP26A1 sequence

(from NCBI) and cloned into pGL3 Basic Luciferase reporter vector (Promega Corp.,

Madison, WI). The Full Length (FL) constructs were generated by PCR amplification (30 cycles) of the upstream region of the translation start site of human CYP26A1 gene

(nucleotides -2131 to +19), using primers 5’-AATAAGCTTCACGAAG

GTGCAGAGCGC-3’ (forward) and 5’-AGCAAGCTTGTACAGATAGATTAA

AACGT-3’ (reverse). After digestion of the PCR product with Hind III, the fragment was isolated and ligated into the pGL3 Basic Luciferase reporter vector. After digestion of FL construct by different restriction enzymes, including Sma I, Bgl II, EcoR I, Apa I and Pst I, in order to eliminate specific parts of the full length promoter, and then self re-ligated to generate eliminated constructs (E1 to E6). The mutant promoter constructs (M1 to M4) were generated by PCR replacing the wild-type RARE with restriction site EcoR V linked selected mutations. Primers were used for mutant constructs: pair1: 5’-

TAGATATCTTTAAAATTGTCTGACCAAGGTAACG-3’ (forward), 5’-TAGATA

TCTAAAATTCTTTAATTGCGGATTGGGCC-3’ (reverse); pair2: 5’-TAGATATCTT

TATGGCCCGAGGATTGGGAATGG-3’ (forward), 5’-TAGATATCAAAATCCTGCA

GGCCGGACCGTG-3’ (reverse); pair3: 5’-TAGATATCTTAAACACGGTCCGGCCT

GCAGG-3’ (forward), 5’-TAGATATCTTAACTGCGGGGCCACCTCGCC-3’ (reverse); pair4: 5’-TTAGATATCCGTCCCCATTCGTCGGCT-3’ (forward), 5’-TTAGATAT

CGGGGCCCATTCCCAATCC-3’ (reverse). Mutated RARE1 was 5’-ATTTTAGAT

ATCTTTAA-3’, mutated RARE2 was 5’-GATTTTGATATCTTTAT-3’, mutated

RARE3 was 5’-TTAAGATATCCTTAAACA-3’, and mutated RARE4 was 5’-

GATATC-3’.

Transient Transfection and Luciferase Assay: Twenty-four hours before transfection, HEK293T or HepG2 cells were split and 5x10 5 cells were seeded in each well of 12-well plates in 1 ml of complete culture medium. Two hours before transfection, the complete culture medium was replaced by 5% FBS medium without antibiotics. Transfection was mediated by LipofectamineTM 2000 (Invitrogen Corp.,

Carlsbad, CA). After 24 hours transfection, cells were treated either with ethanol or with

RA. After a 6 hour or 24 hour treatment, cells were harvested in passive lysis buffer and subjected to luciferase assay according to the manufacturer’s instructions (Promega Corp.,

Madison, WI). TK-Renilla luciferase plasmid (Promega Corp., Madison, WI) was used as internal control for transfection. The ratio of Firefly versus Renilla luciferase activity was defined as the promoter activity.

RNA Isolation : HEK293T cells and HepG2 cells were cultured in 6-well plates

(0.5~1 x 10 6 cells/ well) for various times and treatments, then total cellular RNA was isolated by using Rneasy Mini Kit (Qiagen Inc., Valencia, CA) according to the manufacturer’s instructions.

Reverse Transcription and Real-Time Polymerase Chain Reaction (PCR): One g of total RNA extracted from culture cells as described above and subjected to reverse transcription using Moloney murine leukemia virus reverse transcriptase (Promega Corp.,

Madison, WI). The diluted reaction product (1/20) was used for PCR or real-time PCR analysis using 2 x Real Time SYBR Green/Fluorescein PCR Master Mix (SuperArray

Bioscience, Frederick, MD) in a final volume of 20 µl. Glyeraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was amplified at the same time as an internal control.

PCR conditions were optimized as denaturation at 94OC for 30 sec, annealing at 60 OC for

30 sec, and extension at 72 OC for 30 sec for 30 cycles. PCR products were applied to 1% agarose (Invitrogen Corp., Carlsbad, CA) gel. The primers used for PCR were 5'-

GTGCCAGTGATTGCTGAAGA-3' (forward) and 5'-GGAGGTGTCCTCT GGATGAA-

3' (reverse).

Nuclear Extract Preparation : Nuclear extracts from HepG2 cells were prepared as described by Chen et al [94]. Briefly, cell pellets was homogenized in a hypotonic buffer [10 mM HEPES, pH 7.9, 1.5 mM MgCl 2, 10 mM KCl, 0.2 mM PMSF, 0.5 mM

DTT, 1 mM sodium orthovanadate, 0.5% Nonidet P-40] using a Dounce tissue homogenizer. After centrifugation at 2500g at 4°C for 10 min, the supernatant

(cyptoplasmic fraction) was removed, and then pellets were washed once with hypotonic buffer containing no detergent. Then a hypertonic buffer (20 mM HEPES, pH 7.9, 10% glycerol, 1.5 mM MgCl 2, 400 mM KCl, 0.2 mM ethylenediaminetetraacetic acid, 0.2 mM

PMSF, 0.5 mM DTT, 1 mM sodium orthovanadate) was added to extract nuclear protein.

After 30 min incubation on ice, the mixture was centrifuged at 15 000g for 30 min. The

supernatant was then collected as the nuclear extract, divided into aliquots and stored at

−80°C.

Electrophoretic Mobility Shift Assays (EMSA) : The fragments containing putative

CYP26A1-RAREs (wild type and mutation) were synthesized as probes. The sequences of the probes were: RARE1:5’-taaagaTGAACTttgggTGAACTattgt-3’; mutation of

RARE1:5’-taaagaATTTTAgatatCTTTAAaattgt-3’; RARE2: 5’-ctgcaGGGTCAcaggc

GGGTCAggcccg-3’; mutation of RARE2: 5’-ctgcagGATTTTgatatCTTTATggcccg-3’;

RARE3: 5’-ccgcagAGTTCActcggATGTCAcggtcc-3’; mutation of RARE3: 5’- ccgcagTTAAGAtatctRAAACAcggtcc-3’; RARE4-1: 5’-gattgggaatgggccccAGGTCA cgtccc-3’; mutation of RARE4-1: 5’-gattgggaatgggccccGATATCcgtccc-3’; RARE4-2:

5’-ggccccAGGTCAcgtccccattcgtcggc-3’; mutation of RARE4-2: 5’-ggccccGATATC cgtccccattcgtcggc-3’. The wild type fragments were labeled with [ γ-32 P]dATP and incubated with the nuclear protein extract at room temperature for 30 min. For the supershift assay, nuclear proteins were incubated with antibodies for 15 min. The reactions were separated on a 5% nondenatured polyacrylamide gel by electrophoresis.

After electrophoresis, the gel was dried and subjected to autoradiography.

Statistical analysis : In each experiment, at least two replicate samples were analyzed for each treatment or time point. In addition, most experiments were further repeated. Statistical analysis was performed by using SuperANOVA software (Abacus

Concepts, Berkeley, CA) for student’t t-test, one-way analysis of variance, and simple regression.

Results

Three RAREs and a half site are highly conserved in the upstream region of

CYP26A1.

Sequence analysis showed that three putative RAREs and a half site were embedded in the putative promoter of human CYP26A1 gene (Figure 6). RARE1 is close to the transcription start site (about 100bp), and a TATA box and a guanine-guanine-rich element (SP1/SP3 binding site), are in the proximal region [20]; the other two RAREs and a half RARE are about 2kb upstream away, but close to each other. All the three

RAREs are composed of repeated core hexametric sequences separated by five spacer nucleotides, characterized as DR5. Moreover, the orientation of RARE1 is opposite to the other RAREs’ (Figure 7, orientation indicated by arrows). These RARE including flanking regions are highly conserved among mouse, rat and human (Figure 7).

To investigate the activity of the putative human CYP26A1 promoter, a construct was built up containing 2,150 bp upstream of CYP26A1 translation start site fused to a luciferase reporter gene, and then a series of constructs were generated based on the first construct by eliminating specific parts of the putative promoter (shown in Figure 8). All constructs were transiently transfected into HEK293T or HepG2 cells. Cells were treated with ATRA for certain time and then harvested for the measurement of luciferase activity. The relative luciferase activities represented the promoter activities.

Figur e 6. Sequence analysis shows that three RAREs and a half site are embedded in the putative promoter of human CYP26A1. Shaded box indicates RAREs, and colored nucleotides are core hexametric sequences. One open box is guanine-guanine-rich element (GGRE), the other is TATA box. Numbers are relative to the transcriptional start site.

Figure 7. RAREs in the putative promoter of CYP26A1 are highly conserved among mouse, rat and human. Core hexametric sequences in RARE regions are shaded in boxes, and arrows indicate their orientation. Asterisks indicate perfectly conserved nucleotides. Numbers are relative to the transcriptional start site of human CYP26A1.

Endogenous CYP26A1 gene expression and RA inducibility exhibit cell line specificity .

As shown by RT-PCR analysis (Figure 9A, 9B), both HEK293T and HepG2 cells had endogenous CYP26A1 expression, suggesting that both cell lines are appropriate for

CYP26A1 promoter analysis. CYP26A1 mRNA levels were relatively higher in

HEK293T cells, whereas HepG2 cells showed more induction of transcription of

CYP26A1 after treatment with 1 M ATRA for 6 hours.

CYP26A1 promoter activities show cell line specificity.

The construct FL including all the three RAREs and a half site, exhibited the highest magnitude of RA-inducible activity in both cell lines (Figure 9C, 9D). However, the FL promoter showed more strong inducibility by ATRA in the HepG2 cells (about 20 fold) than in the HEK293T (about 4.5 fold). In the HepG2 cell line, eliminating the upstream fragment from -1885 (E1) to -329 (E3) led to a slight increase in the basal activity of the promoter, but almost lost the response to ATRA. Nevertheless, this phenomenon was not observed in HEK293 cell line. When with the elimination of the region from -1994 to -1652 (E5), in which RARE2 and RARE4 are embedded, the promoter lost RA response in HepG2 cells, implying RARE2/4 might be essential for

RA-induced CYP26A1 expression in HepG2 cells. Moreover, the construct E4, in which the middle fragment was eliminated and RARE1, 2 and 3 were maintained, still exhibited high RA responsiveness. Interestingly, the reverse construct (RFL) had a basal level activity and also strongly responded to RA in HepG2 cells.

Figure 8. Constructs used for transient transfections. A series of constructs containing the region upstream of the CYP26A1 translation start site, fused to a luciferase reporter gene. FL indicates the full-length construct. RAREs are abbreviated to Rs. Es indicate partially eliminated promoter constructs. Lines represent the eliminated regions compared with full-length promoter. Ms indicate the mutated constructs. Black boxes indicate where an element has been mutated. Numbers are relative to the transcriptional start site.

Figure 9. CYP26A1 promoter activities show cell line specificity. Endogenous CYP26A1 mRNA was measured by RT-PCR in HEK293T cells (A) and in HepG2 cells (B) after treated either vehicle or ATRA for 6 hours. RT-PCR products were applied to 1% agarose gel. Transient Transfection analyses were performed in HEK293T cells (C) and HepG2 cells (D). The constructs included Full-length (FL) construct, reverse-full-length (RFL) construct, and the constructs with partially eliminated promoters (E1 to E6). Relative luciferase activities were measured after RA treatment for 24 hours, and renilla activities were measured as internal controls for transfection. And the ratios of firefly to renilla luciferase activities represent promoter activities. Samples from HEK293T cells were duplicated, error bars indicated a range. Samples in HepG2 cells were shown (n=4), error bars indicated the SD.

Relative luciferase activities of the FL construct is correlated with the level of endogenously expressed CYP26A1 mRNA.

HepG2 cells expressed very little endogenous CYP26A1 mRNA prior to retinoid treatment, but they responded rapidly and dose-dependently to 1-1000 nM ATRA (Figure

10B), and therefore closely resembled intact liver [6]. The response to Am580, a specific ligand for RAR α, partially resembled the RA inducibility (Figure 10B), indicating the

RA effect could be through RAR α, and other RARs could be also involved.

The FL construct contained 2150 bp upstream of the translation start site. Relative luciferase activity of the construct FL was dose-dependently regulated by ATRA in

HepG2 cells, and correlated with the level of endogenously expressed CYP26A1 mRNA

(Figure 10A, 10B).

In the time course response in HepG2 cells (Figure 10C, 10D), treated with 0.1

µM ATRA or Am580, the endogenous level of CYP26A1 mRNA was comparable between 6-hour and 24-hour treatments. However, relative luciferase activity was higher after 24-hour treatment than 6-hour treatment. This may result from a time-consuming translation step of the luciferase reporter gene, thus we used 24-hour treatment for further luciferase activity assays.

Figure 10 . Relative luciferase activity for the full-length (FL) construct was correlated with the level of endogenously expressed CYP26A1 mRNA. (A) Relative luciferase activities of FL construct were measured after RA or Am580 treatment for 24 hours in HepG2 cells. (B) Endogenously expressed CYP26A1 mRNA was measured by real-time PCR. HepG2 cells were treated with different doses of RA or Am580 for 24 hours. (C) HepG2 cells were transiently transfected with FL construct. Relative luciferase activities were measured after ATRA or Am580 treatment for 24 hours in HepG2 cells. (D) Endogenously expressed CYP26A1 mRNA in HepG2 cells was measured by real-time PCR. All transient transfection experiments were normalized with the renilla luciferase activities. And all PCR were normalized with GADPH.

All the three RAREs and the half site synergistically regulate RA-mediated

CYP26A1 expression.

To characterize the putative RAREs activities, a number of constructs were generated in which three RAREs and a half site were mutated respectively (Figure 8).

The result of transient transfection showed that mutating any RARE resulted in a decrease of RA-induced promoter activities, implying all the three RAREs and the half site were required for optimal RA-mediated CYP26A1 expression (Figure 11). In

HEK293T cells, the proximal RARE1 was more important for the response to RA compared with either RARE2 or RARE3. This was supported by the lowest RA-mediated induction of promoter activity when RARE1 was mutated (Figure 11A). However, in

HepG2 cells, RARE2 was essential for RA induction compared with the others since in the absence of RARE2 (construct M2) the lowest response to RA was observed (Figure

11B). Therefore, functions of RAREs in the CYP26A1 promoter have cell line specificity. Interestingly, in both cell lines, the half site also contributed to RA induction.

When the half site was mutated (construct M4), the RA induction dramatically decreased

(Figure 11A, 11B).

Figure 11 . All three of the RAREs and the half site synergistically mediate the RA-induced CYP26A1 expression. Three RAREs and the half-site site were mutated respectively, and transiently transfected into HEK293T cells (A) and HepG2 cells (B), and then relative luciferase activities were measured after 1 M RA or vehicle treatment for 24 hours. (C) FL construct and mutated constructs (M1 to M4) were cotransfected with RXR α and each of the individual retinoic acid receptors (RAR α, β, and γ) into HepG2 cells. Relative luciferase activities were measured after 0.1 M ATRA or vehicle treatment for 24 hours. All transient transfection experiments were normalized with the renilla luciferase activities. (On the next page)

Cotransfection of each of the individual retinoic acid receptors (RAR α, β, γ) significantly increases promoter activities .

Am580 treatment suggested that not only RAR α but also RAR β or RAR γ could be involved in RA induction. Therefore, we conducted cotransfection studies. When

RXR α as well as RAR α or RAR β or RAR γ were cotransfected into HepG2 cells, the basal promoter activity of the wild type construct FL was increased. And cotransfection with RXR α and RAR α had less effect on basal activity than either RAR β or RAR γ.

After treatment with 0.1 M RA for 24 hours, the promoter activities, including the wild type and the mutant constructs, were significantly increased. For example, activity of the wild type construct was increased about 9 fold. However, the effects of RAR α, RAR β and RAR γ on the each individual RARE mutated construct were comparable (Figure

11C).

Figure 12 . RAREs have similar binding patterns with nuclear protein and RA increases DNA-binding activities. (A) Set of oligonucleotides used. Core hexameric sequences in RARE regions were shaded. (B) EMSA was performed using radioisotope-labeled oligonucleotides with HepG2 nuclear protein extracts (NPE). HepG2 cells were treated with vehicle or 0.1 M ATRA for 6 hours. Unlabeled oligonucleotides (50x molar excess) were used as competitors. RAR α antibody was pre-incubated with 6-hour RA-treated NPE prior to the addition of labeled probe. (C) EMSA was performed using radioisotope- labeled oligonucleotide containing RARE2 with HepG2 NPE. HepG2 cells were treated with vehicle or 0.1 M ATRA for 6 hours. Unlabeled oligonucleotides (50x molar excess) were used as competitors. RA-treated NPE was used in the competitions. Antibodies were pre-incubated with 6-hour RA-treated NPE prior to the addition of labeled probe.

RAREs from CYP26A1 promoter have similar binding patterns with nuclear proteins and RA increases in DNA-binding activities.

To further determine the physical binding between DNA and RARs, and involvement of RAR subtypes, EMSA was conducted (Figure 12). Nuclear proteins incubated with labeled RARE2 showed stronger binding in the upper bands than both

RARE3 and RARE4 (Figure 12B). The arrows indicate the RARE-RAR/RXR complex.

Specificity is supported by cold probe competition and by antibody reaction. For the

RARE2 complex (Figure 12C), unlabeled wild-type RARE2 and DR5 consensus probes specifically competed with RAR/RXR bound hot RARE2 (comparing lane 4 and lane 6 with lane 3 in Figure 12C), but unlabeled mutated RARE2 and DR5 could not compete with the receptor-bound hot RARE2 (comparing lane 5 and lane 7 with lane 3 in Figure

12C). Moreover, all the unlabeled wild-type RAREs showed competition with the receptor-bound hot RARE2 (shown in lane 8, 9, 10, and 11 in Figure 12C). Pre- incubation with antibodies did not result in supershifted bands, which may be because the complexes with antibodies were too large to enter the gel. However, antibodies did show competition with receptor-bound hot RARE2, reducing the intensities of bands (shown in lane 12 to lane 17 in Figure 12C). Moreover, anti-RAR antibody showed stronger competition than the other antibodies (lane 12 in Figure 12C). In addition, ATRA treatment markedly increased the DNA-protein binding (comparing lane 2 with lane 3 in

Figure 12B and 12C), indicating more of the bound factors were present in RA-treated cells compared to control cells.

CHAPTER 4 DISCUSSION

CYP26A1 is essential for RA homeostasis in normal and clinical conditions.

RA is known to regulate vitamin A metabolism by regulating several enzymes and binding proteins involved in vitamin A metabolism. Figure 13 illustrates the integrated relationship among retinoid metabolism and retinoid-binding proteins.

Depletion of plasma retinol or an increase of intracellular apo-CRBP results in inhibition of LRAT and stimulation of REH, consequently elevating intracellular retinol through mobilization of RE [2]. Moreover, an increase of apo-CRBP also reduces the activity of

RDH, decreasing the oxidation of retinol. RA has been demonstrated to regulate retinol metabolism [95]. RA can increase the gene expression of LRAT [5], and also increase the cellular level of holo-CRBP by increasing transcription of CRBP [9]. Subsequently, an increase of cellular RA level inhibits retinol oxidation, and stimulates retinol esterification. Moreover, CRABPII is induced by RA, facilitating the deliver of RA from cytoplasm to nucleus [57]. In addition, RA induces its own oxidation by increasing

CYP26 expression, preventing cells from overexposing RA [87].

Liver is the major organ for storing vitamin A and providing retinol to peripheral tissues, but in turn, RA production in peripheral tissues regulates hepatic retinoid homeostasis (Figure 14). The ability of RA to regulate hepatic retinol metabolism suggests RA serves as the principal signal of the body’s vitamin A adequacy. Although

the liver contains several enzymes which are capable of oxidizing retinol and retinal to produce RA, the majority of RA (~75%) in liver is produced by peripheral tissues [56]. In the peripheral tissues, the concentrations of RA are about 4-5 times higher than plasma

RA [96], and RA is readily diffusible across the plasma membrane [56]. Hepatic LRAT was decreased in rats fed vitamin A deficient diet, but when these rats were treated with

RA, LRAT level returned back to normal [5]. This suggests that LRAT remains in a constitutively induced state as long as the RA signal is in adequate supply. When vitamin

A intake becomes inadequate and plasma retinol falls [97], then peripheral RA production would be expected to fall, then in turn, the concentration of RA taken by liver would no longer be adequate to maintain expression of hepatic LRAT, consequently less retinol would be converted to retinyl ester for storage, but more retinol would be secreted and transported to target tissues. Supplementation with RA for the vitamin A-adequate rats caused a reduction of plasma retinol with an increased expression of LRAT [87, 98].

CYP26 catabolizes the oxidation of RA. It is important for retinoid homeostasis and prevents tissues from being overexposed to RA. CYP26 gene expression in liver is regulated by vitamin A status: vitamin A deficient < vitamin A marginal < control < vitamin A supplemented << RA treated [6]. CYP26 may be a biochemical mechanism for sensing RA through its RAREs (as we identified) and efficiently catabolizing RA to polar metabolites, leading to rapid clearance plasma RA. Thus, in the vitamin A-deficient state

(Figure 14B), decreased expression of hepatic LRAT and CYP26 results in maximal retinol for secretion, with depletion of storage, and a decrease of oxidation.

Administration of RA (Figure 14C) increases hepatic LRAT and CYP26 expression, leading to increased storage and oxidation but decreased secretion [4-6].

RBP- Plasma Retinol

Apo- CRBP

— + CRBP- RDH CRBP- Polar Retinol Retinal Metabolites

— + RALDH CYP26 + + LRAT REH CRABP- + Retinyl RA Ester

+ RXR RAR

NUCLEUS

Figure 13 . A model of retinoid metabolism that integrates the retinoid binding proteins [2]. CRABP, cellular retinoic acid binding protein; CRBP, cellular retinol binding protein; CYP26, enzymes of cytochrome P450 family; LRAT, lecithin:retinol acyltransferase; RA, retinoic acid; RALDH, retinaldehyde dehydrogenase; REH, retinyl ester hydrolase; RDH, retinol dehydrogenase; RAR/RXR, nuclear retinoic acid and retinoid X receptors. Bold type indicates enzymes. Enzymes or binding proteins that can be positively modulated by RA are labeled with ⊕. The dashed lines indicate the effect of apo-CRBP on LRAT and REH.

Figure 14 . Model of inter-relationship between liver and peripheral tissue retinoid metabolism [4-6]. (A) The possible cross-talk between the liver and peripheral tissues. Retinol circulates several times between plasma and liver before undergoing irreversible disposal, which provide an ideal means for the liver to constantly sample and adjust the concentration of retinol available for the peripheral tissues. Interorgan flux of RA has been proposed as RA is readily diffusible across the plasma membrane. And the preponderance of liver RA is obtained from plasma (indicated by a bold arrow). Liver exports retinol; and then peripheral tissues metabolize retinol to RA, returning excess RA to liver; RA in liver regulates retinoid homeostasis including esterification, secretion, and oxidation. (B) The retinoid homeostasis between liver and peripheral tissues under the vitamin A- deficient status. In the vitamin A deficiency, plasma retinol level is lower than normal. Uptake of retinol in the peripheral tissues is low, and the production of RA is low. Consequently, the signal of RA back to liver is reduced, and the expression of LRAT and CYP26 is decreased, which leads to a decrease of retinol esterification for storage and retinol oxidation. In the vitamin A deficient state, the reaction chain favors providing retinol for secretion as much as possible. In certain peripheral tissues, RE is mobilized to make up their needs. (C) The retinoid homeostasis between liver and peripheral tissues after RA administration. Increased plasma RA is rapidly uptake by liver. RA induces expression of hepatic LRAT and CYP26, increasing the production of RE and the oxidation of RA to polar metabolites. Thus, the reaction chain shifts so that the secretion of retinol is reduced, and low plasma retinol level is observed. In some peripheral tissues, CYP26 is also induced by exogenous RA, which increases local oxidation of RA. Abbreviations: CYP26, enzymes of cytochrome P450 family; LRAT, lecithin:retinol acyltransferase; RA, retinoic acid; RE, retinyl ester. Bold type indicates enzymes. Enzymes or retinoid levels marked with ⊕ if they are increased, and marked with Ө if they are decreased. The bold arrows indicate major reaction directions. (On the next page)

A Liver Peripheral tissues

Retinol Export as Dieta ry input Retinol pool retinol uptake Input from recycling retinol Hepatic sensing of peripheral vitamin A status, and response modulation Plasma Bioactivation [esterification (LRAT), RA RA secretion, oxidation (CYP26), etc.]

B Vitamin A deficiency Liver Peripheral tissues RE _ _ RE _ LRAT Plasma Retinol _ Retinol retinol

Retinal

Plasma _ Plasma _ _ RA RA RA _ CYP26

Polar Polar metabolites metabolites

C RA administration Liver Peripheral tissues RE + RE + + LRAT Plasma Retinol _ Retinol retinol

Retinal

Plasma RA + RA + RA + + CYP26 Polar Polar metabolites metabolites

In the clinical settings, orally administrated all-trans RA has been effective in the treatment of acute promyelocytic leukemia (APL) [52]. APL is a specific type of acute myeloblastic leukemia affecting both children and adults. In most of the cases, APL is characterized by the morphology of blast cells, t(15, 17) translocation and coagulopathy.

In the mid 1980s, the vitamin A derivative all-trans RA was introduced for the treatment of APL. The first patients treated with RA for APL were in Shanghai, China in 1986 [99].

Subsequent clinical trials established the efficiency of RA. Current treatment for APL now consists of RA 45 mg/m 2/day orally for 45–90 days plus anthracycline chemotherapy for induction [100]. However, acquired RA resistance occurs in most patients treated with RA alone and combination RA chemotherapy; the resistance to the

RA-induced clinical remission markedly increases in frequency after second time relapses [52].

Kinetic studies showed that ATRA was very rapidly cleared from plasma following oral administration [101, 102] and chronic i.v. injection [103]. After oral administration of ATRA [102], plasma disappearance of ATRA was characterized by three distinct phases: a brief, initial exponential decline, followed by a relative plateau

(the duration was dose-dependent), and finally a terminal exponential decay. And half- life for ATRA averaged 19 minutes. Chronic i.v. dosing of ATRA induced hypercatabolic response [103]. Plasma RA as measured by the area under the concentration-time curve (AUC) decreased over time when drug was administered on a chronic daily schedule; and a decrease of plasma RA was associated with an increase of plasma RA clearance.

Isotope kinetic studies showed that most of the RA in the liver is derived by uptake from plasma [56]. The liver seems ideally suited by virtue of its anatomic location to act as a sensor of the level of RA. RA enters it both via the portal vein from the intestines and splanchnic circulation and via the hepatic vein from the systemic circulation. It has been shown that CYP1A1, CYP2C8, CYP2C9, and CYP3A4 in microsomes of human liver cells were able to hydroxylate RA to some extent [86], but none of these enzymes at protein and mRNA levels were inducible by RA. In addition, these enzymes usually have low specificity to RA. Therefore, increased RA metabolism after daily administration of RA is likely to be due to expression of CYP26 in metabolically active tissues, especially liver . Moreover, CYP26B1 was regulated by high dose of exogenous RA treatment in the liver [88], but the increase of CYP26B1 was not as dramatic as for CYP26A1. CYP26A1 may provide a biochemical mechanism for both sensing RA via its RARE and responding to RA through its induced enzymatic activity.

In our study, RARE2 region in the promoter of CYP26A1 in the HepG2 cells seems to be more important in sensing RA concentration and increasing CYP26A1 enzyme (Figure

11). Moreover, pretreatment of cells with RA resulted in higher metabolic activity [104].

In addition to decreased plasma RA levels, reduced intracellular RA levels in leukemic cells also contributes to a decreased sensitivity or resistance to the therapy. The treatment of NB4 cells and HL-60 cells (as in vitro APL model systems) with a pharmacological concentration of ATRA (1 M) induced rapid and dose-dependent expression of

CYP26A1 mRNA [104, 105].

Furthermore, in vivo studies, CYP26A1 mRNA expression in liver reached a peak around 10 hours and then decreased rapidly after single dose of RA [6]. And CYP26A1

mRNA expression in HepG2 cells returned to baseline after 48 hours upon removal of

RA in the culture medium [104].

Therefore, different strategies to circumvent the accelerated metabolism of RA, such as encapsulation of RA into liposomes, inhibitors of cytochrome P450 enzymes, intermittent administration of RA and synthetic retinoids have been tested [52].

Understanding the mechanisms RA-induced CYP26A1 expression would help in developing new therapeutic strategies and may significantly improve the rate and long- term maintenance of clinical remission in APL patients.

RAREs in the promoter region of CYP26A1 are highly conserved.

In our study, sequence analysis showed that three RAREs and a half site are embedded in the putative promoter of human CYP26A1 (Figure 6, 7). RARE1 is close to the transcription start site (about 100bp), and a TATA box and a guanine-guanine-rich element (SP1/SP3 binding site), are in the proximal region [20]. A previous study has reported this region in the mouse CYP26A1 promoter [20]. The RARE1 region is perfectly conserved between zebrafish, mouse and human. The other two and a half

RAREs are about 2kb upstream away, but close to each other. All the three RAREs are composed of repeated core hexametric sequences separated by five spacer nucleotides, characterized as DR5. These RARE including flanking regions are highly conserved among mouse, rat and human. After we started our CYP26A1 promoter study, Loudig et al. [106] identified a second putative RARE2 located upstream of RARE1 in the mouse

CYP26A1 promoter, which also supports our hypothesis. They found that RARE2 was highly conserved between mouse and human, and both the RARE1 and RARE2 elements

were required for optimal RA induction of mCYP26A1, which agree with our study. The predicted 497-amino acid human CYP26A1 protein is 93% identical to mouse protein,

91% to rat CYP26A1, and 68% to zebrafish protein. Therefore, both regulation and protein of CYP26A1 have been conserved during evolution, implying that this enzyme is very critical for maintaining RA concentration, and subsequently managing RA-mediated functions. Moreover, the conservation suggests that rats and mice are good models to investigate the regulation of CYP26A1 and enzyme function.

All the three RAREs and the half site synergistically regulated RA- mediated CYP26A1 expression.

In the HepG2 cells, the full-length construct showed the highest RA-induced promoter activity supported by the highest fold increase of reporter activity upon RA treatment. Relative luciferase activity for the construct FL was dose-dependently regulated by RA, and correlated with the level of endogenously expressed CYP26A1 mRNA. Truncated promoter partially lost RA inducibility. Especially when distal regions containing RARE2 were eliminated, the constructs almost entirely lost RA response, suggesting distal regions may be more important for RA response. Individual mutation of

RAREs further confirmed this idea. All the mutations partially lost their RA inducibility, even when only the half site was mutated, indicating all three RAREs and the half site are required for the optimal RA induction. Interestingly, transcriptional synergism between vitamin D responsive elements (VDRE) has been observed in the 25-hydroxyvitamin D 3

24-hydroxylase (CYP24) promoter [107]. In the CYP24 promoter, there are two and a half VDREs, close to each other. Two full VDRE sites showed transcriptional synergism

in response to 1,25-(OH) 2D3. But, the half site of VDRE was controversial; some studies on the half site showed its response to 1,25-(OH) 2D3 [108], but some not [107], which may be due to different cellular contexts. In addition, nuclear receptors are capable of distorting DNA [109]. For example, DNA-binding by estrogen receptor α, heterodimers of RXR with the vitamin D receptor, and the progesterone receptor resulted in distortion angles ranging from 7° to 75°. And RXR could self-associate into homotetramers [109], and binding of RXR tetramers to DNA containing two RXR response elements resulted in a dramatic DNA-looping. Therefore, we propose that RAR/RXR complexes binding to

RARE2 and RARE3 could distort DNA and form looping. Cofactors recruited by

RARE2- and RARE3- bound RAR/RXR complexes, together with cofactors recruited by

RARE1 bound RAR/RXR complexes, facilitate chromosome opening, and consequently favor the docking of mediator and transcription machinery. Since RAR also can interact with the general transcription factor TFIIH [110], additional RAR/RXR complexes could directly enhanced the docking of transcriptional machinery. Taken together, additional

RAREs enhance the RA-mediated induction. This model was supported by the high fold increase of promoter activity of the reverse linked full-length construct (RFL) upon RA treatment (Figure 9).

Although EMSA result showed that RARE4 could bind to RAR/RXR complex to some extent (Figure 12), whether RARE4 could bind to RAR/RXR heterodimer or facilitate RAR/RXR binding RARE2 needs further investigation.

In addition, in HepG2 cells, the further elimination of the region from -1885 to -

329 resulted in an increase of basic promoter activity (Figure 9), suggesting that some

inhibitory factors many bind this region and lower the basic activity, but not influence the

RA induction.

CYP26A1 expression and regulation exhibit cell specificity.

CYP26A1 expression is cell-specific . CYP26A1 mRNA expression in HepG2 cells was highly regulated by RA in a dose and time dependent manner, and closely resembled intact liver [6]. However, in HEK293T cells, untreated CYP26A1 expression level was relatively higher than in HepG2 cells but RA induction in HEK293T cells was not as dramatic as in HepG2 cells. CYP26A1 expression is known for its rapid and strong inducibility by RA in many other cell lines, such as LC-T, MCF7, NB4, and so on, but not in all cell types [17]. For examples, in endothelial cells (HUVEC) a high concentration of RA was required to induce expression of CYP26A1 [104], and MCF10A cells, did not show induction of CYP26A1 [17].

RARE2 is more important for RA induction in HepG2 cells but not in HEK293T .

In HepG2 cells, eliminating the distal region of the promoter resulted in almost complete loss of RA induction. And the RARE2-mutated construct showed a complete loss of RA mediated induction, indicating RARE2 is more responsible for the induction. In our

EMSA experiment (Figure 12), RARE2 showed stronger binding with protein, and other

RAREs partially competed with the RARE2-protein complex. However, in HEK293T cells, although distal region truncated constructs and the RARE2-mutated construct showed partially decreased RA induction, RARE1 mutation caused the most loss of RA induction, suggesting that the proximal RARE1 is more important for RA induction in

HEK293T cells. Moreover, HepG2 cells showed more RA responsiveness than HEK293

cells. Thus, we propose that active RARE2 along with RARE3 and RARE4 may contribute to the rapid and strong RA-mediated induction in HepG2 cells, compared to

HEK293T cells.

RARE3 seemed to have different effects in these two cell lines . In HepG2 cells, mutation of RARE3 resulted in a decrease of RA-mediated promoter activity but an increase in the basal activity; however, in HEK293T cells, this mutation caused a slight increase of promoter activity. This suggests that the region of RARE3 might be bound to an inhibitory factor which might affect the basal promoter activity, and that RARE3 in

HepG2 cells could facilitate RA induction but not in HEK293 cells.

Since the nucleosomes do not impede the binding of RAR/RXR heterodimers to their DNA recognition sequences [3], the different function of the distal region of the promoter may be due to different cell environments.

RAR α is the major RAR subtype involved in the DNA-RAR/RXR complex in

HepG2 cells . Am580, a RAR α agonist, induced CYP26A1 expression, and partially resembled the RA effect in high dose treatment, suggesting that not only RAR α but

RARβ and RAR γ could be involved in the RA-mediated induction. The cotransfection of

RAR α, β, γ significantly increased promoter activity (Figure 11), suggesting that each subtype could mediated RA regulated CYP26A1 expression. In the case of HepG2 cells, in EMSA results, anti-RAR α antibody showed stronger competition than the other antibodies. Moreover, a specific RAR α antagonist inhibited RA-induced expression of

CYP26A1 [104]. Therefore, RAR α plays a major role in CYP26A1 expression in HepG2 cells. The study conducted in F9 cells showed that RAR γ/RXR α heterodimers were responsible for CYP26A1 induction [75], whereas studies in HL-60 and NB4 cells

showed that a RAR α specific antagonist totally abolished the RA-induced expression of

CYP26A1. In vivo studies showed that RAR subtypes (RAR α, β, γ) were expressed in different tissues in adults, and that in embryonic stage RARs were also expressed in a spatial and temporal manner [111]. And RAR subtypes have a partial functional redundancy, but they have been shown involved in the different regulation pathways. Our study showed that the all of the RAR subtypes could mediate RA-induced expression, suggesting that RA can control its own concentration by inducing CYP26A1 through

RARs during the life span. For example, RAR γ is expressed in the tailbud and precartilaginous condensations, where CYP26A1 was upregulated following administration of RA [112].

Overall, in summary, our study has demonstrated that additional RA-responsive elements located upstream of the CYP26A1 coding sequence are both active and necessary for the strong inducibility of CYP26A1 expression by RA in HepG2 cells.

Three RAREs and the half site are embedded in the putative promoter of CYP26A1 and are highly conserved among mouse, rat and human. CYP26A1 promoter activity was significantly higher in HepG2 cells as compared to HEK293T cells, and different RAREs were important for the response to ATRA in these two cell lines. All three of the RAREs and the half site synergistically regulated ATRA-mediated CYP26A1 expression, and were necessary for the strong ATRA inducibility of CYP26A1. Cotransfection of each of the individual retinoic acid receptors (RAR α, β, γ) significantly increased promoter activities, whereas RAR α is the major RAR subtype involved in the DNA-RAR/RXR complex in HepG2 cells.

Future Directions

Our study and other studies showed transcriptional synergism between distal

RAREs and proximal RARE of CYP26A1 promoter, but precise mechanisms are still not clear. It would be interesting to investigate how distal RAREs interact with proximal

RARE, mediator and transcriptional machinery in different cellular environments. This could help understanding chromosome behavior.

Moreover, CYP26A1 exhibits tissue- and cell- specific expression and regulation.

The mechanism of regulation resulting in tissue- and cell- specific expression is not clear.

For APL treatment, synthetic retinoids have been used in clinical trials. Am80 is an example since Am80 (a positional isomer of Am580 which was used in our studies) could induce in vitro differentiation of NB4 cells and is resistant to oxidation [52]. However,

Am80 did not significantly improve the remission compared with RA, after time off RA.

It would be worthwhile to elucidate the factors involved in the repression or induction of

CYP26A1. Based on the mechanism, new therapies could be developed.

Third, RA has been shown to affect the expression of over 500 genes [9]. RAR β has been well characterized as inducible by RA [113], and is often used in experiments as a positive control for RA inducibility. CYP26A1 promoter studies showed that CYP26A1 was more sensitive than RAR β to RA, suggesting that CYP26A1 could be used as a positive control for RA inducibility in the future studies.

In addition, a stable transfected CYP26A1 promoter linked construct could be used for detecting low levels of RA. The transient transfected CYP26A1 promoter in HepG2 cells could be responsive to 1 nM RA (Figure 10). The current common method used to

measure RA in tissues is high performance liquid chromatography (HPLC), but the detection level for RA absorbance is not sensitive enough for small samples. Therefore, a stable transfected CYP26A1 promoter linked construct might be a good option for measuring RA concentration in a reporter cell line.

REFERENCES

1. Samokyszyn, V.M., Gall, W.E., Zawada, G., Freyaldenhoven, M.A., Chen, G., Mackenzie, P.I., Tephly, T.R., and Radominska-Pandya, A. 4-hydroxyretinoic acid, a novel substrate for human liver microsomal UDP- glucuronosyltransferase(s) and recombinant UGT2B7. J Biol Chem. 2000; 275: 6908-14.

2. Napoli, J.L. Interactions of retinoid binding proteins and enzymes in retinoid metabolism. Biochim Biophys Acta. 1999; 1440: 139-62.

3. Bastien, J. and Rochette-Egly, C. Nuclear retinoid receptors and the transcription of retinoid-target genes. Gene. 2004; 328: 1-16.

4. Ross, A.C. and Zolfaghari, R. Regulation of hepatic retinol metabolism: perspectives from studies on vitamin A status. J Nutr. 2004; 134: 269S-275S.

5. Zolfaghari, R. and Ross, A.C. Lecithin:retinol acyltransferase from mouse and rat liver. CDNA cloning and liver-specific regulation by dietary vitamin A and retinoic acid. J Lipid Res. 2000; 41: 2024-34.

6. Wang, Y., Zolfaghari, R., and Ross, A.C. Cloning of rat cytochrome P450RAI (CYP26) cDNA and regulation of its gene expression by all-trans-retinoic acid in vivo. Arch Biochem Biophys. 2002; 401: 235-43.

7. Wei, L.N. Retinoid receptors and their coregulators. Annu Rev Pharmacol Toxicol. 2003; 43: 47-72.

8. Dilworth, F.J. and Chambon, P. Nuclear receptors coordinate the activities of chromatin remodeling complexes and coactivators to facilitate initiation of transcription. Oncogene. 2001; 20: 3047-54.

9. Balmer, J.E. and Blomhoff, R. Gene expression regulation by retinoic acid. J Lipid Res. 2002; 43: 1773-808.

10. Penzes, P., Wang, X., Sperkova, Z., and Napoli, J.L. Cloning of a rat cDNA encoding retinal dehydrogenase isozyme type I and its expression in E. coli. Gene. 1997; 191: 167-72.

11. Zhao, D., McCaffery, P., Ivins, K.J., Neve, R.L., Hogan, P., Chin, W.W., and Drager, U.C. Molecular identification of a major retinoic-acid-synthesizing enzyme, a retinaldehyde-specific dehydrogenase. Eur J Biochem. 1996; 240: 15- 22.

12. Lin, M., Zhang, M., Abraham, M., Smith, S.M., and Napoli, J.L. Mouse retinal dehydrogenase 4 (RALDH4), molecular cloning, cellular expression, and activity in 9-cis-retinoic acid biosynthesis in intact cells. J Biol Chem. 2003; 278: 9856- 61.

13. Mic, F.A., Molotkov, A., Fan, X., Cuenca, A.E., and Duester, G. RALDH3, a retinaldehyde dehydrogenase that generates retinoic acid, is expressed in the ventral retina, otic vesicle and olfactory pit during mouse development. Mech Dev. 2000; 97: 227-30.

14. Roberts, A.B., Lamb, L.C., and Sporn, M.B. Metabolism of all-trans-retinoic acid in hamster liver microsomes: oxidation of 4-hydroxy- to 4-keto-retinoic acid. Arch Biochem Biophys. 1980; 199: 374-83.

15. Roberts, A.B., Nichols, M.D., Newton, D.L., and Sporn, M.B. In vitro metabolism of retinoic acid in hamster intestine and liver. J Biol Chem. 1979; 254: 6296-302.

16. Duell, E.A., Astrom, A., Griffiths, C.E., Chambon, P., and Voorhees, J.J. Human skin levels of retinoic acid and cytochrome P-450-derived 4-hydroxyretinoic acid after topical application of retinoic acid in vivo compared to concentrations required to stimulate retinoic acid receptor-mediated transcription in vitro. J Clin Invest. 1992; 90: 1269-74.

17. White, J.A., Beckett-Jones, B., Guo, Y.D., Dilworth, F.J., Bonasoro, J., Jones, G., and Petkovich, M. cDNA cloning of human retinoic acid-metabolizing enzyme (hP450RAI) identifies a novel family of cytochromes P450. J Biol Chem. 1997; 272: 18538-41.

18. Yamamoto, Y., Zolfaghari, R., and Ross, A.C. Regulation of CYP26 (cytochrome P450RAI) mRNA expression and retinoic acid metabolism by retinoids and dietary vitamin A in liver of mice and rats. Faseb J. 2000; 14: 2119-27.

19. Cifelli, C.J. and Ross, A.C. All-trans-retinoic acid distribution and metabolism in vitamin A-marginal rats. Am J Physiol Gastrointest Liver Physiol. 2006; 291: G195-202.

20. Loudig, O., Babichuk, C., White, J., Abu-Abed, S., Mueller, C., and Petkovich, M. Cytochrome P450RAI(CYP26) promoter: a distinct composite retinoic acid response element underlies the complex regulation of retinoic acid metabolism. Mol Endocrinol. 2000; 14: 1483-97.

21. Wolf, G. A history of vitamin A and retinoids. Faseb J. 1996; 10: 1102-7.

22. Howe, S.B.W.a.P.R. Tissue changes following deprivation of fat-soluble A vitamin. The Journal of Experimental Medicine. 1925; 42: 753-77.

23. Karrer, P., Morf, R., and Schoepp, K. Zur Kenntnis des Vitamins-A aus Fischtranen. Helvetica chimica acta. 1931; 15: 1036.

24. Isler, O., Huber, W., Ronco, A., and Kofler, M. Synthese des Vitamin A. Helvetica chimica acta. 1947; 30: 1911-1927.

25. Wald, G. Molecular basis of visual excitation. Science. 1968; 162: 230-9.

26. Ross, A.C., Zolfaghari, R., and Weisz, J. Vitamin A: recent advances in the biotransformation, transport, and metabolism of retinoids. Curr Opin Gastroenterol. 2001; 17: 184-192.

27. Ross, A.C. Overview of retinoid metabolism. J Nutr. 1993; 123: 346-50.

28. Gropper, S.S., SMith, J.L., and Groff, J.L., Vitamin A and Carotenoids, in Advanced Nutrition and Human Metabolism. p. 325-342.

29. Randolph, R.K. and Ross, A.C. Vitamin A status regulates hepatic lecithin: retinol acyltransferase activity in rats. J Biol Chem. 1991; 266: 16453-7.

30. Matsuura, T. and Ross, A.C. Regulation of hepatic lecithin: retinol acyltransferase activity by retinoic acid. Arch Biochem Biophys. 1993; 301: 221-7.

31. Shimada, T., Ross, A.C., Muccio, D.D., Brouillette, W.J., and Shealy, Y.F. Regulation of hepatic lecithin:retinol acyltransferase activity by retinoic acid receptor-selective retinoids. Arch Biochem Biophys. 1997; 344: 220-7.

32. Yost, R.W., Harrison, E.H., and Ross, A.C. Esterification by rat liver microsomes of retinol bound to cellular retinol-binding protein. J Biol Chem. 1988; 263: 18693-701.

33. Ong, D.E. and Page, D.L. Cellular retinol-binding protein (type two) is abundant in human small intestine. J Lipid Res. 1987; 28: 739-45.

34. Noy, N. and Blaner, W.S. Interactions of retinol with binding proteins: studies with rat cellular retinol-binding protein and with rat retinol-binding protein. Biochemistry. 1991; 30: 6380-6.

35. Ong, D.E. and Chytil, F. Cellular retinol-binding protein from rat liver. Purification and characterization. J Biol Chem. 1978; 253: 828-32.

36. Levin, M.S., Li, E., Ong, D.E., and Gordon, J.I. Comparison of the tissue-specific expression and developmental regulation of two closely linked rodent genes encoding cytosolic retinol-binding proteins. J Biol Chem. 1987; 262: 7118-24.

37. Li, E. Structure and function of cytoplasmic retinoid binding proteins. Mol Cell Biochem. 1999; 192: 105-8.

38. Rajan, N., Blaner, W.S., Soprano, D.R., Suhara, A., and Goodman, D.S. Cellular retinol-binding protein messenger RNA levels in normal and retinoid-deficient rats. J Lipid Res. 1990; 31: 821-9.

39. Ghyselinck, N.B., Bavik, C., Sapin, V., Mark, M., Bonnier, D., Hindelang, C., Dierich, A., Nilsson, C.B., Hakansson, H., Sauvant, P., Azais-Braesco, V., Frasson, M., Picaud, S., and Chambon, P. Cellular retinol-binding protein I is essential for vitamin A homeostasis. Embo J. 1999; 18: 4903-14.

40. E, X., Zhang, L., Lu, J., Tso, P., Blaner, W.S., Levin, M.S., and Li, E. Increased neonatal mortality in mice lacking cellular retinol-binding protein II. J Biol Chem. 2002; 277: 36617-23.

41. Folli, C., Calderone, V., Ottonello, S., Bolchi, A., Zanotti, G., Stoppini, M., and Berni, R. Identification, retinoid binding, and x-ray analysis of a human retinol- binding protein. Proc Natl Acad Sci U S A. 2001; 98: 3710-5.

42. Vogel, S., Mendelsohn, C.L., Mertz, J.R., Piantedosi, R., Waldburger, C., Gottesman, M.E., and Blaner, W.S. Characterization of a new member of the fatty acid-binding protein family that binds all-trans-retinol. J Biol Chem. 2001; 276: 1353-60.

43. Folli, C., Calderone, V., Ramazzina, I., Zanotti, G., and Berni, R. Ligand binding and structural analysis of a human putative cellular retinol-binding protein. J Biol Chem. 2002; 277: 41970-7.

44. Blomhoff, R., Green, M.H., Berg, T., and Norum, K.R. Transport and storage of vitamin A. Science. 1990; 250: 399-404.

45. Fortuna, V.A., Trugo, L.C., and Borojevic, R. Acyl-CoA: retinol acyltransferase (ARAT) and lecithin:retinol acyltransferase (LRAT) activation during the lipocyte phenotype induction in hepatic stellate cells. J Nutr Biochem. 2001; 12: 610-621.

46. Nagy, N.E., Holven, K.B., Roos, N., Senoo, H., Kojima, N., Norum, K.R., and Blomhoff, R. Storage of vitamin A in extrahepatic stellate cells in normal rats. J Lipid Res. 1997; 38: 645-58.

47. Blaner, W.S. Retinol-binding protein: the serum transport protein for vitamin A. Endocr Rev. 1989; 10: 308-16.

48. Pilch, S.M. Analysis of vitamin A data from the health and nutrition examination surveys. J Nutr. 1987; 117: 636-40.

49. Bavik, C.O., Eriksson, U., Allen, R.A., and Peterson, P.A. Identification and partial characterization of a retinal pigment epithelial membrane receptor for plasma retinol-binding protein. J Biol Chem. 1991; 266: 14978-85.

50. Creek, K.E., St Hilaire, P., and Hodam, J.R. A comparison of the uptake, metabolism and biologic effects of retinol delivered to human keratinocytes either free or bound to serum retinol-binding protein. J Nutr. 1993; 123: 356-61.

51. Smith, M.A., Parkinson, D.R., Cheson, B.D., and Friedman, M.A. Retinoids in cancer therapy. J Clin Oncol. 1992; 10: 839-64.

52. Gallagher, R.E. Retinoic acid resistance in acute promyelocytic leukemia. Leukemia. 2002; 16: 1940-58.

53. Smith, J.E., Milch, P.O., Muto, Y., and Goodman, D.S. The plasma transport and metabolism of retinoic acid in the rat. Biochem J. 1973; 132: 821-7.

54. De Leenheer, A.P., Lambert, W.E., and Claeys, I. All-trans-retinoic acid: measurement of reference values in human serum by high performance liquid chromatography. J Lipid Res. 1982; 23: 1362-7.

55. Eckhoff, C. and Nau, H. Identification and quantitation of all-trans- and 13-cis- retinoic acid and 13-cis-4-oxoretinoic acid in human plasma. J Lipid Res. 1990; 31: 1445-54.

56. Kurlandsky, S.B., Gamble, M.V., Ramakrishnan, R., and Blaner, W.S. Plasma delivery of retinoic acid to tissues in the rat. J Biol Chem. 1995; 270: 17850-7.

57. Astrom, A., Pettersson, U., Chambon, P., and Voorhees, J.J. Retinoic acid induction of human cellular retinoic acid-binding protein-II gene transcription is mediated by retinoic acid receptor-retinoid X receptor heterodimers bound to one far upstream retinoic acid-responsive element with 5-base pair spacing. J Biol Chem. 1994; 269: 22334-9.

58. Delva, L., Bastie, J.N., Rochette-Egly, C., Kraiba, R., Balitrand, N., Despouy, G., Chambon, P., and Chomienne, C. Physical and functional interactions between cellular retinoic acid binding protein II and the retinoic acid-dependent nuclear complex. Mol Cell Biol. 1999; 19: 7158-67.

59. Budhu, A.S. and Noy, N. Direct channeling of retinoic acid between cellular retinoic acid-binding protein II and retinoic acid receptor sensitizes mammary carcinoma cells to retinoic acid-induced growth arrest. Mol Cell Biol. 2002; 22: 2632-41.

60. Petkovich, P.M. Retinoic acid metabolism. J Am Acad Dermatol. 2001; 45: S136- 42.

61. Deltour, L., Foglio, M.H., and Duester, G. Metabolic deficiencies in alcohol dehydrogenase Adh1, Adh3, and Adh4 null mutant mice. Overlapping roles of Adh1 and Adh4 in ethanol clearance and metabolism of retinol to retinoic acid. J Biol Chem. 1999; 274: 16796-801.

62. Deltour, L., Foglio, M.H., and Duester, G. Impaired retinol utilization in Adh4 alcohol dehydrogenase mutant mice. Dev Genet. 1999; 25: 1-10.

63. McCaffery, P. and Drager, U.C. Retinoic acid synthesizing enzymes in the embryonic and adult vertebrate. Adv Exp Med Biol. 1995; 372: 173-83.

64. Niederreither, K., McCaffery, P., Drager, U.C., Chambon, P., and Dolle, P. Restricted expression and retinoic acid-induced downregulation of the retinaldehyde dehydrogenase type 2 (RALDH-2) gene during mouse development. Mech Dev. 1997; 62: 67-78.

65. Hoffman, L.M., Garcha, K., Karamboulas, K., Cowan, M.F., Drysdale, L.M., Horton, W.A., and Underhill, T.M. BMP action in skeletogenesis involves attenuation of retinoid signaling. J Cell Biol. 2006; 174: 101-13.

66. Mendelsohn, C., Ruberte, E., LeMeur, M., Morriss-Kay, G., and Chambon, P. Developmental analysis of the retinoic acid-inducible RAR-beta 2 promoter in transgenic animals. Development. 1991; 113: 723-34.

67. Napoli, J.L. and McCormick, A.M. Tissue dependence of retinoic acid metabolism in vivo. Biochim Biophys Acta. 1981; 666: 165-75.

68. Becker, B., Barua, A.B., and Olson, J.A. All-trans-retinoyl beta-glucuronide: new procedure for chemical synthesis and its metabolism in vitamin A-deficient rats. Biochem J. 1996; 314 ( Pt 1): 249-52.

69. Ross, A.C. and Harrison, E.H., Vitamin A: Nutritional Aspects of Retinoids and Carotenoids, in Handbook of Vitamins, J. Zempleni, et al., Editors. 2007. p. 1-41.

70. Frolik, C.A., Roller, P.P., Roberts, A.B., and Sporn, M.B. In vitro and in vivo metabolism of all-trans- and 13-cis-retinoic acid in hamsters. Identification of 13- cis-4-oxoretinoic acid. J Biol Chem. 1980; 255: 8057-62.

71. Fiorella, P.D. and Napoli, J.L. Microsomal retinoic acid metabolism. Effects of cellular retinoic acid-binding protein (type I) and C18-hydroxylation as an initial step. J Biol Chem. 1994; 269: 10538-44.

72. Marikar, Y., Wang, Z., Duell, E.A., Petkovich, M., Voorhees, J.J., and Fisher, G.J. Retinoic acid receptors regulate expression of retinoic acid 4-hydroxylase that specifically inactivates all-trans retinoic acid in human keratinocyte HaCaT cells. J Invest Dermatol. 1998; 111: 434-9.

73. White, J.A., Boffa, M.B., Jones, B., and Petkovich, M. A zebrafish retinoic acid receptor expressed in the regenerating caudal fin. Development. 1994; 120: 1861- 72.

74. White, J.A., Beckett, B., Scherer, S.W., Herbrick, J.A., and Petkovich, M. P450RAI (CYP26A1) maps to human chromosome 10q23-q24 and mouse chromosome 19C2-3. Genomics. 1998; 48: 270-2.

75. Abu-Abed, S.S., Beckett, B.R., Chiba, H., Chithalen, J.V., Jones, G., Metzger, D., Chambon, P., and Petkovich, M. Mouse P450RAI (CYP26) expression and retinoic acid-inducible retinoic acid metabolism in F9 cells are regulated by retinoic acid receptor gamma and retinoid X receptor alpha. J Biol Chem. 1998; 273: 2409-15.

76. Swindell, E.C., Thaller, C., Sockanathan, S., Petkovich, M., Jessell, T.M., and Eichele, G. Complementary domains of retinoic acid production and degradation in the early chick embryo. Dev Biol. 1999; 216: 282-96.

77. Hollemann, T., Chen, Y., Grunz, H., and Pieler, T. Regionalized metabolic activity establishes boundaries of retinoic acid signalling. Embo J. 1998; 17: 7361-72.

78. White, J.A., Ramshaw, H., Taimi, M., Stangle, W., Zhang, A., Everingham, S., Creighton, S., Tam, S.P., Jones, G., and Petkovich, M. Identification of the human cytochrome P450, P450RAI-2, which is predominantly expressed in the adult cerebellum and is responsible for all-trans-retinoic acid metabolism. Proc Natl Acad Sci U S A. 2000; 97: 6403-8.

79. Nelson, D.R. A second CYP26 P450 in humans and zebrafish: CYP26B1. Arch Biochem Biophys. 1999; 371: 345-7.

80. Taimi, M., Helvig, C., Wisniewski, J., Ramshaw, H., White, J., Amad, M., Korczak, B., and Petkovich, M. A novel human cytochrome P450, CYP26C1, involved in metabolism of 9-cis and all-trans isomers of retinoic acid. J Biol Chem. 2004; 279: 77-85.

81. Gu, X., Xu, F., Wang, X., Gao, X., and Zhao, Q. Molecular cloning and expression of a novel CYP26 gene (cyp26d1) during zebrafish early development. Gene Expr Patterns. 2005; 5: 733-9.

82. Ray, W.J., Bain, G., Yao, M., and Gottlieb, D.I. CYP26, a novel mammalian cytochrome P450, is induced by retinoic acid and defines a new family. J Biol Chem. 1997; 272: 18702-8.

83. Abu-Abed, S., Dolle, P., Metzger, D., Beckett, B., Chambon, P., and Petkovich, M. The retinoic acid-metabolizing enzyme, CYP26A1, is essential for normal hindbrain patterning, vertebral identity, and development of posterior structures. Genes Dev. 2001; 15: 226-40.

84. Gu, X., Xu, F., Song, W., Wang, X., Hu, P., Yang, Y., Gao, X., and Zhao, Q. A novel cytochrome P450, zebrafish Cyp26D1, is involved in metabolism of all- trans retinoic acid. Mol Endocrinol. 2006; 20: 1661-72.

85. Lampen, A., Meyer, S., and Nau, H. Effects of receptor-selective retinoids on CYP26 gene expression and metabolism of all-trans-retinoic acid in intestinal cells. Drug Metab Dispos. 2001; 29: 742-7.

86. McSorley, L.C. and Daly, A.K. Identification of human cytochrome P450 isoforms that contribute to all-trans-retinoic acid 4-hydroxylation. Biochem Pharmacol. 2000; 60: 517-26.

87. Ross, A.C. Retinoid production and catabolism: role of diet in regulating retinol esterification and retinoic Acid oxidation. J Nutr. 2003; 133: 291S-296S.

88. Zolfaghari, R., Cifelli, C.J., Lieu, S.O., Chen, Q., Li, N.Q., and Ross, A.C. Lipopolysaccharide opposes the induction of CYP26A1 and CYP26B1 gene expression by retinoic acid in the rat liver in vivo. Am J Physiol Gastrointest Liver Physiol. 2007; 292: G1029-36.

89. Schapira, M., Raaka, B.M., Samuels, H.H., and Abagyan, R. In silico discovery of novel retinoic acid receptor agonist structures. BMC Struct Biol. 2001; 1: 1.

90. de Lera, A.R., Bourguet, W., Altucci, L., and Gronemeyer, H. Design of selective nuclear receptor modulators: RAR and RXR as a case study. Nat Rev Drug Discov. 2007; 6: 811-20.

91. Dolle, P., Fraulob, V., Kastner, P., and Chambon, P. Developmental expression of murine retinoid X receptor (RXR) genes. Mech Dev. 1994; 45: 91-104.

92. Mollard, R., Viville, S., Ward, S.J., Decimo, D., Chambon, P., and Dolle, P. Tissue-specific expression of retinoic acid receptor isoform transcripts in the mouse embryo. Mech Dev. 2000; 94: 223-32.

93. Kagechika, H. [Novel synthetic retinoid agonists and antagonists]. Yakugaku Zasshi. 1994; 114: 847-62.

94. Chen, Q., Ma, Y., and Ross, A.C. Opposing cytokine-specific effects of all trans- retinoic acid on the activation and expression of signal transducer and activator of transcription (STAT)-1 in THP-1 cells. Immunology. 2002; 107: 199-208.

95. Wang, X.D., Krinsky, N.I., and Russell, R.M. Retinoic acid regulates retinol metabolism via feedback inhibition of retinol oxidation and stimulation of retinol esterification in ferret liver. J Nutr. 1993; 123: 1277-85.

96. Kane, M.A., Chen, N., Sparks, S., and Napoli, J.L. Quantification of endogenous retinoic acid in limited biological samples by LC/MS/MS. Biochem J. 2005; 388: 363-9.

97. Cifelli, C.J. and Ross, A.C. Chronic vitamin A status and acute repletion with retinyl palmitate are determinants of the distribution and catabolism of all-trans- retinoic acid in rats. J Nutr. 2007; 137: 63-70.

98. Underwood, B.A., Loerch, J.D., and Lewis, K.C. Effects of dietary vitamin A deficiency, retinoic acid and protein quantity and quality on serially obtained plasma and liver levels of vitamin A in rats. J Nutr. 1979; 109: 796-806.

99. Warrell, R.P., Jr., de The, H., Wang, Z.Y., and Degos, L. Acute promyelocytic leukemia. N Engl J Med. 1993; 329: 177-89.

100. Patatanian, E. and Thompson, D.F. Retinoic acid syndrome: a review. J Clin Pharm Ther. 2008; 33: 331-8.

101. Muindi, J.R., Young, C.W., and Warrell, R.P., Jr. Clinical pharmacology of all- trans retinoic acid. Leukemia. 1994; 8: 1807-12.

102. Adamson, P.C., Balis, F.M., Smith, M.A., Murphy, R.F., Godwin, K.A., and Poplack, D.G. Dose-dependent pharmacokinetics of all-trans-retinoic acid. J Natl Cancer Inst. 1992; 84: 1332-5.

103. Adamson, P.C., Boylan, J.F., Balis, F.M., Murphy, R.F., Godwin, K.A., Gudas, L.J., and Poplack, D.G. Time course of induction of metabolism of all-trans- retinoic acid and the up-regulation of cellular retinoic acid-binding protein. Cancer Res. 1993; 53: 472-6.

104. Ozpolat, B., Mehta, K., and Lopez-Berestein, G. Regulation of a highly specific retinoic acid-4-hydroxylase (CYP26A1) enzyme and all-trans-retinoic acid metabolism in human intestinal, liver, endothelial, and acute promyelocytic leukemia cells. Leuk Lymphoma. 2005; 46: 1497-506.

105. Ozpolat, B., Mehta, K., Tari, A.M., and Lopez-Berestein, G. all-trans-Retinoic acid-induced expression and regulation of retinoic acid 4-hydroxylase (CYP26) in human promyelocytic leukemia. Am J Hematol. 2002; 70: 39-47.

106. Loudig, O., Maclean, G.A., Dore, N.L., Luu, L., and Petkovich, M. Transcriptional co-operativity between distant retinoic acid response elements in regulation of Cyp26A1 inducibility. Biochem J. 2005; 392: 241-8.

107. Kerry, D.M., Dwivedi, P.P., Hahn, C.N., Morris, H.A., Omdahl, J.L., and May, B.K. Transcriptional synergism between vitamin D-responsive elements in the rat 25-hydroxyvitamin D3 24-hydroxylase (CYP24) promoter. J Biol Chem. 1996; 271: 29715-21.

108. Kahlen, J.P. and Carlberg, C. Identification of a vitamin D receptor homodimer- type response element in the rat calcitriol 24-hydroxylase gene promoter. Biochem Biophys Res Commun. 1994; 202: 1366-72.

109. Yasmin, R., Yeung, K.T., Chung, R.H., Gaczynska, M.E., Osmulski, P.A., and Noy, N. DNA-looping by RXR tetramers permits transcriptional regulation "at a distance". J Mol Biol. 2004; 343: 327-38.

110. Gaillard, E., Bruck, N., Brelivet, Y., Bour, G., Lalevee, S., Bauer, A., Poch, O., Moras, D., and Rochette-Egly, C. Phosphorylation by PKA potentiates retinoic acid receptor alpha activity by means of increasing interaction with and

phosphorylation by cyclin H/cdk7. Proc Natl Acad Sci U S A. 2006; 103: 9548- 53.

111. Mangelsdorf, D.J. Vitamin A receptors. Nutr Rev. 1994; 52: S32-44.

112. Iulianella, A., Beckett, B., Petkovich, M., and Lohnes, D. A molecular basis for retinoic acid-induced axial truncation. Dev Biol. 1999; 205: 33-48.

113. de The, H., Vivanco-Ruiz, M.M., Tiollais, P., Stunnenberg, H., and Dejean, A. Identification of a retinoic acid responsive element in the retinoic acid receptor beta gene. Nature. 1990; 343: 177-80.