Role of the Vitamin D Receptor in 1,25-Dihydroxyvitamin D3

Home , Sp1 transcription factor

ROLE OF THE VITAMIN D RECEPTOR IN 1,25-DIHYDROXYVITAMIN D3-

MEDIATED GROWTH ARREST AND APOPTOSIS

A Dissertation

Submitted to the Graduate School

of the University of Notre Dame

In Partial Fulfillment of the Requirements

For the Degree of

Doctor of Philosophy

Meggan E. Valrance, B.S.

JoEllen Welsh, Director

Graduate Program in Biological Sciences

Notre Dame, Indiana

July 2007

Meggan E. Valrance

2007

ROLE OF THE VITAMIN D RECEPTOR IN 1,25-DIHYDROXYVITAMIN D3-

MEDIATED GROWTH ARREST AND APOPTOSIS

Abstract by

Meggan E. Valrance

1,25-dihydroxyvitamin D3 (1,25D) and its cognate nuclear receptor, the

vitamin D receptor (VDR), regulate calcium homeostasis via transcriptional activation of target genes. 1,25D is also known to mediate growth arrest and apoptosis in a variety of cancer cell lines and tumors, however, the role of the

VDR and the role of transcriptional activation in this process remain unclear. To investigate the role of the VDR in mediating the anti-cancer properties of 1,25D,

two murine mammary tumor cell lines with differential VDR expression were characterized.

WT145 cells, which express transcriptionally functional VDR protein, were growth inhibited and rendered apoptotic by 1,25D and synthetic vitamin D analogs. In contrast, KO240 cells, which express no detectable VDR protein or mRNA, were neither growth inhibited nor rendered apoptotic by 1,25D, at doses as high as 1μM. Collectively, these data indicate that functional VDR protein is Meggan E. Valrance required for the anti-cancer effects of 1,25D and structurally related vitamin D based therapeutics in vitro.

To determine whether the anti-cancer actions of vitamin D are mediated via the VDR in vivo, nude mice bearing tumors derived from WT145 or KO240 cells were treated with EB1089, a synthetic analog of 1,25D, or placebo for six

weeks. An untreated subset of tumor-bearing mice was exposed to ultraviolet

(UV) light, to activate endogenous 1,25D production. Both EB1089 and UV light exposure decreased volume of WT145 tumors through decreased tumor cell proliferation and increased tumor cell apoptosis. No effects of either EB1089 or

UV treatments were observed in KO240 tumors, indicating that the vitamin D pathway mediates its anti-tumor effects in vivo via tumor-cell VDR.

KO240 cells stably expressing wild-type VDR and VDR point mutants from hereditary rickets patients were created and characterized, to further examine the mechanism of VDR in growth regulation. KO cells stably expressing VDR were growth inhibited by 1,25D and its structural analogs, indicating that the vitamin D growth regulatory pathway could be reconstituted in VDR null cells. Cells expressing mutant VDRs were differentially affected by 1,25D and analogs; VDR that lack DNA binding ability were growth inhibited by physiological doses of

1,25D while ligand binding domain mutants were not. This suggests that the

anti-cancer effects of 1,25D, while VDR mediated, are in part mediated via novel,

DNA-independent mechanisms.

To Jamie, my source of sanity in an insane world. I could never have done this without your love, your help, your sense of humor, or your understanding. Thank

you for being you, for loving and understanding me, and for putting up with my

many mental breakdowns while writing.

And, to my grandmother Maxine, my mother Kathy, and my aunt Charlie - this is

for you. I couldn’t save you, but I hope beyond hope that my work can save others, so that no other woman has to be as strong and as brave as the three of

you were. You are in my thoughts always.

CONTENTS

LIST OF TABLES ...... vi

LIST OF FIGURES ...... vii

LIST OF ABBREVIATIONS ...... xi

ACKNOWLEDGMENTS ...... xiii

Chapter 1: LITERATURE REVIEW ...... 1

1.1 Introduction ...... 1

1.2 1,25D Biosynthesis ...... 2

1.3 Classical 1,25D/VDR Signaling...... 4

1.4 Non-Classical VDR Signaling Pathways...... 11

1.5 Vitamin D3 and Cancer ...... 14

1.6 Dissertation Objectives ...... 18

Chapter 2: CHARACTERIZATION OF WT145 AND KO240 MURINE MAMMARY TUMOR CELLS IN STEROID-FREE MEDIA ...... 20

2.1 Introduction ...... 20

2.2 Materials and Methods...... 23

2.3 Results...... 27

2.4 Discussion ...... 36

iii

Chapter 3: INDUCTION OF APOPTOSIS BY 1,25D IN MURINE MAMMARY TUMOR CELLS REQUIRES VDR PROTEIN EXPRESSION ...... 39

3.1 Introduction ...... 39

3.2 Materials and Methods...... 43

3.3 Results...... 46

3.4 Discussion ...... 54

Chapter 4: REQUIREMENT OF VDR FOR THE ANTI-CANCER ACTION OF SYNTHETIC STRUCTURAL ANALOGS OF 1,25D AND OTHER VITAMIN D METABOLITES...... 57

4.1 Introduction ...... 57

4.2 Materials and Methods...... 60

4.3 Results...... 63

4.4 Discussion ...... 71

Chapter 5: VDR-DEPENDENT INHIBITION OF MAMMARY TUMOR GROWTH IN VIVO BY EB1089 AND ULTRAVIOLET RADIATION ...... 75

5.1 Introduction ...... 75

5.2 Materials and Methods...... 78

5.3 Results...... 82

5.4 Discussion ...... 94

Chapter 6: CREATION OF VECTORS SUITABLE FOR STABLE EXPRESSION OF WILD-TYPE AND MUTANT VDRS IN MURINE MAMMARY CELLS...... 99

6.1 Introduction ...... 99

6.2 Materials and Methods...... 103

6.3 Results...... 110

6.4 Discussion ...... 120

Chapter 7: CREATION AND CHARACTERIZATION OF KO240 CELLS STABLY EXPRESSING WILD-TYPE HUMAN VDR...... 124

7.1 Introduction ...... 124

7.2 Materials and Methods...... 127

7.3 Results...... 131

7.4 Discussion ...... 143

Chapter 8: CREATION AND CHARACTERIZATION OF KO240 CELLS STABLY EXPRESSING HVDRR-2 POINT-MUTANT VDRS ...... 148

8.1 Introduction ...... 148

8.2 Materials and Methods...... 151

8.3 Results...... 155

8.4 Discussion ...... 163

GENERAL DISCUSSION ...... 169

REFERENCES: ...... 177

LIST OF TABLES

Table 6.1. Nucleotide positions and base changes for VDR mutant constructs...... 107

Table 6.2. A comparison of the wild-type VDR sequence with the point mutations selectively created...... 115

Table 7.1. 1,25D analogs induce growth arrest in KOhVDR cells...... 142

Table 8.1. Transactivation activity of the VDR in KOmutant stable cell lines...... 157

Table 8.2. 1,25D analogs have differential growth effects in KOmutant cells. .... 161

LIST OF FIGURES

Figure 2.1. VDR protein is expressed only in WT145 cells...... 28

Figure 2.2. VDR concentrates in the nucleus following treatment with 1,25D. ... 29

Figure 2.3. WT145 cells express transcriptionally active VDR...... 31

Figure 2.4. Endogenous CYP24 mRNA and protein levels are upregulated in WT145, but not KO240, cells following 1,25D treatment...... 32

Figure 2.5. WT145 cells, but not KO240 cells, are growth inhibited by 1,25D. . 33

Figure 2.6. WT145 cells are sensitive to concentrations of 1,25D in the femtomolar range...... 34

Figure 2.7. The doubling time of WT145 cells is increased by 1,25D...... 35

Figure 3.1. 1,25D induces morphological features of apoptosis in WT145 cells, but not KO240 cells...... 47

Figure 3.2. 1,25D induces caspase 3 activity in WT145 cells, but not KO240 cells...... 49

Figure 3.3. Evidence of protease activation in WT145 cells following 1,25D treatment...... 50

Figure 3.4. 1,25D induces caspase 9 activity in WT145 cells, but not KO240 cells...... 51

Figure 3.5. WT145 and KO240 cells are sensitive to etoposide-mediated apoptosis...... 53

vii

Figure 4.1. WT145 and KO240 cells express CYP27B1...... 64

Figure 4.2. WT145 cells are growth inhibited, but not rendered apoptotic, by 25D...... 65

Figure 4.3. 25D does not induce CYP24 in WT145 cells...... 66

Figure 4.4. EB1089 induces growth arrest and VDR transcriptional activity in WT145 cells...... 68

Figure 4.5. 1,25D analogs induce growth arrest in WT145 cells...... 69

Figure 4.6. 1,25D analogs induce VDR transcriptional activity in WT145 cells... 70

Figure 5.1. EB1089 inhibits growth of xenografted VDR positive tumors...... 83

Figure 5.2. Serum calcium is not altered by EB1089 treatment...... 84

Figure 5.3. Tumor vessel density is not altered by VDR or EB1089 treatment... 84

Figure 5.4. EB1089 inhibits tumor cell proliferation through VDR dependent mechanisms...... 86

Figure 5.5. EB1089 induces apoptosis of tumor cells that express VDR...... 87

Figure 5.6. Chronic UV exposure inhibits proliferation and induces apoptosis in mammary xenografts through VDR dependent mechanisms...... 89

Figure 5.7. Chronic UV exposure increases serum 25D without affecting serum 1,25D or serum calcium...... 91

Figure 5.8. Expression of vitamin D metabolizing enzymes in cells and tumors...... 92

Figure 5.9. UV treatment induces epidermal hyperplasia...... 93

viii

Figure 6.1. PCR amplification of the hygromycin resistance cassette from the pcDNA3.1/hygro(+) vector...... 110

Figure 6.2. Post-digestion of both the destination vector and hygromycin insert...... 111

Figure 6.3. Direct colony PCR for the hygromycin resistance gene insert...... 112

Figure 6.4. EcoRI digest to determine the orientation of the insert within the destination vector...... 113

Figure 6.5. Post-PCR gel for the stable negative control vector...... 114

Figure 6.6. Expression of wild-type and mutant hVDR constructs in KO240 cells...... 116

Figure 6.7. Transactivation activity of wild-type and mutant VDRs in KO240 cells...... 118

Figure 6.8. Transactivation of CYP24 luciferase in cells expressing wild-type VDR and R274L mutant VDR...... 119

Figure 7.1. VDR protein is expressed in KOhVDR stable cell lines...... 132

Figure 7.2. Transactivation activity of the VDR in KOhVDR stable cell lines...... 133

Figure 7.3. Endogenous CYP24 protein is upregulated in KOhVDR stable cell lines following 1,25D treatment...... 134

Figure 7.4. KOhVDR cells, but not KOEV cells, are growth inhibited by 1,25D..... 135

Figure 7.5. KOhVDR cells are sensitive to concentrations of 1,25D in the picomolar range...... 136

Figure 7.6. The doubling time of KOhVDR stable cell lines is increased by 1,25D...... 137

Figure 7.7. 1,25D induces morphological features of apoptosis in KOhVDR cells, but not KOEV cells...... 138

Figure 7.8. KOhVDR cells are growth inhibited by 25D...... 139

Figure 7.9. EB1089 induces growth arrest in KOhVDR cells...... 141

Figure 8.1. VDR protein is present in KOmutant stable cell lines...... 156

Figure 8.2. KOG46DC and KOR274L2 cells are growth inhibited by 1,25D...... 158

Figure 8.3. KOG46DC cells are sensitive to concentrations of 1,25D in the upper picomolar range...... 159

Figure 8.4. KOG46DC and KOR274L2 cells are growth inhibited by EB1089...... 160

Figure 8.5. VDR LBD mutants are differentially transcriptionally activated by 1,25D analogs...... 162

LIST OF ABBREVIATIONS

1,25D 1,25-dihydroxyvitamin D3

25D 25-dihydroxyvitamin D3

ANOVA analysis of variance

CAR constitutive androstane receptor

CB CB1093

CSS charcoal-stripped fetal bovine serum

CYP24 25-hydroxyvitamin D 24-hydroxylase

CYP27A1 vitamin D 25-hydroxylase

CYP27B1 25-hydroxyvitamin D 1α-hydroxylase

DBD DNA-binding domain

DBP vitamin D serum binding protein

DMEM Dulbecco’s modified Eagle’s medium

EB EB1089

ER estrogen receptor

fM femtomolar

HVDRR hereditary vitamin D-resistant rickets

KH KH1230

KO knock-out

LBD ligand binding domain

MC MC1288

μM micromolar nM nanomolar

PARP poly (ADP-ribose) polymerase

PCR polymerase chain reaction pM picomolar

PMSF phenylmethanesulphonylfluoride polyA polyadenylation

PXR pregnane X receptor

RAR retinoic acid receptor

RLU relative luciferase units

RXR retinoid X receptor

TR thyroid hormone receptor

UV ultraviolet

VDR vitamin D receptor

VDRE vitamin D response element

WT wild-type

xii

ACKNOWLEDGMENTS

First and foremost, I would like to thank my advisor, Dr. JoEllen Welsh, for all of her mentorship, guidance, and friendship - our morning coffee chats have come to be a very important part of my day. I would also like to thank Dr. Martin

Tenniswood for his guidance and mentorship, and for allowing me to flex my teaching muscles in his class. For all of her help with the nude mouse studies, I gratefully thank Andrea Brunet, who is a wonderful scientist already, and is going to do great things. Thanks to my entire family: Dad, Jacquie, Grandpa Milton and

Grandma Madeline - you were the first people who believed that I could do this.

Finally, special and emphatic thanks to Jamie, not only for his never-ending emotional support and love, but for his vast knowledge of molecular biology techniques, tips, and tricks, which were invaluable all throughout the course of my work.

xiii

CHAPTER 1:

LITERATURE REVIEW

1.1 Introduction

A great deal of evidence suggests that estrogen drives mammary cell proliferation, and thus aides in the development and progression of mammary carcinoma. This makes blocking estrogen signaling an attractive target for breast cancer treatment and prevention; indeed, several selective estrogen receptor modulators have been developed, and are now commonly used in the treatment of early-stage breast cancer. As breast cancer progresses to later stages, however, estrogen sensitivity is lost, and therefore other therapies, which are not dependent on estrogen receptor signaling, are required.

Another nuclear receptor, the vitamin D receptor (VDR) was discovered in mammary gland in 1980 (Colston et al., 1980), and has become an attractive target for breast carcinoma treatment and prevention. The mechanisms of vitamin D-mediated growth inhibition, differentiation, and apoptosis have been extensively studied, and synthetic analogs of the biologically active form of vitamin D, 1,25-dihydroxyvitamin D3 (1,25D) with good therapeutic profiles have

been developed. Further development of better targeted analogs, however, will

require a better understanding of both the biological activities of 1,25D, and the

role of its cognate nuclear receptor, the VDR, in mediating these activities.

1.2 1,25D Biosynthesis

Vitamin D is a steroid hormone with a diverse array of biological functions.

It was discovered in 1922 as the component of heated cod liver oil which retained anti-rachitic properties (McCollum et al., 1922). The major, and most well- defined, biological effect of vitamin D is on extracellular phosphate and calcium homeostasis, which is necessary for maintenance of bone health. A variety of other biological functions of vitamin D are now recognized, and include maintenance of cell differentiation, induction of apoptosis and/or growth inhibition, and immune regulation.

While vitamin D can be obtained from the diet, from fatty fish oils and fortified dairy products, it is not technically a vitamin, as it is also produced in the skin following exposure to UV light. When human skin is exposed to UVB radiation, 7-dehydrocholesterol is photohydrolyzed to form cholecalciferol

(vitamin D3) (MacLaughlin et al., 1982). The vitamin D3 then enters the

circulation, bound to the vitamin D serum binding protein (DBP), and is

transported throughout the body. Before vitamin D3 can be biologically active, it

undergoes a series of ordered hydroxylation reactions in different organs.

The initial site of vitamin D3 activation is the liver, where the mitochondrial

cytochrome P450 enzyme vitamin D 25-hydroxylase (CYP27A1) catalyzes the

hydroxylation of vitamin D3 to 25-hydroxyvitamin D (25D) (Guo et al., 1993;

Ponchon and DeLuca, 1969). Liver production of 25D is driven by substrate

availability, and therefore serum 25D is directly dependent on the amount of substrate, vitamin D3, delivered to the liver. The lack of regulation of CYP27A1,

and subsequent lack of regulation of 25D production, makes serum levels of 25D

an excellent indicator of vitamin D status.

25D is the major circulating form of vitamin D in the body, and was long

believed to be the metabolically active form. In 1969, a metabolite more polar

than 25D was detected in the intestinal mucosa of chicks given radiolabeled

vitamin D3. This metabolite lacked the radiolabeled hydrogen molecule at the 1α position (Lawson et al., 1969), suggesting this metabolite was hydroxylated at the

1α position. Subsequent studies showed that this metabolite was indeed 1,25D, and was synthesized primarily in the kidney (Fraser and Kodicek, 1970); (Holick et al., 1971; Lawson et al., 1971). The enzyme that catalyzes the conversion of

25D to 1,25D is the 25-hydroxyvitamin D 1α-hydroxylase (CYP27B1), which is highly expressed in renal proximal tubules. The activity of CYP27B1 is negatively regulated by 1,25D; vitamin D deficiency leads to an increase in

CYP27B1 activity, while vitamin D repletion inhibits the production of 1,25D via

CYP27B1 (Henry et al., 1974). This negative feedback loop keeps circulating levels of 1,25D tightly regulated within physiological limits.

While the kidney is the main location of the CYP27B1 enzyme, and thus the main source of circulating 1,25D (Zehnder et al., 1999), a number of other tissues also express the CYP27B1 enzyme. Under normal conditions, functional

CYP27B1 has been localized to macrophages, keratinocytes, placenta, colon, mammary gland, and prostate gland (Townsend et al., 2005b). Non-renal

CYP27B1 can produce local pools of 1,25D; the hypercalcemia seen in sarcoidosis patients is caused by local 1,25D production by macrophages

(Adams et al., 1985). Relative to normal cells, breast and prostate cancer cells express lower levels of CYP27B1, suggesting a possible link between 1,25D and cancer in epithelial tissues. The importance and functional role of CYP27B1 in cancer tissues, however, remains an important area of study.

The catabolism of 1,25D is also an important step in the regulation of vitamin D signaling in relation to calcium homeostasis and the prevention of hypercalcemia. The cytochrome P450 enzyme 25-hydroxyvitamin D 24- hydroxylase (CYP24) is the enzyme responsible for initiating the catabolism of

1,25D, and to a lesser extent, 25D. Although the kidney is the main biological site of 24-hydroxylation, CYP24 is present in every cell that also expresses the

VDR (Burgos-Trinidad and DeLuca, 1991; Jones and Tenenhouse, 2002). The activity of CYP24 is enhanced by 1,25D in a VDR-dependent manner (Omdahl et al., 2002), forming a classic negative feedback loop to suppress activity of 1,25D in target cells. CYP24 expression is increased in carcinomas relative to normal cells, which further suggests a link between vitamin D signaling and cancer

(Bareis et al., 2001; Miller et al., 1995).

1.3 Classical 1,25D/VDR Signaling

1,25D, whether synthesized systemically or locally, exerts its biological functions through its cognate nuclear receptor, the vitamin D receptor (VDR).

The VDR is a member of group one of the nuclear receptor superfamily, along

with the thyroid hormone (TR), retinoic acid (RAR), pregnane X (PXR) , and constitutive androstane receptors (CAR) (Germain et al., 2006). All of these receptors form heterodimer complexes with the retinoid X receptors (RXR) for sequence-specific DNA binding (Escriva et al., 2000). The PXR and CAR are the most closely related nuclear receptors to the VDR, and they share sequence homology, transcriptional targets, and ligands. Both PXR and CAR upregulate cytochrome P450 (CYP) enzymes that mediate xenobiotic detoxification

(Honkakoski et al., 2003) in response to a wide variety of xenobiotics. PXR also binds intestinal bile acids such as lithocholic acid (LCA) (Moore et al., 2002).

Until very recently, 1,25D was believed to be the only ligand for the VDR. In

2002, however, VDR was shown to bind LCA, and a keto derivative, and induce

CYP3A4 (Makishima et al., 2002). These data suggest that the VDR may have evolved initially as a bile acid sensor, like the PXR, and aquired its calcium homeostatic function later.

The VDR was first discovered in chick intestinal extracts, as a saturable binding protein that caused the uptake of 1,25D into the chromatin fraction of the isolated intestinal tissue (Haussler and Norman, 1969). It was next isolated from other known vitamin D target tissues, such as kidney and bone (Chandler et al.,

1979; Kream et al., 1977). Following the discovery that 1,25D could induce differentiation and growth arrest (Abe et al., 1981; Bar-Shavit et al., 1983), the

VDR was discovered in a diverse array of tissues and cell lines, both tumorigenic and nontumorigenic (Haussler et al., 1998). Collectively, these data demonstrate

that the biological functions of the VDR extend far beyond its role as a calcium sensor.

Following the discovery of the VDR, its biochemical properties and structural orientation were delineated. The molecular weight of the VDR ranges from 50-60 kDa, depending on the species, with the human VDR being the smallest (Haussler et al., 1998). VDR binds 1,25D with both high affinity and selectivity, preferring it over other vitamin D3 metabolites or 1,25D degradation

products (Mellon and DeLuca, 1979). Most importantly, the VDR has an affinity

for DNA, which is increased upon ligand binding (Pike and Haussler, 1983).

Following the cloning of the VDR cDNA (Baker et al., 1988), VDR was identified

as a member of the nuclear receptor superfamily, as it exhibited the same

domain structure as and significant homology to the known nuclear receptors at

that time.

Nuclear receptor proteins are composed of six recognized domain

structures, named A-F. The A/B domains are nonconserved, and comprise all of the residues located to the amino end of the DNA binding domain, which is the C domain. The D domain is commonly called the “hinge” region, and lies in between the C domain and the ligand binding/coactivator domains, E/F (Green et al., 1986). In the VDR, the A/B domains are very short in comparison to other members of the nuclear receptor superfamily, the D domain, or hinge region, is longer and more complex than in other superfamily members, and the F domain is absent (McDonnell et al., 1989). Therefore, the VDR is comprised of five

domains, with the DNA and ligand binding domains (domains C and E) being the most important.

The DNA binding domain (C domain) is the most highly conserved domain among the members of the nuclear receptor superfamily, being comprised of two zinc finger DNA-binding motifs common to almost the entire family (Gronemeyer et al., 2004). The zinc atoms are coordinated through four highly conserved cysteine residues, while many other DNA binding proteins contain zinc fingers comprised of two cysteine and two histidine residues (Berg, 1989). In contrast, the ligand binding domain (E domain) is very divergent among the nuclear receptor superfamily, given the wide range of ligands bound by nuclear receptors. The three-dimensional structures of many liganded nuclear receptors have now been solved, and show substantial similarity between superfamily members. Eleven α-helices form the bulk of the ligand binding pocket, the size of which varies greatly from receptor to receptor, and a twelfth α-helix (H12) forms the “lid” for the binding pocket (Gronemeyer et al., 2004). The position of

H12 (i.e. whether the pocket is “open” and the receptor is inactive or “closed” and the receptor is active) is dictated by the ligand bound in the ligand pocket. When no ligand, or an antagonistic ligand, is bound, H12 is open, and corepressor proteins are bound to a hydrophobic groove in between helices 3 and 4. Binding of an agonistic ligand leads to a repositioning of H12 to close the ligand binding pocket. This repositioning makes corepressor binding unfavorable, and coactivators can then bind the LxxLL motifs present in helices 3, 4, and 12

(Gronemeyer et al., 2004). This general nuclear receptor activation scheme

applies to the VDR; the crystal structure of the VDR ligand binding domain bound to 1,25D is known, and is consistent with this mode of action (Rochel et al.,

2000).

The genomic signaling mechanism of the VDR is similar to the genomic signaling of other nuclear receptors. Following binding of an agonist (1,25D) to the ligand binding pocket, and subsequent H12 repositioning as described above, helices 9 and 10 of the ligand binding domain are also repositioned, allowing for association with the RXR heterodimer partner (Thompson et al.,

1998). If the RXR is unliganded, then transcription of vitamin D target genes results; liganded RXR negatively regulates VDR target genes, through diverting

RXRs away from VDR-binding and facilitating binding to other RXR partners, such as TR or RAR (Haussler et al., 1995). Therefore, in the regulation of vitamin D target genes, the VDR is complexed with an unliganded RXR. The binding of VDR to 1,25D has been shown to cause a conformational change in the RXR, leading it to take on a “closed” conformation; this not only allows for coactivator binding to the RXR, but ensures that the RXR remains unliganded

(Bettoun et al., 2003). The next step in vitamin D target gene activation is binding of the VDR-RXR heterodimer to DNA sequences in the promoters of

VDR target genes. These vitamin D response elements, or VDREs, in the context of VDR regulation of calcium homeostasis, are nearly all direct repeats of a six-nucleotide sequence (hexamer), separated by 3 bases (DR3); VDR binds the 3’ hexamer while RXR binds the 5’ hexamer (Haussler et al., 1998).

Although it is known that the calcemic actions of 1,25D are mediated through the

VDR in this manner, it remains unclear whether this genomic signaling is vital for

1,25D-mediated breast cancer cell growth regulation.

VDR also acts as a repressor of gene transcription, both via direct DNA binding and protein-protein interactions with other repressors. The VDR is required for the repression of the CYP27B1 enzyme, in a classic feedback loop mechanism. The p45Skp2 protein, which is involved in the degradation of p27, is transcriptionally repressed by VDR though binding with Sp1 to the Sp1 recognition sites in the p45Skp2 promoter (Huang and Hung, 2006). The

VDR/RXR heterodimer can bind directly to response elements in target genes to mediate transrepression, as in the IEX-1 gene (Im et al., 2002); the half-sites of the response element identified in the IEX-1 promoter are identical to half-sites in known positive VDREs, indicating that other factors may modulate the mechanism of direct VDR transrepression.

VDR genomic signaling is required for the maintenance of skeletal health and calcium and phosphate homeostasis. Perturbation in vitamin D signaling leads to a condition known as rickets, a condition in which the bones do not mineralize properly. Symptoms of rickets in children include: growth retardation, bowing of the legs, bent spine, enlargement of the bone ends, and weak muscles

(Holick, 2006). There are two different classifications of rickets, depending upon the underlying cause: nutritional and hereditary (Nield et al., 2006). Nutritional rickets is the most common form of rickets, resulting from vitamin D deficiency, either due to low dietary vitamin D intake or lack of ultraviolet light exposure. As darker-skinned individuals require more UV light exposure to generate vitamin D,

rickets is more common in African-American children than Caucasian children, particularly in nations where food fortification is not common (Holick, 2006).

Nutritional supplementation with vitamin D and/or calcium successfully cures nutritional rickets cases.

Inherited mutations in genes responsible for the production of 1,25D

(CYP27B1) or in the VDR lead to hereditary vitamin D resistant rickets (HVDRR).

Children with CYP27B1 inactivating mutations (HVDRR-1) respond to treatment with orally administered 1,25D (Holick, 2006). Children with HVDRR caused by

VDR mutations (HVDRR-2), however, show a differential pattern in their response to 1,25D treatment. These patients are distinguishable from both nutritional rickets and HVDRR-1 patients by their elevated levels of circulating

1,25D (Holick, 2006). HVDRR-2 patients also may present with alopecia, which is not a symptom of either dietary rickets or HVDRR-1. It is now clear that differences in HVDRR-2 patients, both in the severity of their rickets and in their response to treatment, correspond to the region of the VDR in which their inherited mutation exists. Mutations in the DNA binding domain (domain C) all present with alopecia, and tend to be more resistant to treatment; these mutant receptors lack VDR genomic activity and DNA binding, but retain the ability to bind 1,25D (Lin et al., 1996; Rut et al., 1994; Saijo et al., 1991; Sone et al.,

1990). Mutations in the ligand binding domain (domain E) generally do not present with alopecia, and tend to be more responsive to treatment, either with supra-physiological doses of 1,25D or with 1,25D analog drugs (Kristjansson et al., 1993; Whitfield et al., 1996). These receptors retain the ability to bind DNA,

but impaired ligand binding precludes significant genomic activity. These differences in phenotype between patients with DNA or ligand binding domain mutations indicate that the VDR may exert DNA- or ligand- independent actions, which have distinct biological effects.

1.4 Non-Classical VDR Signaling Pathways

The actions of the VDR as a transcriptional regulator of 1,25D-target genes involved in calcium homeostasis are relatively well understood. Studies in mice have shown that the calcemic signaling of 1,25D is dependent upon the presence of the nuclear VDR (Erben et al., 2002; Li et al., 1997). However, new evidence for the involvement of VDR in a range of non-classical signaling pathways are emerging that challenge the transcriptional regulatory function of the VDR as the only mechanism by which it mediates the broad range of physiological effects of 1,25D.

Genetically engineered mice lacking the VDR are phenotypically distinct from mice in which production of the ligand has been abrogated; most notably, mice with targeted VDR ablation suffer from alopecia, while CYP27B1 knockout mice do not (Dardenne et al., 2001; Li et al., 1997; Panda et al., 2001). This suggests that the VDR may have ligand-independent actions, or may bind ligands in addition to 1,25D, at least in maintenance of hair cycling. It has now been demonstrated that the interaction of the VDR with the corepressor protein hairless is responsible for the maintenance of the hair cycle, and this binding is

1,25D-independent (Skorija et al., 2005). The VDR has also been shown to

interact with the Rip13Δ1 co-repressor protein in the absence of ligand, and this interaction is also abrogated by 1,25D binding (Dwivedi et al., 1998).

Evidence also exists for actions of the VDR which are either completely

DNA-independent, or are not dependent upon binding to consensus-type VDREs in target gene promoters. The p27 protein is a known vitamin D target, but its promoter contains no identifiable VDREs; studies have shown that VDR induces p27 gene expression in complex with the Sp1 transcription factor at Sp1 binding sites in the p27 promoter (Huang et al., 2004). The transcriptional activity of the

VDR is important in Sp1-mediated upregulation, as mutations in VDR that block coactivator binding abrogate the effect (Cheng et al., 2006a). VDR has also been shown to inhibit the formation of the NFATp/AP-1 transcription factor complex in a DNA-independent manner (Towers et al., 1999).

Following treatment with 1,25D, a range of non-genomic “rapid responses” have been observed in cell and organ culture. These responses include activation of ERK and MAPK, release of intracellular calcium stores, opening of voltage-gated channels, rapid migration of endothelial cells and induction of protein kinases A and C (PKA and PKC). These effects occur too quickly

(seconds to minutes) to require transcriptional activation (Fleet, 2004; Norman,

2006), but the mechanisms that initiate these rapid actions are as yet unknown.

A membrane 1,25D receptor termed MARRS (also known as ERp57/GRp58) has been implicated in some studies (Nemere et al., 2004). Interestingly, chondrocytes from VDRKO mice retain some 1,25D rapid responses, which may be mediated through MARRS (Boyan et al., 2003). However, cells from HVDRR-

2 rickets patients fail to exhibit these rapid responses, indicating that the nuclear

VDR may be required for DNA-independent signaling (Nguyen et al., 2004).

Osteoblast cells from VDRKO mice are also lack the rapid regulation of ion channels in response to 1,25D, suggesting that the rapid response is also dependent on the nuclear VDR (Zanello and Norman, 2004). Recent evidence has shown that a pool of VDR is localized to the caveolar fraction of the plasma membrane, where it retains its specific 1,25D-binding capacity (Huhtakangas et al., 2004). Thus, current evidence suggests that the nuclear VDR may modulate rapid responses when localized to the plasma membrane and genomic signaling when localized in the nucleus. In some cells, MARRS may mediate a subset of rapid responses.

Perhaps the best evidence for the multiple and separable actions of the

VDR has been generated through the study of structural analogs of 1,25D.

Although 1,25D has potent anti-cancer effects, the doses necessary for these effects induce hypercalemia and soft tissue calcification (Vieth, 1990).

Therefore, considerable effort has been expended on the development and testing of synthetic analogs of 1,25D that retain anti-growth, pro-differentiation, and/or pro-apoptotic properties, with reduced side effects. Several well-studied compounds, including EB1089 (Leo Pharmaceuticals), exhibit increased potency for growth inhibition in vitro with reduced calcemia in vivo (Mathiasen et al., 1993;

VanWeelden et al., 1998). The fact that the growth regulatory and calcemic effects of VDR can be separated indicates that dinstinct mechanisms underlie the calcemic activities and anti-cancer activities.

Analogs have also been developed that specifically trigger either the genomic or rapid effects of VDR (Norman, 2006). Utilizing one of these analogs,

JN, an agonist for VDR rapid-responses, computational analysis indicated the possible presence of a second ligand binding pocket in VDR (Mizwicki et al.,

2004). This second pocket, known as the “A” or alternate pocket, overlaps the known ligand binding pocket (called the “G” or genomic pocket) at several residues, but appears to have a completely separate entrance from the G pocket, so that ligand binding is not dependent upon the conformation of helix 12

(Norman, 2006). This work is theoretical, and experimental research is necessary before any conclusions can be drawn; however, binding of 1,25D to the A pocket represents one possible mechanism by which anti-cancer effects of the VDR could be mediated.

1.5 Vitamin D3 and Cancer

The first suggestion of an anti-cancer effect of vitamin D came in 1980,

where an epidemiological study found that colon cancer death rates tended to

increase with decreasing sunlight exposure at increasing latitudes (Garland and

Garland, 1980). It was later discovered that incidence of colon cancer is

inversely proportional to the serum concentration of 25D in patient serum

(Garland et al., 1989). A link between UV light exposure, vitamin D status, and

prostate cancer has also been suggested (Garland et al., 1990; Hanchette and

Schwartz, 1992). In addition to epidemiological associations, many in vitro and in vivo studies support a role for vitamin D signaling in growth arrest and

differentiation of cancer cells. 1,25D has been shown to stimulate differentiation of myeloid leukemia cells (Miyaura et al., 1981), inhibit the growth of melanoma cells (Colston et al., 1981), inhibit the growth of retinoblastoma tumors (Albert et al., 1992), induce differentiation of osteosarcoma cells (Tsuchiya et al., 1993), and inhibit prostate cancer metastasis (Lokeshwar et al., 1999).

Vitamin D status has also been linked to breast cancer in a number of studies. Two epidemiological studies have shown an inverse correlation between intakes of dairy products, calcium, and vitamin D and breast cancer risk

(Knekt et al., 1996; Shin et al., 2002). Another study included an evaluation of sunlight exposure in addition to measures of dietary vitamin D intake, and showed that increased sun exposure or dietary vitamin D intake were associated with a decreased risk of breast cancer (John et al., 1999); this finding has been corroborated by another recent study (Knight et al., 2007). Serum levels of 25D have also been inversely correlated with breast cancer risk (Garland et al., 2007).

Low serum levels of 1,25D are associated with poorer prognosis and disease progression (Mawer et al., 1997b), as well as increased risk of breast cancer development (Janowsky et al., 1999). Breast cancer survival has been shown to be dependent upon the season of diagnosis, with cancers diagnosed in the summer and autumn, when sunlight exposure would be greatest, showing the best survival rates (Lim et al., 2006).

A role for vitamin D in breast cancer prevention has been suggested, based upon some of the studies discussed above, and on the observation that the VDR is expressed in normal mammary gland in a developmentally regulated

manner (Colston et al., 1988). Previous studies from our lab have highlighted the importance of the VDR in the mammary gland, showing that mammary gland development and ductal branching are accelerated in VDRKO mice, as compared to their WT littermates (Zinser et al., 2002). Glands from VDRKO mice also exhibit enhanced growth stimulation following estrogen or progesterone, both in vivo and in vitro organ culture.

Human population studies have also underscored a link between VDR polymorphisms and breast cancer development and/or progression. An example is the FokI restriction site polymorphism in VDR exon 2, which codes for receptors of slightly different sizes – M1 VDR with 427 amino acids or M4 VDR with 424 amino acids. The M4 variant VDR is more transcriptionally active than the M1, and also interacts more strongly with the TFIIB transcription factor

(Jurutka et al., 2000). The M1 VDR has been associated with an increased risk of breast cancer (Chen et al., 2005), as well as an increase in risk for African-

American women (Ingles et al., 1997). The Apa1 restriction site polymorphism in exon 9 has been linked with an increase in risk for bone metastasis following breast cancer (Schondorf et al., 2003a). These data suggest that 1,25D signaling pathways in general, and VDR structure in particular, may play a role in breast cancer development and progression.

The actions of vitamin D on breast cancer cells and tumors are three-fold: vitamin D treatment has been shown to arrest cell growth, induce apoptotic cell death, and decrease invasion and metastasis in a variety of models of breast cancer. Treatment of MCF-7 breast cancer cells with nanomolar concentrations

of 1,25D leads to cell cycle arrest in association with dephosphorylation of the cell cycle promoter retinoblastoma (Fan and Yu, 1995; Simboli-Campbell et al.,

1997; Wu et al., 1997). Vitamin D signaling also leads to increases in the cyclin dependent kinase inhibitors p21 and p27. The p21 gene promoter contains a functional VDRE, suggesting that it is a direct transcriptional target of the 1,25D-

VDR complex (Carlberg et al., 2007; Liu et al., 1996). Upregulation of p21 and p27 by vitamin D signaling causes inhibition of CDK activity, which inhibits G1/S phase transition (Verlinden et al., 1998). Thus, vitamin D causes accumulation of cancer cells in G1 phase, and thus blocks entrance into S phase.

Vitamin D signaling, either via 1,25D or synthetic vitamin D analogs, has been shown to induce the hallmarks of apoptosis – cell shrinking, chromatin condensation, DNA fragmentation, PARP cleavage, and phosphatidylserine reorientation – in breast cancer cells (Narvaez and Welsh, 2001; Simboli-

Campbell et al., 1996; Welsh, 1994). Studies have shown that the Bcl-2 family of pro-apoptotic proteins is important in 1,25D-mediated apoptosis. 1,25D treatment leads to redistribution of pro-apoptotic Bax from the cytosol to the mitochondria, and also downregulates the level of the anti-apoptotic protein Bcl-

2. 1,25D-mediated Bax translocation triggers the production of reactive oxygen species (ROS), decreases mitochondrial membrane potential, and leads to the release of cytochrome c into the cytosol (Narvaez et al., 2003; Narvaez and

Welsh, 2001), all of which are features of the intrinsic, or mitochondrial, apoptotic pathway (Scorrano and Korsmeyer, 2003). 1,25D treatment leads to the activation of caspase 9, a component of the intrinsic pathway, but not caspase 8,

a caspase activated via cell-surface receptor signaling (Guzey et al., 2002).

Vitamin D signaling also activates the endoplasmic reticulum (ER) apoptotic pathway, via calcium release from the ER and subsequent activation of μ- calpain, an apoptotic protease (Mathiasen et al., 2002). The ER apoptotic pathway also involves caspase 12, which can then activate the intrinsic apoptotic pathway (Nakagawa and Yuan, 2000; Nakagawa et al., 2000); thus vitamin D- mediated apoptosis involves at least two inter-connected protease pathways.

Metastasis, the process by which tumor cells leave their primary sites, travel through the bloodstream, and invade and proliferate in secondary sites, is facilitated by angiogenesis, or the growth of new blood vessels into tumors.

1,25D and the synthetic analog EB1089 both inhibit the invasion of late-stage breast cancer cells in vitro (Hansen et al., 1994), and this effect may be caused by VDR-mediated upregulation of extracellular matrix protease inhibitors (Koli and Keski-Oja, 2000). 1,25D was also shown to inhibit angiogenesis both in in vitro assays and in animal models (Bernardi et al., 2002; Majewski et al., 1996;

Mantell et al., 2000; Oikawa et al., 1990). The anti-angiogenic properties of

1,25D, therefore, represent a valuable therapeutic target.

1.6 Dissertation Objectives

Ample evidence indicates that 1,25D and the VDR are important regulators of mammary gland growth and development, and also mediate growth arrest and apoptosis in mammary tumors and tumor cells. While much is known about the mechanisms of VDR-mediated transcriptional regulation of target

genes, much less is known about the other functions of VDR, including transrepression and the rapid responses. In order to facilitate the development of new, more targeted synthetic 1,25D analogs, we must better understand the activities of the VDR in mediating the anti-cancer actions of 1,25D.

1. Characterize the 1,25D/VDR signaling axis and examine the anti-cancer effects of 1,25D in the WT145 and KO240 murine mammary tumor cell model system in vitro.

2. Determine the effects of structural analogs of vitamin D on growth, VDR signaling, and apoptosis in the WT145 and KO240 cell lines.

3. Examine the anti-tumor effects of EB1089, a vitamin D analog, and UV- exposure (to enhance endogenous vitamin D production) on WT145 and KO240 tumors in nude mouse xenografts.

4. Investigate whether DNA- and/or ligand- independent actions of VDR may contribute to its anti-tumor activities.

CHAPTER 2:

CHARACTERIZATION OF WT145 AND KO240 MURINE MAMMARY TUMOR

CELLS IN STEROID-FREE MEDIA

2.1 Introduction

1,25-dihydroxyvitamin D3 (1,25D) is the biologically active form of the

steroid vitamin D3. The best characterized biological action of 1,25D is

maintenance of extracellular calcium homeostasis, which is mediated via

genomic signaling through the vitamin D receptor (VDR), a member of the

nuclear receptor superfamily. It is well accepted that the 1,25D-VDR complex

acts as a transcription factor through direct contact of its DNA binding domain to vitamin D response elements (VDREs) in the promoter regions of target genes.

Mutations in the amino acids within the DNA binding domain that make direct

contact with DNA abolish the transcriptional activity of VDR, causing end-organ

resistance to 1,25D and hypocalcemic rickets (Malloy and Feldman, 2003). Thus,

the transcriptional activity of the VDR is crucial for mediating the calcemic effects of vitamin D.

In addition to its calcemic effects, 1,25D can induce growth arrest, apoptosis, and differentiation in a variety of normal and transformed cells in vitro, and synthetic vitamin D analogs can mediate tumor regression in vivo (Colston et al., 2003; Gombart et al., 2006; Ordonez-Moran et al., 2005; VanWeelden et al.,

1998; Yee et al., 2005). Furthermore, studies with VDRKO mice have demonstrated that VDR status modulates proliferation and carcinogenesis in normal tissues including mammary gland, lymphoid tissue, colon and skin

(Colston and Welsh, 2005; Kallay et al., 2001; Li et al., 1997; Meindl et al., 2005;

Zinser et al., 2005). Multiple pathways and effectors have been implicated in the cell regulatory effects of vitamin D compounds, including inhibition of mitogenic growth factors, induction of apoptosis, up-regulation of cell cycle inhibitors such as p21 and p27, and induction of oxidative stress (Lowe et al., 2003a; Simboli-

Campbell et al., 1996).

Although VDR is required for 1,25D-mediated growth regulation, the targets of genomic signaling and the possible contribution of non-genomic pathways in mediating growth arrest, differentiation and apoptosis remain to be clarified. The development of VDR agonists, such as the vitamin D analog

EB1089, with enhanced anti-proliferative properties and minimal calcemic effects in vivo, indicates that the cell regulatory and calcemic actions of VDR can be dissociated. At the genomic level, this dissociation could be achieved via distinct response elements in growth regulatory vs calcemic regulatory genes

(Danielsson et al., 1997). Alternatively, VDR modulation of growth regulatory genes may be mediated via novel mechanisms, as has been demonstrated for induction of the cyclin dependent kinase inhibitor p27 by 1,25D. On this gene promoter, ligand-occupied VDR does not contact DNA directly, but interacts with the Sp1 transcription factor to enhance its binding at GC rich regions (Cheng et al., 2006b; Huang et al., 2004) Thus, VDR may target gene promoters involved

in growth regulation through mechanisms distinct from those involved in calcemic regulation. It is also likely that non-transcriptional regulation of signal transduction pathways at the membrane or in the cytosol by 1,25D and/or the VDR contribute to growth regulation (Capiati et al., 2004; Huhtakangas et al., 2004; Rochel and

Moras, 2006; Zanello and Norman, 2004).

To study the association between VDR genomic signaling and the anti- cancer effects of 1,25D, murine mammary tumor cell lines that differentially express the VDR were developed from VDRKO mice and their wild-type littermates (Zinser et al., 2003). In the current studies these cell lines were adapted to steroid-free culture conditions to further examine the role of the VDR in mediating the effects of 1,25D on growth inhibition.

2.2 Materials and Methods

2.2.1 Cell Culture

WT145 and KO240 cells were maintained in Dulbecco’s Modified Eagle’s

Medium (DMEM)/F12 medium (Sigma Aldrich, St. Louis, MO) containing 5% charcoal-stripped fetal bovine serum (CSS) (HyClone, Logan, UT). Cells were routinely passaged twice weekly using trypsin/EDTA.

2.2.2 Western blotting

Cells were seeded in 150mm dishes (Corning, Corning, NY) at a density of 500,000 cells per dish, and treated 24 hours later with 100nM 1,25D (gift of

Leo Pharmaceuticals, Ballerup, Denmark) or vehicle control. For whole cell lysates, cells were harvested 48 hours post-treatment by scraping into 2x

Laemmli buffer containing the following protease and phosphatase inhibitors:

10mM benzamdidne, 10mM sodium fluoride, 100mM sodium vanadate, 1mM phenylmethanesulphonylfluoride (PMSF), 25 μg/mL leupeptin, 25 μg/mL aprotinin, and 25 μg/mL pepstatin. Lysates were sonicated, and protein concentration determined using the BCA protein assay (Pierce Biotechnology,

Rockford, IL). Subcellular fractions were prepared as previously described

(Narvaez and Welsh, 2001). 50μg of protein lysates and subcellular fractions were separated via SDS-PAGE, transferred to nitrocellulose filters, blocked in skim milk, and incubated overnight with primary antibodies including mouse

monoclonal VDR Clone D-6, 1:100 (Santa Cruz Biotechnology, Santa Cruz, CA), mouse monoclonal CYP24, 1:200 (Cytochroma, Markham, Ontario, Canada), and goat polyclonal actin, 1:100 (Santa Cruz Biotechnology). Appropriate horseradish-peroxidase conjugated secondary antibodies (obtained from

Amersham Biosciences, Piscataway, NJ) were incubated with filters for 1 hour.

Specific bands were detected via chemiluminescence (SuperSignal West Dura,

Pierce, Rockford, IL) and exposure to x-ray film. Films were scanned with a flatbed computer scanner.

2.2.3 Immunofluorescence

Cells were seeded in 2-well chamber slides (Nalge Nunc, Naperville, IL) at a density of 10,000 cells per chamber, and treated 24 hours later with 100nM

1,25D or vehicle control. 24 hours post-treatment, cells were fixed, permeabilized with ice cold methanol and incubated with VDR antibody clone D-

6, 1:200 (Santa Cruz Biotechnology, Santa Cruz, CA), followed by anti-mouse

Alexa Fluor 488 (Invitrogen, Carlsbad, CA) secondary antibody, 1:100. DNA was counterstained using Hoechst 33342 (Invitrogen, Carlsbad, CA). Coverslips were mounted with anti-fade mounting medium, and slides viewed on an

Olympus Provis AX70 microscope with a Spot RT Slider digital camera.

2.2.4 VDR Transactivation Assay

Cells were seeded in 12-well plates (Corning, Corning, NY) at a density of

30,000 cells per well. After 24 hours, cells were transfected with the 1.6μg pGL3-

24 hydroxylase luciferase reporter vector which contains approximately 300bp of

the human CYP24 gene promoter with its two DR3 VDRE regions (gift of the late

Dr. Omdahl). A pRL-TK driven luciferase plasmid (400ng) (Promega, Madison,

WI) was co-transfected to normalize for transfection efficiency. TransFast transfection reagent (Promega, Madison, WI) was used according to manufacturer’s recommendations. 24 hours post-transfection, cells were treated with the indicated doses of 1,25D. After 24 hours of treatment, cells were harvested with 1x Passive Lysis Buffer and fluorescence was quantitated via the

Dual Luciferase system (Promega, Madison, WI). Data are presented as the ratio of pGL/pRL, which is called relative luciferase units (RLU).

2.2.5 Quantitative Real-Time Polymerase Chain Reaction (PCR)

Cells were seeded in 100mm dishes (BD Biosciences, San Jose, CA) at a density of 500,000 cells per dish, and treated 24 hours later with 100nM 1,25D.

Cells were pelleted via centrifugation and RNA was harvested using the RNeasy

Mini Kit (Qiagen, Valencia, CA) according to manufacturer’s recommendations.

Three cDNA replicates were made for each RNA, using the TaqMan Reverse

Transcription Reagents kit (Applied Biosystems, Foster City, CA). Gene expression analysis was performed using SYBR green (ABGene, Rochester,

NY), and values were normalized against 18S RNA. Primer sequences are as follows: Mus CYP24 Forward – AAGTCATGGACTTGGCCTTCA; Mus CYP24

Reverse – GCTCCGCCTTCTCGTTGA; 18S rRNA Forward -

AGTCCCTGCCCTTTGTACACA; 18S rRNA Reverse -

GTTCCGAGGGCCTCACTAAAC . Triplicate plates were run on the ABI-Prism

7700 (Applied Biosystems, Foster City, CA).

2.2.6 Crystal Violet Cell Growth Assay and Doubling Time

Cells were seeded in 24-well plates (ICN Biomedicals, Aurora, OH) at a density of 2,000 cells per well and treated 24 hours later with the indicated doses of 1,25D. 96 hours post-treatment, cells were fixed with 1% glutaraldehyde in

PBS for 20 minutes, and then stained with 0.1% crystal violet dye (Fisher

Scientific, Pittsburgh, PA) for 15 minutes. Dye was resuspended in 0.2% Triton-

X100 for 15 minutes, and absorbance was measured at 590nm. For doubling time, cells were fixed and stained for five consecutive days, and doubling time calculated according to measured absorbance.

2.2.7 Statistical Analysis

Data are expressed as mean +/- standard error. ANOVA or student’s t-test were performed using GraphPad Prism software (San Diego, CA), and means were considered statistically significant when p-values less than 0.05 were obtained. Statistical significance is indicated on all data figures as letters or asterisks above bars; bars are labeled with different letters for means that are significantly different by ANOVA, and asterisks signify statistical difference from control by t-test.

2.3 Results

2.3.1 VDR protein expression in WT and VDRKO cells

The cells used in these studies were isolated from mammary tumors that developed in WT and VDRKO mice (Zinser et al., 2003) and were adapted to steroid-free medium containing stripped serum. Western blotting was used to confirm the differential expression of VDR in these cells under these conditions.

As shown in Figure 2.1, the 50kDa VDR protein was expressed in WT145 cells but was undetectable in KO240 cells. The level of VDR protein in WT145 cells was increased following treatment with 1,25D, but was not induced in KO240 cells. Subcellular fractionation of WT145 cells (Figure 2.2a) indicated that the

VDR protein was detected in both cytosolic and nuclear fractions under control conditions, and was markedly increased in both compartments following treatment with 1,25D. By immunofluorescence microscopy (Figure 2.2b), VDR expression was low in WT145 cells in the absence of 1,25D, and accumulated in the nucleus upon hormone treatment. In KO240 cells, negligible VDR staining was detected even in the presence of 1,25D, a finding that was confirmed with several distinct antibodies to VDR (data not shown). Thus, these VDRKO cells, derived from the colony generated by Demay’s group (Li et al., 1997), do not express a truncated VDR protein that retains ligand binding ability as reported for another strain of VDRKO mice (Bula et al., 2005).

WT145 KO240

VDR

actin

Con1,25D Con 1,25D

Figure 2.1. VDR protein is expressed only in WT145 cells. Whole cell lysates of WT145 and KO240 cells treated for 48 hours with 100nM 1,25D or vehicle control were immunoblotted with antibodies directed against VDR (top) and actin (bottom) as a loading control. Blot is representative of five independent whole cell lysate harvests.

2.3.2 VDR transcriptional activity in WT145 cells

To assess VDR transcriptional activity, transient transfection assays were conducted with a VDR-responsive luciferase construct that contains the promoter region of the human CYP24 gene. In the WT145 cell line, CYP24 promoter activity was low under basal conditions, and was significantly increased after

1,25D treatment. In the KO240 cell line, basal CYP24 promoter activity level was

a) Nuclear Cytosol VDR

Con1,25D Con 1,25D

b) Phase Nuclei VDR

WT145 Con

20μ

WT145 1,25D

KO240 Con

KO240 1,25D

Figure 2.2. VDR concentrates in the nucleus following treatment with 1,25D. a) Western blot for VDR in subcellular fractions from WT145 cells treated for 48 hours with 100nM 1,25D or vehicle control. b) Phase contrast and fluorescence images of WT145 and KO240 cells treated for 24 hours with 100nM 1,25D or vehicle control. Cells were incubated with Hoechst (nuclei) or anti- VDR antibody (VDR) prior to imaging.

low and was not induced by 1,25D treatment (Figure 2.3a). Dose-response experiments (Figure 2.3b) indicated that CYP24 promoter activity was significantly increased in WT145 cells following treatment with concentrations of

1,25D as low as 1nM. Real-time PCR analysis (Figure 2.4a) indicated that 1,25D also induced the endogenous CYP24 gene and protein (Figure 2.4b) in WT145 cells, but not KO240 cells. These data confirm that VDR remains functional in

WT145 cells selected for growth in stripped serum devoid of 1,25D, and that mammary cells derived from VDRKO mice do not express any proteins other than VDR that can mediate transcription of the CYP24 gene in response to

1,25D.

2.3.3 Differential growth inhibition by 1,25D in WT and VDRKO cells

Growth assays were conducted to determine whether the VDR present in

WT145 cells could mediate growth inhibitory effects of 1,25D. WT145 cell cultures were growth inhibited by decreasing doses of 1,25D (Figure 2.5); the lower limit for this effect was found to be 100fM 1,25D (Figure 2.6). Under the same conditions, 1,25D had no effect on growth of KO240 cells, up to a concentration of 1μM 1,25D (Figure 2.5). The doubling time of WT145 cells was significantly lengthened by treatment with 100nM 1,25D, by almost 3 fold, while

KO240 cells were unaffected (Figure 2.7). Thus, even after selection for growth in CSS, WT145 cells remained sensitive, and KO240 cells remained insensitive, to 1,25D-mediated growth inhibition.

a WT145 KO240 5 b 4

RLU 2

1 ac c a 0 Con1,25D Con 1,25D

b WT 145 10

c bc b 5 RLU

a 0 Con 1nM 10nM 100nM

Figure 2.3. WT145 cells express transcriptionally active VDR. a) CYP24 reporter gene activity in WT145 and KO240 cells treated for 24 hours with 100nM 1,25D or vehicle control. b) CYP24 reporter gene activity in WT145 cells treated for 24 hours with indicated doses of 1,25D. In both graphs, data were normalized for transfection efficiency measured by co-transfected pRL-TK, and are expressed as relative luciferase units (RLU). Data points represent mean ± SEM of six values; bars with different letters are statistically different.

WT145 KO240 a)

3 b

a a a 1

CYP24 Gene Expression 0 Con1,25D Con 1,25D

b) WT145 KO240 CYP24

Con1,25D Con 1,25D

Figure 2.4. Endogenous CYP24 mRNA and protein levels are upregulated in WT145, but not KO240, cells following 1,25D treatment. a) Real time PCR for endogenous CYP24 mRNA in WT145 and KO240 cells after 6 hours of treatment with 100nM 1,25D or vehicle control. Bars with different letters are statistically different. b) Western blot of CYP24 in WT145 and KO240 cells after 48 hours of treatment with 100nM 1,25D or vehicle control. Blot is representative of three independent whole cell lysate harvests.

WT145 1.5

a 1.0

b b bbb 0.5

Absorbance 590nm Absorbance

0.0 Con 1 10 100 500 1000 nM 1,25D

KO240 1.5

1.0

0.5

590nm Absorbance

0.0 Con 1 10 100 500 1000 nM 1,25D

Figure 2.5. WT145 cells, but not KO240 cells, are growth inhibited by 1,25D. Crystal violet growth assay in WT145 and KO240 cells treated for 96 hours with 1,25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values; bars with different letters are statistically different.

WT145 2

a b c d 1 e f

590nm Absorbance 0 Con 100fM 1pM 10pM 100pM 1nM 1,25D

Figure 2.6. WT145 cells are sensitive to concentrations of 1,25D in the femtomolar range. Crystal violet growth assay in WT145 cells treated for 96 hours with 1,25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values; bars with different letters are statistically different.

WT145 133.8 150 *

100

Time (h) 48.4 50

0 Con 1,25D

KO240

50 37.4 38.8 40

Time (h) 20

0 Con 1,25D

Figure 2.7. The doubling time of WT145 cells is increased by 1,25D. Graphs of average doubling times of WT145 and KO240 cells following 100nM 1,25D treatment. Numbers above bars are the average doubling time, in hours, for that treatment. Data points represent mean ± SEM of three independent time course trials, each with at least 8 replicates. Asterisks indicate statistical significance as determined by t-test.

2.4 Discussion

In these studies, we have utilized a pair of mammary tumor cell lines with differential VDR expression to examine the role of the VDR in mediating the effects of vitamin D steroids on growth arrest and apoptosis. These cells were isolated from tumors that developed in VDR WT and KO mice. Previous studies indicated that both cell lines are epithelial in origin, estrogen receptor positive, and have differential expression of the VDR protein (Zinser et al., 2003). In these follow-up studies, the cells were adapted to CSS-containing media, free of exogenous steroid compounds including 1,25D, and characterized to ensure that these culture conditions did not alter the basic characteristics of either cell line.

In CSS media, the WT145 cell line expresses cytoplasmically-distributed

VDR protein at a low basal level, which is upregulated and translocated to the nucleus following treatment with 1,25D. This VDR protein is transcriptionally active, as indicated by upregulation of a human VDRE-containing VDR- responsive promoter at concentrations of 1,25D as low as 1nM, and the upregulation of the endogenous CYP24 protein and mRNA following 1,25D treatment. Thus, in these cells, VDR activation measured by an exogenous transiently transfected reporter gene correlates well with endogenous VDR target gene and protein expression. The KO240 cell line does not express detectable

VDR and no changes in CYP24 promoter activity, mRNA, or protein expression occur following treatment with 1,25D. This confirms that the expression of VDR protein is required for the 1,25D-mediated activation of this target gene, and suggests that the KO240 cell line does not express any other protein which can

function as a mediator of 1,25D-responsive gene expression through typical consensus DR3 VDREs.

The growth-inhibitory effects of 1,25D were examined in the WT145 and

KO240 cell lines, and found to be dependent upon the presence of VDR protein.

The WT145 cell line is growth inhibited following treatment with 1,25D at the physiologically relevant concentration of 100fM, whereas the KO240 cell line is not affected by treatment with 1,25D at concentrations up to 1μM. The overall growth rates of both cell lines are slower in medium containing CSS, as compared to previous results in FBS medium (Zinser et al., 2003). This may be due to downregulation of the endogenous estrogen receptor, which we have observed in both cell lines after adaption to CSS (data not shown). Loss of estrogen receptor and slower growth, however, do not impact the responsiveness to 1,25D treatment. WT145 cells are actually more sensitive to the growth inhibitory effects of 1,25D in CSS vs. FBS, while KO240 cells remain insensitive.

Overall, our results indicate that the growth-inhibitory effects of 1,25D require expression of the VDR protein, although whether the transcriptional activity of the VDR is required for these anti-growth actions is unclear from the studies above. It is clear, however, that the KO240 cell line is resistant to the growth inhibitory actions of 1,25D, at supra-physiological concentrations as high as 1μM. The WT145 and KO240 cell lines, therefore, comprise a uniquely useful model system, which will be utilized in this thesis to dissect the mechanisms by

which VDR mediates diverse actions of endogenous and synthetic vitamin D steroids.

CHAPTER 3:

INDUCTION OF APOPTOSIS BY 1,25D IN MURINE MAMMARY TUMOR

CELLS REQUIRES VDR PROTEIN EXPRESSION

3.1 Introduction

Apoptosis, the ordered process of programmed cell death, is mediated through activation of specific proteases, known as caspases. Caspase enzymes are constitutively expressed in cells in an inactive proform. Once activated, they are auto-catalytic, and can subsequently cleave other caspase family members to propagate the apoptotic signal. The caspases that are activated early in the cascade are known as initiator caspases, which in turn cleave and activate downstream, or effector, caspases. These effector caspases then cleave downstream targets, such as cell structural proteins and anti-apoptotic factors, and thus lead to the death of the cell (Thornberry and Lazebnik, 1998).

The initiation of apoptosis is mediated through both extracellular and intracellular pathways. The extrinsic apoptotic pathway is induced via the binding of apoptotic peptides to their cell surface receptors; once these receptors are liganded, they recruit procaspase 8 molecules, which auto-activate themselves and induce the rest of the apoptotic cascade (Raff, 1998). Damage to organelles can also activate apoptosis, although the mechanisms through which this occurs are not completely understood. It is known that mitochondrial damage causes

the release of cytochrome c from the mitochondrial membrane, which can then activate procaspase 9, to induce apoptosis through the intrinsic pathway (Raff,

1998). A large family of intracellular proteins called the Bcl-2 family is also involved in modulation of the apoptotic cascade. Some Bcl-2 family members are pro-apoptotic, and some are anti-apoptotic - the ratio of pro- to anti- apoptotic

Bcl-2 proteins determine’s a cells susceptibility to apoptosis (Raff, 1998). The

Bcl-2 family member Bax binds to the mitochondrial membrane and assists in the release of cytochrome c in response to apoptotic signals (Raff, 1998).

In addition to its calcemic effects, 1,25D has been shown to cause growth arrest and apoptosis in a variety of breast cancer cell lines in vitro and in vivo.

Previous studies have shown that 1,25D treatement causes MCF-7 breast cancer cells to exhibit characteristic morphological features of apoptosis, such as chromatin condensation, pyknotic nuclei, and cytoplasmic condensation (Simboli-

Campbell et al., 1996). Treatment with EB1089, a synthetic 1,25D analog, was later shown to exert effects on apoptosis indistinguishable from those caused by

1,25D itself, and both compounds downregulated the anti-apoptotic protein Bcl-2

(Simboli-Campbell et al., 1997). Overexpression of the Bcl-2 protein blocked

1,25D-mediated apoptosis (Mathiasen et al., 1999). Blocking the activation of the caspase signaling cascade was shown to have no effect on the induction of

1,25D-mediated apoptosis in MCF-7 cells (Mathiasen et al., 1999; Narvaez and

Welsh, 2001), indicating that the commitment to apoptosis induced by 1,25D is caspase independent. However, apoptosis still occurred after caspase blockade, it was merely delayed in onset, which indicates that caspase activation is not

necessary for the execution of apoptosis by 1,25D. One mechanism for apoptotic activation in MCF-7 cells was shown to be disruption of mitochondrial membrane potential, via Bax protein translocation to the mitochondria, which leads to cytochrome c release from the mitochondrial membrane and production of reactive oxygen species (Narvaez and Welsh, 2001). 1,25D and its structural analogs are catabolized by CYP450 enzymes within the mitochondria, which may contribute to non-specific mitochondrial stress. 1,25D was also shown to activate calpain, a calcium-dependent apoptotic protease, in MCF-7 cells; the authors suggest that calpain may act as the main protease in MCF-7 cells (Mathiasen et al., 2002).

While these studies highlighted the importance of the mitochondrial apoptotic pathway in 1,25D-mediated apoptosis, they were performed in MCF-7 cells, which express mutated, and consequently non-functional, caspase 3 protein (Kurokawa et al., 1999). In prostate cells with functional caspase 3, activation of the caspase cascade occurred following 1,25D treatment, and the caspase 3 enzyme was activated as a part of this cascade (Guzey et al., 2002).

1,25D-mediated apoptosis was also shown to proceed via activation of caspase

3 in squamous cell carcinoma (Alagbala et al., 2006). Therefore, the observation that 1,25D-mediated apoptosis is caspase-independent may not be applicable to all cancer cell types, and may, instead, represent an alternative, possibly redundant mechanism, by which 1,25D induces apoptosis in the absence of functional caspase 3.

With this in mind, we set out to utilize our model system to clarify the role of caspases in 1,25D-induced apoptosis. Use of both the WT145 and KO240 cell lines allowed us to determine whether the VDR is required for the apoptotic actions of 1,25D, or whether a non-specific mechanism involving redox pathways or 1,25D metabolism in the mitochondria is involved. We have also examined whether the presence of the VDR affects caspase activity, and which apoptotic enzymes participate in 1,25D-mediated apoptosis.

3.2 Materials and Methods

3.2.1 Phase Contrast Microscopy

Cells were seeded in 2-well chamber slides (Nalge Nunc, Naperville, IL) at a density of 10,000 cells per chamber, and treated 24 hours later with 100nM

1,25D or vehicle control. 96 hours post-treatment, cells were fixed, permeabilized with ice cold methanol and incubated with Hoechst 33342

(Invitrogen, Carlsbad, CA) to highlight nuclear morphology. Coverslips were mounted with anti-fade mounting medium, and slides viewed on an Olympus

Provis AX70 microscope with a Spot RT Slider digital camera.

3.2.2 Western Blotting

Cells were seeded in 150mm dishes (Corning, Corning, NY) at a density of 500,000 cells per dish, and treated 24 hours later with indicated doses of

1,25D (gift of Leo Pharmaceuticals, Ballerup, Denmark), etoposide (EMD

Chemicals, San Diego, CA) or vehicle control. Cells were harvested 48 hours post-treatment by scraping into 2x Laemmli buffer containing protease and phosphatase inhibitors, as described in chapter 2. Lysates were sonicated, and protein concentration determined using the BCA protein assay (Pierce

Biotechnology, Rockford, IL). 50μg of lysate was separated via SDS-PAGE, transferred to nitrocellulose filters, blocked in skim milk, and incubated overnight with primary antibodies including mouse specific Apoptosis Antibody Sampler Kit,

1:1000 (Cell Signaling Technology, Danvers, MA), rabbit polyclonal caspase 7,

1:500 (Cell Signaling Technology), and rabbit polyclonal mu-calpain, 1:200

(Affinity Bioreagents, Golden, CO). The Apoptosis Antibody Sampler Kit contained the appropriate secondary antibody for all primary antibodies used in the kit, and blotting was performed according to manufacturer’s instructions. For caspase 7 and mu-calpain, appropriate secondary antibodies (Amersham

Biosciences, Piscataway, NJ) were incubated with filters for 1 hour, and specific bands were detected via chemiluminescence (SuperSignal West Dura, Pierce,

Rockford, IL) and exposure to x-ray film. Films were scanned with a flatbed computer scanner.

3.2.3 Caspase Activity Assay

Cells were seeded in 150mm dishes (Corning, Corning, NY) at a density of 300,000 cells per dish, and treated 24 hours later with the indicated doses of test compounds or vehicle controls. 48 hours post-treatment, cells were harvested via trypsinization. One million cells per treatment were assayed using commercially available caspase 3 and 9 activity assays (BD Biosciences, San

Jose, CA), according to manufacturer’s recommendations. Caspase activity was measured via cleavage of a synthetic caspase peptide target, and release of fluorescence. Assays were performed at least three times with three separate cell preparations.

3.2.4 Crystal Violet Cell Growth Assay

Cells were seeded in 24-well plates (ICN Biomedicals, Aurora, OH) at a density of 2,000 cells per well and treated 24 hours later with the indicated doses of etoposide (EMD Chemicals) or vehicle control. 96 hours post-treatment, cells were fixed using 2% glutaraldehyde in PBS and stained with crystal violet dye.

Dye was resuspended in 0.1% Triton-X100 and absorbance was measured at

590nm. Assays were performed at least three times.

3.3 Results

3.3.1 Morphological features of apoptosis in WT145 cells, but not KO240 cells,

following 1,25D treatment

To assess whether 1,25D induces apoptotic morphology in our transformed murine mammary tumor cells, we examined WT145 and KO240 cells after 96h treatment with 1,25D or vehicle control. Phase contrast microscopy shows that the morphology of KO240 cells is not altered by treatment with 100nM 1,25D, when compared with vehicle control (Figure 3.1). WT145 cells treated with 1,25D are less densely packed, appear elongated, and appear less well adhered to the culture surface as compared with vehicle control (Figure

3.1). WT145 cells treated with 1,25D also show small, dense nuclei, and nuclear irregularity (indicated by arrows), which are both hallmarks of apoptosis.

Phase Nuclei

WT145

Con

20μ

WT145 1,25D

KO240 Con

KO240 1,25D

Figure 3.1. 1,25D induces morphological features of apoptosis in WT145 cells, but not KO240 cells. Phase contrast and fluorescent images of WT145 and KO240 cells treated for 4 days with 100nM 1,25D or vehicle control, and incubated with Hoechst dye to visualize nuclear morphology.

3.3.2 Activation of caspase 3 by 1,25D requires VDR protein expression

We next examined whether expression and/or activation of caspase 3 was altered by VDR ablation or 1,25D treatment. Western blotting was used to monitor expression of the inactive caspase 3 zymogen as well as the cleaved, catalytically active mature form (Figure 3.2a). Although VDR ablation per se did not alter basal caspase 3 expression, treatment with 100nM 1,25D induced caspase 3 cleavage in WT145 cells but not in KO240 cells. Measurement of caspase 3 activity (Figure 3.2b) indicated that the low basal level of caspase 3 activity was significantly increased in WT145 cells treated with 100nM 1,25D. In

KO240 cells, basal caspase 3 activity was lower than that of WT145 cells and was unaffected by 1,25D.

3.3.3 1,25D activates apoptotic protease pathways in tumor cells that express

VDR

To further characterize cell death induction by 1,25D, we examined caspases 9 and 12, which act as initiator caspases in the mitochondrial and endoplasmic reticular pathways, respectively. Activation of both caspase 9 and

12 (monitored as disappearance of the pro-caspase forms due to lack of antibodies that recognize the cleaved fragments) was observed in WT145 cells treated with 100nM 1,25D, but not in KO240 cells (Figure 3.3). We also assessed the downstream effector caspases that propagate apoptotic signals, and found that procaspases 7 and 10, but not procaspase 6, were activated by

1,25D in WT145 cells but not KO240 cells. Consistent with evidence of caspase activation in WT145 cells, cleavage of its substrate poly (ADP-ribose)

a Proenzyme (33kDa)

Caspase 3 Active (19kDa)

b WT 145 KO240 7500 b

5000

a 2500 c c Caspase 3 Activity Units 0 Con1,25D Con 1,25D

Figure 3.2. 1,25D induces caspase 3 activity in WT145 cells, but not KO240 cells. a) Western blot for caspase 3 in WT145 and KO240 cells after 48 hours of treatment with 100nM 1,25D or vehicle control. Blot is representative of three independent whole cell lysate harvests. b) Caspase 3 activity in WT145 and KO240 cells following 48 hours of treatment with 100nM 1,25D or vehicle control. Bars represent mean ± SEM of 3 values; different letters are statistically different. polymerase (PARP) was enhanced in WT145 cells but not KO240 cells following

1,25D treatment. The calcium dependent protease calpain was also cleaved to its active form in 1,25D treated WT145 cells but not in KO240 cells (Figure 3.3).

To corroborate the western blot data, we also performed caspase 9 activity assays in both WT145 and KO240 cells. Both cell lines show basal levels of activated caspase 9 (Figure 3.4), but only WT145 cells show increased caspase 9 activity following 1,25D treatment.

WT145 KO240 Proenzyme (49kDa) Caspase 9

Caspase 12 Proenzyme (55kDa)

Proenzyme (35kDa) Caspase 6

Caspase 7 Proenzyme (35kDa)

Proenzyme (66kDa) Caspase 10

Proenzyme (80kDa)

Calpain Active (50kDa)

Full-length (116kDa) PARP Cleaved (89kDa)

Con 1,25D Con 1,25D

Figure 3.3. Evidence of protease activation in WT145 cells following 1,25D treatment. Western blots for various caspases, calpain, and the caspase substrate poly(ADP-ribose)polymerase (PARP) in WT145 and KO240 cells after 48 hours of treatment with 100nM 1,25D or vehicle control. Blots are representative of three independent whole cell lysate harvests.

WT145 KO240

30000 *

20000

10000

Caspase 9 Activity Units 0 Con1,25D Con 1,25D

Figure 3.4. 1,25D induces caspase 9 activity in WT145 cells, but not KO240 cells. Caspase 9 activity in WT145 and KO240 cells following 48 hours of treatment with 100nM 1,25D or vehicle control. Bars represent mean ± SEM of 3 values; *p<0.05, con vs. 1,25D.

3.3.4 VDR ablation does not alter sensitivity to other apoptotic agents

To determine whether VDR ablation had effects on cellular sensitivity to apoptosis in general, we examined the induction of apoptosis in WT145 and

KO240 cells following treatment with etoposide, a known DNA-damaging apoptotic agent. Both WT145 and KO240 cells were comparably growth inhibited by increasing concentrations of the DNA-damaging agent etoposide

(Figure 3.5a). Although caspase 3 activation occurred at lower etoposide doses in WT145 cells than KO240 cells, both cell lines exhibited comparable responses to concentrations of 1μM and higher. Caspase 3 activity was increased, and the cleaved caspase 3 enzyme form was present, in both WT145 and KO240 cells treated with 3μM etoposide (Figure 3.5b,c). Thus, cells lacking VDR retain a

functional DNA damage response pathway leading to activation of caspase 3 and cell death. Collectively, these data indicate that 1,25D induces apoptosis in murine mammary cells through VDR dependent pathways that disrupt intracellular organelles and trigger activation of multiple proteases.

a WT145 KO240 1.0 a a 2 b ab a b

0.5 1 c c d d Absorbance 590nm e e Absorbance 590nm e e

0.0 0 Con 0.05 0.1 0.5 1 3 5 Etop Con 0.05 0.1 0.5 1 3 5 Etop b WT145 KO240 60000 60000 c

40000 40000 b b

20000 20000 b Caspase 3 Activity Caspase Units a Caspase3 Activity Units a 0 0 Con 0.5 3 Etop Con 0.5 3 Etop

c WT145 KO240 Proenzyme (33kDa)

Caspase 3 Active (19kDa) Con 0.5 3 Con 0.5 3

Figure 3.5. WT145 and KO240 cells are sensitive to etoposide- mediated apoptosis. a) Crystal violet growth assay in WT145 and KO240 cells following 96 hours of treatment with indicated concentrations of etoposide (μM) or vehicle control. b) Caspase 3 activity in WT145 and KO240 cells following 48 hours of treatment with indicated concentrations of etoposide (μM) or vehicle control. c) Western blot for caspase 3 in WT145 and KO240 cells after 48 hours of treatment with indicated concentrations of etoposide (μM) or vehicle control. Blot is representative of three independent whole cell lysate harvests. In a and b, bars represent mean ± SEM of 6 values; different letters are significantly different.

3.4 Discussion

In these studies, we examined the activation of the caspase-mediated apoptotic cascade following 1,25D treatment of mammary tumor cells.

Treatment of WT145 cells with 100nM 1,25D lead to induction of apoptosis, as shown by induction of apoptotic morphology and activation of proteolytic caspase enzymes. In previous studies, approximately 15% of WT145 cells exhibited DNA fragmentation following 1,25D treatment (Zinser et al., 2003), and in the current studies, activated caspase 3 was detected in cell lysates, and the cleavage activity of caspase 3 was increased following treatment with 1,25D. 1,25D mediated apoptosis in WT145 cells was associated with activation of the upstream caspases 9 and 12. No evidence for activation of caspase 8, which transduces membrane death receptor signals, by 1,25D was obtained (data not shown). The activation of caspase 3 in WT145 cells, therefore, was through both the mitochondrial (via caspase 9) and endoplasmic reticular (via caspase 12) pathways (Thornberry and Lazebnik, 1998). A variety of effector caspases, including 7 and 10, were also activated. Activated calpain, a ubiquitously expressed protease activated by disruptions in calcium homeostasis, has been correlated to 1,25D-mediated apoptosis in human breast cancer cells (Mathiasen et al., 2002). We observed calpain activation in WT145 cells following 1,25D treatment. Activated calpain can activate caspase 12, leading to the propagation of the caspase cascade (Nakagawa and Yuan, 2000). It is unclear from our data whether caspase 12 was activated by calpain solely, or whether endoplasmic reticular stress via calcium release following 1,25D treatment was activating the

protease. Further studies will be required to determine whether caspase 12 activation by 1,25D is a primary event, or secondary to the activation of calpain.

Regardless of the specific mechanism, it is clear that protease activation during 1,25D mediated apoptosis requires the VDR, since KO240 cells contained no detectable active caspase 3 protein, and their caspase 3 activity levels did not rise above basal, following treatment with 1,25D. There was also no detectable initiator or effector caspase activation, calpain activation, or PARP cleavage in

KO240 cells following treatment with 1,25D. This suggests that triggering of apoptosis by 1,25D is dependent upon the presence of the VDR protein in cells, and that non-specific redox-mediated perturbation of the mitochondria by 1,25D catabolism is not responsible for 1,25D mediate apoptosis. However, lack of VDR did not alter sensitivity of KO240 cells to the DNA-damaging agent etoposide.

This indicates that KO240 cells are not globally resistant to apoptosis; rather they are selectively non-responsive to 1,25D-mediated apoptosis, suggesting that the absence of the VDR does not negatively impact on common apoptotic pathways.

Many previous studies, often in MCF-7 cells, have examined the importance of Bcl-2 family members in apoptosis mediated by 1,25D; we chose to focus on caspase signaling instead in our model system, and whether the absence of the VDR lead to perturbation of caspase signaling. Our studies indicate that in breast cancer cells with functional caspase 3, 1,25D activates both the intrinsic caspase pathway, through caspase 9, and the endoplasmic reticular caspase pathway, through caspase 12 and calpain. Our studies are in good agreement with studies in prostatic cancer cell lines that showed 1,25D-

mediated activation of the intrinsic caspase pathway (Guzey et al., 2002). Our studies also agree with those showing that 1,25D activates calpain in MCF-7 cells (Mathiasen et al., 2002); while the activation of calpain may mediate apoptosis in the absence of caspase 3 in MCF-7 cells, it is unclear from our studies whether it is activated directly by 1,25D, or via caspase-mediated cleavage and activation. In either case, it suggests that 1,25D activates apoptosis in cells expressing VDR through a variety of caspase-dependent and - independent mechanisms, which may serve as a sort of “fail-safe”.

CHAPTER 4:

REQUIREMENT OF VDR FOR THE ANTI-CANCER ACTION OF SYNTHETIC

STRUCTURAL ANALOGS OF 1,25D AND OTHER VITAMIN D METABOLITES

4.1 Introduction

It is well accepted that the 1,25D-VDR complex acts as a transcription factor through direct contact of its DNA binding domain to vitamin D response elements (VDREs) in the promoter regions of target genes, many of which are involved in calcium homeostasis. Most 1,25D-inducible genes contain VDREs composed of a direct repeat of two separate six-base pair elements with a three nucleotide spacer (DR3) (Haussler et al., 1998). The best example of this type of

VDRE has been identified in the proximal promoter of the 24-hydroxylase

(CYP24) gene, which contains two well defined DR3-type VDREs. In other VDR target genes, functional VDREs composed of inverted palindromes with 9-base pair spacers (IP9), everted repeats with 6 or 9-base pair spacers (ER6, ER9), and direct repeats with 4 or 6-base pair spacers (DR4, DR6) have been described (Quack and Carlberg, 2000; Saramaki et al., 2006; Schrader et al.,

1997; Thompson et al., 2002; Thummel et al., 2001; Xie and Bikle, 1997).

Mutations in the amino acids within the DNA binding domain that make direct contact with DNA abolish the transcriptional activity of VDR, causing end-organ resistance to 1,25D and hypocalcemic rickets (Malloy and Feldman, 2003). Thus,

the transcriptional activity of the VDR is crucial for mediating the calcemic effects of vitamin D. Although VDR is required for 1,25D-mediated growth regulation, the mechanism and targets of genomic signaling and the possible contribution of non-genomic pathways in mediating the anti-cancer actions of 1,25D remain to be clarified.

The development of VDR agonists, such as the synthetic vitamin D analog

EB1089, with enhanced growth regulatory properties and minimal calcemic effects in vivo indicates that the anti-cancer and calcemic actions of VDR can be dissociated. At the genomic level, this dissociation could be achieved via distinct response elements in growth regulatory vs calcemic regulatory genes. In support of this concept, induction of IP9-type VDREs correlated better with 1,25D- induced apoptosis than did induction of DR3-type VDREs (Danielsson et al.,

1997), and VDR bound to EB1089 exhibited increased affinity for IP9 VDREs as compared to DR3 VDREs (Nayeri et al., 1995). Another study identified specific mutations in VDR that impair 1,25D induction of relB, a gene required for dendritic cell maturation, without impairing CYP24 activation (Nguyen et al.,

2006). Alternatively, VDR modulation of growth regulatory genes may be mediated via novel mechanisms, as has been demonstrated for induction of the cyclin dependent kinase inhibitor p27 by 1,25D. On this gene promoter, ligand- occupied VDR directly interacts with the Sp1 transcription factor which binds and activates gene expression at Sp1 sites (Huang et al., 2004). It is also likely that non-transcriptional regulation of signal transduction pathways at the membrane or in the cytosol by 1,25D and/or the VDR contribute to growth regulation (Capiati

et al., 2004; Huhtakangas et al., 2004; Rochel and Moras, 2006; Zanello and

Norman, 2004). Collectively, these data support the concept that VDR may mediate growth regulation through mechanisms distinct from those involved in calcemic regulation.

While we have shown that VDR expression is required for the growth inhibitory and apoptotic effects of 1,25D in mammary tumor cells, the data above suggest that non-1,25D agonists of the VDR mediate their anti-cancer effects in a manner which is mechanistically separable from the genomic signaling mechanism of the VDR. Therefore, in these studies, we examined the effects of synthetic 1,25D analogs as well as the 1,25D precursor molecule 25D on gene regulation, growth arrest, and apoptosis in the WT145 and KO240 cell lines.

4.2 Materials and Methods

4.2.1 Western Blotting

Cells were seeded in 150mm dishes (Corning, Corning, NY) at a density of 500,000 cells per dish, and treated 24 hours later with 100nM 1,25D, 25D (gift of Leo Pharmaceuticals, Ballerup, Denmark), or vehicle control. Cells were harvested 48 hours post-treatment by scraping into 2x Laemmli buffer containing protease and phosphatase inhibitors, as described in chapter 2. Lysates were sonicated, and protein concentration determined using the BCA protein assay

(Pierce Biotechnology, Rockford, IL). 50μg of lysate was separated via SDS-

PAGE, transferred to nitrocellulose filters, blocked in skim milk, and incubated overnight with mouse monoclonal CYP24, 1:200 (Cytochroma, Markham,

Ontario, Canada), sheep polyclonal CYP27B1, 1:100 (The Binding Site, San

Diego, CA) and goat polyclonal actin, 1:100 (Santa Cruz Biotechnology) primary antibodies. Appropriate horseradish-peroxidase conjugated secondary antibodies (obtained from Amersham Biosciences, Piscataway, NJ) were incubated with filters for 1 hour. Specific bands were detected via chemiluminescence (SuperSignal West Dura, Pierce, Rockford, IL) and exposure to x-ray film. Films were scanned with a flatbed computer scanner.

4.2.2 Quantitative Real-Time PCR

Cells were seeded in 100mm dishes (BD Biosciences, San Jose, CA) at a density of 500,000 cells per dish, and treated 24 hours later with 100nM 1,25D.

Cells were pelleted via centrifugation and RNA was harvested using the RNeasy

Mini Kit (Qiagen, Valencia, CA) according to manufacturer’s recommendations.

Three cDNA replicates were made for each RNA, using the TaqMan Reverse

Transcription Reagents kit (Applied Biosystems, Foster City, CA). Gene expression analysis was performed using SYBR green (ABGene, Rochester,

NY), and values were normalized against 18S RNA. Primer sequences are as follows: Mus CYP24 Forward – AAGTCATGGACTTGGCCTTCA; Mus CYP24

Reverse – GCTCCGCCTTCTCGTTGA; Mus CYP27B1 Forward –

CAGAGCGCTGTAGTTTCTCATCA; Mus CYP27B1 Reverse –

CGTTAGCAATCCGCAAGCA; 18S rRNA Forward -

AGTCCCTGCCCTTTGTACACA; 18S rRNA Reverse -

GTTCCGAGGGCCTCACTAAAC . Triplicate plates were run on the ABI-Prism

7700 (Applied Biosystems, Foster City, CA).

4.2.3 Crystal Violet Cell Growth Assay

Cells were seeded in 24-well plates (ICN Biomedicals, Aurora, OH) at a density of 2,000 cells per well and treated 24 hours later with the indicated doses of 25D, EB1089, MC1288, KH1230, or CB1093 (a gift from Leo Pharmaceuticals,

Ballerup, Denmark) or vehicle control. 96 hours post-treatment, cells were fixed using 2% glutaraldehyde in PBS and stained with crystal violet dye. Dye was

resuspended in 0.1% Triton-X100 and absorbance was measured at 590nm.

Assays were performed at least three times.

4.2.4 Caspase Activity Assay

Cells were seeded in 150mm dishes (Corning, Corning, NY) at a density of 300,000 cells per dish, and treated 24 hours later with the indicated doses of test compounds or vehicle controls. 48 hours post-treatment, cells were harvested via trypsinization. One million cells per treatment were assayed using a commercially available caspase 3 activity assay (BD Biosciences, San Jose,

CA), according to manufacturer’s recommendations. Assays were performed at least three times with independent cell preparations.

4.2.5 VDR Transactivation Assay

Cells were seeded in 12-well plates (Corning, Corning, NY) at a density of

30,000 cells per well. After 24 hours, cells were transfected with the pGL3-24 hydroxylase luciferase reporter vector which contains approximately 300bp of the human CYP24 gene promoter with its two DR3 VDRE regions (gift of the late Dr.

Omdahl). A pRL-TK driven luciferase plasmid (Promega, Madison, WI) was co- transfected to normalize for transfection efficiency. TransFast transfection reagent (Promega, Madison, WI) was used according to manufacturer’s recommendations. 24 hours post-transfection, cells were treated with the indicated doses of 1,25D. After 24 hours of treatment, cells were harvested with

1x Passive Lysis Buffer and fluorescence was read via the Dual Luciferase system (Promega, Madison, WI).

4.3 Results

4.3.1 WT145 and KO240 cells express CYP27B1

CYP27B1, the cytochrome P450 enzyme that generates 1,25D from 25- hydroxyvitamin D3 (25D), has been localized to human mammary cells as well as mouse mammary gland (Townsend et al., 2005a; Zinser and Welsh, 2004a).

Western blotting indicated that the WT145 and KO240 cells, which were derived from carcinogen induced mouse mammary tumors, retain expression of

CYP27B1 protein (Figure 4.1a) and RNA (Figure 4.1b). In vitro, CYP27B1 expression was not consistently altered by exposure to 25D or 1,25D in either cell line, after correction for protein loading determined by blotting for actin.

4.3.2 Growth inhibition by 25D is dependent upon the VDR

To test whether the presence of CYP27B1 sensitized cells to growth inhibition by 25D, as reported in human mammary cells (Friedrich et al., 2000;

Kemmis et al., 2006), growth was assessed in WT145 and KO240 cells treated with increasing doses of 25D for 96 hours. As shown in Figure 4.2a, 25D inhibited growth of WT145 cells at the physiologically relevant concentration of

100nM, but had no effect on growth of KO240 cells. In contrast to 1,25D, which increased caspase 3 activity in WT145 cells, 25D did not activate caspase 3

(Figure 4.2b), suggesting that 25D inhibited growth without induction of apoptosis.

a WT145 KO240 CYP27B1

Con1,25D25D Con 1,25D 25D

Actin

Con1,25D25D Con 1,25D 25D

b WT145 KO240 2

CYP27B1 Expression

0 - + - + 1,25D

Figure 4.1. WT145 and KO240 cells express CYP27B1. a) Whole cell lysates of WT145 and KO240 cells treated for 48 hours with 100nM 1,25D or vehicle control were immunoblotted with antibodies directed against CYP27B1 (top) and actin (bottom) as a loading control. Blot is representative of three independent whole cell lysate harvests. b) Real time PCR for endogenous CYP27B1 mRNA in WT145 and KO240 cells after 6 hours of treatment with 100nM 1,25D or vehicle control.

4.3.3 25D does not transcriptionally activate CYP24 in WT145 cells

Reporter gene assays were conducted to determine whether the concentration of 25D that inhibited growth of WT145 cells was sufficient to activate VDR transcription. In contrast to 1,25D, 25D did not activate the heterologous CYP24 promoter (Figure 4.3a), and did not induce the endogenous murine CYP24 protein (Figure 4.3b), or mRNA (Figure 4.3c) in WT145 cells.

These experiments were repeated in media which was not charcoal stripped, and

no activation of either the endogenous or heterologous CYP24 promoters was observed (data not shown). A detailed quantitative real time PCR time course showed no elevation of CYP24 mRNA in WT145 cells treated with 25D for up to

96 hours (Figure 4.3d). Collectively, these data suggest that the presence of both CYP27B1 and VDR are permissive for growth inhibition by 25D, but that activation of the VDR target gene CYP24 is not required.

a WT145 KO240

2 2 a a a b

1 1 c c Absorbance 590nm Absorbance 590nm Absorbance

0 0 Con 1 10 100 500 1000 nM 25D Con 1 10 100 500 1000 nM 25D

b WT145 KO240

3 b

2 a a a a a 1 (Fold Induction) CaspaseActivity 3

0 Con25D1,25D Con 25D 1,25D Figure 4.2. WT145 cells are growth inhibited, but not rendered apoptotic, by 25D. a) Crystal violet growth assay in WT145 and KO240 cells treated for 96 hours with 25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of six values. b) Caspase 3 activity in WT145 and KO240 cells following 48 hours of treatment with 100nM 25D or 1,25D, or vehicle control. Data points represent mean ± SEM of four values. For all graphs, bars with different letters are significantly different.

a WT 145 KO240 4 b b WT145 KO240 3 CYP24

2 RLU Con 1,25D25D Con 25D 1,25D

a a a a a 0 Con1,25D25D Con 1,25D 25D

c WT145 KO240 d 3 b 2

1 a a a a a 1 (Fold Change) (Fold Induction) CYP24 Expression Con 25D CYP24 Gene Expression Gene CYP24

0 0 Con1,25D25D Con 1,25D 25D 0 20 40 60 80 100 Time, h

Figure 4.3. 25D does not induce CYP24 in WT145 cells. a) CYP24 reporter gene activity in WT145 and KO240 cells treated for 24 hours with 100nM 1,25D, 100nM 25D, or vehicle control. Data were normalized for transfection efficiency measured by co- transfected pRL-TK, and are expressed as relative luciferase units (RLU). b) Whole cell lysates of WT145 and KO240 cells treated for 48 hours with 100nM 1,25D or vehicle control were immunoblotted with antibodies directed against CYP24. Blot is representative of three independent whole cell lysate harvests. c) Real time PCR for endogenous CYP24 mRNA in WT145 and KO240 cells treated for 6 hours with 100nM 1,25D, 25D, or vehicle control. d) Real time PCR time course of CYP24 mRNA in WT145 and KO240 cells treated with 100nM 25D or vehicle control for up to 96 hours. For a, c, and d, data points represent mean ± SEM of six values, and bars with different letters are significantly different.

4.3.4 EB1089 mediates growth arrest via VDR-dependent pathways

We examined whether growth of murine mammary cells was affected by

EB1089, a synthetic vitamin D analog that exerts less calcemic activity than

1,25D in vivo. As shown in Figure 4.4a, EB1089 inhibited growth of WT145 cells at concentrations as low as 1nM. In contrast, the potency of EB1089 to upregulate transcription of the human CYP24 promoter was significantly less than that of 1,25D (Figure 4.4b). Similarly, EB1089 induced the endogenous murine CYP24 gene in WT145 cells, but the time course of induction was slower than that of 1,25D (Figure 4.4c). As expected, EB1089 did not inhibit growth, activate the CYP24 promoter or induce the endogenous CYP24 gene (data not shown) in KO240 cells.

4.3.5 Other synthetic 1,25D analogs induce growth arrest and mediate CYP24

transcription in WT145 cells

Crystal violet assays were performed using three other synthetic 1,25D analogs - CB1093, MC1288, and KH1230 - also known to be more potent in growth regulation but less calcemic than 1,25D. All three analogs inhibited growth of WT145 cells at concentrations as low as 1nM (Figure 4.5), while

KO240 cells were unaffected by all three compounds at concentrations as high as 1μM. Each analog exhibited a similar potency for CYP24 transcriptional activation in WT145 cells (Figure 4.6), and none induced a transcriptional response in KO240 cells.

a WT145 KO240 1.5 1.5

1.0 1.0

b b b b b 0.5 0.5 Absorbance 590nm Absorbance 590nm

0.0 0.0 Con 1 10 100 500 1000 EB Con 1 10 100 500 1000 EB

WT145 KO240 c b 6 hour 12 hour 15 20 b c

10 c

RLU 10

5 c (Fold Control)

CYP24 ExpressionCYP24 b a a a a a 0 0 Con1,25DEB Con 1,25D EB 1,25DEB 1,25D EB

Figure 4.4. EB1089 induces growth arrest and VDR transcriptional activity in WT145 cells. a) Crystal violet growth assay in WT145 and KO240 cells treated for 96h with EB1089 (EB) at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of six values. b) CYP24 reporter gene activity in WT145 and KO240 cells treated for 24 hours with 100nM 1,25D, 100nM EB1089, or vehicle control. Data were normalized for transfection efficiency measured by co-transfected pRL-TK, and are expressed as relative luciferase units (RLU). c) Real time PCR for endogenous CYP24 mRNA in WT145 and KO240 cells treated for 6 or 12 hours with 100nM 1,25D, EB1089, or vehicle control. Data are expressed as mean fold increase over control values. For all graphs, bars with different letters are significantly different; for c, letters signify difference from control as well as other bars.

WT145 KO240

0.7 1.0 0.6 a 0.8 0.5

0.4 0.6 b b b b 0.3 b 0.4 0.2

Absorbance 590nm Absorbance 0.2 0.1 Absorbance 590nm

0.0 0.0 Con 1 10 100 500 1000 CB Con 1 10 100 500 1000 CB

a 0.5 1.2

0.4 1.0 b 0.8 0.3 b b b b 0.6 0.2 0.4

Absorbance 590nm 0.1 Absorbance 590nm 0.2

0.0 0.0 Con 1 10 100 500 1000 MC Con 1 10 100 500 1000 MC

0.5 a 1.0

0.4 0.8 0.3 b b b b b 0.6

0.2 0.4

Absorbance 590nm 0.1 Absorbance 590nm 0.2

0.0 0.0 Con 1 10 100 500 1000 KH Con 1 10 100 500 1000 KH

Figure 4.5. 1,25D analogs induce growth arrest in WT145 cells. Crystal violet growth assays in WT145 and KO240 cells treated for 96h with CB1093 (CB), MC1288 (MC), or KH1230 (KH) at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values. Bars with different letters are significantly different.

WT145 KO240 15 * * * 10

(Fold Contol) RLU

0 CBMCKH CB MC KH 100nM

Figure 4.6. 1,25D analogs induce VDR transcriptional activity in WT145 cells. CYP24 reporter gene activity in WT145 and KO240 cells treated for 24 hours with 100nM CB1093 (CB), 100nM MC1288 (MC), 100nM KH1230 (KH) or vehicle control. Data were normalized for transfection efficiency measured by co-transfected pRL-TK, and are expressed as relative luciferase activity (RLU). Bars represent mean ± SEM of six values, and are expressed as the fold increase over ethanol control treated cells.

4.4 Discussion

In these studies, we have utilized mammary tumor cell lines from mice with targeted deletion of VDR and their wild type littermates to examine cellular mechanisms involved in vitamin D mediated growth inhibition. We specifically compared the efficacy of selected non-1,25D VDR agonists to inhibit growth and to activate the CYP24 gene in these cells. The CYP24 promoter contains several DR3 type VDREs (Vaisanen et al., 2005), is highly induced by 1,25D, and is commonly used as an indicator of VDR transcriptional activity. However, the suitability of CYP24 as a measure of sensitivity to other cellular processes mediated by vitamin D compounds has not been established. In these studies, we demonstrate that although VDR is absolutely required for growth arrest and apoptosis by VDR agonists, the potency of a vitamin D metabolite or analog to induce the VDR target gene CYP24 does not predict its potency in mediating growth regulation.

The biggest discrepancy between growth inhibition and CYP24 induction was observed when cells were treated with physiological concentrations of the natural vitamin D metabolite, 25D. We and others have demonstrated that human mammary cells express CYP27B1 and are growth inhibited by 25D

(Kemmis et al., 2006; Townsend et al., 2005a) and in the present study we detected CYP27B1 in both WT145 and KO240 murine mammary cell lines. The presence of CYP27B1 conferred sensitivity to 25D mediated growth arrest in

WT145, but not KO240 cells, indicating that growth inhibition by 25D was VDR dependent. We therefore predicted that 25D mediated growth arrest in WT145

cells would be associated with CYP24 induction. Surprisingly, three different assays (human CYP24 luciferase reporter, real time PCR for the endogenous murine gene and western blot for CYP24 protein) failed to detect induction of

CYP24 in WT cells treated with 100nM 25D. Assays conducted in parallel demonstrated that 1,25D activated the CYP24 reporter gene and increased both

CYP24 mRNA and protein in WT145 cells (Figures 2.3, 2.4). In fact, a five-fold induction of the CYP24 promoter was consistently observed at 1nM 1,25D

(Figure 2.3), a dose which induced growth inhibition of WT145 cells comparable to that achieved with 100nM 25D.

The observed dissociation between growth inhibition and CYP24 induction by 25D was unexpected, and clarification of the mechanisms and significance of this finding will require additional study. It is possible that the amount of 1,25D generated from 25D in WT145 cells is sufficient to inhibit growth but is too low to activate CYP24 transcription. Mechanistically, this would presume that growth arrest of cells treated with 25D is mediated by VDR target genes that are responsive to much lower 1,25D concentrations than CYP24 and/or that low concentrations of 1,25D impact on growth via novel mechanisms. Our data is also consistent with the possibility that 25D mediates growth inhibition of WT145 cells by directly binding VDR, independent of its conversion to 1,25D. Indeed, precedent for VDR activation by ligands other than 1,25D exists, as bile acids have been identified as a new class of VDR ligands that participate in regulation of metabolic gene expression in the intestine (Jurutka et al., 2005; McCarthy et al., 2005; Thompson et al., 2002). Although the affinity of 25D for VDR is

approximately 60 fold less than that of 1,25D (Skowronski et al., 1995), it is conceivable that the intracellular concentrations of 25D are sufficient for VDR binding. If so, the 25D-VDR complex could trigger either genomic or non- genomic effects. Recent modeling studies have suggested the existence of an alternate binding pocket in VDR that binds 25D and generates a receptor conformation that does not support genomic signaling but could likely support non-genomic effects (Mizwicki et al., 2005). Examples of non-genomic effects of

VDR ligands implicated in cell growth and differentiation include modulation of phosphorylation cascades in leukemic cells (Berry et al., 1996; Bhatia et al.,

1995), activation of protein phosphatases in colon carcinoma cells (Bettoun et al., 2002; Bettoun et al., 2004) and rapid effects on multiple kinase pathways in breast and squamous carcinoma cells (Capiati et al., 2004; Ma et al., 2006).

Further studies will be necessary to determine whether 25D directly binds VDR in

WT145 cells and if so, whether the resulting 25D-VDR complex exerts genomic or non-genomic effects that can be linked to growth arrest.

We also observed dissociation between growth inhibition and CYP24 induction in WT145 cells treated with the synthetic vitamin D analog EB1089, which was equipotent to 1,25D in triggering growth inhibition, but less potent than

1,25D in CYP24 activation. These observations mirror those obtained in MCF-7 human breast cancer cells (Danielsson et al., 1997), where EB1089 more potently induced growth arrest and apoptosis than 1,25D but was approximately

10-fold less potent than 1,25D in activation of the DR3 VDRE of the osteopontin

gene. In our study, no effects of EB1089 were observed in KO240 cells, indicating that VDR is required for growth regulation by EB1089.

Three other 1,25D-analogs, CB1093, KH1230, and MC1288, were also examined in our model system. WT145 cells were growth inhibited by all three compounds, and no definitive differences in sensitivity of WT cells to any of the compounds were noted. All three compounds were able to transcriptionally activate the CYP24 promoter in WT145 cells, with approximately equal potency.

In contrast to the results obtained with EB1089, the levels of CYP24 transcription induced by these three compounds were approximately equipotent to that induced by 1,25D. KO240 cells were not impacted by treatment with any of the synthetic vitamin D compounds.

Collectively, these data are consistent with the concept that synthetic vitamin D analogs exert unique effects by promoting VDR conformations that differentially activate target gene expression. Notably, in MCF-7 cells, EB1089 preferentially activated IP9 type VDREs compared to DR3 type VDREs (Nayeri et al., 1995). Since EB1089 displayed an activity profile intermediate between

25D and 1,25D, further comparison of the potency of these three vitamin D compounds to activate additional VDR responsive promoters in WT145 cells will be of particular interest.

CHAPTER 5:

VDR-DEPENDENT INHIBITION OF MAMMARY TUMOR GROWTH IN VIVO BY

EB1089 AND ULTRAVIOLET RADIATION

5.1 Introduction

1α,25-dihydroxyvitamin D (1,25D) is the biologically active form of vitamin

D3 (cholecalciferol), a steroid that can be obtained in the diet or through

endogenous synthesis in skin exposed to UV light. The metabolism and

biological effects of dietary and UV generated vitamin D are indistinguishable.

For biological activity, vitamin D is sequentially hydroxylated to 25-hydroxyvitamin

D (25D, the major circulating form) and 1,25D (the biologically active form).

1,25D is the ligand for the vitamin D receptor (VDR), a transcription factor that

regulates tissue specific gene expression, and also exerts non-receptor

mediated effects on intracellular calcium and signaling pathways (Fleet, 2004). In

vitro, 1,25D induces growth arrest, differentiation and apoptosis in transformed

cell lines derived from breast, prostate, colon and other tissues (Lowe et al.,

2003b) and these effects are mediated by VDR (Narvaez and Welsh, 2001;

Simboli-Campbell et al., 1996). The VDR is expressed in 80% of human breast

cancers, and tumor VDR has been correlated with prognosis (Berger et al., 1991;

Berger et al., 1987; Eisman et al., 1986).

Despite the beneficial growth inhibitory effects of 1,25D in vitro, its use in cancer therapy is precluded by its potent calcemic effects. However, synthetic vitamin D analogs that display enhanced anti-cancer activity with reduced calcemic activity have been developed for potential therapeutic applications

(O'Kelly and Koeffler, 2003). Although the growth inhibitory effects of these vitamin D analogs in vitro are VDR mediated (Zinser et al., 2003), the mechanism of the dissociation between their anti-proliferative and calcemic effects in vivo is still unclear. EB1089 (seocalcitol) is a well-characterized vitamin D analog which is more potent than 1,25D in regulating growth and differentiation, but 50% less calcemic than 1,25D (Hansen et al., 2001). The efficacy of vitamin D analogs to induce growth arrest, activate apoptosis, inhibit angiogenesis and mediate tumor regression in animal models of cancer with minimal calcemic toxicity has been documented (Mehta et al., 2000; Narvaez and Welsh, 2001; Oades et al., 2002;

VanWeelden et al., 1998). However, previous studies have not addressed the relative contributions of genomic and non-genomic vitamin D signaling, or identified the specific cellular targets of vitamin D signaling in vivo.

More recently, evidence has also accumulated to support a role for vitamin

D in breast cancer prevention. Studies in knockout mice have demonstrated that the VDR impacts on normal mammary gland development and tumorigenesis

(Zinser et al., 2002; Zinser et al., 2005; Zinser and Welsh, 2004b; Zinser and

Welsh, 2004c). Epidemiological studies have demonstrated that breast cancer incidence and mortality inversely correlate with incident UV radiation and occupational or recreational sun exposure (Gorham et al., 1990; Grant, 2003). It

has been proposed that these correlations might reflect the ability of UV rays present in sunlight to stimulate conversion of 7-dehydrocholesterol to vitamin D in skin. Exposure of human skin to sub-erythemal (non-reddening) doses of UV rapidly generates in excess of 20,000 units of vitamin D, an amount much higher than can be obtained from the diet, and approximately 100 times the established

“adequate intake” for adults of 200 units/day (Holick, 2003). It is well accepted that UV stimulation of vitamin D synthesis can both prevent and cure vitamin D deficiency, indicating that vitamin D generated in the skin is fully functional.

Despite evidence that vitamin D impacts on mammary gland biology and may protect against breast cancer development or progression (Mawer et al., 1997a;

Schondorf et al., 2003b), no studies have addressed whether UV-generated vitamin D influences the growth of normal or transformed breast cells.

In the studies reported here, we used a unique model system comprised

of cells derived from VDR knockout (KO) and wild type (WT) mice (Zinser et al.,

2003) to determine whether endogenous (UV generated) or exogenous (EB1089)

stimulation of the vitamin D pathway altered growth of tumors in a VDR

dependent manner.

5.2 Materials and Methods

5.2.1 Cell Culture

WT145 and KO240 murine mammary tumor cells were cultured in

DMEM/F12 (Sigma Aldrich) containing 5% charcoal-dextran treated fetal bovine serum (Hyclone, Logan UT). Cells were grown in T-150 flasks, and routinely yielded 5-10x106 cells per flask, depending on confluence. For inoculation into

nude mice, cells were washed twice with PBS, trypsinized, resuspended in

DMEM/F12, and counted. After centrifugation, cells were resuspended in

Matrigel (BD Biosciences, San Jose, CA):DMEM/F12 (4:1).

5.2.2 Nude Mouse Xenografts

Female, athymic, ovariectomized NCr nu/nu nude mice were housed in

sterile isolator cages and maintained on sterilized water and irradiated low (0.1%)

calcium diet (Purina Test Diets, Richmond, IN) ad libitum. Mice were implanted

subcutaneously (sc) with a 90-day extended release 1.7mg 17β-estradiol pellet

(Innovative Research, Sarasota, FL). Two weeks later, mice were inoculated sc

in the flank with 2x106 WT145 or KO240 tumor cells, (Valrance and Welsh, 2004;

Zinser et al., 2003) suspended in 300μL Matrigel (BD Bioseicences, San Jose,

CA) :DMEM/F12 (4:1). Mice were weighed and tumor volumes were measured

bi-weekly via caliper measurement, and volume calculated using the formula for

a semi-ellipsoid ((W/2*L/2*H*4/3π)/2). After two weeks, when tumors were

palpable, mice were randomized into the following treatment groups: Placebo

(n=8), EB1089 (Leo Pharmaceuticals, Ballerup, Denmark; n=8), and UV light

(n=4). EB1089-treated animals were injected intraperitoneally (ip) with 45 pmoles EB1089 in a total volume of 20μL three times weekly for six weeks.

Placebo controls received vehicle injections on the same schedule. UV light treated mice were exposed for ten minutes, three times weekly, for six weeks, utilizing cage-top lamps designed for reptile aquaria (Big Apple Herpetological,

Hauppauge, NY). UV light exposure was calculated to be approximately

100uW/cm2. Two hours prior to sacrifice, selected animals were injected sc with

1mg bromodeoxyuridine (BrdU) in 0.9% saline. Animals were sacrificed via CO2 asphyxiation and cervical dislocation. Tumors were removed, weighed and measured, and fixed in 4% formalin for histological analysis. Blood was collected via cardiac puncture. Skin biopsies were removed from the mid-dorsal region, and fixed in 4% formalin. All procedures were performed following institutional regulations regarding the care and use of laboratory animals.

5.2.3 Serum Analysis

Blood was fractionated using serum separator tubes (BD Biosciences,

San Jose, CA), and serum stored at -80C. Serum calcium was assayed using the QuantiChrom Calcium Assay Kit (BioAssay Systems, Hayward, CA) according to manufacturer’s recommendations. Serum 25D and 1,25D levels were measured using the OCTEIA 25-hydroxyvitamin D kit and OCTEIA 1,25- dihydroxyvitamin D3 kit (Immunodiagnostic Systems Ltd., Boldon, UK) according

to manufacturer’s recommendations. 79

5.2.4 Histological Analysis

Formalin fixed tumor and skin sections were dehydrated through a graded series of ethanols, paraffin embedded, and sectioned at 5μ. Sections were stained using hematoxylin and eosin (H&E). Sections from animals injected with

BrdU prior to sacrifice were analyzed with a BrdU staining Kit (Zymed

Laboratories, South San Francisco, CA) according to manufacturer’s recommendations. TUNEL staining was performed using the In Situ Cell Death

Detection Kit, POD (Roche Applied Science, Penzberg, Germany) according to manufacturer’s recommendations.

5.2.5 Blood Vessel Density

Vessel density was calculated on H&E stained tumor sections. The number of vessels per field at 20X magnification was counted in four random fields per tumor section, and averaged. Slides were viewed using brightfield microscopy with an Olympus AX70 microscope equipped with a Spot-RT digital camera.

5.2.6 Quantitative RT-PCR

Tumor tissue (100mg) was homogenized in Trizol (Invitrogen, Carlsbad,

CA), according to manufacturer’s instructions. Three cDNA replicates were made for all RNA, using the TaqMan Reverse Transcription Reagents kit (Applied

Biosystems, Foster City, CA). Gene expression analysis was performed using

SYBR green (ABGene, Rochester, NY), and values were normalized against 18S

RNA. Primer sequences are as follows: Mus CYP24 Forward –

AAGTCATGGACTTGGCCTTCA; Mus CYP24 Reverse –

GCTCCGCCTTCTCGTTGA; Mus CYP27B1 Forward –

CAGAGCGCTGTAGTTTCTCATCA; Mus CYP27B1 Reverse –

CGTTAGCAATCCGCAAGCA; 18S rRNA Forward -

AGTCCCTGCCCTTTGTACACA; 18S rRNA Reverse -

GTTCCGAGGGCCTCACTAAAC . Plates were run in triplicate on the ABI-Prism

7700 (Applied Biosystems, Foster City, CA), and mean values of representative results are shown.

5.2.7 Statistical Analysis

Data are expressed as mean ± standard error. One way ANOVA or

Students t-tests were performed, as appropriate, using GraphPad software.

Means were considered statistically significant when p-values less than 0.05 were obtained.

5.3 Results

5.3.1 EB1089 inhibits growth of murine xenografts via tumor cell VDR.

The effect of chronic EB1089 treatment on growth of established xenografts derived from murine mammary tumor cells was examined over a five week period. Tumor volumes were assessed twice weekly via caliper measurement, and the fold change in tumor size over the final two weeks of the study was calculated (Figure 5.1). The growth rate of tumors derived from

WT145 cells, which express functional VDR, was significantly reduced in EB1089 treated mice compared to the placebo group (p<0.05). In contrast, EB1089 treatment did not affect the growth of tumors derived from KO240 cells, which were derived from a VDR knockout mouse and are unresponsive to EB1089 in vitro (Valrance et al., 2007; Valrance and Welsh, 2004).

The overall health of the tumor bearing mice was assessed by monitoring body weight and serum calcium at the end of the study. Corrected animal weight

(total weight minus tumor weight) was calculated to control for the differences in tumor weight. No differences in body weight were observed between placebo and EB1089 treated mice or between mice bearing WT145 or KO240 tumors

(data not shown). EB1089 did not significantly elevate serum calcium as measured 6 or more hours after dosing (Figure 5.2). Serum 1,25D levels were significantly lower in EB1089 treated animals (data not shown), indicating that

EB1089 effectively suppressed the formation of endogenous 1,25D as expected.

WT145 KO240 3 a a a 2 b

(fold change) 1 Tumor Volume

0 - + - + EB1089

Figure 5.1. EB1089 inhibits growth of xenografted VDR positive tumors. Nude mice bearing murine mammary xenografts derived from VDR positive (WT145) or VDR negative (KO240) cell lines were treated with 45 pmole EB1089 three times weekly for six weeks. Fold change in tumor volume was calculated for the last two weeks of the study. Bars represent mean ± standard error of 6 or more mice per group. Bars with different letters are statistically significant (p<0.05).

5.3.2 Angiogenesis in mammary tumors is not inhibited by EB1089.

Previous data (Bernardi et al., 2002; Blutt et al., 2000; Mantell et al., 2000) has suggested that the anti-tumor activities of vitamin D analogs such as EB1089 may be through inhibition of tumor angiogenesis. To assess the potential contribution of anti-angiogenic effects in our model system, blood vessel density was quantitated in tumor sections (Figure 5.3). The average number of blood vessels per field was similar in tumors consisting of WT145 and KO240 cells, and no significant effects of EB1089 treatment on vessel density were observed in either tumor type.

Serum CalciumSerum (mg/dL) 0 Con EB1089

Figure 5.2. Serum calcium is not altered by EB1089 treatment. Serum total calcium was measured by colorimetric assay in mice after six weeks placebo or EB1089 treatment. Bars represent mean ± standard error of 6 or more mice per group.

20 WT145 KO240

(number/field) Vessel Density

0 - + - + EB1089

Figure 5.3. Tumor vessel density is not altered by VDR or EB1089 treatment. Blood vessel density was quantitated on sections of xenografts derived from VDR positive (WT145) and VDR negative (KO240) tumor cells. Four random fields of view for each mouse were counted at 20X magnification. Bars represent mean ± standard error of 6 or more mice per group.

5.3.3 EB1089 induces growth arrest and apoptosis in VDR-positive mammary

tumor cells

BrdU uptake was assessed on tumor sections as a measure of active cell proliferation (Figure 5.4). The average percentage of BrdU positive cells in control mice bearing tumors derived from WT145 cells was approximately 17%

(Figure 5.4a), and staining was restricted to the epithelial cells (Figure 5.4b). The percentage of BrdU positive cells was significantly decreased to less than 5% in

WT145 tumors derived from mice treated with EB1089 (p<0.05). BrdU incorporation in KO240 tumors was also restricted to epithelial cells, and was not significantly different from that in WT145 tumors. No effect of EB1089 treatment on the percentage of BrdU positive cells was observed in tumors derived from

KO240 cells.

TUNEL assays were used to quantitate apoptotic index in tumors as a function of treatment and cell type. Figure 5.5a shows the average percentage of

TUNEL positive cells per tumor (apoptotic index), and Figure 5.5b shows representative images from each group. Basal apoptotic index in WT145 and

KO240 tumors was similar (approximately 2%), and TUNEL positivity was restricted to the tumor cells. In WT145 tumors, apoptotic index increased approximately 2.5 fold (p<0.05) following EB1089 treatment. No effect of

EB1089 treatment on apoptotic index of KO240 tumors was detected.

a WT145 KO240

a 20 a

10 b BrdU Positive (%)

0 - + - + EB1089

b Con EB1089

WT145

KO240

Figure 5.4. EB1089 inhibits tumor cell proliferation through VDR dependent mechanisms. a) The percentage of cells positive for the proliferation marker BrdU was assessed on sections of xenografts derived from VDR positive (WT145) and VDR negative (KO240) tumor cells. Counts were made on four random fields of view for each mouse at 20X magnification. Bars represent mean ± standard error of 6 or more mice per group. Bars with different letters are statistically significant (p<0.05). Panel b shows representative histological sections for each group.

a WT145 KO240 8

a a a TUNEL Positive (%)

0 - + - + EB1089

b Con EB1089

WT145

KO240

Figure 5.5. EB1089 induces apoptosis of tumor cells that express VDR. a) The percentage of TUNEL positive cells was assessed on sections of xenografts derived from VDR positive (WT145) and VDR negative (KO240) tumor cells. Counts were made on four random fields of view for each mouse at 20X magnification. Bars represent mean ± standard error of 6 or more mice per group. Bars with different letters are statistically significant (p<0.05). Panel b shows representative histological sections for each group.

5.3.4 Chronic UV treatment elevates vitamin D status and inhibits growth of

VDR positive xenografts.

To assess whether endogenously produced vitamin D could elicit anti- tumor activities comparable to those of EB1089, a subset of mice bearing WT145 and KO240 tumors were exposed to UV light three times weekly for six weeks.

Chronic UV exposure did not affect final body weight in either group of tumor bearing mice (data not shown). As reported in Figure 5.6a, chronic UV treatment decreased the size of WT145 tumors, from approximately 2000 mm3 to

approximately 1200 mm3 (p<0.05). In contrast, KO240 tumor size was not

significantly different from untreated WT145 tumors, and was not significantly

affected by UV light treatment.

Proliferation was measured by BrdU staining in tumor sections from all

four treatment groups (Figure 5.6b). The average percentage of BrdU positive

epithelial cells in WT145 tumors was significantly decreased following chronic UV

light treatment, from approximately 17% to approximately 2% (p<0.05). No

significant effect was observed in KO240 tumors. Apoptosis was measured via

TUNEL analysis in tumor sections (Figure 5.6c). A low basal level of apoptosis

was observed in WT145 tumors from untreated mouse hosts, and this level was

increased approximately 3 fold (p< .05), in response to UV light treatment.

KO240 tumors also had a low basal level of apoptosis, but this was not affected

by chronic UV light exposure.

a 4000 WT145 KO240

) 3 a 3000 a a 2000 b

1000 Tumor Volume (mm Volume Tumor

0 - + - + UV

b c WT145 KO240 WT145 KO240 a 20 10 b

10 c 5 a c a BrdU Positive (%) Positive BrdU a b TUNEL Positive (%)

0 0 - + - + UV - + - + UV

Figure 5.6. Chronic UV exposure inhibits proliferation and induces apoptosis in mammary xenografts through VDR dependent mechanisms. a) Nude mice bearing murine mammary xenografts derived from VDR positive (WT145) or VDR negative (KO240) cell lines were chronically exposed to low dose UV radiation three times weekly for six weeks. Bars represent mean ± standard error of tumor volume calculated at the end of the study. The percentage of BrdU (b) and TUNEL (c) positive cells was assessed on sections of xenografts derived from VDR positive (WT145) and VDR negative (KO240) tumor cells as described in Figures 4 and 5. Bars represent mean ± standard error of 6 or more mice per group. Bars with different letters are statistically significant (p<0.05).

The effect of UV treatment on vitamin D status was assessed by measurement of 25D, the major circulating form of vitamin D which increases in proportion to epidermal synthesis or dietary intake. Figure 5.7a shows that circulating 25D was significantly elevated in response to chronic UV treatment, with levels approximately double that of non-exposed controls (p<0.05).

Elevation in serum 25D was not associated with increases in either circulating

1,25D (Figure 5.7b) or calcium (Figure 5.7c), suggesting the possibility that the anti-tumor effects of UV might be secondary to the conversion of 25D to 1,25D within the tumor tissue. In support of this concept, we detected CYP27B1, the

25D-1α-hydroxylase enzyme, in WT145 and KO240 cells (Figure 5.8a) .and tumors (Figure 5.8b). No significant effects of 1,25D treatment or UV exposure on CYP27B1 expression were detected. CYP24, a well characterized VDR target gene, was examined in tumor tissue as a biomarker of vitamin D action. As demonstrated in Figure 8C, chronic UV exposure significantly elevated CYP24 gene expression in WT tumors.

5.3.5 Chronic UV exposure increases epidermal thickness and basal cell

proliferation.

Although no erythema was observed in UV exposed mice, histological sections of skin taken from the mid-dorsal region were examined to determine whether the UV treatment regimen adversely affected epidermal cell populations

(Figure 5.9). H&E staining to examine the overall architecture of the skin suggested thickening of the epidermal layer in response to UV exposure, and this was confirmed by direct measurement. Quantitation of BrdU incorporation on

300

* 200

100 Serum 25D(nmol/L)

0 Con UV

b c 300 10

200

5 100 Serum 1,25D(pmol/L)

(mg/dL) Calcium Serum 0 0 Con UV Con UV

Figure 5.7. Chronic UV exposure increases serum 25D without affecting serum 1,25D or serum calcium. Serum 25D (a) and serum 1,25D (b) were measured by ELISA after six weeks of UV exposure. c. Total calcium was measured by colorimetric assay in the same animals. Graphs are mean ± standard error of 6 or more mice per group. Bars are statistically significant (p<0.05).

a WT145 KO240 2

ExpressionCYP27B1 0 - + - + 1,25D

b c

3 3 *

2 2

1 1 CYP24 Expression CYP27B1 ExpressionCYP27B1

0 0 Con UV Con UV

Figure 5.8. Expression of vitamin D metabolizing enzymes in cells and tumors. (a) CYP27B1 mRNA expression in WT145 and KO240 cells was measured via SYBR green analysis following 1,25D treatment. (b) CYP27B1 mRNA expression in WT145 tumors was measured via SYBR green analysis after six weeks of UV exposure. (c) CYP24 mRNA expression in WT145 tumors was measured by SYBR green analysis after six weeks of UV exposure. Bars are statistically significant (p<0.05).

a a Con UV

b Con UV Thickness (μ) 5.8 ± 0.5a 25.8 ± 3.0b BrdU Positive (%) 3.3 ± 0.7a 7.7 ± 1.7b

TUNEL Positive (%) 32.0 ± 4.7 25.2 ± 3.0

Figure 5.9. UV treatment induces epidermal hyperplasia. After six weeks of UV exposure, dorsal skin biopsies were processed for H&E analysis, and representative figures are shown in A. Quantitative data is summarized in B. Skin thickness was measured on H&E sections, and BrdU and TUNEL assays were quantitated as described in Figures 4 and 5. a,bData with different letters are statistically significant (p<0.05).

skin sections indicated that UV treatment increased the percentage of proliferating cells in the epidermis without altering the rate of apoptosis as measured by TUNEL staining.

5.4 Discussion

In these studies, we have exploited a novel model system comprised of mammary tumor cell lines with differential VDR expression to distinguish the effects of VDR agonists that are mediated directly in tumor cells versus those mediated systemically or indirectly via stromal-epithelial interactions or through inhibition of angiogenesis. Furthermore, we have manipulated the vitamin D pathway via both exogenous (treatment with the analog drug EB1089) and endogenous (treatment with chronic UV light) approaches, and shown that in both cases, the anti-tumor effects of 1,25D absolutely require VDR expression in the tumor epithelial cell.

Although previous studies (Blutt et al., 2000; Milliken et al., 2005; Zhang et al., 2005) have consistently demonstrated anti-tumor effects of EB1089, the precise mechanism of action of this analog is not known. EB1089 binds the VDR with lower affinity than 1,25D (Carlberg et al., 1994) and does not inhibit growth of tumor cells lacking VDR in vitro (Valrance and Welsh, 2004). However, in vivo evidence suggests that the anti-tumor effects of EB1089 may be mediated, at least partially, by direct effects on endothelial cells to inhibit angiogenesis

(Bernardi et al., 2002; Mantell et al., 2000). In these studies, we created heterogeneous xenografts containing VDR positive stromal and endothelial cells

(originating from the host mouse), and either VDR positive (WT145) or VDR negative (KO240) epithelial tumor cells (inoculated subcutaneously). With this system, we have clearly demonstrated that EB1089 exhibits anti-tumor effects in vivo, but only in xenografts containing WT145 cells (VDR positive). Thus, the effects of EB1089 are not mediated extra-tumorally, and absolutely require expression of the VDR in the tumor epithelial cells. Histological analysis confirmed that both WT145 and KO240 tumors were comparably infiltrated with

VDR positive stromal and endothelial cells (not shown). Thus, if any effects of

EB1089 were indirectly mediated, for example via disruption of stromal cell growth factor production, we would anticipate some growth inhibition in KO240 tumors in response to EB1089, but no significant growth inhibition was observed.

We also found no evidence for direct effects of EB1089 on angiogenesis, which would have been mediated through the VDR positive endothelial cells present in both WT145 and KO240 xenografts. These data suggest that the anti- proliferative effect of EB1089 on mammary tumors is mediated directly via tumor

R>Kirkland, J. B.Meckling-Gill, K.

These are the first studies to demonstrate proof of principle that enhancement of the endogenous vitamin D synthesis pathway via UV exposure can exert anti-tumor effects. A protective effect of sunlight exposure against a variety of human cancers, including breast, prostate, and melanoma, has been

suggested by epidemiological studies (Gorham et al., 1990; Grant, 2003; Kricker and Armstrong, 2006), but the role of UV mediated vitamin D synthesis has yet to be experimentally tested. Because UV exposure exerts pleiotropic effects on biological systems (such as immunosuppression, tissue hyperplasia and generation of photoproducts), it has been difficult to specifically distinguish the subset of UV effects which are mediated secondary to vitamin D synthesis. We therefore exploited the xenograft system described above to determine whether tumors derived from VDR positive epithelial cells (WT145) would respond differently to UV exposure than tumors derived from VDRKO epithelial cells

(KO240). Our data indicate that chronic UV exposure elevated circulating 25D and induced the VDR target gene CYP24 in association with growth inhibition of

VDR positive tumor xenografts. No effects of UV on growth of xenografts lacking

VDR were observed. Although unlikely, it is formally possibly that WT and

VDRKO cells differ in their responsiveness to some other UV mediated effects unrelated to vitamin D status; further in vivo studies with VDRKO cells stably expressing VDR (in which vitamin D mediated growth inhibition has been reconstituted) would be necessary to completely rule out this possibility.

Our findings also suggest that higher vitamin D status in UV exposed mice translated to growth inhibitory effects on VDR positive tumor cells via local generation of 1,25D within mammary tumors. In support of this concept, we detected CYP27B1 gene expression in WT145 and KO240 cells and xenografts.

Of particular interest, the increase in 25D associated with chronic UV exposure did not elevate serum calcium, indicating that the anti-tumor effects are not

related to changes in calcium homeostasis. Thus, it is tempting to speculate that elevating 25D via UV exposure could reduce breast cancer risk via autocrine generation of 1,25D in the mammary gland. This notion is supported by our recent demonstration that normal human mammary epithelial cells express

CYP27B1 and are growth inhibited by 25D.

Our results demonstrating beneficial effects of UV exposure on breast tumors should be interpreted with caution, since UV is an established risk factor for the development of skin cancer (Gallagher and Lee, 2006). In our study, skin of the host nude mice exhibited enhanced proliferation and epidermal thickening in response to the UV regimen. While this clearly demonstrates that our UV exposure conditions were sufficient to impact on skin, and support the notion that the elevated serum 25D is derived from conversion of epidermal precursors to cholecalciferol, further studies will be necessary to determine whether optimum vitamin D status can be induced by chronic UV exposure in the absence of skin pathology. Alternatively, it is possible that a high dose topical or oral vitamin D regimen could be designed to mimic the anti-tumor effects of UV without epidermal hyperplasia or calcemic toxicity.

In summary, we have shown that enhancement of vitamin D signaling via either endogenous or exogenous pathways elicits anti-proliferative and pro- apoptotic activity in mammary xenografts, but only in tumors composed of VDR expressing epithelial cells. No evidence for inhibitory effects of EB1089 on accessory cells or angiogenesis was obtained. The comparable effects of

EB1089 and UV treatments on proliferation and apoptosis in WT145 tumor cells,

and lack of effect in KO240 tumor cells, suggest that similar VDR dependent pathways are activated by these distinct endogenous and exogenous regimens.

This study is the first, to our knowledge, to demonstrate that 25D generated via

UV exposure exerts VDR-dependent anti-cancer effects, and that these effects are comparable to those mediated by a synthetic vitamin D analog. These results provide support to the hypothesis that the observed correlations between sunlight exposure and cancer risk are related to the vitamin D endocrine system.

CHAPTER 6:

CREATION OF VECTORS SUITABLE FOR STABLE EXPRESSION OF WILD-

TYPE AND MUTANT VDRS IN MURINE MAMMARY CELLS

6.1 Introduction

The genomic signaling mechanisms of the VDR are well-characterized and well-understood. It is also well-accepted that 1,25D, through the VDR, causes growth arrest, apoptosis, and differentiation in a variety of cancer cell types. The mechanisms by which the VDR mediates these effects, however, are less well understood. The observation that 1,25D causes some effects too rapidly to require transcription (Fleet, 2004), and that these effects are abrogated in VDRKO mice (Nguyen et al., 2004; Zanello and Norman, 2004), clearly suggests that the VDR has non-transcriptional modes of action. While many of these rapid effects involve calcium mobilization (Fleet, 2004), rapid activation of the mitogen-activated protein kinase (MAPK) pathway by 1,25D through the VDR has also been described (Song et al., 1998). Thus, it is tempting to speculate that the VDR impacts on cell growth and differentiation via non-genomic mechanisms. In order to study this possibility, our goal was to create a model system comprised of VDR naïve cells stably expressing wild-type and mutant

VDRs, in an attempt to determine which functional domains of the VDR were required for the anti-cancer actions of the VDR.

Inherited mutations of the VDR, seen in patients with HVDRR-2, represent a unique means for studying the functionality of the VDR protein in a variety of biological processes. Patients with HVDRR present clinically with severe symptoms of rickets, hypocalcemia, and elevated parathyroid hormone and

1,25D serum levels (Holick, 2006), and it is now known that mutations in the VDR are responsible for HVDRR-2 (Malloy et al., 1999). Mutated VDRs discovered in

HVDRR-2 patients lack the ability to regulate calcium homeostasis and skeletal health, and have been previously characterized for DNA and ligand binding abilities, making them extremely useful for experimental manipulation and testing. For the purposes of our study, we have selected three previously characterized HVDRR-2 mutations: G46D, R274L, and W286R.

The G46D mutation occurs in the DNA binding domain of the VDR, at the base of the first zinc finger. It was discovered in a two-year-old Saudi Arabian boy, who presented with alopecia, severe rickets, hypocalcemia, hypophosphatemia, and elevated serum 1,25 levels (Lin et al., 1996). In vitro, the G46D mutant receptor was found to bind 1,25D normally, with an affinity comparable to that of the wild-type receptor. However, the mutant receptor was unable to mediate transcription of the osteocalcin promoter, even following

100nM 1,25D treatment, and exhibited greatly reduced affinity for DNA. The glycine residue at position 46 is not absolutely conserved across steroid hormone receptor superfamily members, but is conserved across all nuclear receptors which heterodimerize with the RXR (Lin et al., 1996).

100

R274L and W286R are both residues in the ligand binding domain of the

VDR. The R274L mutation was discovered in a young Middle Eastern boy with bowed legs (rickets), elevated PTH, and high levels of serum 1,25D (Kristjansson et al., 1993). In vitro, the R274L receptor did not induce the CYP24 promoter following treatment of 1,25D as high as 1μM, but did induce transcriptional activation of the osteocalcin promoter, at doses of 1,25D of 50nM or above. It was calculated, therefore, that the R274L mutation decreases the affinity of the

VDR for 1,25D by approximately 1000-fold (Kristjansson et al., 1993). The crystal structure for the VDR bound to 1,25D showed that the arginine at position

274 is the contact point for the 1-hydroxyl group of 1,25D, and that mutation of this arginine to leucine reduces 1,25D binding affinity through altering this contact point (Rochel et al., 2000).

The W286R mutant receptor was discovered in two Algerian siblings, who each suffered from rickets with hypocalcemia and elevated serum PTH, but no evidence of alopecia (Nguyen et al., 2002). It is interesting to note that W286 is the only tryptophan residue in the VDR, and it is highly conserved across the nuclear receptor superfamily. W286 also makes direct contact with 1,25D, and is involved in the formation of a required hydrophobic channel in the ligand binding domain (Rochel et al., 2000). The VDR protein translated from the W286R mutant was the same size as wild-type VDR (50kDa), and was localized in both the nucleus and the cytoplasm by immunocytochemistry (Nguyen et al., 2002).

The W286R VDR did not bind 1,25D, and did not upregulate transcription of the

CYP24 promoter.

101

All three of these VDR mutations were associated with rickets, hypocalcemia, and other alterations in calcium homeostatic signaling. The G46D

DNA binding mutant binds ligand with normal affinity, but lacks a transcriptional response to 1,25D treatment, and patients with this mutation exhibit alopecia.

Interestingly, both ligand binding mutants also lack 1,25D-mediated transcriptional activity, but are not associated with alopecia in affected patients, which suggests that the repressive functions of VDR in the hair follicle are intact in these patients (Skorija et al., 2005). Because all of these VDR mutations generate VDR protein of normal size that appears to be stable, we chose to replicate these mutations for study of the functional domains of the VDR.

102

6.2 Materials and Methods

6.2.1 Insertion of a hygromycin resistance gene into the pSG5-hVDR

expression vector

The pSG5-hVDR human VDR expression vector, obtained as a gift from

Dr. Paul MacDonald (Case Western Reserve University, Cleveland OH), was unsuitable for stable transfection and selection, as it lacked a selection marker.

We therefore utilized the pSG5-hVDR and a commercially available hygromycin resistance cassette-containing vector, pcDNA3.1/Hygro(+), to create a stable expression vector.

PCR was performed on the pcDNA3.1/Hygro(+) vector (Invitrogen,

Carlsbad, CA) using the primer sequences listed below, and the Pfx high fidelity polymerase (Invitrogen):

Forward:

AAAAAAAACCAGACGTCTGGTGTGGAATGTGTGTCAGTTAGGGTGTG

Reverse:

AAAAAAAACCAGACGTATGGAGCTCACTCATTAGGCACCCCAGG

Both of these primers contain an AatII restriction enzyme restriction site within a BstXI restriction enzyme recognition site, for ease of insertion into the pSG5-hVDR vector. Following PCR, reactions were run on a 1% agarose gel,

103

and the bands corresponding to the hygromycin resistance gene and its SV40 origin and polyadenylation (polyA) sequences were excised and purified using the Wizard SV PCR and Gel Purification Kit (Promega, Madison, WI).

The hygromycin insert DNA was digested with BstXI, and the destination pSG5-hVDR vector was digested with AatII. Following digestion, the destination vector was treated with Antarctic Phosphatase (New England Biolabs, city, state). The digested insert and phosphatase treated vector were then ligated overnight. Following ligation, DH5α transformation-competent E. coli were transformed with the ligation mixture, and spread onto Luria broth agar plates containing 100μg/mL ampicillin.

The following day, a direct PCR on colonies from the ligation plates was performed. Colonies were picked from the agar plate using a 20μL pipettor, and added to PCR mix in each well of a 96-well PCR plate. The hygromycin primers listed above were utilized in each reaction. Once the PCR cycles were complete, reactions were run on a 1% agarose gel containing ethidium bromide. The gel was imaged using the Bio-Rad Gel Doc (Hercules, CA). Colonies corresponding to gel lanes with high expression levels of the hygromycin insert were grown overnight in liquid Luria broth containing 100μg/mL ampicillin.

DNA was purified from bacterial cultures using the Wizard Plus SV miniprep kit (Promega). Purified DNA was digested with EcoRI, in order to determine the orientation of the insert within the vector. The culture which had the banding pattern corresponding to the correct hygromycin orientation was selected for use, and named pSG5-hVDR-hygro.

104

6.2.2 Removal of the VDR coding region to create the pSG5 stable parent

vector

The pSG5-hVDR-hygro vector was used to create the pSG5-hygro negative control vector. This negative control vector utilizes the same backbone, pSG5 (Stratagene, LaJolla, CA), and contains the hygromycin resistance gene, but lacks any VDR expression sequence. Primers were designed against the pSG5-hVDR-hygro vector, and were designed to bind to either side of the VDR coding sequence. Primer sequences are listed below:

Forward: CCCCGGTACCAGATCTTATTAAAGCAGAACTTGTTTATTGC

Reverse: TTTTGGTACCGCCCTATAGTGAGTCGTATTACAATTCTTTG

Both of these primers contain an Acc65I restriction site, flanking the vector sequence. Following PCR, the DNA was phenol-chloroform extracted and ethanol precipitated. Purified DNA was digested with the DpnI restriction enzyme, to remove the input DNA, and then digested with Acc65I, to remove the blunt ends. The DNA was then ligated overnight, electroporated into DH10B electrocompetent E. coli, and spread onto Luria broth agar plates containing

100μg/mL ampicillin.

The following day, colonies were picked from the plate and grown overnight in Luria broth containing 100μg/mL ampicillin. DNA was purified from bacterial cultures using the Wizard Plus SV miniprep kit (Promega). Purified

DNA was digested with the EcoRI restriction enzyme, then run out on a 1%

105

agarose gel containing ethidium bromide. The gel was imaged using the Bio-

Rad Gel Doc. One culture had the banding pattern indicating that it was the negative control vector, and not residual pSG5-hVDR-hygro, so it was selected for use, and named pSG5-hygro.

6.2.3 Site-directed mutagenesis of the pSG5-hVDR-hygro vector

The three mutations (G46D, R274L, W286R) required for this project were created in the pSG5-hVDR-hygro vector by the SeqWright company, according to our instructions. We specified, by nucleotide position within the sequence, which bases were to be mutated, and what base they should be mutated to. A table containing this information is provided (Table 6.1). Once made, the DNA obtained was sequenced, and all point mutations were found to be accurate, with no changes to the rest of the vector sequence.

106

TABLE 6.1.

NUCLEOTIDE POSITIONS AND BASE CHANGES FOR VDR MUTANT

CONSTRUCTS

Nucleotide Nucleotide Position Change

G46D 1295 G -> A R274L 1979 G -> T

W286R 2014 T - >A

The pSG5-hVDR-hygro vector was numbered, beginning at the multiple cloning sequence, and the nucleotide position above corresponds to the exact number placement of each nucleotide within the sequence. G= guanine, A= adenine, T= thymine.

6.2.4 Western Blotting

KO240 cells were seeded at a density of 30,000 cells per well into each well of a 12-well plate. The following day, cells were transfected with 10ng/well of each VDR construct using TransFast transfection reagent (Promega) according to manufacturer’s instructions. Twenty-four hours post-transfection, cells were harvested in 1x Laemmli Sample Buffer containing protease and phosphatase inhibitors, as described in chapter 2. Lysates were sonicated, and protein concentration determined using the BCA protein assay (Pierce

Biotechnology, Rockford, IL). 50μg of protein were run on 10% polyacrylamide gels, transferred to nitrocellulose, Ponceau stained to confirm equal loading, and blocked overnight with 5% skim milk in PBS, containing 0.02% sodium azide. A primary antibody against the VDR (VDR clone D-6, Santa Cruz Biotechnology,

107

Santa Cruz, CA) was utilized at a 1:100 dilution in 5% skim milk in PBS containing 0.1% Tween-20 (Sigma Aldrich, St. Louis, MO), followed by anti- mouse secondary antibody (Amersham Biosciences, Buckinghamshire, UK) at a

1:5000 dilution in 5% skim milk in PBS containing 0.1% Tween-20. The blot was stripped using 0.1% glycerol and 1mM Tris pH 7.5, then re-probed with a primary antibody against actin (Santa Cruz Biotechnology) at a 1:100 dilution in 5% skim milk in PBS containing 0.1% Tween-20, followed by anti-goat secondary antibody

(Santa Cruz Biotechnology) at a 1:5000 dilution in 5% skim milk in PBS containing 0.1% Tween-20.

6.2.5 Transient Transfection Luciferase Reporter Assays

KO240 cells were seeded at a density of 30,000 cells per well into each well of a 12-well plate (Costar, Corning, NY). Twenty-four hours post plating, cells were transfected using TransFast transfection reagent (Promega, Madison,

WI) according to manufacturer’s instructions. A 1:1 ratio of micrograms of DNA to microliters of TransFast was utilized. Ten ng of each VDR construct (or negative control), 1.6μg of 24-hydroxylase VDRE-luciferase promoter plasmid

(Dr. John Omdahl, University of New Mexico, Albuquerque, NM) and 400ng of pRL-TK Renilla luciferase normalization plasmid (Promega) were transfected per well. Twenty-four hours post transfection, cells were treated with either the indicated doses of 1,25D or ethanol vehicle control. Twenty-four hours post treatment, cells were harvested using Passive Lysis Buffer (Promega), and luciferase activities were read using the Dual Luciferase Reporter System

108

(Promega). Results are presented as Relative Luciferase Units (RLU), representing luciferase values normalized to renilla luciferase values.

6.2.6 Statistical Analysis

Data are expressed as mean ± standard error. One way ANOVA was performed using GraphPad software. Means were considered statistically significant when p-values less than 0.05 were obtained.

109

6.3 Results

6.3.1 Creation of a stable expression vector for the human VDR

In order to create a stable VDR expression vector, the hygromycin resistance gene with its promoter and polyA sequence was PCR amplified from within the pcDNA3.1/hygro(+) vector. Four identical PCR reactions were performed, and the DNA was subjected to agarose gel electrophoresis, to check for the appropriate banding pattern. As shown in Figure 6.1, all four PCRs were successful, and all contained bands of the appropriate size, approximately 1600 base pairs.

Figure 6.1. PCR amplification of the hygromycin resistance cassette from the pcDNA3.1/hygro(+) vector. Following PCR amplification, DNA was run out on an agarose gel to ensure that the hygromycin resistance cassette had been successfully PCR amplified from the pcDNA3.1/hygro(+) vector. All four unmarked lanes contain identical PCR reaction mixtures. M = 1Kb Plus Marker Ladder (Invitrogen).

110

Following purification of the DNA from the gel, restriction enzyme digests were performed on the hygromycin resistance cassette and the destination vector, pSG5-hVDR. The hygromycin resistance cassette was digested with

BstXI, and the pSG5-hVDR vector was digested with AatII. DNA from each digestion reaction was run on an agarose gel, to ensure that no contamination had taken place. Figure 6.2 shows an image of the agarose gel, and indicates that both digestions were successful, although some DNA degradation occurred in the BstXI digestion.

The hygromycin resistance cassette was ligated into the destination vector, and transformed into bacteria. After growth on selection agar, colonies that grew were selected for a direct colony screening PCR. Colonies were

BstXI AatII

hygromycin vector

Figure 6.2. Post-digestion of both the destination vector and hygromycin insert. Agarose gel of digested DNA shows the presence of one major band in each lane, with a small amount of degradation in the hygromycin cassette.

111

picked off of the growth agar and placed directly into PCR reaction mix containing the hygromycin cassette-specific primers. Following PCR, DNA from each colony was run on an agarose gel, to check for the presence of the hygromycin resistance cassette, and control for self-ligation of the destination vector. Many of the colonies picked were positive for the hygromycin resistance gene (Figure 6.3), but three (#17, 22, 29) had much higher expression than the others, and were chosen for further analysis.

DNA from each of these colonies was purified, and digested with the

EcoRI restriction enzyme. After digestion, if the hygromycin cassette was ligated into the vector in the forward direction, bands of 3034, 2877, and 2001 base pairs would be expected. If the cassette was ligated into the vector in the reverse

1 2 3 4 5 6 7 8 9 10 - M 11 12 13 14 15 16 17 18 19 20 21 22

23 24 25 26 27 28 29 30 31 32 - M 33 34 35 36 37 38 39 40 41 42 43 44

Figure 6.3. Direct colony PCR for the hygromycin resistance gene insert. Post-ligation colonies were picked and subjected to direct PCR for the presence of the hygromycin insert; bands at approximately 1800 base pairs (arrows) represent insert. M = 1Kb Plus Marker Ladder (Invitrogen).

112

orientation, bands of 3456, 2455, and 2001 base pairs would be expected.

Figure 6.4 shows the results of the digest, and indicates that colony number 29 contains the hygromycin resistance cassette in the correct orientation. This culture was propagated, and the DNA within named pSG5-hVDR-hygro.

6.3.2 Creation of stable negative control vector

A negative control vector, utilizing the pSG5 backbone and containing the hygromycin resistance cassette, was created using the pSG5-hVDR-hygro vector. PCR primers were designed to bind the pSG5-hVDR-hygro vector sequence on either side of the hVDR coding region. Successful PCR would,

Colonies M 1729 22

Figure 6.4. EcoRI digest to determine the orientation of the insert within the destination vector. The resulting banding patterns indicated that colonies #17 and 22 had the insert in the reverse orientation while #29 was in the correct orientation. M = 1Kb Plus Marker Ladder (Invitrogen).

113

therefore, amplify the pSG5 backbone and inserted hygromycin resistance cassette, while not amplifying the hVDR sequence. The PCR reactions were run out on an agarose gel, to check for the presence of the appropriate band (Figure

6.5). All three reaction lanes show the presence of the appropriately sized band

(lower band), but also a band of a larger size (upper band).

The DNA was digested with DpnI restriction enzyme, in order to degrade any residual input DNA, and then ligated. Ligated DNA was transformed into E. coli, and colonies which grew on selection agar were picked from the plate. DNA from these colonies was purified and digested with EcoRI. A banding pattern with three bands indicated input DNA (pSG5-hVDR-hygro) carryover, whereas

Figure 6.5. Post-PCR gel for the stable negative control vector. The upper band (arrow) represents successful PCR of the pSG5- hVDR-hygro vector from either side of the hVDR coding region. M = 1Kb Plus Marker Ladder (Invitrogen).

114

one single band at approximately 5000 base pairs indicated successful PCR of the negative control. Figure 6.6 shows the results of the digestion, and indicates that one colony, number 17, contained the negative control vector. This culture was propagated and renamed pSG5-hygro.

6.3.3 Creation of stable expression vectors for mutated human VDRs

Site-directed mutagenesis reactions for the three mutations of interest were performed off-site, by the SeqWright Corporation (Houston, TX). The pSG5-hVDR-hygro vector was used as the parent vector, and mutations were performed according to our specifications. All mutation reactions performed at

SeqWright were sequenced upon completion, to ensure that only the desired mutation had been induced. A comparison of the wild-type and mutant VDR sequences is shown in Table 6.2.

TABLE 6.2.

A COMPARISON OF THE WILD-TYPE VDR SEQUENCE WITH THE POINT

MUTATIONS SELECTIVELY CREATED

Wild-type Mutated G46D TGC AAA GGC TTC TTC TGC AAA GAC TTC TTC R274L ATG TTG CGC TCC AAT ATG TTG CTC TCC AAT W286R ATG TCC TGG ACC TGT ATG TCC AGG ACC TGT

115

6.3.4 Protein expression and transcriptional activity of created vectors

To determine whether the newly created vectors would direct VDR protein expression, each vector was transiently transfected into KO240 cells in culture.

Cell lysates were subjected to western blotting with an antibody that recognizes

VDR at 48kDa. As demonstrated in Figure 6.6, all of the constructs directed expression of a 48kDa protein that was recognized by the VDR antibody. While pSG5-hVDR-hygro (“hVDR”) protein expression appeared higher than any of the mutant constructs (Figure 6.6), all of the mutant constructs were expressed at easily detectable levels. Figure 6.6 also shows an actin loading control, and indicates that the lane containing the hVDR construct contained more protein than any of the lanes in which mutant hVDR constructs were run.

The transcriptional activity of each of the hVDR expression constructs was determined via transient transfection into KO240 cells. A luciferase reporter construct, driven by the promoter of the VDRE-containing CYP24 promoter, was used as a measure of VDR transcriptional activity. Co-transfection of a Renilla

VDR

actin

hVDRG46D F47I R274L W286R

Figure 6.6. Expression of wild-type and mutant hVDR constructs in KO240 cells. Vectors expressing wild-type (hVDR) and mutant (G46D, R274L, and W286R) VDRs were transiently expressed in KO240 cells, and cell lysates were immunoblotted with antibodies against VDR (top) or actin (bottom).

116

luciferase reporter, driven by the thymidine kinase promoter, was used to control for transfection efficiency. KO240 cells transiently transfected with the pSG5- hygro “empty vector” control show no activation of the CYP24 promoter following treatment with 1,25D (Figure 6.7). Transient transfection of the pSG5-hVDR- hygro vector, however, increased CYP24 promoter activity approximately 45 fold over control. No increase in CYP24 luciferase activity was detected in cells transfected with the pSG5-G46D or W286R-hygro vectors, indicating that no functional VDR protein is produced by these vectors. The pSG5-R274L-hygro, however, did induce CYP24 promoter activity approximately 20 fold above control values when transfected into KO240 cells treated with 100nM 1,25D. A dose-response analysis of both pSG5-hVDR-hygro and pSG5-R274L-hygro demonstrated that while the hVDR construct induced CYP24 promoter activity at doses of 1,25D as low as 1pM (Figure 6.8a), the R274L construct did not induce

CYP24 promoter activity at doses of 1,25D lower than 100nM (Figure 6.8b).

117

3 b

RLU c 1

a a a a a a a a 0 - + - + - + - + - + 1,25D EV hVDR G46D R274L W286R

Figure 6.7. Transactivation activity of wild-type and mutant VDRs in KO240 cells. KO240 cells transiently transfected with empty vector (EV), wild-type VDR (hVDR), or various mutant constructs (G46D, R274L, or W286R) were treated with vehicle control or 100nM 1,25D for 24 hours. Data were normalized for transfection efficiency measured by co-transfected pRL-TK, and are expressed as relative luciferase activity (RLU) of the 24- hydroxylase promoter. Bars represent mean ± SEM of six values, and different letters signify statistical difference.

118

a hVDR

100 c c

75 b ab b

50 a a 25

RLU (Fold Increase) 0 1pM 10pM 100pM 1nM 10nM 50nM 100nM

b R274L VDR

30 b

10 a

RLU (Fold Increase) a a a a a 0 1pM 10pM 100pM 1nM 10nM 50nM 100nM

Figure 6.8. Transactivation of CYP24 luciferase in cells expressing wild-type VDR and R274L mutant VDR. KO240 cells transiently transfected with wild-type VDR (a), or R274L mutant VDR (b) were treated with vehicle control or the indicated concentrations of 1,25D for 24 hours. Data were normalized for transfection efficiency measured by co-transfected pRL-TK, and are expressed as relative luciferase units (RLU). Bars represent mean ± SEM of four values, and are expressed as the fold increase over ethanol control treated cells.

119

6.4 Discussion

The aim of the experiments described in this chapter was to create mammalian expression vectors suitable for stable transfection of wild-type human VDR, VDRs with selected point mutations, and a negative control “empty” vector. The wild-type human VDR construct was made first, and served as the basis for all of the other constructs.

Given that our laboratory had already tested the pSG5-hVDR transient expression vector, and found that it restored the vitamin D signaling pathway in

KO240 cells, we felt that it was the best candidate template vector for the proposed studies. The pSG5 backbone, however, is designed for transient transfection, and therefore lacks any form of stable selection cassette. In lieu of excising the hVDR coding region from pSG5-hVDR and moving it to a hygromycin selection vector, we chose instead to move a hygromycin selection cassette into pSG5-hVDR. The reason for this was two-fold: firstly, this would lessen the chances of introducing unwanted point mutations into the hVDR coding region, and secondly, it eliminated the need for further testing of hygromycin selection vectors, as we already had evidence that the pSG5-hVDR was transcribed by, and functional in, the KO240 cell line.

The pcDNA3.1/hygro(+) vector and pSG5-hVDR vectors lacked any common single- or double- cutting restriction enzymes, so primers had to be designed that would allow the hygromycin resistance fragment to be ligated into the pSG5-hVDR vector with a high degree of success. pSG5-hVDR contains a single AatII restriction site, in a backbone area of the vector, well separated from

120

the hVDR coding region, so this was chosen as a destination for ligation. The

PCR primers for the hygromycin cassette were designed with this in mind, adding a BstXI restriction site, containing an AatII restriction site, to each end of the insert. The BstXI enzyme recognizes two, three-base pair sequences, separated by six base pairs. The six base pair spacer can contain any bases, and it was here that the AatII recognition site was added. In this way, the destination vector could be digested with AatII, to create “sticky ends” for ligation, and the insert could be digested with BstXI to create complementary ends. The strategy, while complex, had a high level of success, as shown by the large number of positive colonies in Figure 6.3.

In order to make a negative control stable selection vector, with all of the same properties as pSG5-hVDR-hygro, save for the VDR coding region, the pSG5-hVDR-hygro vector was used as the template for a PCR reaction designed to amplify the entire vector except for the VDR coding region. This amplification destroyed two of the three EcoRI recognition sites in the pSG5-hVDR-hygro vector, so that pSG5-hygro could be easily distinguished, as shown in Figure 6.6.

Finally, the mutant VDR stable expression vectors were created via site- directed mutagenesis of the pSG5-hVDR-hygro vector. Each of the mutations created was chosen based upon its identification as a mutation that causes hereditary vitamin D resistant rickets (HVDRR). The G46D mutation in the DBD, at the base of the first zinc finger region, retains ligand binding, but cannot bind

DNA (Lin et al., 1996; Rut et al., 1994). The R274L and W286R mutations, in the

LBD, can bind DNA, but not 1,25D (Kristjansson et al., 1993; Nguyen et al.,

121

2002). Since these mutations all cause rickets, each resulting VDR protein is clearly deficient in vitamin D-driven calcium homeostatic signaling, even though the mutants chosen are located in different functional domains of the VDR. The three VDR mutations were produced via site-directed mutagenesis at the

SeqWright company, sequenced, and found to contain only the intended point mutation within the VDR coding region.

Once the vectors had been produced, their ability to direct VDR protein expression in the KO240 cell line was tested. Western blotting of lysates from transiently transfected cells indicated that the wild-type and all mutant VDRs could be expressed at the protein level in KO240 cells (Figure 6.9). Furthermore, all VDR proteins are the appropriate molecular weight, and no truncated proteins were detected. Furthermore, these mutations did not appear to alter stability of the protein, as all VDRs were expressed at similar levels, and also approximately equivalent to the wild-type VDR in WT145 cells.

The wild-type VDR protein, when expressed in KO240 cells, is transcriptionally active, as shown in Figure 6.10. The transiently expressed wild- type VDR is capable of initiating transcription of the CYP24 luciferase, following

1pM or higher 1,25D treatment. The G46D and W286R mutants are unable to mediate transcription of the CYP24 promoter following 100nM 1,25D treatment

(Figure 6.10), as expected based on the phenotype of humans with these mutations, who exhibit HVDRR. The R274L mutant is able to mediate transcription of CYP24, but only at concentrations of 1,25D of 100nM or higher

(Figure 6.11). The R274L mutant is approximately 2.5 times less potent than

122

wild-type VDR, and is not responsive to pM concentrations of 1,25D. This correlates with the decreased affinity for 1,25D noted when the mutant was discovered (Kristjansson et al., 1993). Therefore, at physiologically relevant doses of 1,25D, the R274L mutant is not capable of transactivation, which correlates with the exhibition of the HVDRR phenotype in individuals with this mutation.

In summary, we have created a series of mammalian expression vectors, capable of stably expressing WT and selected point-mutant VDR proteins, with the anticipated transcriptional activities. All expressed VDR proteins are of the proper size (50kDa), and while the WT VDR transcriptionally activates CYP24, none of the HVDRR point-mutant VDRs is able to mediate CYP24 transcriptional upregulation; given that these are HVDRR mutations, this finding was expected.

We have also created a negative control stable selection vector, with the same backbone as the VDR expression vectors. In all, these vectors are necessary components for the next objectives of this dissertation, and will be discussed further in the following chapters.

123

CHAPTER 7:

CREATION AND CHARACTERIZATION OF KO240 CELLS STABLY

EXPRESSING WILD-TYPE HUMAN VDR

7.1 Introduction

While it is now well-accepted that 1,25D mediates its growth inhibitory actions in both ER+ and ER- breast cells (Flanagan et al., 2003; Flanagan et al.,

1999), the mechanisms underlying these effects, and the pathways involved, remain unclear. In ER+ cells, transcription of cell cycle inhibitory and pro- apoptotic genes was upregulated following 1,25D treatment, while growth factors and cytokines were downregulated; in ER- cells, few cell cycle regulatory genes were upregulated (Swami et al., 2003). Therefore, while the growth inhibitory effects of 1,25D do not depend on estrogen signaling, they appear to be modulated by the presence of the ER. It is also well-accepted that the growth inhibitory actions of 1,25D target the G1/S phase transition of the cell cycle in most cell lines, however G2/M phase transition arrest has also been documented in leukemic and monocytic cell lines (Godyn et al., 1994; Harrison et al., 1999).

Therefore, the growth inhibitory effects of 1,25D, and the mechanisms through which they act, appear to be cell-type specific. This highlights the need for model systems derived from common lineages in which to study the cell-type-specific mechanisms of 1,25D-VDR signaling. To meet this need, we developed a

124

mammary tumor-specific model system from one well-characterized parental cell line, derived from VDRKO mice. To better understand the mammary-specific growth inhibitory actions of 1,25D mediated by the VDR, VDRKO cells were developed that stably expressed the VDR.

Prior attempts to recapitulate the VDR-mediated growth inhibitory effects of 1,25D in cell lines with low or no detectable VDR protein expression have had mixed results. Stable transfection of VDR into JCA-1 prostate adenocarcinoma cells, which are unresponsive to 1,25D-mediated growth inhibition, increased

CYP24 transcription and sensitized cells to growth inhibition by 1,25D (Hedlund et al., 1996). Reintroduction of VDR in C6 rat glioma cells was found to be necessary, but not sufficient, for induction of growth arrest following 1,25D treatment; all stable cell clones had detectable VDR protein and mRNA, but one clone was not growth inhibited in spite of this expression (Davoust et al., 1998).

Most recently, stable expression of VDR in both ER + and ER- breast cancer cells did not lead to growth inhibition in either cell line following treatment with the vitamin D analog 1α-hydroxyvitamin D5, even though CYP24 gene expression was increased in both cell lines (Peng et al., 2007). Therefore, while recapitulation of VDR-mediated growth arrest is possible via stable transfection of the VDR into cultured cells, this has yet to be successfully accomplished in mammary tumor cell lines. Furthermore, no previous models have been generated on a VDRKO background.

We set out, therefore, to create a model system of mammary tumor cells with differential VDR expression from one parental cell line. Our data with the

125

WT145 and KO240 cell lines demonstrated proof of principle that the VDR is required in many cellular processes, and that transient transfection of the human

VDR into the KO240 cell line could restore 1,25D-mediated transcription (Zinser et al., 2003). Therefore, KO240 cell lines stably expressing human VDR, or empty vector, were generated. The objectives of these studies were two-fold:

• To determine whether the VDR signaling pathway leading to growth inhibition is conserved across species

• To determine whether VDR is necessary and/or sufficient to restore 1,25D-mediated growth inhibition and apoptosis in VDRKO murine mammary cells

126

7.2 Materials and Methods

7.2.1 Cell Culture

WT145 and KO240 cells were maintained in DMEM/F12 medium (Sigma

Aldrich, St. Louis, MO) containing 5% charcoal-stripped fetal bovine serum (CSS)

(HyClone, Logan, UT). For stable selection, KO240 cells were seeded in 12-well plates (Corning, Corning, NY) at a density of 30,000 cells per well. 24 hours later, cells were transfected with 10ng pSG5-hVDR-hygro (described in Chapter

6), using TransFast transfection reagent (Promega, Madison, WI) according to manufacturer’s recommendations. 48 hours post-transfection, cells were passaged using trypsin/EDTA, and two wells were pooled together in T-25 culture flasks (Corning, Corning, NY) containing DMEM/F12 medium with 5%

CSS (Hyclone) and 500μg/mL hygromycin B (HyClone). After 24 hours of selection, medium containing dead cells was removed, and DMEM/F12 with 5%

CSS was added. Cells were grown to 80% confluency, then passaged using trypsin/EDTA. Each T-25 was split into one T-75, containing DMEM/F12 medium with 5% CSS and 500mg/mL hygromycin B. Cells were selected to 20% confluence, then medium containing dead cells was removed and replaced with fresh DMEM/F12 with 5%CSS. Once cells reached 80% confluence, they were again passaged for selection. At the end of three rounds of selection, cells were considered stably transfected. KOEV, KOhVDRC, KOhVDRE, and KOhVDRF were maintained in DMEM/F12 medium (Sigma Aldrich) containing 5% CSS (HyClone)

127

and 500μg/mL hygromycin B (HyClone). All cells were routinely passaged twice weekly using trypsin/EDTA.

7.2.2 Western blotting

Cells were seeded in 150mm dishes (Corning, Corning, NY) at a density of 500,000 cells per dish, and treated 24 hours later with 100nM 1,25D (gift of

Leo Pharmaceuticals, Ballerup, Denmark) or vehicle control. For whole cell lysates, cells were harvested 48 hours post-treatment by scraping into 2x

Laemmli buffer containing protease and phosphatase inhibitors, as described in

Chapter 2. Lysates were sonicated, and protein concentration determined with the BCA protein assay (Pierce Biotechnology, Rockford, IL). 50μg of lysate was separated via SDS-PAGE, transferred to nitrocellulose filters, blocked in skim milk, and incubated overnight with primary antibodies including mouse monoclonal VDR Clone D-6, 1:100 (Santa Cruz Biotechnology, Santa Cruz, CA), mouse monoclonal CYP24, 1:200 (Cytochroma, Markham, Ontario, Canada), and goat polyclonal actin, 1:100 (Santa Cruz Biotechnology). Appropriate horseradish-peroxidase conjugated secondary antibodies (obtained from

Amersham Biosciences, Piscataway, NJ) were incubated with filters for 1 hour.

Specific bands were detected via chemiluminescence (SuperSignal West Dura,

Pierce, Rockford, IL) and exposure to x-ray film. Films were scanned with a flatbed computer scanner.

128

7.2.3 VDR Transactivation Assay

Cells were seeded in 12-well plates (Corning, Corning, NY) at a density of

1x Passive Lysis Buffer and fluorescence was read via the Dual Luciferase system (Promega, Madison, WI).

7.2.4 Crystal Violet Cell Growth Assay and Doubling Time

Cells were seeded in 24-well plates (ICN Biomedicals, Aurora, OH) at a density of 2,000 cells per well and treated 24 hours later with the indicated doses of test compounds. 96 hours post-treatment, cells were fixed using 1% glutaraldehyde in PBS for 20 minutes, and then stained with 0.1% crystal violet dye (Fisher Scientific, Pittsburgh, PA) for 15 minutes. Dye was resuspended in

0.2% Triton-X100 for 15 miutes, and absorbance was measured at 590nm. For doubling time, cells were fixed and stained for five consecutive days, and doubling time calculated according to measured absorbance.

129

7.2.5 Immunofluorescence

Cells were seeded in 2-well chamber slides (Nalge Nunc, Naperville, IL) at a density of 10,000 cells per chamber, and treated 24 hours later with 100nM

1,25D or vehicle control. 24 hours post-treatment, cells were fixed, permeabilized with ice cold methanol and incubated with VDR antibody clone D-6

(Santa Cruz Biotechnology, Santa Cruz, CA), followed by anti-mouse Alexa Fluor

488 (Invitrogen, Carlsbad, CA) secondary antibody. DNA was counterstained with Hoechst 33342 (Invitrogen, Carlsbad, CA). Coverslips were mounted with anti-fade mounting medium, and slides viewed on an Olympus Provis AX70 microscope with a Spot RT Slider digital camera.

7.2.6 Statistical Analysis

130

7.3 Results

7.3.1 VDR protein expression in KOhVDR stable cell lines

Western blotting was used to confirm the expression of VDR in the newly created KOhVDR stable cell lines under steroid-free media conditions. As shown in Figure 7.1, the 50kDa VDR protein was expressed at detectable levels in all three KOhVDR clones, at levels roughly approximate to those seen in WT145 cells.

The level of VDR protein was decreased following treatment with 1,25D in all three cell clones. As expected, the KOEV control stably transfected cell line expressed no detectable VDR protein.

7.3.2 VDR transcriptional activity in KOhVDR stable cells

To assess VDR transcriptional activity, transient transfection assays were conducted with a VDR-responsive luciferase construct that contains the promoter region of the human CYP24 gene. In all three KOhVDR cell clones, CYP24 promoter activity was significantly increased after 1,25D treatment (Figure 7.2).

In the KOEV cell line, basal CYP24 promoter activity level was low and was not induced by 1,25D treatment (Figure 7.2). Western blot analysis (Figure 7.3) indicated that 1,25D also induced the endogenous CYP24 protein in all three clones of the KOhVDR stable cells, at levels approximating those seen in WT145 cells. These data confirm that VDR retains its functional transcriptional signaling capability, both on endogenous and exogenous promoters, when stably transfected into KO240 cells.

131

WT145KO240 KOhVDRC KOhVDRE

VDR

actin

- + - + - + - + 1,25D

WT145 KOhVDRFKOEV

VDR

actin

- + - + - + 1,25D

Figure 7.1. VDR protein is expressed in KOhVDR stable cell lines. Whole cell lysates of WT145, KO240, KOhVDRC, KOhVDRE, KOhVDRF, and KOEV cells treated for 48 hours with 100nM 1,25D or vehicle control were immunoblotted with antibodies directed against VDR (top) and actin (bottom) as a loading control. Blot is representative of two independent whole cell lysate preps for each cell line.

132

75 b

50 b

RLU, Fold Control Fold RLU, a 0 EV hVDR hVDR hVDR KO KO CKO EKO F

Figure 7.2. Transactivation activity of the VDR in KOhVDR stable cell lines. CYP24 reporter gene activity in KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells treated with vehicle control or 100nM 1,25D for 24 hours. Data were normalized for transfection efficiency measured by co-transfected pRL-TK, and are expressed as relative luciferase activity (RLU). Bars represent mean ± SEM of six values, and are expressed as the fold increase over ethanol control treated cells. Bars with different letters are statistically different from control.

7.3.3 Growth inhibition by 1,25D in KOhVDR stable cells

Growth assays were conducted to determine whether the VDR present in

KOhVDR stable cells could mediate growth inhibitory effects of 1,25D. 96 hour treatment with 1,25D inhibited growth of all three KOhVDR cell clones (Figure 7.4); doses of 1,25D as low as 10pM caused growth inhibition of all three cell clones

(Figure 7.5). Under the same conditions, 1,25D had no effect on growth of KOEV cells, up to a concentration of 1μM 1,25D (Figure 7.4). The doubling time of all three KOhVDR cell clones was significantly lengthened by 1,25D: KOhVDRC by over two-fold, KOhVDRE by three-fold, and KOhVDRF by 2.5-fold. KOEV cells were

133

unaffected (Figure 7.6). Thus, stable selection of VDR successfully conferred cellular sensivity to 1,25D mediated growth inhibition, and the stable selection process did not cause sensitivity to 1,25D treatment, as KOEV cells were not affected. VDR in the absence of ligand also had no effect on cell growth, as the doubling times for KOEV and KOhVDR cells under control conditions were not significantly different (Figure 7.6).

WT145 KOhVDRCKOhVDREKOhVDRF CYP24

- + - + - + - + 1,25D

Figure 7.3. Endogenous CYP24 protein is upregulated in KOhVDR stable cell lines following 1,25D treatment. Western blot of CYP24 in WT145, KOhVDRC, KOhVDRE, and KOhVDRF cells after 48 hours of treatment with 100nM 1,25D or vehicle control. Blot is representative of two independent whole cell lysate harvests.

7.3.4 Morphological features of apoptosis in KOhVDR cells, but not KOEV cells,

following 1,25D treatment

To assess whether 1,25D-mediated apoptosis occurs in KO cells stably expressing VDR, we monitored morphological indices of apoptosis in the presence and absence of 1,25D. Phase contrast microscopy shows that while the morphology of KOEV cells is not altered by treatment with 100nM 1,25D, all three KOhVDR cell clones are less densely packed, appear elongated, and less well adhered to the culture surface after 96h treatment with 1,25D (Figure 7.7).

KOhVDR cells treated with 1,25D also exhibit small, dense nuclei, and nuclear irregularity, which are both signs of apoptosis.

134

KOEV KOhVDRC 2 2 a

1 1 b b b b b Absorbance 590nm Absorbance 590nm Absorbance

0 0 EtOH 1 10 100 500 1000 nM 1,25D EtOH 1 10 100 500 1000 nM 1,25D

KOhVDRE KOhVDRF 2 2 a a

1 b 1 b c b b b b c c c Absorbance 590nm Absorbance 590nm

0 0 EtOH 1 10 100 500 1000 nM 1,25D EtOH 1 10 100 500 1000 nM 1,25D

Figure 7.4. KOhVDR cells, but not KOEV cells, are growth inhibited by 1,25D. Crystal violet growth assay in KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells treated for 96 hours with 1,25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values; bars with different letters are statistically different.

135

a a 2 a b c d 1 KOhVDRC

Absorbance 590nm 0 Con 100fM 1pM10pM100pM 1nM 1,25D a 2 a a b c

d 1

KOhVDRE Absorbance 590nm

0 Con 100fM 1pM10pM 100pM 1nM 1,25D

a a 2 ab b c

d 1

hVDR

Absorbance 590nm KO F

0 Con 100fM 1pM10pM 100pM 1nM 1,25D

Figure 7.5. KOhVDR cells are sensitive to concentrations of 1,25D in the picomolar range. Crystal violet growth assay in KOhVDRC, KOhVDRE, and KOhVDRF cells treated for 96 hours with 1,25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values; bars with different letters are statistically different.

136

KOEV KOhVDRC 200 200

82.9* 100 100 Time (h) Time (h) 42.6 41.4 38.2

0 0 Con 1,25D Con 1,25D

KOhVDRE KOhVDRF

200 200 137.5* 106.7*

100 100 Time (h)

Time (h) 44.3 42.5

0 0 Con 1,25D Con 1,25D

Figure 7.6. The doubling time of KOhVDR stable cell lines is increased by 1,25D. Graphs of average doubling times of KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells following 1,25D treatment. Numbers above bars are the average doubling time, in hours, for that treatment. Data points represent mean ± SEM of three independent time course trials, each with at least 8 replicates. Asterisks indicate statistical significance as determined by t-test, 1,25D vs. control for each cell line.

137

PhaseNuclei Phase Nuclei

KOEV

20μ

KOhVDRC

KOhVDRE

KOhVDRF

Con 1,25D

Figure 7.7. 1,25D induces morphological features of apoptosis in KOhVDR cells, but not KOEV cells. Phase contrast and fluorescent images of KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells treated for 4 days with 100nM 1,25D or vehicle control, and incubated with Hoechst dye to visualize nuclear morphology.

138

7.3.5 Growth inhibition of KOhVDR cells by 25D is dependent upon the VDR

Growth was assessed in KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells treated with increasing doses of 25D for 96 hours. As shown in Figure 7.8, 25D inhibited growth of KOhVDRF cells at the physiologically relevant concentration of

100nM, and both KOhVDRC and KOhVDRE cells at 500nM. 25D treatment had no effect on growth of KOEV cells.

KOEV KOhVDRC 1.5 2 a a a a 1.0

1 b 0.5 b Absorbance 590nm Absorbance Absorbance 590nm 0.0 0 Con 1 10 100 500 1000 nM 25D Con 1 10 100 500 1000 nM 25D

KOhVDRF KOhVDRE

2 2 a a a a a a b b

1 b b 1

c c Absorbance 590nm Absorbance 590nm Absorbance

0 0 Con 1 10 100 500 1000 nM 25D Con 1 10 100 500 1000 nM 25D

Figure 7.8. KOhVDR cells are growth inhibited by 25D. Crystal violet growth assay in KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells treated for 96 hours with 25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values. For all graphs, bars with different letters are significantly different.

139

7.3.6 EB1089 mediates growth arrest in KOhVDR cells via VDR-dependent

pathways

We examined whether growth of the KOhVDR stable cells was affected by

EB1089, a synthetic vitamin D analog that is known to exert less calcemic activity than 1,25D in vivo. As shown in Figure 7.9, EB1089 inhibited growth of all three

KOhVDR stable cell clones at concentrations as low as 1nM. In contrast, EB1089 did not inhibit growth of KOEV cells. Therefore, stable selection of VDR was permissive for the growth inhibitory effects of EB1089.

7.3.7 Other synthetic 1,25D analogs induce growth arrest in KOhVDR cell clones

Crystal violet assays were performed using three other synthetic 1,25D analogs - CB1093, MC1288, and KH1230 - also known to be more potent and less calcemic than 1,25D. All three analogs inhibited growth of WT145 cells at concentrations as low as 1nM (Table 7.1), while KOEV cells were unaffected by all three compounds at concentrations as high as 1μM. Each analog exhibited similar potency of growth inhibition in KOhVDRC and F cells, while KH1230 was more potent, and MC1288 less potent, in KOhVDRE cells.

140

KOEV KOhVDRC 0.6 2 a

0.4

1 b b b b b 0.2 Absorbance 590nm Absorbance 590nm

0.0 0 Con 1 10 100 500 1000 nM EB Con 1 10 100 500 1000 nM EB

KOhVDRE KOhVDRF 2 2 a a

1 1 b b b b b b b b b b Absorbance 590nm Absorbance 590nm

0 0 Con 1 10 100 500 1000 nM EB Con 1 10 100 500 1000 nM EB

Figure 7.9. EB1089 induces growth arrest in KOhVDR cells. Crystal violet growth assay in KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells treated for 96h with EB1089 (EB) at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values, and bars with different letters are significantly different.

141

TABLE 7.1.

1,25D ANALOGS INDUCE GROWTH ARREST IN KOHVDR CELLS

KOEV 1nM 1μM KOhVDRC 1nM 1μM CB1093 94.6 92.4 CB1093 47.4* 46.9* MC1288 95.9 91.4 MC1288 45.8* 40.9*

KH1230 92.2 110.1 KH1230 45.9* 45.1*

hVDR hVDR KO E 1nM 1μM KO F 1nM 1μM CB1093 33.4* 35.5* CB1093 35.5* 33.5*

MC1288 37.7* 26.8* MC1288 39.9* 32.8* KH1230 28.3* 23.7* KH1230 34.6* 34.4*

Crystal violet growth assays in KOEV, KOhVDRC, KOhVDRE, and KOhVDRF cells treated for 96h with CB1093, MC1288, or KH1230 at indicated concentrations or vehicle control. Data are expressed as percentage of control absorbance of crystal violet dye, which was set to 100%. Data points represent mean ± SEM of four values. Asterisks signify significant difference from control.

142

7.4 Discussion

In these studies, we created and characterized a new model system consisting of three murine VDRKO cell lines stably expressing the human VDR.

These cell lines, named KOhVDR, were created from the KO240 parental cell line, which was previously shown to be completely resistant to the growth inhibitory and apoptotic effects of 1,25D (Chapters 2-4). Reconstitution of 1,25D-mediated transcriptional activation of CYP24 was achieved via transient transfection of the human VDR into KO240 cells (Zinser et al., 2003), therefore we chose to utilize the human VDR in development of this model system.

All three KOhVDR cell clones express the 50kDa VDR protein, at levels comparable to that of WT145 cells. Each of the clones express VDR at differing levels, with the KOhVDRC having the highest expression. No detectable VDR protein expression was observed in either the KO240 parental cell line, or the

KOEV negative control cell line, as anticipated. Interestingly, each KOhVDR clone exhibited a decrease in VDR protein expression following 100nM 1,25D treatment. As the KO240 cell line was very difficult to stably select, we know that each KOhVDR cell population is heterogeneous; it is possible, therefore, that this decrease in VDR protein level following 1,25D treatment can be attributed to a decreased percentage of cells expressing VDR making up the total cell population following 1,25D treatment. This possibility will require further investigation.

143

We utilized the exogenous CYP24 promoter luciferase construct described in previous chapters to examine the transactivation potential of the human VDR stably expressed in the KO240 cells. All KOhVDR cell lines induced the CYP24 promoter following 100nM 1,25D treatment, at levels greater than those observed in the WT145 cell line - WT145 cells exhibited an approximate 25 fold increase in

CYP24 over control following 1,25D treatment, while the KOhVDR cells exhibited increases of approximately 40-70 fold. The endogenous CYP24 protein was also upregulated in each of the KOhVDR cell clones following 1,25D treatment, indicating that the stably expressed VDRs are transcriptionally competent on both exogenous transfected DNA and endogenous mouse cellular DNA. This suggests that the 1,25D-VDR signaling axis is highly conserved between species, even though differences in VDR protein, nuclear cofactor, and target gene promoters exist.

Growth inhibition in the KOhVDR cells by 1,25D was examined via crystal violet assay. All three KOhVDR cell lines were strongly growth inhibited by 1,25D, and all were sensitive to 1,25D-mediated growth arrest at the physiologically relevant concentration of 10pM 1,25D. The lowest effective concentration of

1,25D that inhibited WT145 cell growth was 100fM, which indicates that the

KOhVDR cell clones are less sensitive to 1,25D than the companion cell line expressing murine VDR. Despite this difference, the KOhVDR cells were still growth inhibited by physiologically achievable levels of 1,25D, however, which validates them as a model system for studying the effects of 1,25D and its underlying mechanisms. The KOEV cells, as expected, were not growth inhibited

144

by 1,25D at doses as high as 1μM. Therefore, even in hygromycin-selected stable cell lines, the growth inhibitory effects of 1,25D require VDR expression.

All three KOhVDR cell cultures also exhibited morphological features of apoptosis following 1,25D treatment, similar to those observed in WT145 cells, while KOEV cells did not. These data suggest that the heterologous VDR is sufficicent to trigger apoptosis.

The doubling time of each of the KOhVDR cell clones was also examined, and found to be comparable to the results obtained in WT145 cells - treatment with 100nM 1,25D significantly slowed the growth of each of the KOhVDR cell clones. The growth of KOEV cells was not altered by 100nM 1,25D treatment, as anticipated. The doubling time of each of the KOhVDR clones under control conditions was not significantly different from that of the KOEV cells, indicating that stable transfection of hVDR in the absence of ligand does not impact on cell growth. The clones, while all growth inhibited, showed slight differences in sensitivity to 1,25D-mediated growth inhibiton, with the KOhVDRE cells most sensitive, and the KOhVDRC cells least sensitive. While this is most likely due to clonal differences within the cell populations, it is interesting to note that the sensitivity of these cells to 1,25D-mediated growth inhibition may be inversely correlated with the degree of VDR expression, given that the KOhVDRC cells had the highest VDR protein expression of the three clones. It has been our experience that transient transfection of KO240 cells with increasing amounts of

VDR leads to peak transactivation activity at 10ng VDR, which steadily decreased when more VDR was transfected in (data not shown). The

145

significance of this observation to growth regulation by 1,25D, if any, will require additional study.

The sensitivity of the KOhVDR cell clones to growth inhibition by other VDR agonists was also examined. While KOhVDR C, E, and F cells were growth inhibited by concentrations of 25D greater than 500nM, they were less sensitive than WT145 cells, which responded to 100nM 25D. The KO240 parental and

KOEV control cell lines did not respond to 25D at doses as high as 1μM, suggesting that the growth inhibition seen in the KOhVDR cell lines is VDR- mediated. This also suggests that CYP27B1 expression and activity is retained in the KOhVDR cell clones, although this has yet to be fully tested. All three

KOhVDR cell clones were also growth inhibited by EB1089, to a magnitude consistent with that observed in WT145 cells. When the analogs CB1093 (CB),

MC1288 (MC), and KH1230 (KH) were tested, it was observed that while all

KOhVDR cells were growth inhibited by doses as low as 1nM, the KOhVDRC cell line were less sensitive to the analogs than the other clones. All three KOhVDR cell lines were more sensitive to CB, MC, and KH than the WT145 cells, which showed an approximate 50% reduction in cell number following treatment.

Overall, these data strongly suggest that, save for the absence of the

VDR, the 1,25D signaling axis in KO240 cells is intact, and can be successfully reconstituted. The data also indicate that VDR signaling pathways are highly conserved between species, as human VDR was transcriptionally active and competent for signaling growth inhibition in murine cells. Taken together, the data obtained with the KOhVDR cell lines also validate the data obtained

146

previously with the WT145 cell line; although there are differences in response between the KOhVDR and WT145 cell lines, the patterns of response are highly similar, and suggest similar pathways and mechanisms are involved. Therefore, the KOhVDR cell lines, along with the KOEV control cell line, comprise a separate model system, derived on a homogenous cellular background, useful for examining the mechanisms of VDR signaling in a variety of cellular pathways and processes.

147

CHAPTER 8:

CREATION AND CHARACTERIZATION OF KO240 CELLS STABLY

EXPRESSING HVDRR-2 POINT-MUTANT VDRS

8.1 Introduction

Both VDR and its ligand, 1,25-dihydroxyvitamin D (1,25D), have been implicated in regulation of proliferation, apoptosis, differentiation and invasion of breast cancer cells, but the specific molecular mechanisms remain undefined.

The development of VDR agonists, such as synthetic vitamin D analogs with enhanced growth regulatory properties and minimal calcemic effects, indicates that the anti-cancer and calcemic actions of VDR can be dissociated. How this dissociation is achieved at the molecular level is as yet unknown, but based on the known modes of action of the VDR, several possible mechanisms exist. The

VDR can interact with a variety of structurally distinct vitamin D response elements (VDREs) in target genes. The direct-repeat with a 3 base pair spacer

(DR3) is the best characterized, and has been identified in the promoters of many 1,25D-inducible genes (Haussler et al., 1998). The CYP24 gene, involved in the catabolism of 1,25D, contains two DR3 elements, and is highly induced by

1,25D. In other VDR target genes, functional VDREs composed of inverted palindromes with 9-base pair spacers (IP9), everted repeats with 6 or 9-base pair spacers (ER6, ER9), and direct repeats with 4 or 6-base pair spacers (DR4,

148

DR6) have been described (Quack and Carlberg, 2000; Saramaki et al., 2006;

Schrader et al., 1997; Thompson et al., 2002; Thummel et al., 2001; Xie and

Bikle, 1997). Evidence exists that VDR may exert distinct functions based upon

VDRE selection; IP9 VDREs has been correlated with induction of apoptosis, and some vitamin D analogs preferentially activate IP9 VDREs (Danielsson et al.,

1997; Nayeri et al., 1995). The VDR is also known to induce some target gene promoters via interaction with other transcription factors at their response elements. For example, 1,25D upregulates the p27 promoter, but this regulation is not dependent upon the presence of a VDRE. Instead, VDR binds to the Sp1 transcription factor, and this complex then binds and activates p27 through Sp1 transcription sites (Huang et al., 2004). The presence of the VDR at the plasma membrane, and the known “rapid responses” of 1,25D suggest alternative, DNA independent mechanisms by which growth inhibition could be achieved (Capiati et al., 2004; Huhtakangas et al., 2004; Rochel and Moras, 2006; Zanello and

Norman, 2004).

Development of VDR modulators as breast cancer therapeutics is dependent on identification of the specific molecular events that underlie the effects of 1,25D on mammary cells. Although the VDR has been identified as an important negative growth regulator, no studies have examined the impact of polymorphic VDRs or mutant VDRs in a cell biology context in general or in breast cancer in particular. This deficit is directly related to the lack of a model system to differentiate between ligand dependent, ligand independent and DNA independent effects of VDR in a cellular context. This project seeks to develop

149

and test a model system for determining whether VDR can exert effects that are independent of DNA binding and/or ligand binding. To accomplish this, we have selected VDR point mutantions, discovered in hereditary vitamin D resistant rickets type 2 (HVDRR-2) patients, which are deficient in either ligand or DNA binding. Because patients bearing these mutations exhibit clinical rickets, these mutations clearly render VDR deficient in calcemic signaling regulation. Thus, we have chosen to stably express these mutant receptors in KO240 cells, to test whether they impact on cell growth in the presence or absence of 1,25D. These

KOmutant stable cell lines comprise a new model system, uniquely suited to study the ligand dependent, ligand independent and DNA independent effects of VDR on breast cancer cell signaling.

150

8.2 Materials and Methods

8.2.1 Cell Culture

WT145 cells were maintained in DMEM/F12 medium (Sigma Aldrich, St.

Louis, MO) containing 5% charcoal-stripped fetal bovine serum (CSS) (HyClone,

Logan, UT). For stable selection, KO240 cells were seeded in 12-well plates

(Corning, Corning, NY) at a density of 30,000 cells per well. 24 hours later, cells were transfected with 10ng pSG5-hVDR-hygro (described in Chapter 6), using

TransFast transfection reagent (Promega, Madison, WI) according to manufacturer’s recommendations. 48 hours post-transfection, cells were passaged using trypsin/EDTA, and two wells were pooled together in T-25 culture flasks (Corning, Corning, NY) containing DMEM/F12 medium with 5%

CSS (Hyclone) and 500μg/mL hygromycin B (HyClone). After 24 hours of selection, medium containing dead cells was removed, and DMEM/F12 with 5%

CSS was added. Cells were grown to 80% confluency, then passaged using trypsin/EDTA. Each T-25 was split into one T-75, containing DMEM/F12 medium with 5% CSS and 500mg/mL hygromycin B. Cells were selected to 20% confluence, then medium containing dead cells was removed and replaced with fresh DMEM/F12 with 5%CSS. Once cells reached 80% confluence, they were again passaged for selection. At the end of three rounds of selection, cells were considered stably transfected. KOEV, KOG46DC, KOR274L2, and KOW286R6 were maintained in DMEM/F12 medium (Sigma Aldrich) containing 5% CSS (HyClone)

151

and 500μg/mL hygromycin B (HyClone). All cells were routinely passaged twice weekly using trypsin/EDTA.

8.2.2 Western blotting

Cells were seeded in 150mm dishes (Corning, Corning, NY) at a density of 500,000 cells per dish, and treated 24 hours later with 100nM 1,25D (gift of

Leo Pharmaceuticals, Ballerup, Denmark) or vehicle control. For whole cell lysates, cells were harvested 48 hours post-treatment by scraping into 2x

Laemmli buffer containing protease and phosphatase inhibitors, as described in chapter 2. Lysates were sonicated, and protein concentration determined using the BCA protein assay (Pierce Biotechnology, Rockford, IL). 50μg of lysate was separated via SDS-PAGE, transferred to nitrocellulose filters, blocked in skim milk, and incubated overnight with the following primary antibodies: mouse monoclonal VDR Clone D-6, 1:100 (Santa Cruz Biotechnology, Santa Cruz, CA) and goat polyclonal actin, 1:100 (Santa Cruz Biotechnology). Appropriate horseradish-peroxidase conjugated secondary antibodies (obtained from

Amersham Biosciences, Piscataway, NJ) were incubated with filters for 1 hour.

Specific bands were detected via chemiluminescence (SuperSignal West Dura,

Pierce, Rockford, IL) and exposure to x-ray film. Films were scanned with a flatbed computer scanner.

8.2.3 VDR Transactivation Assay

Cells were seeded in 12-well plates (Corning, Corning, NY) at a density of

30,000 cells per well. After 24 hours, cells were transfected with the pGL3-24

152

hydroxylase luciferase reporter vector which contains approximately 300bp of the human CYP24 gene promoter with its two DR3 VDRE regions (gift of the late Dr.

1x Passive Lysis Buffer and fluorescence was read via the Dual Luciferase system (Promega, Madison, WI).

8.2.4 Crystal Violet Cell Growth Assay and Doubling Time

8.2.5 Statistical Analysis

153

obtained. Statistical significance is indicated on all data figures as letters or asterisks above bars; bars are labeled with different letters for means that are significantly different by ANOVA, and asterisks signify statistical difference from control by t-test.

154

8.3 Results

8.3.1 VDR protein expression in KOmutant stable cell lines

Western blotting was used to confirm the expression of VDR in the newly created KOmutant stable cell lines under steroid-free media conditions. As shown in Figure 8.1, the 50kDa VDR protein was expressed at detectable levels in

KOG46DC, KOR274L2, and KOW286R6 stable cell lines. No evidence for truncated products was obtained (data not shown). When protein loading is taken into account, no consistent changes in VDR protein expression were evident following treatment with 100nM 1,25D. As previously reported, the KOEV control stably transfected cell line did not express detectable VDR protein.

8.3.2 VDR transcriptional activity in KOhVDR stable cells

To assess transcriptional activity of mutant VDRs, transient transfection assays were conducted with a VDR-responsive luciferase construct that contains the promoter region of the human CYP24 gene. CYP24 promoter activity was not induced in KOW286R6 cells, but was increased approximately 3 to 4-fold in

KOG46DC and KOR274L2 cells following 100nM 1,25D treatment. This 4-fold increase is not statistically significant in comparison with the approximately 40- fold increase induced in KOhVDRF cells, shown for comparison (Table 8.1). At lower, physiologically-relevant doses of 1,25D (1pM-1nM), no CYP24 promoter activation was observed in either KOG46DC or KOR274L2 cells (Table 8.1). Thus, at

155

WT145 KOEV KOG46DC VDR

actin

- + - + - + 1,25D

KOEV KOR274L2KOW286R6

VDR

actin

- + - + - + 1,25D

Figure 8.1. VDR protein is present in KOmutant stable cell lines. Whole cell lysates of WT145, KO240, KOEV, KOG46DC, KOR274L2 and KOW286R6 cells treated for 48 hours with 100nM 1,25D or vehicle control were immunoblotted with antibodies directed against VDR (top) and actin (bottom) as a loading control. Blot is representative of two independent whole cell lysate preps for each cell line. physiological concentrations of 1,25, the selected HVDRR mutant VDRs are transcriptionally inactive in modulation of CYP24, a known 1,25D target gene driven by DR3 promoter elements.

8.3.3 Growth inhibition by 1,25D in KOmutant stable cell lines

Growth assays were conducted to determine whether the HVDRR mutant

VDRs expressed in KOG46DC, KOR274L2, or KOW286R6 stable cells could mediate growth inhibitory effects of 1,25D. 96 hour treatment with 1,25D inhibited growth

156

of the KOG46DC and KOR274L2 cell lines, but not the KOW286R6 cell line (Figure

8.2); KOR274L2 cells were inhibited by doses of 1,25D as low as 500nM (Figure

8.2), whereas KOG46DC cells were inhibited by doses of 1,25D as low as 100pM

(Figure 8.3). Therefore, the DBD mutant, G46D, retains the ability to mediate

1,25D growth inhibition at physiological doses of 1,25D, whereas the LBD mutants are either completely non-responsive to 1,25D (W286R), or only responsive to supra-physiological levels of 1,25D (R274L).

TABLE 8.1.

TRANSACTIVATION ACTIVITY OF THE VDR IN KOMUTANT STABLE CELL

LINES

1pM 1nM 100nM KOG46DC 0.90 ± .10 2.16 ± .91 4.78 ± .15

KOR274L2 1.37 ± .13 1.42 ± .25 3.55 ± .06 KOW286R6 1.02 ± .18 0.95 ± .10 0.94 ± .11 KOhVDRF 1.20 ± .28 19.4 ± 4.2* 41.0 ± 8.9*

CYP24 reporter gene activity in KOG46DC, KOR274L2, and KOW286R6 cells treated with vehicle control or indicated doses of 1,25D for 24 hours. KOhVDRF cells are shown for comparison. Data were normalized for transfection efficiency measured by co-transfected pRL- TK, and are expressed as relative luciferase activity (RLU). Bars represent mean ± SEM of six values, and are expressed as the fold increase over ethanol control treated cells. Asterisks signify statistical difference from corresponding control.

157

KOG46DC

1.5 a b b b b b 1.0

0.5 Absorbance 590nm Absorbance

0.0 Con 1 10 100 500 1000 nM 1,25D

KOR274L2 KOW286R6

0.75 1.0 a a a a

0.50 b b 0.5

0.25 Absorbance 590nm Absorbance 590nm

0.00 0.0 Con 1 10 100 500 1000 nM 1,25D Con 1 10 100 500 1000 nM 1,25D

Figure 8.2. KOG46DC and KOR274L2 cells are growth inhibited by 1,25D. Crystal violet growth assay in KOG46DC, KOR274L2, and KOW286R6 cells treated for 96 hours with 1,25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values; bars with different letters are statistically different.

158

G46D KO C

a a 2 a a b b

Absorbance 590nm

0 Con 100fM 1pM10pM100pM 1nM 1,25D

Figure 8.3. KOG46DC cells are sensitive to concentrations of 1,25D in the upper picomolar range. Crystal violet growth assay in KOG46DC cells treated for 96 hours with 1,25D at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values; bars with different letters are statistically different.

8.3.4 EB1089 mediates growth arrest in KOG46DC and KOR274L2 cells

We examined whether growth of the KOmutant stable cells was affected by

EB1089, a synthetic vitamin D analog that is known to exert less calcemic activity than 1,25D in vivo. As shown in Figure 8.4, EB1089 inhibited growth of KOG46DC cells at a dose as low as 1nM, and inhibited KOR274L2 cell growth at a dose of

500nM. In contrast, EB1089 did not inhibit growth in KOW286R6 cells.

159

KOG46DC 1.5

a 1.0 b b c b b

0.5 Absorbance 590nm

0.0 Con 1 10 100 500 1000 nM EB

KOR274L2 KOW286R6 1.5 1.5

a a a a 1.0 b 1.0 c

0.5 0.5 Absorbance 590nm Absorbance Absorbance 590nm

0.0 0.0 Con 1 10 100 500 1000 nM EB Con 1 10 100 500 1000 nM EB

Figure 8.4. KOG46DC and KOR274L2 cells are growth inhibited by EB1089. Crystal violet growth assay in KOG46DC, KOR274L2, and KOW286R6 cells treated for 96h with EB1089 (EB) at indicated concentrations or vehicle control. Data are expressed as absorbance of crystal violet dye, which is proportional to cell density under the conditions used. Data points represent mean ± SEM of four values, and bars with different letters are significantly different.

8.3.5 Other synthetic 1,25D analogs induce growth arrest and mediate CYP24

transcription in KOmutant cell lines

Crystal violet assays were performed using three other synthetic 1,25D analogs - CB1093, MC1288, and KH1230 - which are also known to be more potent and less calcemic than 1,25D. KOG46DC cells were growth inhibited by all three analogs at 1nM concentrations, with approximately equivalent efficacy

160

(Table 8.2). KOR274L2 cells were also growth inhibited by all three compounds, although each compound exhibited a different growth inhibition profile (Table

8.1). Each analog also exhibited a similar potency of CYP24 transcriptional activation in KOR274L2 cells (Figure 8.7). KOW286R6 cells were also growth inhibited by all three compounds, but only at doses of 1μM (Table 8.1), and none of the analogs caused CYP24 transcriptional activation (Figure 8.5).

TABLE 8.2.

1,25D ANALOGS HAVE DIFFERENTIAL GROWTH EFFECTS IN KOMUTANT

CELLS

KOG46DC 1nM 1μM CB1093 68.3* 64.8*

MC1288 72.8* 66.6* KH1230 70.0* 71.5*

KOR274L2 1nM 1μM KOW286R6 1nM 1μM

CB1093 100.0 44.4* CB1093 95.1 84.8* MC1288 74.5* 46.7* MC1288 95.1 75.8* KH1230 89.4* 51.9* KH1230 96.6 89.0*

Crystal violet growth assays in KOG46DC, KOR274L2, and KOW286R6 cells treated for 96h with CB1093, MC1288, or KH1230 at indicated concentrations or vehicle control. Data are expressed as percentage of control absorbance of crystal violet dye, which was set to 100%. Data points represent mean ± SEM of four values. Asterisks signify significant difference from control.

161

R274L W286R KO 2KO6 30 * *

20 *

Control Fold RLU,

0 CB MC KHCB MC KH

Figure 8.5. VDR LBD mutants are differentially transcriptionally activated by 1,25D analogs. CYP24 reporter gene activity in KOR274L2 and KOW286R6 cells treated with vehicle control or 100nM CB1093, MC1288, or KH1230 for 24 hours. Data were normalized for transfection efficiency measured by co-transfected pRL-TK, and are expressed as relative luciferase activity (RLU). Bars represent mean ± SEM of three values, and are expressed as the fold increase over ethanol control treated cells. Asterisks signify statistical significance from control.

162

8.4 Discussion

The purpose of these studies was to further develop our novel model system to provide a tool for dissecting the impact of specific functional domains of the VDR on breast cancer cell growth. We stably expressed VDRs containing point mutations which abolished the functionality of either the DNA- or ligand- binding domains, in KO240 murine mammary tumor cells. These mutant VDRs were originally characterized in HVDRR-2 patients, who exhibit rickets and end- organ resistance to 1,25D, due to mutations which render their VDRs non- functional for calcemic signaling. The resulting KOmutant cell lines were then characterized in terms of VDR signaling and growth inhibition after treatment with vitamin D steroids.

Cells transfected with the KOG46DC DBD mutant, and the KOR274L2 and

KOW286R6 LBD mutants, all expressed the 50kDa VDR protein, with no obvious alteration of protein expression following 1,25D treatment. The CYP24 transactivation potential of each mutant receptor was shown to be negligible when compared with the WT receptor, stably expressed in the same background

(ie the KOhVDR clones). Both the KOG46DC and KOR274L2 cell lines did upregulate the CYP24 promoter 3- 5-fold following treatment with the supra-physiological concentration of 100nM 1,25D, compared to KOhVDR cells which upregulated

CYP24 almost 20-fold following 1nM 1,25D treatment. These data confirm that both the G46D and R274L mutant VDRs are unable to mediate transcriptional regulation of VDR target genes at physiologically relevant doses of 1,25D. This

163

was anticipated, given that these mutations were identified and characterized from HVDRR-2 patients, who have known defects in the 1,25D-VDR calcemic signaling axis, which results in rickets.

The ability of each of the three mutant VDRs to mediate the growth inhibitory effects of 1,25D was also examined. While the KOW286R6 stable cell line is resistant to 1,25D-mediated growth inhibition at doses as high as 1μM, both the KOG46DC and KOR274L2 cells show partial growth inhibition relative to the response seen in KOhVDR cells (Chapter 7). The KOR274L2 cell line responds to treatment with 500nM 1,25D, which is a supra-physiological concentration. In transient transfection assays, the R274L mutant VDR was shown to activate the osteocalcin promoter at 50nM 1,25D, and was therefore calculated to be approximately 1000-fold less transcriptionally active than WT VDR (Kristjansson et al., 1993). Therefore, the growth inhibition of the KOR274L2 cells by 500nM

1,25D appears to be related to the ability of supra-physiological concentrations of

1,25D to bind the mutant receptor.

Growth inhibition of the KOG46DC cell line was detected at concentrations of 1,25D as low as 100pM, a physiologically achievable concentration. The magnitude of growth inhibition in these cells was approximately 30% of control, across the entire range of 1,25D doses from 100pM to 100nM. In contrast, the magnitude of growth inhibition in KOhVDR cells was dose-dependent, and ranged from 10% at 10pM to over 50% at 1μM 1,25D. Lin et al. (1996) reported that the

G46D mutant VDR was able to bind 1,25D with high affinity, but was not able to mediate transcription of the osteocalcin promoter, even at supra-physiological

164

doses. Our studies indicate that cells stably expressing the G46D mutant VDR were growth inhibited by concentrations of 1,25D that did not activate CYP24 transcription. This, when coupled with the lack of a dose-response, suggests that the growth inhibition of the KOG46DC cell line by 1,25D does not correlate with transcriptional activation of known VDR target genes.

The 1,25D analog EB1089 also showed differential effects on growth of the KOmutant cells, similar to those noted following treatment with 1,25D. The

KOR274L2 cell line was growth inhibited by EB1089 at supra-physiological doses of 500nM or higher, suggesting that the R274L mutant can mediate growth inhibition but only when bound to ligand. The KOW286R6 cell line was unaffected by treatment with EB1089, up to 1μM. The KOG46DC stable cell line responded to

EB1089 at concentrations as low as 1nM, although lower doses were not tested.

The magnitude of growth inhibition was approximately 25% throughout the range of EB1089 concentrations, which is much less than observed in KOhVDR cells, though still significant. This again suggests that the G46D mutant VDR can partially mimic WT VDR in mediating growth inhibition by EB1089.

A panel of 1,25D analogs - CB1093 (CB), MC1288 (MC), and KH1230

(KH) - was tested on the KOmutant cell lines. The KOG46DC cell line was growth inhibited by approximately 30% by all three analogs, at 1nM and 1μM doses.

KOR274L2 was also growth inhibited by all three analogs, but the potency and efficacy of each of the mutants in inducing growth inhibition was different. CB induced a substantial growth inhibition at the 1μM concentration, but 1nM CB was ineffective in KOR274L2 cells; 100nM CB induced the CYP24 promoter in

165

KOR274L2 cells by approximately 13-fold. MC and KH both induced growth inhibition of KOR274L2 cells at both 1nM and 1μM doses, but MC was a more potent inhibitor of KOR274L2 cell growth; both compounds were equipotent in inducing the CYP24 promoter (approximately 22-fold) in KOR274L2 cells.

KOW286R6 cells were growth inhibited by 1mM concentrations of all three analogs, with MC being the most potent; KOW286R6 cells were the least sensitive to the growth inhibition induced by these three analogs, and none of the three analogs induced CYP24 activation in KOW286R6 cells. This suggests that while differences in potency between analogs exist, in general, 1,25D synthetic analogs mediate growth inhibition in KOR274L2 and KOW286R6 cells only at concentrations that overcome their weak ligand binding capacity. This is especially evident in the

W286R mutant cells, where growth inhibition only occurred with supra- physiological doses of the analog compounds.

Taken together, the above data suggest that the growth-inhibitory effects of 1,25D are mediated in part through non-classical mechanisms of VDR signaling. Each of the analogs tested was designed to impart maximal anti- growth effects with minimal calcemic side effects (Mathiasen et al., 1993); the fact that these two actions of the VDR can be separated is strong evidence for mechanisms of VDR action which are outside the classical transcriptional signaling through DR3 VDREs. Most 1,25D-inducible genes involved in calcium homeostasis contain DR3 VDREs (Haussler et al., 1998), including CYP24 and osteocalcin. However, other functional types of VDREs are now recognized, and include IP9, ER6, ER4, DR4, and DR6 elements (Quack and Carlberg, 2000;

166

Saramaki et al., 2006; Schrader et al., 1997; Thompson et al., 2002; Thummel et al., 2001; Xie and Bikle, 1997). It is interesting to note that some of these alternate VDREs occur in 1,25D-target genes which are not involved in calcium homeostasis, such as p21 (Saramaki et al., 2006) and c-fos (Schrader et al.,

1997). In support of this, IP9 VDREs were shown to correlate better with induction of apoptosis by 1,25D than DR3 VDREs (Danielsson et al., 1997);

VDR bound to EB1089 also exhibited increased affinity for IP9 VDREs as compared to DR3 VDREs (Nayeri et al., 1995). While HVDRR-2 mutations in the

DBD are known to abolish ligand binding to DR3 VDREs, no studies to our knowledge have assessed their ability to bind to any of the alternate VDREs, therefore it is possible that the G46D mutant VDR retains the ability to bind other

VDREs, even though its transcriptional activation activity on DR3 VDREs is lost.

Alternatively, the G46D VDR may mediate its anti-growth properties via other

DNA-independent actions of the VDR, such as binding in concert with other transcription factors on their respective response elements, such as on the p27 promoter (Huang et al., 2004), or inhibiting formation of other transcription factor complexes (Towers et al., 1999). It is also possible that the rapid response pathways initiated at the plasma membrane or in the cytosol may contribute to growth regulation, and thse pathways are retained in certain mutants (Capiati et al., 2004; Huhtakangas et al., 2004; Nguyen et al., 2004; Rochel and Moras,

2006; Zanello and Norman, 2004).

Elucidation of the specific non-classical mechanisms of VDR signaling impact on breast cancer cell growth regulation will require additional study,

167

however our data suggest that at least a portion of VDR-mediated growth inhibition occurs through a non-DR3 transcriptional activation-mediated mechanism. The KOmutant cell lines developed in this thesis project comprise a novel and useful model system for detailed study of these non-DR3 effects, and of other possible non-classical VDR mechanisms.

168

GENERAL DISCUSSION

In addition to its calcium regulatory role, 1,25D, the biologically active form of vitamin D3, also induces differentiation, growth arrest, and apoptosis in a variety of cancer cell types. Treatment with 1,25D was shown to differentiate leukemia and osteosarcoma cells (Miyaura et al., 1981; Tsuchiya et al., 1993), inhibit growth of retinoblastomas and melanomas (Albert et al., 1992; Colston et al., 1981), and inhibit metastasis in prostate tumors (Lokeshwar et al., 1999).

1,25D has also been shown to play a dual role in breast cancers, both as a preventive agent and as a potential therapeutic. Epidemiological associations have shown inverse correlations between vitamin D status and breast cancer risk

(John et al., 1999; Shin et al., 2002), and mice without VDR exhibit accelerated mammary gland development (Zinser et al., 2002), suggesting a preventive role for 1,25D-VDR signaling in breast cancer. As the VDR is expressed in over 80% of human breast cancers, it represents an excellent therapeutic target; however its role in mediating the anti-cancer effects of 1,25D remains unclear.

The best-studied mechanism of action of the VDR is transcriptional upregulation of target genes via direct DNA binding. Liganded VDR heterodimerizes with the RXR on VDREs in its target gene promoters; DR3 elements are the most common type of VDRE in calcium regulation (Haussler et al., 1998). VDR can also mediate repression of target genes, both as a

169

heterodimer with RXR and as a binding partner for other transcription factors

(Huang and Hung, 2006; Im et al., 2002). Other mechanisms by which VDR mediates effects on target genes have been recently described, and these mechanisms challenge the paradigm that transcriptional regulation of target genes is the only mechanism by which VDR mediates the effects of 1,25D. In the absence of ligand, VDR mediates repression of target genes; this repression is abrogated by ligand binding (Dwivedi et al., 1998; Skorija et al., 2005). VDR also mediates transcription of target genes in the absence of direct VDRE binding (Huang et al., 2004), and can inhibit formation of other transcriptional complexes in a DNA-independent manner (Towers et al., 1999). A pool of VDR has been localized to caveolae in the plasma membrane (Huhtakangas et al.,

2004), and may be responsible for the so-called “rapid responses” mediated by

1,25D, which are known to be transcription-independent (Fleet, 2004; Norman,

2006).

In an attempt to elucidate whether any of these non-transcriptional mechanisms of VDR signaling are important in 1,25D-mediated anti-cancer actions, we have characterized a model system comprised of WT and VDRKO murine mammary cell lines for their responses to 1,25D, its precursor, and several synthetic analogs. We have also created a new model system, comprised of VDRKO cell lines stably expressing WT and point-mutant VDRs, to allow the DNA- and ligand- independent actions of the VDR to be studied in greater detail.

170

Comparison of the WT145 VDRWT and KO240 VDRKO cell lines conclusively demonstrates that the VDR is required for the growth inhibitory and apoptotic effects of 1,25D. KO240 cells are neither growth inhibited nor rendered apoptotic by 1,25D at supra-physiological doses, while WT145 cells are growth inhibited by fM concentrations of 1,25D, which is a physiologically achievable dose, and rendered apoptotic by 100nM 1,25D, a dose which does not induce apoptosis in KO240 cells. Studies utilizing synthetic analogs of 1,25D, designed to maximize growth and apoptotic effects while minimizing calcemic side effects, also indicated that the VDR is necessary to mediate the anti-cancer effects of

1,25D. KO240 cells were not growth inhibited by any of the analogs tested, even when supra-physiological concentrations were used. This implies that 1,25D and its analogs mediate their growth inhibitory and pro-apoptotic effects via common pathways, which require the VDR.

When 25D, the biological precursor to 1,25D, was tested in the VDRWT and KO cells, we found that, while the presence of the VDR protein was required for the mediation of growth inhibition by 25D, transcriptional activation of the

CYP24 gene, which contains two DR3 VDREs, was not associated with 25D treatment. Both WT145 and KO240 cells express CYP27B1, the enzyme that metabolizes 25D to 1,25D, and both cell lines upregulate CYP24 following treatment with 1,25D and synthetic analogs, suggesting a mechanistic difference between 1,25D- and 25D-mediated growth inhibition via the VDR. A variety of mechanisms, discussed above, could mediate this non-transcriptional response, and further research is necessary to delineate the significance of this finding.

171

However, it presents proof of principle that the VDRE-mediated transcriptional actions of the VDR are not its only biologically significant mechanisms.

While the VDR expressed in the WT145 cells was necessary to mediate the growth inhibitory and apoptotic effects of 1,25D in vitro, several studies have indicated that in vivo, the mechanism of 1,25D-mediated tumor cell growth inhibition and apoptosis is an indirect mechanism, occurring via stromal cell signaling and inhibition of angiogenesis (Bernardi et al., 2002; Mantell et al.,

2000). Therefore, the role of tumor-expressed VDR was studied in WT145 and

KO240 cells xenografted into nude mice, and treated with EB1089, a synthetic vitamin D analog, or UV light, to induce endogenous generation of vitamin D.

Both treatment regimens significantly decreased tumor volume, via tumor cell growth inhibition and tumor cell apoptosis in WT145 tumors, but not KO240 tumors. This again indicates the requirement of the VDR in mediating the effects of vitamin D, and suggests that EB1089 and UV-light generated vitamin D mediate growth arrest and apoptosis in tumors via the same, or very similar, pathways. This is further supported by the similarities in growth arrest induced by 1,25D and its synthetic analogs in vitro, discussed above. Given that no effects of EB1089 or UV light treatment were observed in KO240 cells, the anti- tumor effects of 1,25D in our model system appear to be directly tumor-mediated, as opposed to mediated via stromal signaling. Previous studies have used either stromal cells in isolation (Bernardi et al., 2002), or xenografted VDRWT MCF-7 cells (Mantell et al., 2000) to study the effect of 1,25D on angiogenesis. Our model system, with its ability to clearly separate the VDR-dependent and VDR-

172

independent effects of 1,25D, conclusively demonstrates that the anti-tumor effects of 1,25D are mediated via tumor-expressed VDRs, and are not mediated extra-tumorally.

To our knowledge, our use of UV light to raise vitamin D status is the first study of its kind, and provides proof of principle that UV light exerts VDR- dependent anti-tumor effects. This also provides the first experimental link between the epidemiological studies connecting vitamin D status and breast cancer risk, and the known anti-tumor effects of 1,25D and the VDR.

In an effort to study the effects of VDR ablation and mutation on breast cancer cell growth, and in an attempt to minimize the cell-to-cell variation that occur in many studies, we created a series of KO240 cell lines stably expressing

WT human VDR and VDRs from HVDRR-2 patients; these mutant VDRs are known to bear point-mutations in either their DNA- or ligand- binding domains which render them incompetent for 1,25D-mediated calcemic transcriptional regulation. KO240 cells stably expressing transcriptionally-active WT VDR

(KOhVDR cells) were growth inhibited by 1,25D at physiologically achievable concentrations, and also exhibited apoptotic morphology following 100nM 1,25D treatment. All three KOhVDR cell clones were also growth inhibited by 25D and by several synthetic 1,25D analogs, including EB1089. Therefore, we have created the first mammary-specific model system to study the effects of VDR ablation in cells sharing the same background. We have also provided further evidence for the requirement of VDR in 1,25D-mediated anti-growth signaling; the KOEV control stable cell line was not growth inhibited by any vitamin D compounds,

173

even at supra-physiological doses, indicating both that stable selection does not alter sensitivity to 1,25D, and that the VDR is absolutely required to mediate the growth inhibitory effects of 1,25D. Since the VDR stably transfected was human

VDR, and these are murine cells, we also provide evidence that 1,25D-VDR signaling pathways are highly evolutionarily conserved. The human VDR was not only transcriptionally functional in murine cells, it was able to mediate the transcription of endogenous murine genes, and cause a phenotype in murine cells similar to that created by human VDR in human breast cancer cells treated with 1,25D (Narvaez and Welsh, 2001; Simboli-Campbell et al., 1996).

The KO240 cells stably expressing either DBD mutant VDR (KOG46DC) or

LBD mutant VDRs (KOR274L2 and KOW286R6) were differentially affected by 1,25D treatment. When crystal violet assays were used to examine the growth response of each mutant following 1,25D treatment, both the KOG46DC and

KOR274L2 mutant VDRs were found to confer sensitivity to 1,25D-mediated growth arrest. The G46D mutant VDR was growth inhibited by physiologically achievable levels of 1,25D, and also by 25D and various 1,25D synthetic ligands; these concentrations of 1,25D and synthetic analogs did not, however, activate

CYP24 transcription. This loss of transcriptional regulation was anticipated, as each mutant was selected from HVDRR-2 rickets patients, who are known to be deficient in calcium homeostasis. Therefore, growth inhibition of the KOG46DC cell line by 1,25D does not correlate with transcriptional activation of known VDR target genes. Several DR3-independent modes of action for the VDR have been

174

discussed, and further studies will be required to determine which, if any, plays a role in 1,25D-mediated growth inhibition.

KOR274L2 stable cell lines were growth inhibited by 1,25D, but only at supra-physiological concentrations (500nM). This concentration of 1,25D was also associated with CYP24 transcriptional upregulation. The R274L mutant

VDR also mediated growth inhibition following treatment with a panel of synthetic vitamin D analogs, but, again, only at concentrations associated with CYP24 transcriptional upregulation. The KOW286R6 mutant VDR stable cell line was insensitive to 1,25D and EB1089 at all doses tested, and was growth inhibited by other synthetic vitamin D analogs only at the supra-physiological concentration of

1μM; the W286R mutant VDR mediated no detectable CYP24 transcriptional upregulation following any of the VDR ligands tested. Given that R274L mutant

VDR is approximately 1000-fold less sensitive to 1,25D, but can still bind ligand

(Kristjansson et al., 1993), and that the KOW286R6 cells were only growth inhibited by supra-physiological levels of synthetic 1,25D analogs designed to maximally mediate growth inhibition, the growth-inhibitory actions of VDR ligands in LBD

VDRs appear to be associated with the binding of ligands to these mutant VDRs at concentrations high enough to overcome their greatly reduced binding capacity.

In summary, we have conclusively demonstrated that the growth inhibitory and pro-apoptotic effects of 1,25D on breast cancer cells are dependent upon

VDR protein expression, and involve the activation of the intrinsic caspase cascade. In vivo, EB1089, a vitamin D analog, directly mediates its anti-tumor

175

effects via tumor-cell expressed VDR, and not via extra-tumorally expressed

VDR, as previous studies have suggested. We also provide proof of principle that UV-generated vitamin D exerts anti-cancer effects in vivo, via a tumor cell expressed VDR-dependent mechanism, which is highly similar to the mechanism employed by endogenous vitamin D sterols. Data generated in the KOhVDR cell lines support the requirement of the VDR in mediating the effects of 1,25D, and also show conclusive evidence that the 1,25D-VDR signaling pathways are highly evolutionarily conserved, at least between human and murine species.

Finally, KO240 cells stably expressing HVDRR-2 mutant VDRs provide evidence that the growth inhibitory effects of 1,25D may be mediated, at least in part, through DNA-independent mechanisms of VDR signaling.

176

REFERENCES

Abe E, Miyaura C, Sakagami H, Takeda M, Konno K, Yamazaki T, Yoshiki S, Suda T (1981): Differentiation of mouse myeloid leukemia cells induced by 1 alpha,25-dihydroxyvitamin D3. Proc Natl Acad Sci U S A 78:4990-4994.

Adams JS, Singer FR, Gacad MA, Sharma OP, Hayes MJ, Vouros P, Holick MF (1985): Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured alveolar macrophages in sarcoidosis. J Clin Endocrinol Metab 60:960-966.

Alagbala AA, Johnson CS, Trump DL, Posner GH, Foster BA (2006): Antitumor effects of two less-calcemic vitamin D analogs (Paricalcitol and QW- 1624F2-2) in squamous cell carcinoma cells. Oncology 70:483-492.

Albert DM, Marcus DM, Gallo JP, O'Brien JM (1992): The antineoplastic effect of vitamin D in transgenic mice with retinoblastoma. Invest Ophthalmol Vis Sci 33:2354-2364.

Baker AR, McDonnell DP, Hughes M, Crisp TM, Mangelsdorf DJ, Haussler MR, Pike JW, Shine J, O'Malley BW (1988): Cloning and expression of full- length cDNA encoding human vitamin D receptor. Proc Natl Acad Sci USA 85:3294-3298.

Bareis P, Bises G, Bischof MG, Cross HS, Peterlik M (2001): 25-hydroxy-vitamin d metabolism in human colon cancer cells during tumor progression. Biochem Biophys Res Commun 285:1012-1017.

Bar-Shavit Z, Teitelbaum SL, Reitsma P, Hall A, Pegg LE, Trial J, Kahn AJ (1983): Induction of monocytic differentiation and bone resorption by 1,25- dihydroxyvitamin D3. Proc Natl Acad Sci USA 80:5907-5911.

Berg JM (1989): DNA binding specificity of steroid receptors. Cell 57:1065-1068.

Berger U, McClelland RA, Wilson P, Greene GL, Haussler MR, Pike JW, Colston K, Easton D, Coombes RC (1991): Immunocytochemical determination of estrogen receptor, progesterone receptor, and 1,25-dihydroxyvitamin D3 receptor in breast cancer and relationship to prognosis. Cancer Research 51:239-244.

177

Berger U, Wilson P, McClelland RA, Colston K, Haussler MR, Pike JW, Coombes RC (1987): Immunocytochemical detection of 1,25-dihydroxyvitamin D3 receptor in breast cancer. Cancer Res 47:6793-6799.

Bernardi RJ, Johnson CS, Modzelewski RA, Trump DL (2002): Antiproliferative effects of 1alpha,25-dihydroxyvitamin D(3) and vitamin D analogs on tumor-derived endothelial cells. Endocrinol 143:2508-2514.

Berry DM, Antochi R, Bhatia M, Meckling-Gill KA (1996): 1,25-Dihydroxyvitamin D3 stimulates expression and translocation of protein kinase Calpha and Cdelta via a nongenomic mechanism and rapidly induces phosphorylation of a 33-kDa protein in acute promyelocytic NB4 cells. J Biol Chem 271:16090-16096.

Bettoun DJ, Buck DW, 2nd, Lu J, Khalifa B, Chin WW, Nagpal S (2002): A vitamin D receptor-Ser/Thr phosphatase-p70 S6 kinase complex and modulation of its enzymatic activities by the ligand. J Biol Chem 277:24847-24850.

Bettoun DJ, Burris TP, Houck KA, Buck DW, 2nd, Stayrook KR, Khalifa B, Lu J, Chin WW, Nagpal S (2003): Retinoid X receptor is a nonsilent major contributor to vitamin D receptor-mediated transcriptional activation. Mol Endocrinol 17:2320-2328.

Bettoun DJ, Lu J, Khalifa B, Yee Y, Chin WW, Nagpal S (2004): Ligand modulates VDR-Ser/Thr protein phosphatase interaction and p70S6 kinase phosphorylation in a cell-context-dependent manner. J Steroid Biochem Mol Biol. 89-90:195-198.

Bhatia M, Kirkland JB, Meckling-Gill KA (1995): Monocytic differentiation of acute promyelocytic leukemia cells in response to 1,25-dihydroxyvitamin D3 is independent of nuclear receptor binding. J Biol Chem 270:15962-15965.

Blutt SE, Polek TC, Stewart LV, Kattan MW, Weigel NL (2000): A calcitriol analogue, EB1089, inhibits the growth of LNCaP tumors in nude mice. Cancer Res 60:779-782.

Boyan BD, Sylvia VL, McKinney N, Schwartz Z (2003): Membrane actions of vitamin D metabolites 1alpha,25(OH)2D3 and 24R,25(OH)2D3 are retained in growth plate cartilage cells from vitamin D receptor knockout mice. J Cell Biochem 90:1207-1223.

Bula CM, Huhtakangas J, Olivera C, Bishop JE, Norman AW, Henry HL (2005): Presence of a truncated form of the vitamin D receptor (VDR) in a strain of VDR-knockout mice. Endocrinol 146:5581-5586.

178

Burgos-Trinidad M, DeLuca HF (1991): Kinetic properties of 25-hydroxyvitamin D- and 1,25-dihydroxyvitamin D-24-hydroxylase from chick kidney. Biochim Biophys Acta 1078:226-230.

Capiati DA, Rossi AM, Picotto G, Benassati S, Boland RL (2004): Inhibition of serum-stimulated mitogen activated protein kinase by 1alpha,25(OH)2- vitamin D3 in MCF-7 breast cancer cells. J Cell Biochem 93:384-397.

Carlberg C, Dunlop TW, Saramaki A, Sinkkonen L, Matilainen M, Vaisanen S (2007): Controlling the chromatin organization of vitamin D target genes by multiple vitamin D receptor binding sites. J Steroid Biochem Mol Biol 103:338-343.

Carlberg C, Mathiasen IS, Saurat JH, Binderup L (1994): The 1,25- dihydroxyvitamin D3 (VD) analogues MC903, EB1089 and KH1060 activate the VD receptor: homodimers show higher ligand sensitivity than heterodimers with retinoid X receptors. J Steroid Biochem Mol Biol 51:137-142.

Chandler JS, Pike JW, Haussler MR (1979): 1,25-Dihydroxyvitamin D3 receptors in rat kidney cytosol. Biochem Biophys Res Commun 90:1057-1063.

Chen WY, Bertone-Johnson ER, Hunter DJ, Willett WC, Hankinson SE (2005): Associations between polymorphisms in the vitamin D receptor and breast cancer risk. Cancer Epidemiol Biomarkers Prev 14:2335-2339.

Cheng HT, Chen JY, Huang YC, Chang HC, Hung WC (2006a): Functional role of VDR in the activation of p27Kip1 by the VDR/Sp1 complex. J Cell Biochem 98:1450-1456.

Cheng HT, Chen JY, Huang YC, Chang HC, Hung WC (2006b): Functional role of VDR in the activation of p27Kip1 by the VDR/Sp1 complex. J Cell Physiol 98:1450-1456.

Colston K, Colston MJ, Feldman D (1981): 1,25-dihydroxyvitamin D3 and malignant melanoma: the presence of receptors and inhibition of cell growth in culture. Endocrinol 108:1083-1086.

Colston K, Hirt M, Feldman D (1980): Organ distribution of the cytoplasmic 1,25- dihydroxycholecalciferol receptor in various mouse tissues. Endocrinol 107:1916-1922.

Colston KW, Berger U, Wilson P, Hadcocks L, Naeem I, Earl HM, Coombes RC (1988): Mammary gland 1,25-dihydroxyvitamin D3 receptor content during pregnancy and lactation. Mol Cell Endocrinol 60:15-22.

179

Colston KW, Pirianov G, Bramm E, Hamberg KJ, Binderup L (2003): Effects of Seocalcitol (EB1089) on nitrosomethyl urea-induced rat mammary tumors. Breast Cancer Res Treat 80:303-311.

Colston KW, Welsh J (2005): Vitamin D and Breast Cancer. In Feldman D, Pike JW, Glorieux FH (eds): "Vitamin D, 2nd Edition." Elsevier, pp 1663-1677.

Danielsson C, Mathiasen IS, James SY, Nayeri S, Bretting C, Hansen CM, Colston KW, Carlberg C (1997): Sensitive induction of apoptosis in breast cancer cells by a novel 1,25-dihydroxyvitamin D3 analogue shows relation to promoter selectivity. J Cell Biochem 66:552-562.

Dardenne O, Prud'homme J, Arabian A, Glorieux FH, St-Arnaud R (2001): Targeted inactivation of the 25-hydroxyvitamin D(3)-1(alpha)-hydroxylase gene (CYP27B1) creates an animal model of pseudovitamin D-deficiency rickets. Endocrinol 142:3135-3141.

Davoust N, Wion D, Chevalier G, Garabedian M, Brachet P, Couez D (1998): Vitamin D receptor stable transfection restores the susceptibility to 1,25- dihydroxyvitamin D3 cytotoxicity in a rat glioma resistant clone. J Neurosci Res 52:210-219.

Dwivedi PP, Muscat GE, Bailey PJ, Omdahl JL, May BK (1998): Repression of basal transcription by vitamin D receptor: evidence for interaction of unliganded vitamin D receptor with two receptor interaction domains in RIP13delta1. J Mol Endocrinol 20:327-335.

Eisman JA, Suva LJ, Martin TJ (1986): Significance of 1,25-dihydroxyvitamin D3 receptor in primary breast cancers. Cancer Res 46:5406-5408.

Erben RG, Soegiarto DW, Weber K, Zeitz U, Lieberherr M, Gniadecki R, Moller G, Adamski J, Balling R (2002): Deletion of deoxyribonucleic acid binding domain of the vitamin D receptor abrogates genomic and nongenomic functions of vitamin D. Mol Endocrinol 16:1524-1537.

Escriva H, Delaunay F, Laudet V (2000): Ligand binding and nuclear receptor evolution. Bioessays 22:717-727.

Fan FS, Yu WC (1995): 1,25-Dihydroxyvitamin D3 suppresses cell growth, DNA synthesis, and phosphorylation of retinoblastoma protein in a breast cancer cell line. Cancer Invest 13:280-286.

Flanagan L, Packman K, Juba B, O'Neill S, Tenniswood M, Welsh J (2003): Efficacy of Vitamin D compounds to modulate estrogen receptor negative breast cancer growth and invasion. J Steroid Biochem Mol Biol 84:181- 192.

180

Flanagan L, Van Weelden K, Ammerman C, Ethier SP, Welsh J (1999): SUM- 159PT cells: a novel estrogen independent human breast cancer model system. Breast Cancer Res Treat 58:193-204.

Fleet JC (2004): Rapid, membrane-initiated actions of 1,25 dihydroxyvitamin D: what are they and what do they mean? J Nutr 134:3215-3218.

Fraser DR, Kodicek E (1970): Unique biosynthesis by kidney of a biological active vitamin D metabolite. Nature 228:764-766.

Friedrich M, Reichrath J, Chen T, al. e (2000): Expression of 25-hydroxyvitamin D3 1- hydroxylase in breast tissue. In Norman A, Bouillon R, Thomasset M (eds): "Vitamin D endocrine system: structural, biological, genetic and clinical aspects." Riverside, CA: University of Riverside Press, pp 189– 191.

Gallagher RP, Lee TK (2006): Adverse effects of ultraviolet radiation: a brief review. Prog Biophys Mol Biol 92:119-131.

Garland CF, Comstock GW, Garland FC, Helsing KJ, Shaw EK, Gorham ED (1989): Serum 25-hydroxyvitamin D and colon cancer: eight-year prospective study. Lancet 2:1176-1178.

Garland CF, Garland FC (1980): Do sunlight and vitamin D reduce the likelihood of colon cancer? Int J Epidemiol 9:227-231.

Garland CF, Gorham ED, Mohr SB, Grant WB, Giovannucci EL, Lipkin M, Newmark H, Holick MF, Garland FC (2007): Vitamin D and prevention of breast cancer: pooled analysis. J Steroid Biochem Mol Biol 103:708-711.

Garland FC, Garland CF, Gorham ED, Young JF (1990): Geographic variation in breast cancer mortality in the United States: a hypothesis involving exposure to solar radiation. Prev Med 19:614-622.

Germain P, Staels B, Dacquet C, Spedding M, Laudet V (2006): Overview of nomenclature of nuclear receptors. Pharmacol Rev 58:685-704.

Godyn JJ, Xu H, Zhang F, Kolla S, Studzinski GP (1994): A dual block to cell cycle progression in HL60 cells exposed to analogues of vitamin D3. Cell Prolif 27:37-46.

Gombart AF, Luong QT, Koeffler HP (2006): Vitamin D compounds: activity against microbes and cancer. Anticancer Res 26:2531-2542.

Gorham ED, Garland FC, Garland CF (1990): Sunlight and breast cancer incidence in the USSR. Int J Epidemiol 20:1145-1146.

181

Grant WB (2003): Ecologic studies of solar UV-B radiation and cancer mortality rates. Recent Results in Cancer Res 164:371-377.

Green S, Walter P, Kumar V, Krust A, Bornert JM, Argos P, Chambon P (1986): Human oestrogen receptor cDNA: sequence, expression and homology to v-erb-A. Nature 320:134-139.

Gronemeyer H, Gustafsson JA, Laudet V (2004): Principles for modulation of the nuclear receptor superfamily. Nat Rev Drug Discov 3:950-964.

Guo YD, Strugnell S, Back DW, Jones G (1993): Transfected human liver cytochrome P-450 hydroxylates vitamin D analogs at different side-chain positions. Proc Natl Acad Sci USA 90:8668-8672.

Guzey M, Kitada S, Reed J (2002): Apoptosis induction by 1alpha,25- dihydroxyvitamin D3 in prostate cancer. Mol Cancer Ther 1:667-677.

Hanchette CL, Schwartz GG (1992): Geographic patterns of prostate cancer mortality. Evidence for a protective effect of ultraviolet radiation. Cancer 70:2861-2869.

Hansen CM, Binderup L, Hamberg KJ, Carlberg C (2001): Vitamin D and cancer: effects of 1,25(OH)2D3 and its analogs on growth control and tumorigenesis. Front Biosci 6:D820-848.

Hansen CM, Frandsen TL, Brunner N, Binderup L (1994): 1 alpha,25- Dihydroxyvitamin D3 inhibits the invasive potential of human breast cancer cells in vitro. Clin Exp Metastasis 12:195-202.

Harrison LE, Wang QM, Studzinski GP (1999): 1,25-dihydroxyvitamin D(3)- induced retardation of the G(2)/M traverse is associated with decreased levels of p34(cdc2) in HL60 cells. J Cell Biochem 75:226-234.

Haussler MR, Jurutka PW, Hsieh JC, Thompson PD, Selznick SH, Haussler CA, Whitfield GK (1995): New understanding of the molecular mechanism of receptor-mediated genomic actions of the vitamin D hormone. Bone 17:33S-38S.

Haussler MR, Norman AW (1969): Chromosomal receptor for a vitamin D metabolite. Proc Natl Acad Sci USA 62:155-162.

Haussler MR, Whitfield GK, Haussler CA, Hsieh JC, Thompson PD, Selznick SH, Dominguez CE, Jurutka PW (1998): The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res 13:325-349.

182

Hedlund TE, Moffatt KA, Miller GJ (1996): Stable expression of the nuclear vitamin D receptor in the human prostatic carcinoma cell line JCA-1: evidence that the antiproliferative effects of 1 alpha, 25-dihydroxyvitamin D3 are mediated exclusively through the genomic signaling pathway. Endocrinol 137:1554-1561.

Henry HL, Midgett RJ, Norman AW (1974): Regulation of 25-hydroxyvitamin D3- 1-hydroxylase in vivo. J Biol Chem 249:7584-7592.

Holick MF (2003): Evolution and function of vitamin D. Recent Res Cancer Res 164:3-28.

Holick MF (2006): Resurrection of vitamin D deficiency and rickets. J Clin Invest 116:2062-2072.

Holick MF, Schnoes HK, DeLuca HF, Suda T, Cousins RJ (1971): Isolation and identification of 1,25-dihydroxycholecalciferol. A metabolite of vitamin D active in intestine. Biochem 10:2799-2804.

Honkakoski P, Sueyoshi T, Negishi M (2003): Drug-activated nuclear receptors CAR and PXR. Ann Med 35:172-182.

Huang Y-C, Chen J-Y, Hung W-C (2004): Vitamin D3 receptor/Sp1 complex is Kip1 required for the induction of p27 expression by vitamin D3. Oncogene 23:4856-4861.

Huang YC, Hung WC (2006): 1,25-dihydroxyvitamin D3 transcriptionally represses p45Skp2 expression via the Sp1 sites in human prostate cancer cells. J Cell Physiol 209:363-369.

Huhtakangas JA, Olivera CJ, Bishop JE, Zanello LP, Norman AW (2004): The vitamin D receptor is present in caveolae-enriched plasma membranes and binds 1 alpha,25(OH)2-vitamin D3 in vivo and in vitro. Mol Endocrinol 18:2660-2671.

Im HJ, Craig TA, Pittelkow MR, Kumar R (2002): Characterization of a novel hexameric repeat DNA sequence in the promoter of the immediate early gene, IEX-1, that mediates 1alpha,25-dihydroxyvitamin D(3)-associated IEX-1 gene repression. Oncogene 21:3706-3714.

Ingles SA, Haile R, Henderson B, Kolonel L, Coetzee G (1997): Association of vitamin D receptor genetic polymorphism with breast cancer risk in African-American and Hispanic women.: "Vitamin D: chemistry, biology, and clinical applications of the steroid hormone." Riverside, CA: University of California, pp 813-814.

183

Janowsky EC, Lester GE, Weinberg CR, Millikan RC, Schildkraut JM, Garrett PA, Hulka BS (1999): Association between low levels of 1,25-dihydroxyvitamin D and breast cancer risk. Pub Health Nutr 2:283-291.

John EM, Schwartz GG, Dreon DM, Koo J (1999): Vitamin D and breast cancer risk: the NHANES I Epidemiologic follow-up study, 1971-1975 to 1992. National Health and Nutrition Examination Survey. Cancer Epidemiol Biomarkers Prev 8:399-406.

Jones G, Tenenhouse HS (2002): 1,25(OH)2D, the preferred substrate for CYP24. J Bone Miner Res 17:179-81.

Jurutka PW, Remus LS, Whitfield GK, Thompson PD, Hsieh JC, Zitzer H, Tavakkoli P, Galligan MA, Dang HT, Haussler CA, Haussler MR (2000): The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol Endocrinol 14:401-20.

Jurutka PW, Thompson PD, Whitfield GK, Eichhorst KR, Hall N, Dominguez CE, Hsieh JC, Haussler CA, Haussler MR (2005): Molecular and functional comparison of 1,25-dihydroxyvitamin D(3) and the novel vitamin D receptor ligand, lithocholic acid, in activating transcription of cytochrome P450 3A4. J Cell Biochem. 94:917-943.

Kallay E, Pietschmann P, Toyokuni S, Bajna E, Hahn P, Mazzucco K, Bieglmayer C, Kato S, Cross HS (2001): Characterization of a vitamin D receptor knockout mouse as a model of colorectal hyperproliferation and DNA damage. Carcinogen 22:1429-1435.

Kemmis CM, Salvador SM, Smith KM, Welsh J (2006): Human Mammary Epithelial Cells Express CYP27B1 and Are Growth Inhibited by 25- Hydroxyvitamin D-3, the Major Circulating Form of Vitamin D-3. J Nutr 136:887-892.

Knekt P, Jarvinen R, Seppanen R, Pukkala E, Aromaa A (1996): Intake of dairy products and the risk of breast cancer. Br J Cancer 73:687-691.

Knight JA, Lesosky M, Barnett H, Raboud JM, Vieth R (2007): Vitamin D and reduced risk of breast cancer: a population-based case-control study. Cancer Epidemiol Biomarkers Prev 16:422-429.

Koli K, Keski-Oja J (2000): 1alpha,25-dihydroxyvitamin D3 and its analogues down-regulate cell invasion-associated proteases in cultured malignant cells. Cell Growth Differ 11:221-229.

Kream BE, Jose M, Yamada S, DeLuca HF (1977): A specific high-affinity binding macromolecule for 1,25-dihydroxyvitamin D3 in fetal bone. Science 197:1086-1088.

184

Kricker A, Armstrong B (2006): Does sunlight have a beneficial influence on certain cancers? Prog Biophys Mol Biol 92:132-139.

Kristjansson K, Rut AR, Hewison M, O'Riordan JL, Hughes MR (1993): Two mutations in the hormone binding domain of the vitamin D receptor cause tissue resistance to 1,25 dihydroxyvitamin D3. J Clin Invest 92:12-16.

Kurokawa H, Nishio K, Fukumoto H, Tomonari A, Suzuki T, Saijo N (1999): Alteration of caspase-3 (CPP32/Yama/apopain) in wild-type MCF-7, breast cancer cells. Oncol Rep 6:33-37.

Lawson DE, Fraser DR, Kodicek E, Morris HR, Williams DH (1971): Identification of 1,25-dihydroxycholecalciferol, a new kidney hormone controlling calcium metabolism. Nature 230:228-230.

Lawson DE, Wilson PW, Kodicek E (1969): Metabolism of vitamin D. A new cholecalciferol metabolite, involving loss of hydrogen at C-1, in chick intestinal nuclei. Biochem J 115:269-277.

Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB (1997): Targeted ablation of the vitamin D receptor: an animal model of vitamin D- dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831- 9835.

Lim HS, Roychoudhuri R, Peto J, Schwartz G, Baade P, Moller H (2006): Cancer survival is dependent on season of diagnosis and sunlight exposure. Int J Cancer 119:1530-1536.

Lin NU, Malloy PJ, Sakati N, al-Ashwal A, Feldman D (1996): A novel mutation in the deoxyribonucleic acid-binding domain of the vitamin D receptor causes hereditary 1,25-dihydroxyvitamin D-resistant rickets. J Clin Endocrinol Metab 81:2564-2569.

Liu M, Lee MH, Cohen M, Bommakanti M, Freedman LP (1996): Transcriptional activation of the Cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev 10:142- 153.

Lokeshwar BL, Schwartz GG, Selzer MG, Burnstein KL, Zhuang SH, Block NL, Binderup L (1999): Inhibition of prostate cancer metastasis in vivo: a comparison of 1,23-dihydroxyvitamin D (calcitriol) and EB1089. Cancer Epidemiol Biomarkers Prev 8:241-248.

Lowe L, Hansen CM, Senaratne S, Colston KW (2003a): Mechanisms implicated in the growth regulatory effects of vitamin D compounds in breast cancer cells. Recent Res Cancer Res 164:99-110.

185

Lowe L, Hansen CM, Senaratne S, Colston KW (2003b): Mechanisms implicated in the growth regulatory effects of vitamin D compounds in breast cancer cells. Recent Res in Cancer Res. 164:99-110.

Ma Y, Yu WD, Kong RX, Trump DL, Johnson CS (2006): Role of Nongenomic Activation of Phosphatidylinositol 3-Kinase/Akt and Mitogen-Activated Protein Kinase/Extracellular Signal-Regulated Kinase Kinase/Extracellular Signal-Regulated Kinase 1/2 Pathways in 1,25D3-Mediated Apoptosis in Squamous Cell Carcinoma Cells. Cancer Res 66:8131-8138.

MacLaughlin JA, Anderson RR, Holick MF (1982): Spectral character of sunlight modulates photosynthesis of previtamin D3 and its photoisomers in human skin. Science 216:1001-1003.

Majewski S, Skopinska M, Marczak M, Szmurlo A, Bollag W, Jablonska S (1996): Vitamin D3 is a potent inhibitor of tumor cell-induced angiogenesis. J Investig Dermatol Symp Proc 1:97-101.

Makishima M, Lu TT, Xie W, Whitfield GK, Domoto H, Evans RM, Haussler MR, Mangelsdorf DJ (2002): Vitamin D receptor as an intestinal bile acid sensor. Science 296:1313-1316.

Malloy PJ, Feldman D (2003): Hereditary 1,25-Dihydroxyvitamin D-resistant rickets. Endocr Dev 6:175-199.

Malloy PJ, Pike JW, Feldman D (1999): The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D-resistant rickets. Endocr Rev 20:156-188.

Mantell DJ, Owens PE, Bundred NJ, Mawer EB, Canfield AE (2000): 1 alpha,25- dihydroxyvitamin D(3) inhibits angiogenesis in vitro and in vivo. Circ Res 87:214-220.

Mathiasen IS, Colston KW, Binderup L (1993): EB 1089, a novel vitamin D analogue, has strong antiproliferative and differentiation inducing effects on cancer cells. J Steroid Biochem Mol Biol 46:365-371.

Mathiasen IS, Lademann U, Jaattela M (1999): Apoptosis induced by vitamin D compounds in breast cancer cells is inhibited by Bcl-2 but does not involve known caspases or p53. Cancer Res 59:4848-4856.

Mathiasen IS, Sergeev IN, Bastholm L, Elling F, Norman AW, Jaattela M (2002): Calcium and calpain as key mediators of apoptosis-like death induced by vitamin D compounds in breast cancer cells. J Biol Chem 277:30738- 30745.

Mawer E, B,, Walls J, Howell A, Davies M, Ratcliffe WA, Bundred NJ (1997a): Serum 1,25-dihydroxyvitamin D may be related inversely to disease

186

activity in breast cancer patients with bone metastases. J Clin Endocrinol Metab 82:118-122.

Mawer EB, Walls J, Howell A, Davies M, Ratcliffe WA, Bundred NJ (1997b): Serum 1,25-dihydroxyvitamin D may be related inversely to disease activity in breast cancer patients with bone metastases. J Clin Endocrinol Metab 82:118-22.

McCarthy TC, Li X, Sinal CJ (2005): Vitamin D receptor-dependent regulation of colon multidrug resistance-associated protein 3 gene expression by bile acids. J Biol Chem. 2005 280:23232-23242.

McCollum E, Simmonds N, Becker J, Shipley P (1922): Studies on experimental rickets; and experimental demonstration of the existence of a vitamin which promotes calcium deposition. J Biol Chem 53:293-312.

McDonnell DP, Scott RA, Kerner SA, O'Malley BW, Pike JW (1989): Functional domains of the human vitamin D3 receptor regulate osteocalcin gene expression. Mol Endocrinol 3:635-644.

Mehta R, Hawthorne M, Uselding L, Albinescu D, Moriarty R, Christov K (2000): Prevention of N-methyl-N-nitrosourea-induced mammary carcinogenesis in rats by 1alpha-hydroxyvitamin D(5). J Natl Cancer Inst 92:1836-1840.

Meindl S, Rot A, Hoetzenecker W, Kato S, Cross HS, Elbe-Burger A (2005): Vitamin D receptor ablation alters skin architecture and homeostasis of dendritic epidermal T cells. Br J Dermatol. 152:231-241.

Mellon WS, DeLuca HF (1979): An equilibrium and kinetic study of 1,25- dihydroxyvitamin D3 binding to chicken intestinal cytosol employing high specific activity 1,25-dehydroxy[3H-26, 27] vitamin D3. Arch Biochem Biophys 197:90-95.

Miller GJ, Stapleton GE, Hedlund TE, Moffat KA (1995): Vitamin D receptor expression, 24-hydroxylase activity, and inhibition of growth by 1alpha,25- dihydroxyvitamin D3 in seven human prostatic carcinoma cell lines. Clin Cancer Res 1:997-1003.

Milliken EL, Zhang X, Flask C, Duerk JL, MacDonald PN, Keri RA (2005): EB1089, a vitamin D receptor agonist, reduces proliferation and decreases tumor growth rate in a mouse model of hormone-induced mammary cancer. Cancer Lett 229:205-215.

Miyaura C, Abe E, Kuribayashi T, Tanaka H, Konno K, Nishii Y, Suda T (1981): 1 alpha,25-Dihydroxyvitamin D3 induces differentiation of human myeloid leukemia cells. Biochem Biophys Res Commun 102:937-943.

187

Mizwicki MT, Bula CM, Bishop JE, Norman AW (2005): A perspective on how the Vitamin D sterol/Vitamin D receptor (VDR) conformational ensemble model can potentially be used to understand the structure-function results of A-ring modified Vitamin D sterols. J Steroid Biochem Mol Biol 97:69-82.

Mizwicki MT, Keidel D, Bula CM, Bishop JE, Zanello LP, Wurtz JM, Moras D, Norman AW (2004): Identification of an alternative ligand-binding pocket in the nuclear vitamin D receptor and its functional importance in 1alpha,25(OH)2-vitamin D3 signaling. Proc Natl Acad Sci USA 101:12876- 12881.

Moore LB, Maglich JM, McKee DD, Wisely B, Willson TM, Kliewer SA, Lambert MH, Moore JT (2002): Pregnane X receptor (PXR), constitutive androstane receptor (CAR), and benzoate X receptor (BXR) define three pharmacologically distinct classes of nuclear receptors. Mol Endocrinol 16:977-986.

Nakagawa T, Yuan J (2000): Cross-talk between Two Cysteine Protease Families: Activation of Caspase-12 by Calpain in Apoptosis. J Cell Biol 150:887-894.

Nakagawa T, Zhu H, Morishima N, Li E, Xu J, Yankner BA, Yuan J (2000): Caspase-12 mediates endoplasmic-reticulum-specific apoptosis and cytotoxicity by amyloid-beta. Nature 403:98-103.

Narvaez CJ, Byrne BM, Romu S, Valrance M, Welsh J (2003): Induction of apoptosis by 1,25-dihydroxyvitamin D3 in MCF-7 Vitamin D3-resistant variant can be sensitized by TPA. J Steroid Biochem Mol Biol 84:199-209.

Narvaez CJ, Welsh J (2001): Role of mitochondria and caspases in vitamin D- mediated apoptosis of MCF-7 breast cancer cells. J Biol Chem 276:9101- 9107.

Nayeri S, Danielsson C, Kahlen JP, Schrader M, Mathiasen IS, Binderup L, Carlberg C (1995): The anti-proliferative effect of vitamin D3 analogues is not mediated by inhibition of the AP-1 pathway, but may be related to promoter selectivity. Oncogene 11:1853-1858.

Nemere I, Safford SE, Rohe B, DeSouza MM, Farach-Carson MC (2004): Identification and characterization of 1,25D3-membrane-associated rapid response, steroid (1,25D3-MARRS) binding protein. J Steroid Biochem Mol Biol 89-90:281-285.

Nguyen M, d'Alesio A, Pascussi JM, Kumar R, Griffin MD, Dong X, Guillozo H, Rizk-Rabin M, Sinding C, Bougneres P, Jehan F, Garabedian M (2006): Vitamin D-resistant rickets and type 1 diabetes in a child with compound heterozygous mutations of the vitamin D receptor (L263R and R391S):

188

dissociated responses of the CYP-24 and rel-B promoters to 1,25- dihydroxyvitamin D3. J Bone Miner Res. 21:886-894.

Nguyen TM, Adiceam P, Kottler ML, Guillozo H, Rizk-Rabin M, Brouillard F, Lagier P, Palix C, Garnier JM, Garabedian M (2002): Tryptophan missense mutation in the ligand-binding domain of the vitamin D receptor causes severe resistance to 1,25-dihydroxyvitamin D. J Bone Miner Res 17:1728-1737.

Nguyen TM, Lieberherr M, Fritsch J, Guillozo H, Alvarez ML, Fitouri Z, Jehan F, Garabedian M (2004): The rapid effects of 1,25-dihydroxyvitamin D3 require the vitamin D receptor and influence 24-hydroxylase activity: studies in human skin fibroblasts bearing vitamin D receptor mutations. J Biol Chem 279:7591-7597.

Nield LS, Mahajan P, Joshi A, Kamat D (2006): Rickets: not a disease of the past. Am Fam Physician 74:619-626.

Norman AW (2006): Minireview: vitamin D receptor: new assignments for an already busy receptor. Endocrinol 147:5542-5548.

Oades GM, Dredge K, Kirby RS, Colston KW (2002): Vitamin D receptor- dependent antitumour effects of 1,25-dihydroxyvitamin D3 and two synthetic analogues in three in vivo models of prostate cancer. Brit J Urol Int 90:607-616.

Oikawa T, Hirotani K, Ogasawara H, Katayama T, Nakamura O, Iwaguchi T, Hiragun A (1990): Inhibition of angiogenesis by vitamin D3 analogues. Eur J Pharmacol 178:247-250.

O'Kelly J, Koeffler HP (2003): Vitamin D analogs and breast cancer. Recent Res Cancer Res 164:333-348.

Omdahl JL, Morris HA, May BK (2002): Hydroxylase enzymes of the vitamin D pathway: expression, function, and regulation. Annu Rev Nutr 22:139-166.

Ordonez-Moran P, Larriba MJ, Pendas-Franco N, Aguilera O, Gonzalez-Sancho JM, Munoz A (2005): Vitamin D and cancer: an update of in vitro and in vivo data. Front Biosci 10:2723-2749.

Panda D, Miao D, Tremblay M, Sirois J, Farookhi R, Hendy G, Goltzman D (2001): Targeted ablation of the 25-hydroxyvitamin D 1alpha -hydroxylase enzyme: evidence for skeletal, reproductive, and immune dysfunction. Proc Natl Acad Sci USA 98:7498-7503.

Peng X, Jhaveri P, Hussain-Hakimjee EA, Mehta RG (2007): Overexpression of ER and VDR is not sufficient to make ER-negative MDA-MB231 breast

189

cancer cells responsive to 1alpha-hydroxyvitamin D5. Carcinogen 28:1000-1007.

Pike JW, Haussler MR (1983): Association of 1,25-dihydroxyvitamin D3 with cultured 3T6 mouse fibroblasts. Cellular uptake and receptor-mediated migration to the nucleus. J Biol Chem 258:8554-8560.

Ponchon G, DeLuca HF (1969): The role of the liver in the metabolism of vitamin D. J Clin Invest 48:1273-1279.

Quack M, Carlberg C (2000): Ligand-triggered stabilization of vitamin D receptor/retinoid X receptor heterodimer conformations on DR4-type response elements. J Mol Biol 296:743-756.

Raff M (1998): Cell suicide for beginners. Nature 396:119-122.

Rochel N, Moras D (2006): Ligand binding domain of vitamin D receptors. Curr Top Med Chem. 6:1229-1241.

Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D (2000): The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173-179.

Rut AR, Hewison M, Kristjansson K, Luisi B, Hughes MR, O'Riordan JL (1994): Two mutations causing vitamin D resistant rickets: modelling on the basis of steroid hormone receptor DNA-binding domain crystal structures. Clin Endocrinol 41:581-590.

Saijo T, Ito M, Takeda E, Huq AH, Naito E, Yokota I, Sone T, Pike JW, Kuroda Y (1991): A unique mutation in the vitamin D receptor gene in three Japanese patients with vitamin D-dependent rickets type II: utility of single-strand conformation polymorphism analysis for heterozygous carrier detection. Am J Hum Genet 49:668-673.

Saramaki A, Banwell CM, Campbell MJ, Carlberg C (2006): Regulation of the human p21(waf1/cip1) gene promoter via multiple binding sites for p53 and the vitamin D3 receptor. Nucleic Acids Res 34:543-554.

Schondorf T, Eisberg C, Wassmer G, Warm M, Becker M, Rein DT, Gohring UJ (2003a): Association of the vitamin D receptor genotype with bone metastases in breast cancer patients. Oncol 64:154-159.

Schondorf T, Eisberg C, Wassmer G, Warm M, Becker M, Rein DT, Gohring UJ (2003b): Association of the vitamin D receptor genotype with bone metastases in breast cancer patients. Oncol 64:154-159.

190

Schrader M, Kahlen J-P, Carlberg C (1997): Functional characterization of a novel type of 1a,25-dihydroxyvitamin D3 response element identified in the mouse c-fos promoter. Biochem Biophys Res Commun 230:646-651.

Scorrano L, Korsmeyer SJ (2003): Mechanisms of cytochrome c release by proapoptotic BCL-2 family members. Biochem Biophys Res Commun 304:437-444.

Shin MH, Holmes MD, Hankinson SE, Wu K, Colditz GA, Willett WC (2002): Intake of dairy products, calcium, and vitamin d and risk of breast cancer. J Natl Cancer Inst 94:1301-1311.

Simboli-Campbell M, Narvaez CJ, Tenniswood M, Welsh J (1996): 1,25- Dihydroxyvitamin D3 induces morphological and biochemical markers of apoptosis in MCF-7 breast cancer cells. J Steroid Biochem Mol Biol 58:367-376.

Simboli-Campbell M, Narvaez CJ, van Weelden K, Tenniswood M, Welsh J (1997): Comparative effects of 1,25(OH)2D3 and EB1089 on cell cycle kinetics and apoptosis in MCF-7 breast cancer cells. Breast Cancer Res Treat 42:31-41.

Skorija K, Cox M, Sisk JM, Dowd DR, MacDonald PN, Thompson CC, Demay MB (2005): Ligand-independent actions of the vitamin D receptor maintain hair follicle homeostasis. Mol Endocrinol 19:855-862.

Skowronski RJ, Peehl DM, Feldman D (1995): Actions of vitamin D3, analogs on human prostate cancer cell lines: comparison with 1,25-dihydroxyvitamin D3. Endocrinol 136:20-26.

Sone T, Marx SJ, Liberman UA, Pike JW (1990): A unique point mutation in the human vitamin D receptor chromosomal gene confers hereditary resistance to 1,25-dihydroxyvitamin D3. Mol Endocrinol 4:623-631.

Song X, Bishop JE, Okamura WH, Norman AW (1998): Stimulation of phosphorylation of mitogen-activated protein kinase by 1alpha,25- dihydroxyvitamin D3 in promyelocytic NB4 leukemia cells: a structure- function study. Endocrinology 139:457-465.

Swami S, Raghavachari N, Muller UR, Bao YP, Feldman D (2003): Vitamin D growth inhibition of breast cancer cells: gene expression patterns assessed by cDNA microarray. Breast Cancer Res Treat 80:49-62.

Thompson PD, Jurutka PW, Haussler CA, Whitfield GK, Haussler MR (1998): Heterodimeric DNA binding by the vitamin D receptor and retinoid X receptors is enhanced by 1,25-dihydroxyvitamin D3 and inhibited by 9-cis- retinoic acid. Evidence for allosteric receptor interactions. J Biol Chem 273:8483-8491.

191

Thompson PD, Jurutka PW, Whitfield GK, Myskowski SM, Eichhorst KR, Dominguez CE, Haussler CA, Haussler MR (2002): Liganded VDR induces CYP3A4 in small intestinal and colon cancer cells via DR3 and ER6 vitamin D responsive elements. Biochem Biophys Res Commun 299:730-738.

Thornberry NA, Lazebnik Y (1998): Caspases: enemies within. Science 281:1312-1316.

Thummel KE, Brimer C, Yasuda K, Thottassery J, Senn T, Lin Y, Ishizuka H, Kharasch E, Schuetz J, Schuetz E (2001): Transcriptional control of intestinal cytochrome P-4503A by 1alpha,25-dihydroxy vitamin D3. Mol Pharmacol 60:1399-1406.

Towers TL, Staeva TP, Freedman LP (1999): A two-hit mechanism for vitamin D3-mediated transcriptional repression of the granulocyte-macrophage colony-stimulating factor gene: vitamin D receptor competes for DNA binding with NFAT1 and stabilizes c-Jun. Mol Cell Biol 19:4191-4199.

Townsend K, Banwell CM, Guy M, Colston KW, Mansi JL, Stewart PM, Campbell MJ, Hewison M (2005a): Autocrine metabolism of vitamin D in normal and malignant breast tissue.T. Clin Cancer Res 11:3579-3586.

Townsend K, Evans KN, Campbell MJ, Colston KW, Adams JS, Hewison M (2005b): Biological actions of extra-renal 25-hydroxyvitamin D-1alpha- hydroxylase and implications for chemoprevention and treatment. J Steroid Biochem Mol Biol 97:103-109.

Tsuchiya H, Morishita H, Tomita K, Ueda Y, Tanaka M (1993): Differentiating and antitumor activities of 1 alpha,25-dihydroxyvitamin D3 in vitro and 1 alpha- hydroxyvitamin D3 in vivo on human osteosarcoma. J Orthop Res 11:122- 130.

Vaisanen S, Dunlop TW, Sinkkonen L, Frank C, Carlberg C (2005): Spatio- temporal activation of chromatin on the human CYP24 gene promoter in the presence of 1alpha,25-Dihydroxyvitamin D3. J Mol Biol 350:65-77.

Valrance ME, Brunet AH, Acosta A, Welsh J (2007): Dissociation of growth arrest and CYP24 induction by VDR ligands in mammary tumor cells. J Cell Biochem.

Valrance ME, Welsh J (2004): Breast cancer cell regulation by high-dose Vitamin D compounds in the absence of nuclear vitamin D receptor. J Steroid Biochem Mol Biol 89-90:221-225.

VanWeelden K, Flanagan L, Binderup L, Tenniswood M, Welsh J (1998): Apoptotic regression of MCF-7 xenografts in nude mice treated with the vitamin D3 analog, EB1089. Endocrinology 139:2102-2110.

192

Verlinden L, Verstuyf A, Convents R, Marcelis S, Van Camp M, Bouillon R (1998): Action of 1,25(OH)2D3 on the cell cycle genes, cyclin D1, p21 and p27 in MCF-7 cells. Mol Cell Endocrinol 142:57-65.

Vieth R (1990): The mechanisms of vitamin D toxicity. Bone Miner 11:267-272.

Welsh J (1994): Induction of apoptosis in breast cancer cells in response to vitamin D and antiestrogens. Biochem Cell Biol 72:537-45.

Whitfield GK, Selznick SH, Haussler CA, Hsieh JC, Galligan MA, Jurutka PW, Thompson PD, Lee SM, Zerwekh JE, Haussler MR (1996): Vitamin D receptors from patients with resistance to 1,25-dihydroxyvitamin D3: point mutations confer reduced transactivation in response to ligand and impaired interaction with the retinoid X receptor heterodimeric partner. Mol Endocrinol 10:1617-1631.

Wu G, Fan RS, Li W, Ko TC, Brattain MG (1997): Modulation of cell cycle control by vitamin D3 and its analogue, EB1089, in human breast cancer cells. Oncogene 15:1555-1563.

Xie Z, Bikle DD (1997): Cloning of the human phospholipase C-gamma1 promoter and identification of a DR6-type vitamin D-responsive element. J Biol Chem 272:6573-6577.

Yee YK, Chintalacharuvu SR, Lu J, Nagpal S (2005): Vitamin D receptor modulators for inflammation and cancer. Mini Rev Med Chem. 5:761-768.

Zanello LP, Norman AW (2004): Rapid modulation of osteoblast ion channel responses by 1alpha,25(OH)2-vitamin D3 requires the presence of a functional vitamin D nuclear receptor. Proc Natl Acad Sci USA 101:1589- 1594.

Zehnder D, Bland R, Walker EA, Bradwell AR, Howie AJ, Hewison M, Stewart PM (1999): Expression of 25-hydroxyvitamin D3-1alpha-hydroxylase in the human kidney. J Am Soc Nephrol 10:2465-2473.

Zhang X, Jiang F, Li P, Li C, Ma Q, Nicosia SV, Bai W (2005): Growth suppression of ovarian cancer xenografts in nude mice by vitamin D analogue EB1089. Clin Cancer Res 11:323-328.

Zinser G, Packman K, Welsh J (2002): Vitamin D(3) receptor ablation alters mammary gland morphogenesis. Develop 129:3067-3076.

Zinser G, Welsh J (2004a): Accelerated mammary gland development during pregnancy and delayed postlactational involution in vitamin D3 receptor null mice. Mol Endocrinol. 18:2208-2223.

193

Zinser GM, McEleney K, Welsh J (2003): Characterization of mammary tumor cell lines from wild type and vitamin D3 receptor knockout mice. Mol Cell Endocrinol 200:67-80.

Zinser GM, Suckow M, Welsh J (2005): Vitamin D receptor (VDR) ablation alters carcinogen-induced tumorigenesis in mammary gland, epidermis and lymphoid tissues. J Steroid Biochem Mol Biol 97:153-164.

Zinser GM, Welsh J (2004b): Accelerated mammary gland development during pregnancy and delayed postlactational involution in vitamin D3 receptor null mice. Mol Endocrinol 18:2208-2223.

Zinser GM, Welsh J (2004c): Vitamin D receptor status alters mammary gland morphology and tumorigenesis in MMTV-neu mice. Carcinogen 25:2361- 2372.

194