An Investigation Into the Role of the MARRS Receptor in Murine Mammary Gland Growth and Development by Allison Wilkin a Thesis

An Investigation into the Role of the MARRS Receptor in Murine Mammary Gland Growth and Development

Allison Wilkin

A Thesis

presented to

The University of Guelph

In partial fulfillment of requirements

for the degree of

Doctor of Philosophy

Human Health and Nutritional Sciences

Guelph, Ontario, Canada

ABSTRACT

An Investigation into the Role of the MARRS Receptor in Murine Mammary Gland Growth and Development

Allison Wilkin Advisor: Dr. Kelly A. Meckling University of Guelph, 2019

Vitamin D has been implicated in mammary gland development, however nothing is known about the role of the second 1,25-dihydroxyvitamin D3 (1,25D3) receptor MARRS and its effects on post-natal mammary gland development or interaction with 1,25D3 in this tissue.

Therefore, the objectives of this thesis were to investigate these unknowns in a tissue-specific knockout mouse model. In study 1, MARRS was knocked out in murine mammary gland epithelial cells of 4-week old mice (n=30). Several markers of mammary gland (MG) growth were measured, including number of terminal end buds (TEB), ductal coverage of the fat pad, and ductal extension. It was found that the knockout animals had decreased numbers of TEBs

(p=0.019), and decreased ductal extension (p=0.018). Study 2 aimed to expand the investigation to MARRS’s actions with varying vitamin D3 dietary dose. Abdominal MGs were collected from 6-week old MARRS knockout female mice (n=94) on diets of 10,000 IU/kg (replete), 1,000

IU/kg (sufficient) or 0 IU/kg (deficient) of D3. There was a significant interaction between genotype and diet regarding TEBs (p=0.001) and ductal coverage (p=0.03). Knockout mice on the sufficient diet had significantly fewer TEBs (p=0.001) compared to wildtypes on the same diet, but the opposite effect was seen in mice on the replete diet. The increased growth in knockout mice on replete diets was unexpected; therefore in Study 3, a novel double knockout mouse lacking both MARRS and VDR was created to study the actions of both pathways

simultaneously. There was a significant overall interaction between genotype groups (p=0.004) regarding mean ductal coverage, yet number of TEBs remained similar across all genotypes.

Overall, this thesis demonstrates that MARRS does indeed play a role in mammary gland development providing a growth-positive effect, and that effect can be altered depending on the amount of vitamin D3 available in the diet. Although neither MARRS nor VDR are essential for mammary gland growth, their signalling pathways and gene-diet interactions provide valuable information regarding dietary intakes during puberty and influences on tissue growth. These results now merit further studies in human populations to investigate dietary trends and long- term consequences in human adolescents.

TABLE OF CONTENTS ACKNOWLEDGEMENTS ...... VII LIST OF TABLES ...... IX LIST OF FIGURES ...... IX LIST OF ABBREVIATIONS ...... XI CHAPTER 1: LITERATURE REVIEW ...... 1 1.1 INTRODUCTION ...... 1 1.2 MURINE MAMMARY GLAND DEVELOPMENT ...... 1 1.2.1 Prenatal Development ...... 1 1.2.2 Terminal End Bud Structure ...... 2 1.2.3 Hormonal Regulation of Pubertal Mammary Gland Development ...... 4 1.3 ROLE OF VITAMIN D IN MAMMARY GLAND DEVELOPMENT ...... 8 1.3.1 Vitamin D Metabolism ...... 8 1.3.2 Classic Role of 1,25D3 ...... 9 1.3.3 Regulation of Mammary Gland Development by VDR ...... 10 1.4 NON-GENOMIC ACTIONS OF 1,25D3 ...... 11 1.4.1 Evidence of a Second Receptor ...... 11 1.4.2 Calcium and Phosphate Uptake ...... 13 1.4.3 Chondrocytes and Osteoblasts ...... 14 1.4.4 Haematopoeisis and Leukemia ...... 15 1.4.5 Lifespan ...... 16 I.5 VITAMIN D AND BREAST CANCER ...... 16 1.5.1 Epidemiological Studies ...... 16 1.5.2 1,25D3 and Cell Growth Inhibition in Cell Culture ...... 17 1.5.3 VDR Knockouts and Tumorigenesis ...... 19 1.5.4 MARRS and Breast Cancer ...... 20 1.5.5 Connection between TEBs and Breast Cancer ...... 21 1.6 CONCLUSION ...... 23 CHAPTER 2: RATIONALE AND OBJECTIVES ...... 24 CHAPTER 3: ROLE OF THE MARRS RECEPTOR IN MURINE MAMMARY GLAND GROWTH AND DEVELOPMENT ...... 25 3.1 ABSTRACT ...... 25 3.2 INTRODUCTION ...... 26 3.3 METHODS ...... 28 3.3.1 Animals ...... 28 3.3.2 Genotyping ...... 28 3.3.3 Necropsy and Tissue Collection ...... 29 3.3.4 Whole mounting Mammary Glands ...... 29 3.3.5 Western Blotting ...... 30 3.3.6 Immunohistochemistry ...... 31

3.3.7 Statistical Analysis ...... 31 3.4 RESULTS ...... 32 3.4.1 Phenotype Characteristics ...... 32 3.4.2 Mammary gland characteristics ...... 32 3.4.3 Western immunoblots ...... 33 3.4.4 Immunohistochemistry ...... 33 3.5 DISCUSSION ...... 38 3.6 CONCLUSION ...... 41 CHAPTER 4: DIFFERENTIAL EFFECTS OF THE 1,25D3-MARRS RECEPTOR (ERP57/PDIA3) ON MURINE MAMMARY GLAND DEVELOPMENT DEPEND ON VITAMIN D3 DOSE ...... 43 4.1 ABSTRACT ...... 43 4.2 INTRODUCTION ...... 44 4.3 METHODS ...... 46 4.3.1 Animals ...... 46 4.3.2 Diets ...... 46 4.3.3 Genotyping ...... 48 4.3.4 Necropsy and Tissue Collection ...... 48 4.3.5 Whole mounting Mammary Glands ...... 49 4.3.6 Western Blotting ...... 49 4.3.7 Immunohistochemistry ...... 50 4.3.8 Statistical Analysis ...... 51 4.4 RESULTS ...... 51 4.4.1 Phenotype characteristics ...... 51 4.4.2 Blood tests ...... 51 4.4.3 Mammary gland growth variables ...... 52 4.4.4 Western immunoblots ...... 53 4.4.5 Immunohistochemistry ...... 53 4.5 DISCUSSION ...... 61 4.6 CONCLUSIONS ...... 64 CHAPTER 5: EFFECTS OF KNOCKING OUT BOTH THE 1,25D3-MARRS RECEPTOR AND VITAMIN D RECEPTOR IN MURINE PUBERTAL MAMMARY GLAND DEVELOPMENT ...... 66 5.1 ABSTRACT ...... 66 5.2 INTRODUCTION ...... 67 5.3 METHODS ...... 68 5.3.1 Animals and Diets ...... 68 5.3.2 Genotyping ...... 69 5.3.3 Necropsy and Tissue Collection ...... 69 5.3.3 Whole mounting Mammary Glands ...... 70 5.3.4 Western immunoblotting ...... 70 5.3.5 Immunohistochemistry ...... 71 5.3.6 RNA analysis ...... 71

5.3.7 Statistics ...... 72 5.4 RESULTS ...... 74 5.4.1 Physical Characteristics and Breeding ...... 74 5.4.2 Mammary gland growth variables ...... 74 5.4.3 Western immunoblots ...... 75 5.4.4 Immunohistochemistry ...... 75 5.4.5 RNA expression ...... 76 5.5 DISCUSSION ...... 85 5.6 CONCLUSIONS ...... 88 CHAPTER 6: INTEGRATED DISCUSSION ...... 89 6.1 GENERAL SUMMARY ...... 89 6.1.1 Study 1 ...... 89 6.1.2 Study 2 ...... 90 6.1.3 Study 3 ...... 91 6.2 FUTURE DIRECTIONS ...... 93 6.3 CONCLUSIONS ...... 95

ACKNOWLEDGEMENTS

First I would like to thank my advisor, Dr. Meckling. Kelly, you allowed me to begin this journey by first accepting me as an undergraduate research student. Thank you for placing me in charge of an animal study when I began my Master’s, and thank you for fast-tracking me into the

PhD program. I don’t think I would have continued my education had it not been for your belief in my abilities as a researcher. Thank you for all our discussions; I believe I learned something new – about research, MARRS, vitamin D – every time we had a chat! And above all, I am so grateful to have an advisor that was as compassionate and understanding as you were during the last year of my PhD. To be allowed to write at my own pace was a huge reason that I have finished today.

Thank you to my advisory committee, Dr. Jim Kirkland, Dr. David Ma, and Dr. Roger

Moorehead. With every committee meeting, email, and discussion, your invaluable insights and suggestions improved my work and shaped the way I thought about approaching research.

Thank you to my qualifying exam committee, Dr. Geoff Wood and Dr. Alison Duncan.

You pushed me out of my comfort zone during my qualifying exam and instilled me with an improved, rounded education on animal studies and human nutrition.

Thank you to the Ma lab, especially Lyn Hillyer and Dr. Salma Abdelmagid for teaching me the protocols I needed and for use of your equipment. Thank you to the Simpson Lab for letting me use your microscope. Thank you to Dr. Rob Jones for your advice on protocols. A big thank you to the faculty, staff, and students of the Histoprep Lab and RHBL lab in OVC, especially Jodi Morrison, for providing a welcoming atmosphere to conduct my lab work.

vii

Thank you very much to the irreplaceable A-team of HHNS. Ann, Andra and Anne, your help and guidance made my years at HHNS seamless and panic-free, and I appreciate every time

I needed your help and you were there. Also to Diana and Andy, your hard work and help in the animal wing made a mouse project seem less daunting. And to all the students, faculty and staff of HHNS thank you for creating an energetic and friendly environment to work in.

Thank you to all my lab mates over the years. Teaching students who love to learn made coming into lab everyday not seem like work. Thank you for sacrificing your weekends to look after the mice – I appreciate it! I have enjoyed our friendship, our lab lunches, outings, and travels to conferences. Especially thank you to Cheryl Cragg, who tirelessly and selflessly helped me all through my graduate degree – from teaching lab techniques, to being a friend and a confidante.

Finally, thank you to my family. Your unconditional support was instrumental in me finishing this degree. Thank you to my father for your advice, words of support, and for offering to edit my papers even though you don’t understand them! Thank you to my mum who dropped everything and supported me in a time of crisis, and pushed me to write when I wanted to quit.

And thank you to my sister, whose late-night texts, Snaps, and visits kept me sane all these years.

I truly would not have finished this degree without you, so thank you for being there!

viii

LIST OF TABLES

CHAPTER 4 Table 4.1. Nutrient composition of modified animal diets. 47 CHAPTER 5 Table 5.1. List of primers used in genotyping. 73 Table 5.2. List of primers used in RNA analysis. 73

LIST OF FIGURES CHAPTER 1 Figure 1.1. Overview of mammary gland development. 3

CHAPTER 3 Figure 3.1. Mean number of TEBs and median ductal extension of mammary 34 glands in 4-week old mice. Figure 3.2. Representative whole mounts of mammary glands of 4-week old mice. 35 Figure 3.3. Western blot and densitometry analysis results for MARRS and β-Actin 36 expression in mammary epithelial cells. Figure 3.4. Immunohistochemistry of mammary glands depicting anti-MARRS 37 staining. CHAPTER 4 Figure 4.1. Hind foot abnormalities between deficient and replete mice at 6 wks 54 of age. Figure 4.2. Mean plasma 25(OH)D levels of mice at 6 wks of age. 55 Figure 4.3. Mammary gland growth variables: average number of TEBs and 56 area of ductal branching coverage (% of total fat pad) of whole mounts of mammary glands at 6 wks of age. Figure 4.4. Wholemount representations of 6-week old female mammary glands. 57 Figure 4.5. Western imunoblot of mammary gland epithelial cells depicting 59

MARRS protein expression. Figure 4.6. Immunohistochemistry results for right abdominal mammary glands 60 depicting anti-MARRS antibody Chapter 5 Figure 5.1. Mean body weight of 6-week old mice. 77 Figure 5.2. Mean % area of branching coverage throughout the mammary gland. 78 Figure 5.3. Mean ductal extension of alveolar branching through mammary glands. 79 Figure 5.4. Whole mounts of female mammary glands at 6 weeks of age. 80 Figure 5.5. Mean TEBs per genotype of 6-week old mice. 81 Figure 5.6. Western blot and densitometry analysis of MARRS and VDR in 82 mammary gland epithelial cells. Figure 5.7. Immunohistochemistry results of anti-MARRS and anti-VDR staining 83 in 6-week old female mammary glands. Figure 5.8. Reverse-transcribed RNA analysis of TGFβ and ERα in mammary 84 epithelial cells.

LIST OF ABBREVIATIONS

1,25D3 1,25-dihydroxyvitamin D3, calcitriol 24,25(OH)D 24,25-dihydroxyvitamin D3 25(OH)D 25-hydroxyvitamin D3, calcidiol Akt protein kinase B ANOVA analysis of variance Areg amphiregulin BMI body mass index DAB 3,3’-Diaminobenzidine DBP vitamin D binding protein DD vitamin D3 deficient diet (0 IU/kg) DMBA dimethylbenzanthracene DRI dietary reference intake DR vitamin D3 replete diet (10,000 IU/kg) DS vitamin D3 sufficient diet (1,000 IU/kg) EGF epidermal growth factor EGFR epidermal growth factor receptor EMT epithelial to mesenchymal transition ERα estrogen receptor alpha ERK extracellular signal-regulated kinase ERKO estrogen receptor knockout GH growth hormone IGF-1 insulin-like growth factor IU international unit MARRS membrane associated rapid response steroid MG mammary gland MMP matrix metalloproteinases

NF-κB nuclear factor kappa B PBS phosphate buffered saline PI3K phosphatidylinositol-3-kinase PKA protein kinase A PKC protein kinase C PR progesterone receptor RANKL receptor activator of nuclear factor kappa-b ligand TBS-T tris buffered saline with Tween TEB terminal end bud TGFβ transforming growth factor-beta TRPV6 transient receptor potential vanilloid member 6 UV ultra violet VDR vitamin D receptor VDRE vitamin D response element

xii

Chapter 1: Literature Review

1.1 Introduction

The mammary gland is a unique organ in that it does not develop into a fully differentiated tissue in utero. At birth, only a rudimentary tissue and ductal network is established, with full development occurring post-natally during puberty and pregnancy. At the onset of puberty, the mammary gland is susceptible to a plethora of hormones, growth factors, and other regulatory signals that maintain controlled and organized development. At this time, proper signalling is crucial to avoid long-term consequences, including tumorigenesis. This chapter will review the process by which the mammary gland develops during puberty, the main hormones that contribute to regulated growth, and the role of vitamin D in pubertal morphology through both the nuclear and membrane vitamin D receptors.

1.2 Murine Mammary Gland Development

1.2.1 Prenatal Development

In mice, mammary glands begin to form from the ectoderm and mesoderm at embryonic day 10, driven by signalling contained within the tissue [1,2]. The ectoderm forms two multilayered structures that run down the length of the embryo termed the milk lines, which eventually recede into five pairs of placodes on the ventral side [1]. Placodes are thickened clusters formed by the migration of ectodermal cells driven by signalling factors from the surrounding mesenchyme [3]. Later, these placodes form mammary buds along with some of the mesenchyme [4]. In male embryos, these buds are abolished when exposure to testosterone around day 14 causes the mesenchyme surrounding the gland to condense, resulting in necrosis

[5]. Meanwhile in females, the mammary bud undergoes cellular proliferation into the mesenchyme under the influence of estrogen, forming a rudimentary ductal tree [6]. By the end of gestation, the mammary gland consists of the fat pad with minimal ductal branching consisting of single-layered epithelial cells forming the lumen. The resulting organ remains essentially unchanged from birth until puberty [7].

1.2.2 Terminal End Bud Structure

The main goal of puberty, with regards to the mammary gland, is to establish a fully completed ductal network of alveolar branches within the fat pad with room for lobuloalveolar units when needed during pregnancy (See Figure 1.1). This is accomplished through systemic and paracrine signalling of several ovarian and pituitary hormones triggered for release at the onset of puberty.

The driving force of pubertal morphology is the terminal end bud (TEB). TEBs are club- like structures comprised of rapidly dividing epithelial cells; through bifurcation, TEBs work to lay the foundation of the mammary tree [8]. The TEB contains two main types of cells: cap cells at the edge of the TEB, and multiple layers of unorganized body cells within the club-structure.

Cap cells generally give rise to the basal layer of the ducts – the myoepithelial cells, while body cells differentiate into the luminal cells [9]. However, the cap cell compartment also contains mammary stem cells that can give rise to each cell type in the mammary tree [9]. In fact, mammary stem cells have been identified as those cells lacking CD31 and CD45 (endothelial and haematopoietic respectively) antigens, as well as containing high levels of CD29 and CD24, and can re-populate a cleared mammary fat pad into a fully-functioning organ from just a single cell [10]. The mechanism by which the TEB achieves forward movement through the fat pad is

Figure 1.1 Overview of mammary gland development with schematic of the terminal end bud showing organization of epithelial cells. From Paine & Lewis (2017) published in the Journal of Mammary Gland Biology and Neoplasia [8] with open access at Springerlink.com. Used with permission under the terms of the Creative Commons Attribution 4.0 International License.

(http://creativecommons.org/licenses/by/4.0/)

still under investigation, but there are a few theories in the literature. Firstly, unlike other migratory cells in other mammals, TEBs do not extend any protrusions to pull the cells forward

[11]. Instead, one theory, termed collective migration, states that all cells move together, but remain adherent with tight junctions [11]. One study found that while luminal and basal-laminal cells remained polarized and adherent, the interior cells of the TEB remain adherent only through desmosomes, and do exhibit some migratory protrusions [12]. Another theory states that duct cells seem to undergo an epithelial to mesenchymal transition (EMT) as they express transcription factors common in EMT including Slug (member of Snail transcription factors)

[13,14], which aid in their motility. Another theory is the actions of matrix metalloproteinases

(MMP) in front of the TEB breakdown the extracellular matrix to allow easy passage forward.

MMP2-/- mice lack proper invasion of the ductal branching into the fat pad [15]. However, it has been shown that MMP14-/- and MMP15-/- mice develop normal mammary glands [16].

Therefore it is likely a combination of mechanisms that allow TEBs to move through the fat pad.

Ductal elongation via TEBs usually concludes around 8 weeks of age [7].

1.2.3 Hormonal Regulation of Pubertal Mammary Gland Development

Estrogen is a critical signal in TEB formation. Early studies investigating the role of estrogen on mammary morphology using ovariectomized pre-pubertal mice found that TEBs failed to form, the diameter of the ductal tree was decreased [17] and ductal elongation was stunted [18]. Additional studies adding estradiol implants showed the results of an ovariectomy could be reversed, with histologically normal TEBs reappearing after only 3 days of treatment

[19]. Further studies explored the relationship between epithelial and stromal estradiol signalling with tissue recombination experiments. Injecting epithelial mammary cells from neo-natal estrogen-receptor knockout (ERKO) mice into fat-pads of wild-type mice with the epithelial

structures removed, still resulted in properly growing mammary trees. ERKO mice who received epithelial mammary cells from wildtype mice showed no ductal growth [20]. Thus it was concluded that stromal estrogen signalling was more critical to mammary development than epithelial signalling. However, later studies found that the timing of these recombination studies was critical, and that neo-natal mice do express Estrogen Receptor alpha (ERα) in the mammary gland, but it is not responsive until puberty [21]. Subsequent studies performing tissue recombination surgery on mice at least 3 weeks of age, found that although both compartments of estrogen signalling are important, the epithelial estrogen receptors play a more significant role in TEB formation and ductal elongation [18]. The mechanism by which estrogen acts upon epithelial cells is in a paracrine fashion, as only ~40% of epithelial cells express the ERα [22].

17-β-estradiol upregulates the epidermal growth factor-like ligand amphiregulin (Areg). Areg -/- mice do not develop TEBs, and the ductal tree does not extend past its rudimentary pre-pubertal state [23]. Areg is secreted by ERα+ cells, and binds the epidermal growth factor receptor

(EGFR) in the stroma and thus drives cellular proliferation [24]. This theory is strengthened by the fact that TEBs do not proliferate if the stroma is EGFR-/- [25].

Another important mediator in TEB formation is growth hormone (GH). GH is a pituitary hormone, but can also be secreted by the ovary, and mRNA expression of GH and GH receptor have been found in the mammary gland during puberty [26,27]. GH is essential, as GH-/- mice do not grow proper mammary glands [28]. The effects of GH are carried out by insulin-like growth factor (IGF-1), which when secreted by the liver travels to the mammary gland and binds

IGF receptors in the stroma [29]. It is known that the effects of GH act upon the mammary gland entirely through IGF-1, as IGF-1-/- mice do not develop proper mammary glands, and growth is not restored with estrogen treatment [30]. Early studies on animals found that GH and estradiol

work together to provide optimal growth: mammary glands from oophorectomized animals were stimulated with IGF-1 implants with, or without, added estrogen implants. Mammary glands with IGF-1 only experienced some development, but those animals with the added estrogen experienced the most enhanced growth [31,32].

TEB formation and ductal elongation is not the only process occurring during mammary development. After most of the ductal elongation has been completed, the ductal tree continues to grow laterally with the production of side branches during the mouse’s newly established estrous cycles [33], mainly under the influence of progesterone. Early studies using intact mammary glands with in situ progesterone implants found significant increases in ductal side- branching, more-so than estrogen implants alone [34]. Mammary glands of progesterone receptor

(PR) -/- mice completely lack interductal lobular structures, even when given high doses of estrogen and progesterone, but otherwise develop normal TEBs and complete ductal elongation

[35]. Tissue recombination studies were also conducted between PR +/+ and PR-/- mice to gain insight into stromal vs epithelial signalling. Mammary glands lacking PR in the epithelium developed similarly to PR knockout mice, while stromal PR-/- developed normally [36]. This information coincides with the fact that ~80% of all PR in the mammary gland are expressed on epithelial cells [37], making a strong argument that the target for PR is found in the epithelium, and stromal PR is not require for side branching.

The mechanism by which progesterone elicits its proliferation effects also depends on paracrine signalling, as like ERα, not all epithelial cells express PR [38]. Interestingly, progesterone can produce an effect in the absence of estrogen, demonstrated by tissue recombination studies [21] and ovariectomized mouse studies [38]. Firstly, PR+ cells utilize a

Cyclin D1-dependent mechanism to drive initial proliferation [38]. Estrogen alone can induce

Cyclin D1 expression, but the addition of progesterone significantly increases the response, which is not seen in PR knockout mice [39]. Later, PR- cells depend on receptor activator of nuclear factor kappa-B ligand (RANKL) signalling [38]. RANKL expression is observed in luminal epithelial cells regardless of PR status, and exposing pubescent mice to RANKL results in early side-branching [40]. RANKL has been shown to be driven by progesterone signalling in mouse studies where RANKL-/- mice treated with progesterone showed very little epithelial proliferation [38].

Overall progesterone seems to have a much more distinct role in the production of alveoli during pregnancy, which this review will not discuss.

While puberty is a time for upregulated cellular proliferation, there are a number of signals in place to ensure growth does not continue un-checked. Transforming growth factor-beta

(TGFβ) is the most well documented signal of growth control in the mouse mammary gland, and has recognized roles in apoptosis and proliferation [41]. First reported using in vivo studies,

TGFβ implants were inserted into mouse mammary glands and left for 4 days in a pubescent mouse [42]. In this time, number of TEBs was reduced by 75% and there was no elongation of the ducts. The TGFβ implants were then removed, and after 11 days it was found that the number of TEBs had increased to the same number as control mice, and ductal elongation had resumed – although they had only travelled a third of the distance of the control mice [42]. Studies using tissue recombination, TGFβ-/- mice [43], and overexpression of TGFβ in vivo [44] confirm a role in ductal elongation. TGFβ signalling is essential enough that TGFβ +/- animals also experience reduced proliferation in mammary epithelium [45]. TGFβ may be regulated by ovarian hormones, as the number of epithelial cells expressing TGFβ are reduced during the proliferative stages of the estrous cycle [45]. TGFβ initiates cell signalling by activating serine-

threonine kinase receptors and the signalling proteins Smad2 and 3, ultimately affecting gene transcription [41,46]. In the pubertal mammary gland, Smad3 is essential for TGFβ-mediated apoptosis [47], but Smad 2 is not [46]. Unlike the other growth factors and hormones responsible for pubertal growth, TGFβ seems to work in an autocrine fashion, as TGFβ+ cells also express activated Smad proteins [48]. TGFβ also inhibits proliferation by arresting the cell cycle in the

G1 phase by suppressing c-Myc [49] and altering the binding capabilities of cyclin-dependent kinases 4 and 6 [50]. There is also evidence of TGFβ-induced apoptosis during involution after pregnancy [51], but this process will not be reviewed here.

1.3 Role of Vitamin D in Mammary Gland Development

1.3.1 Vitamin D Metabolism

Contrary to many other vitamins and nutrients, vitamin D is not readily available in the human diet. Foods with some vitamin D include fatty fish, egg yolks and mushrooms [52]. Due to Canada’s limited and varying UV exposure during the year, the government of Canada has fortified many foods with extra vitamin D to help Canadians achieve the daily-reference intake

(DRI). Foods such as orange juice, milk and other dairy products contain approximately 100 international units (IU) of vitamin D per cup [53].

When uncovered skin is exposed to UVB light, 7-dehydrocholesterol is converted to cholecalciferol (vitamin D3) [54]. Food sources of vitamin D are either in the D2 (ergocalciferol) or D3 form, and absorbed into the lymphatic system by chylomicrons [52]. Vitamin D3 travels through the bloodstream bound to a vitamin D-binding protein (DBP) to the liver, where 25-α- hydroxylase (CYP2R1) will convert it to 25-hydroxyvitamin D (25(OH)D, or calcidiol). [54].

25(OH)D is a stable, yet largely inactive compound, and able to travel around the body bound to

DBP to be delivered to a target tissue – usually the kidney. 25(OH)D is the most abundant form of vitamin D in circulation, and is therefore used as a biomarker of vitamin D status [55]. At the kidney, CYP27B1, or 1-α-hydroxylase will hydroxylate the molecule for a second time creating the active 1,25-dihydroxyvitamin D3 (1,25D3, or calcitriol) [56]. Although 1-α-hydroxylase has been extensively documented in the kidney, there is evidence for its expression in other tissues with 1,25D3 actions, including the mammary gland [57]. Both 25(OH)D and 1,25D3 can be converted to the inactive 24,25-dihydroxyvitamin D (24,25(OH)D) at the kidney by CYP24A1 when 1,25D3 levels are too high [55].

1.3.2 Classic Role of 1,25D3

The classic role of 1,25D3 involves binding to the nuclear vitamin D receptor (VDR) at target cells. After binding, it forms a heterodimer with the retinoid X receptor, then binds to

Vitamin D Response Elements (VDREs) found within the genome, which are responsible for the up- or down-regulation of gene transcription [55]. The most well understood actions of 1,25D3 are calcium and phosphate regulation at the intestine, kidney, and bone.

The main action of 1,25D3 is calcium uptake at the intestine [55]. When plasma calcium is low, activation of 1,25D3 by parathyroid hormone is increased to ensure adequate expression of the calcium membrane channel transient receptor potential vanilloid member 6 (TRPV6), as well as binding protein calbindin-D9k for facilitated diffusion of calcium across the intestinal membranes [58,59]. However, both calbindin-D9k and TRPV6 knockout mice do not suffer from any reduction in intestinal calcium absorption, and treatment with 1,25D3 resulted in the same calcium absorption as wildtypes [60,61], therefore providing evidence for the existence of

another absorption mechanism. Similar proteins, TRPV5 and calbindin-D28k, work to increase calcium reabsorption at the kidney [55].

At the bone, 1,25D3 also has direct actions in response to low plasma calcium. Activation of 1,25D3 in bone leads to direct transcriptional regulation of RANKL expression on osteoblasts

[62] , and the direct cell-cell contact of RANKL to RANK receptors on osteoclasts triggers their activation, leading to bone demineralization [63]. 1,25D3 is also critical for healthy bone, evidenced by the high incidence of rickets in populations with vitamin D deficiencies [52], and

VDR knockout mice developing rickets, among other abnormalities [64]. VDR signalling is crucial for osteoblasts to transition into healthy osteocytes [65], and a deficiency in vitamin D will result in up to a 90% reduction in dietary calcium absorption [52], therefore resulting in low plasma calcium and the subsequent calcium-saving actions of demineralizing bone.

Currently, it is these actions of calcium uptake and bone health that are the only effects

Health Canada recognizes when setting vitamin D DRIs for Canadians [53].

1.3.3 Regulation of Mammary Gland Development by VDR

Exploring the role of 1,25D3 and VDR in mouse models is usually completed by using

VDR ablation. The VDR is commonly knocked out by removing the second zinc finger of the

DNA-binding site. First used by Li et al. [64], they found that the percentage of VDR null pups was high enough to rule out embryonic death due to loss of the receptor. Overall, these mice develop a phenotype similar to human patients with rickets: instances of slow growth, truncated leg bones, hypocalcemia, hypophosphatemia, and alopecia [64]. Therefore these mice need to be fed special diets containing extra calcium, phosphorous, lactose and vitamin D3 to maintain proper health and reach age of maturity. Mammary glands of VDR null mice have been studied

at 4, 5, 6, 7, 8, and 10 weeks of age, and growth has been assessed by whole mounting abdominal glands and examining the length of ductal branching and the number of TEBs. VDR null mice showed enhanced gland development during puberty. There was increased ductal branching, increased numbers of TEBs, and mammary fat pads were significantly larger compared to whole body size [66]. Thus the VDR plays a role in inhibiting over-growth of the mammary gland during puberty.

A subsequent study explored the role of VDR signalling in the epithelium versus adipocytes of the mammary gland using conditional knockout mice [67]. Results of this study were not as obvious as whole-body VDR ablation, as TEBs remained relatively the same between knockouts and wildtypes, yet ductal extension was still increased in VDR knockouts regardless of which compartment was targeted. Epithelial VDR knockouts experienced increased alveolar budding – which normally occurs during pregnancy and lactation, not pubertal development. This was explained by a heightened sensitivity to ovarian hormones compared to wildtypes. However, loss of VDR in either compartment did not change estrogen or progesterone receptor expression within the tissue [67].

1.4 Non-Genomic Actions of 1,25D3

1.4.1 Evidence of a Second Receptor

There has been controversy in the literature as to whether the MARRS (Membrane

Associated Rapid Response Steroid) receptor is indeed a separate molecule from the VDR, or if the VDR is simply translocating to the plasma membrane and causing rapid effects through intercellular signalling pathways.

Suggesting a plasma membrane receptor was capable of binding 1,25D3 was first proposed by Nemere and Szego in 1981 [68]. Rat intestinal cells were exposed to 1,25D3, and calcium release was measured in vitro in as little as 5 minutes. They suggested that the result was due to a receptor in the plasma membrane, however, they believed it was not due to the classical receptor working from the membrane [69]. Since then, a number of researchers have found evidence for a second receptor, and studying its signalling effects in various tissues. For example, Nemere et al. [70] studied duodenal mucosa of chicks and demonstrated specific binding to a basal-lateral membrane receptor. This receptor seemed to be mostly responsible for rapid intestinal calcium transport and signalling, and preferred to bind cis-analogs of 1,25D3.

The closed configuration of this analog is incapable of binding the classical VDR which requires

1,25D3 to be in the trans conformation [71].

Furthermore, when studying calcium uptake in intestinal cells, it was noted that this second receptor bound to antibody 099, while the classical VDR bound antibody 9A7, and in using both antibodies different results were seen [72]. Eventually, this new receptor was found to be the 57 kDa protein ERp57, also named GRp58, or PDIA3 [71], and when mentioned with respect to vitamin D signalling, is referred to in the literature as MARRS (or 1,25D3-MARRS).

The presence and function of MARRS has since been studied in multiple cell types, including the intestinal epithelium, chondrocytes, hematopoietic cells and human breast cancer cell lines. Although MARRS has not yet been studied in an animal model in relation to breast cancer, important functions have been described for it in other tissues concerning classical VDR targets.

1.4.2 Calcium and Phosphate Uptake

The role that MARRS plays in calcium and phosphate homeostasis in the intestine has mostly been studied using chicks. Early studies looked at calcium and phosphate uptake in vitro from cells taken from vitamin-deficient chicks. A vitamin deficiency in the animal resulted in an increase in 1,25D3 binding to MARRS in vitro, as well as increased calcium ion uptake [73].

These results suggest that MARRS may have a more important role when an individual is vitamin D deficient, and attempts to extract as much calcium from the diet as possible [73].

When MARRS was studied in vivo, conditional intestinal cell knockouts were created using the flx/Cre system and were then treated with 1,25D3 [74]. Results indicated that by knocking out MARRS, rapid calcium uptake was diminished, suggesting that it is indeed

MARRS that facilitates rapid calcium uptake in the intestines of mice rather than the classic

VDR [74]. Treatment with 300 pM 1,25D3 quickly resulted in phosphate uptake in the enterocytes of controls, but no phosphate uptake was reported in the knockouts [75]. Regarding the signalling pathway, there were differences seen between sexes. Female mice exhibit calcium uptake through both the protein kinase A (PKA) and protein kinase C (PKC) signalling cascades, whereas males activate PKA only [74]. Similar to calcium, MARRS seems to activate the PKA pathway for phosphate uptake. However, adult male mice lost the ability to uptake phosphate with age whereas females did not. Therefore, estrogen may also be involved in this signalling pathway [75].

Similar studies of phosphate and calcium uptake have also been conducted on chick kidney cells with remarkably similar results. Initial phosphate uptake occurs within seconds of exposure to small concentrations of 1,25D3, and is completely eliminated when cells are

incubated with Antibody 099 against MARRS [76]. Overall, although PKC signalling has generally been deemed responsible for calcium channel control, studies involving MARRS suggest that PKA signalling supersedes when this receptor is activated [77].

1.4.3 Chondrocytes and Osteoblasts

Study of the MARRS receptor in bone health has been done on both human prenatal craniofacial samples and mouse models. Human fetal craniums were found to possess the

MARRS protein when tested against the chicken duodenal Ab 099 antibody [78]. During the 4th week in utero, MARRS was detected in abundant concentrations within chondrocytes, osteoclasts and osteoblasts. Regarding tooth formation, MARRS could also be found in the dental epithelium as early as week 7 of development, and was found in high concentrations in the areas of future tooth mineralization [78], suggesting an important role for MARRS in bone formation and health. Indeed, MARRS is thought to control the movement of bone-mineralizing vesicles in the growth plate, as MARRS is missing from areas without vesicles [78,79].

Cell culture studies in animal chondrocytes have shown that 1,25D3 mediates its rapid effects through the phospholipase C signalling cascade that ultimately activates PKC and the extracellular signal-regulated kinase (ERK) pathway to facilitate cartilage growth [80]. MARRS has also been shown to activate calcium ion channels in rat osteoblasts [81]. To provide evidence for this MARRS signalling pathway, Boyan et al. [80] used the VDR knockout mouse to show that PKC activation increased in response to 1,25D3 in both the transgenic mouse and the control. Furthermore, the Ab 099 antibody against MARRS reduced the PKC activity when chondrocytes were treated with 1,25D3. Therefore, chondrocytes were able to respond to 1,25D3 without the classic nuclear receptor, and were most likely working through MARRS. The full

pathway, as described by Boyan et al. [82] involves 1,25D3 binding MARRS, activating phospholipase A, releasing arachidonic acid and activating PKC.

1.4.4 Haematopoeisis and Leukemia

Myeloid cells require the activation of phosphatidylinositol-3-kinase (PI3K) signalling cascades to activate protein kinase B (Akt) and nuclear factor kappa B (NF-kB). Ultimately, the enzyme sulphatase is activated, which is required to transform inactive estrogen and androgens into their active form, as well as mediate myeloid cell differentiation [83,84]. When studying

NB4 myeloid cells in vitro, Hughes et al. [84] noticed that NF-kB translocation to the nucleus was required for sulphatase activation. 1,25D3 is known to activate this signalling pathway, but it was always assumed that the actions were mediated entirely through the nuclear VDR. Even though the pathway is activated too rapidly for genomic responses to be responsible, researchers believed that the VDR was initiating the signalling events in the cytosol. It was not until 2010 when Wu et al. [71] demonstrated that the MARRS protein was expressed in NB4 cells. When these cells were treated with 1,25D3, both MARRS and NF-kB translocated to the nucleus. Since the time frame for movement was comparable for both molecules, it may actually be MARRS that is responsible for activating this signalling cascade, as opposed to the nuclear VDR [71].

Further studies involving MARRS knockout and knockdown cells in the myeloid line have been carried out in mice. C57Bl mice had MARRS knocked out using the flx/Cre system with Cre being expressed in the Lyz2 gene, affecting myeloid cell lineage. Mice lacking the

MARRS receptor had fewer differentiated white blood cells than wildtypes. Knockdowns also had increased occurrences of splenomegaly and lymphomas, which in humans is indicative of

myeloid leukemia. Although further research is needed in this area, it seems MARRS plays a critical role in myeloid cell differentiation (unpublished results, Meckling lab, 2013).

1.4.5 Lifespan

Caloric restriction and low body mass index (BMI) has been linked to increased lifespan

[85]. It has also been noted that MARRS knockout mice are smaller in size compared to wildtype littermates, leading researchers to wonder if MARRS played a role in caloric restriction or increased lifespan. Mice with floxed MARRS gene were bred to mice with the cre-recombinase under the villin promoter. It was found the female knockouts had less fat mass than littermate females, and the interesting discovery that estradiol competes with 1,25D3 for binding to the

MARRS receptor [86]. Furthermore, MARRS knockout animals have a 30% increased lifespan compared to littermates [86].

I.5 Vitamin D and Breast Cancer

1.5.1 Epidemiological Studies

In Canada, it is estimated that 1 in every 2 people will develop cancer in their lifetime, and 1 in 4 will die from cancer. Breast cancer accounts for 13% of all cancers diagnosed in

Canada, and 25% of all cancers diagnosed in women, making it the most common type of cancer in women [87].

Meta-analyses of 25(OH)D status and breast cancer, although sometimes contradictory, often show an inverse relationship between 25(OH)D status, breast cancer risk and survival post- diagnosis [88]. Breast cancer patients with higher concentrations of vitamin D receptors in the

breast tissue [89], and higher concentrations of 25(OH)D [90] experience higher survival rates.

Increasing vitamin D intake through the diet, has also been shown to decrease breast cancer risk in pre-menopausal women [91]. A more recent study investigating 25(OH)D status and breast cancer risk, depending on receptor type, found there was no risk to having adequate 25(OH)D levels [92]. In contrast, The Women’s Health Initiative trial found no association between 400IU daily of vitamin D3 and breast cancer incidence in postmenopausal women [93]. However, some researchers believe that the optimal 25(OH)D serum level is between 75 and 80nmol/L which requires an intake of 800 to 1000 IU/day [94] yet even higher intakes are believed to be required to produce anti-cancer effects [95].

Historically, observational studies looking at rates of breast cancer in relation to geographic location indicated that increased exposure to ultra-violet (UV) rays resulted in a decrease in breast cancer mortality, leading to the hypothesis that 25(OH)D is related to cancer protection [96]. However newer observational studies such as Zamoiski et al. (2016) are finding less of an association between UV exposure/time spent outdoors and reduced cancer risk [97].

However, this latter study did not account for actual time spent in direct sunlight (as opposed to in the shade), or sunscreen use – which has been consciously integrated more into everyday cosmetic products. Other reasons for the reduction in correlation between the 1950s and now could be the increased contribution of other risk-factors [98] including obesity and physical activity, which also affect 25(OH)D status [88].

1.5.2 1,25D3 and Cell Growth Inhibition in Cell Culture

1,25D3 has potent growth inhibition effects on many tissues, including breast cancer cells, as demonstrated in many in vitro studies. When MCF-7 human breast tumour cells were

exposed to differing concentrations of 1,25D3, they experienced a dose-dependent growth inhibition, seen even at picomolar concentrations [99]. 1,25D3 is known to cause these effects in breast cancer cells by inhibiting insulin-like growth factor, which is often upregulated in cancer cell lines [100]. 1,25D3 can induce apoptosis through DNA fragmentation, nuclear matrix reduction and upregulating genes associated with programmed cell death such as clusterin and cathepsin B [101], as well as stop the cell cycle from progressing past the G1 phase by activating several cyclin-dependent kinase inhibitors [102]. Although using 1,25D3 as a treatment for cancer in humans would require concentrations that may induce hypercalcemia [101], the same apoptotic results are seen with vitamin D analogs such as EB1089 [100]. In some studies, use of an analog has induced stronger results than 1,25D3 alone [101] and are being studied in relation to their effects on blood calcium and phosphate homeostasis.

Although it was originally thought that 25(OH)D could only be converted to the active

1,25D3 form in the kidney, more tissues have been found to possess their own locally-acting vitamin D hydroxylases [103]. Non-cancerous human mammary cells, when exposed to

100nmol/L of 1,25D3 in vitro, significantly upregulated the vitamin D hydroxylase CYP27B1 gene and protein expression, demonstrating these cells’ ability to locally activate 1,25D3 and its receptor, and therefore auto-regulating the tissue’s cell proliferation cycle [103]. Furthermore, when human mammary cells have their vitamin D pathway disrupted, CYP27B1 expression,

VDR expression and growth inhibition via 1,25D3 is reduced [104].

Dimethylbenzanthracene (DMBA)-induced mammary tumours have been tested for their sensitivity to 1,25D3 ex-vivo. Cell lines from both wildtype mice and VDR knockouts were treated with 100 nM 1,25D3 for 96 hours [105]. Control cells experienced significant growth suppression when exposed to 1,25D3 while knockouts had no change in their growth curve.

Furthermore, wildtype cells experienced an increased level of apoptosis, while knockouts did not experience any changes in cell cycle development or programmed cell death [106]. Similar results were also found when mammary tumour cells were exposed to Vitamin D analogs such as

EB1089, CB1093, MC1288 and KH1230 [107]. However, at some higher µM concentrations of

1,25D3, knockout cells were found to experience some growth inhibition, suggesting that there are VDR-independent pathways operating in this tumour type [107]. Overall, these results show that tumorigenesis in mammary epithelium can be slowed with the presence of the VDR [105].

1.5.3 VDR Knockouts and Tumorigenesis

The effects of carcinogen-induced tumorigenesis have been studied in VDR knockout mice as well. VDR whole-body knockouts on a C57Bl6 background were exposed to DMBA in vivo to examine its carcinogenic effects. In vivo administration was done at a dose of 1mg/10g body weight at both 5.5 and 7 weeks of age, and mice were euthanized at 7 months of age [108].

DMBA was found to be a reliable method of inducing mammary tumours, despite the

VDR status. Tumours were found to be of epithelium origin due to the expression of cytokeratins

[105]. Mice with no expression of VDR had increased instances of lobular and alveolar hyperplasia, however, the percentage of mice that developed tumours (74%) was identical to the percentage of control mice with tumours (75%). Yet, the type of tumour differed between groups. VDR null mice were more prone to squamous cell carcinomas than controls [108]. These results suggest that VDR knockouts transformed mammary epithelium into a squamous cell type

[108]. Unexpectedly, knockout mice were also prone to developing tumours in other cell types:

VDR knockouts had high incidence of skin tumours, thymic lymphomas, lymphoblastic lymphomas and lung adenomas [108].

VDR ablation has also been executed in MMTV-neu mice, a model of human breast cancer. These mice again expressed pubertal mammary glands with changes to their growing structure, with increased budding and thickened ducts. Mammary fat pads of knockout mice at

10 months old showed signs of increased hyperplasia compared to wildtypes; tumours were detected in the majority of knockout animals in the study, and were detected earlier than in wildtypes. Finally, VDR-/- mice also exhibited decreased survival rates and minimal abdominal fat stores. [109].

1.5.4 MARRS and Breast Cancer

The MARRS receptor has recently been studied in relation to 1,25D3 and breast epithelium in vitro. In a study by Richards et al. [110], MCF-7, MDA-MB 231, and MCF-10A cells were grown in vitro. MCF-7 and MDA-MB 231 cells are both human breast cancer cells, while MCF-10A are non-cancerous human breast epithelial cells used as a control. All cell lines were shown to express the MARRS receptor in all parts of the cell, but less so in the nucleus

[110]. Furthermore, there was more MARRS protein present in the cancer cell lines compared to control. Using plasmids with an anti-MARRS ribozyme, MCF-7 cells had their MARRS protein effectively knocked-down by 50%. It was found that VDR expression did not change in relation to MARRS expression. All cells were then exposed to treatments of 1,25D3. The MCF-7 cells with knocked down MARRS expression experienced a greater sensitivity to 1,25D3 treatments: they exhibited greater inhibited growth, and more cells paused in G0 growth phase than MCF-7 cells with normal expression of MARRS. Similar findings were also noted with vitamin D analogs KH1060 and MC903. These results show that breast cancer cells experience growth inhibition in reaction to a subtle change in MARRS protein, and may provide a novel approach to treating breast cancer in humans.

MARRS may also play a role in the epidermal growth factor (EGF) signalling pathway, which ultimately also joins with the PLC-PKC pathway [111]. MARRS was knocked down in

MDA-MB-468 human breast cancer cells using RNA interference, and this resulted in a decrease in EGF phosphorylation and deactivation. Since EGF has been known to increase in human breast cancer cases, an increase in MARRS activity may also be beneficial in the prevention and treatment of breast cancer [111].

1.5.5 Connection between TEBs and Breast Cancer

TEBs possess many qualities similar to developing tumours including rapid cellular proliferation, and EMT capabilities. Due to their vulnerability to hormones and growth signals, as well as their low degree of differentiation, they have become a site of interest when studying breast cancer, as aggressive tumours often develop from less differentiated structures [112].

Early animal experiments studying the relationship between TEBs and cancer exposed pubescent rats to DMBA and monitored them for breast tumorigenesis for 8 weeks thereafter. Treated rats had increased TEB density within the developing mammary gland, larger TEBs, increased mitotic activity within TEBs, and fewer differentiated alveolar structures post-puberty [113].

Importantly, they also found that incidence of tumorigenesis correlated with the number of TEBs at the time of DMBA treatment [113]. Because TEBs are highly susceptible to carcinogenesis and mutagenesis due to the increased mitotic activity, and are highly undifferentiated structures, the higher number of TEBs correlated with increased mammary tumorigenesis [114]. And similarly, as observed in rat studies, fewer TEBs during puberty and more fully differentiated alveolar structures were seen in adult rats with lower incidences of tumours [114]. Similar animal studies have been repeated over the years, testing dietary components and chemicals often associated with breast cancer risk in humans. Each study compares the influence of the

treatment on TEB numbers or density and results repeatedly show that a decrease in TEBs results in the animals experiencing reduced tumorigenesis [115–118]. Proving a connection between

TEBs and breast cancer in humans is much more difficult due to the tissue sampling required.

Finding a connection between adolescent behaviour and adult risk of cancer is also a lengthy process few studies have explored. Some studies have found a link between adolescent soy intake and decreased breast cancer risk [119,120], and some have proposed a mechanism linking soy to estrogen and altered mammary gland development [121], but the shift to investigating adolescent intakes for vitamin D is more recent and therefore not substantially explored. Past observational studies usually assess vitamin D intake and status at time of breast cancer diagnosis, or in post-menopausal women, although there are a few recent studies looking at vitamin D exposure in adolescents. Data within the Nurse’s Health Study II found that women with the highest vitamin D intake during adolescence (>500 IU/day) had a 21% lower risk of benign breast disease (a marker of breast cancer risk) [122], and similar results were found in the

Growing Up Today Study of NHS participants’ children [123]. Anderson et al. (2011) conducted a case-control study of Ontario breast cancer patients and age-matched controls. They found that time spent outdoors during teenage years was associated with reduced breast cancer

[124], although associations were stronger in older age groups. Conversely, Knight et al (2007) found consistent inverse associations between both sun exposure and increased vitamin D dietary intake during adolescence and breast cancer [125]. One difference between these studies is

Knight’s questionnaire was focused on evaluating vitamin D exposure as opposed to total time in sun, as spending 30 minutes outdoors during summer would likely allow for maximum 25(OH)D production in their Caucasian population [125].

Further evidence towards mammary gland development having a great impact on adult breast cancer risk are the studies linking lower age of first menarche to increased risk. Modeled by Pike [126] and Colditz [127], they hypothesized that the time between menarche and first full- term pregnancy has significant impact on breast cancer risk due to the susceptibility to DNA damage, and that damage being carried out through the development period until the mammary gland is fully differentiated after pregnancy [128]. Epidemiological studies have confirmed this hypothesis by finding associations between age of first menarche and breast cancer risk continuously over the years [129,130]. Data from the Nurses’ Health Study II have also noted links between alcohol consumption before first pregnancy and increased breast cancer risk, with the associations being stronger in women with 10+ years between the start of menstruation and first pregnancy [131] demonstrating the susceptibility to carcinogenic events at this time.

1.6 Conclusion

Although MARRS has been studied in a variety of tissues, its role has yet to be studied in breast tissue and mammary glands in an animal model. Since 1,25D3 and VDR have been shown to play an influential role in mammary gland development and breast cancer inhibition, and

MARRS has been shown to have important roles in the same tissues as 1,25D3, it is therefore possible that MARRS also plays a part in mammary gland development. Based on previous in vitro experiments regarding MARRS signalling in breast cancer cells, MARRS seems to have independent actions from the nuclear VDR, and warrants study in an in vivo model.

Chapter 2: Rationale and Objectives

Although there is a growing amount of data supporting a new receptor for 1,25D3, this remains a controversial topic within vitamin D research, and many researchers believe it is the nuclear vitamin D receptor translocating to the surface membrane and triggering non-genomic cell signalling from there. Further, MARRS has so far been studied in tissues with known VDR actions (intestine, bone), yet nothing is known about the actions of MARRS in the pubertal mammary gland. There is some data on the actions of MARRS in breast cancer cells, but no in vivo data. Therefore, the objectives of this thesis were to determine: 1) if MARRS plays a significant role in mammary gland development during the pubertal phase, 2) if changes in

25(OH)D status would affect the actions of MARRS (as it does in cell culture [110]), and 3) if either the MARRS or VDR pathway supersedes the other within the mammary gland in regards to epithelial growth. All these objectives were studied in a tissue-specific knockout mouse model.

We hypothesize that knocking out MARRS will have the opposite effect on epithelial growth than VDR knockouts; specifically, a reduction in MARRS epithelial expression will result in hindered or slowed growth. We also anticipate that the loss of MARRS receptors in epithelial cells will enhance the anti-growth effect of 1,25D3 (based on a previous in vitro study

[110]), and if the MARRS pathway has a more dominant role in epithelial growth, this will manifest in double knockouts displaying the phenotype of the MARRS knockout.

Chapter 3: Role of the MARRS receptor in murine mammary gland growth and development

3.1 Abstract

The protein disulfide isomerase MARRS (GRp58/PDIA3/ERp57) has been implicated in a multitude of signaling pathways throughout the entire body. Most thoroughly studied for its protein-folding role, MARRS has also been found to have multiple binding partners, and have significant effects on cellular growth. MARRS has been studied in the context of several neurodegenerative disorders, metabolic conditions, and can be used as a prognosis marker in certain cancers. One role, as an alternate vitamin D binding receptor (VDR), has prompted research in tissues with known vitamin D activity, such as the intestine and bone. Vitamin D has been studied in relation to mammary gland growth and development, but it is not yet known if

MARRS plays an independent role in this tissue. In this study, MARRS was knocked out in murine mammary gland epithelial cells of 30, 4-week old mice. Several markers of mammary gland growth were measured, including number of terminal end buds (TEB), ductal coverage of the fat pad, and ductal extension. It was found the knockout animals had decreased numbers of

TEBs (p=0.019), and decreased ductal extension (p=0.018) compared to wildtype animals, with no differences in gross body weight. Immunohistochemistry analysis of mammary glands showed MARRS localized to the apical side of alveolar branches, and on leading edges of TEBs.

These results provide further evidence for MARRS functioning separately from VDR, and insights into the roles of MARRS.

3.2 Introduction

The disulfide isomerase ERp57 (also known as GRp58/PDIA3/1,25-D3 membrane- associated rapid response steroid binding receptor - MARRS) was first discovered as an endoplasmic reticulum protein aiding in the folding of glycoproteins [76]. While most of a cell’s

ERp57 protein is located within the endoplasmic reticulum, it can also be found within the nucleus [71] or on the plasma membrane [132]. ERp57 is found in all tissues of the body [133], and is embryonic lethal in total-body knockouts [134].

One of the earliest roles attributed to ERp57 was as a chaperone protein, aiding in protein-folding through thiol-oxidoreductase reactions in proteins bound to calreticulin or calnexin [76,135]. A major responsibility of ERp57 includes participating in the folding, stabilization and loading of MHC class I heavy chains [136,137] and therefore plays a pivotal role in the immune system.

But Erp57 has been implicated in roles extending beyond its abilities at the endoplasmic reticulum. In vitro studies in NB4 acute promyelocytic leukemia cells revealed ERp57 complexes with nuclear factor kappa-B (NFκB) and translocates to the nucleus [71], a step that may be imperative for fate determination of these cells. Additionally, ERP57 is essential for the nuclear localization of Retinoic acid receptor-α (RARA) and aids in binding all-trans retinoic acid with RARA in Sertoli cells [138]. Changes in ERp57 activity are associated with cancer progression via several signaling pathways such as EGFR [111], and STAT3 [139], and can be used as a biomarker for certain cancer prognoses [139,140].

A novel role for ERp57 is by acting as a plasma membrane receptor for calcitriol resulting in rapid cell signaling effects differing from actions of the classic vitamin D receptor

(VDR), and from hereon referred to as MARRS [110,132]. Evidence for a receptor independent of the VDR includes studies performed in the basolateral membrane of intestinal epithelium in chicks showing the rapid uptake of calcium and phosphate was eradicated after tissue-specific knockout of the MARRS receptor [132]. Studies in chondrocytes demonstrate ligand binding to

MARRS elicits signal transduction that results in the activation of protein kinase C (PKC), protein kinase A (PKA) [141]. Therefore, a new avenue of research regarding MARRS is studying its activity in tissues with known calcitriol effects to elucidate any roles it may be playing.

One such tissue with reported calcitriol effects attributed to the VDR is the mammary gland. Mammary glands of both humans and mice begin development in utero, but complete the majority of development post-natally. Initial epithelial structures arise from the ectoderm, surrounded by mesenchymal cells which will ultimately become the fat pad [142]. In puberty, a rush of ovarian hormones and other growth factors stimulate epithelial cell proliferation, producing a framework of branched epithelial ducts [142]. These ducts are laid by a cluster of rapidly dividing cells in the formation of a Terminal End Bud (TEB). TEBs are highly influenced by extra-cellular signals, hormones, growth factors, and carcinogens [114], and are therefore an origin for tumorigenesis [115,143]. Studies using VDR knockout mice have confirmed an anti- proliferative role for the VDR within the mammary gland. In the absence of a functional VDR, mammary gland growth is explosive, resulting in high numbers of TEBs, elongated ducts, and widespread secondary branching [66].

This study’s purpose was to determine if MARRS played a role in murine mammary gland development. In the present study, we generated a tissue-specific knockout mouse model

with reduced expression of MARRS in the epithelial cells and monitored several variables in development to determine any changes in morphology.

3.3 Methods

3.3.1 Animals

C57Bl/6-Pdia mice with the MARRS gene floxed were obtained from Utah State

University and Tg(MMTV-cre)4Mam/J mice were purchased from Jackson Laboratories.

MARRS-floxed mice were crossed with MMTV Cre+ mice, and resulting progeny were backcrossed to the Tg(MMTV-cre)4Mam/J 4 times to reduce C57Bl/6 background. Resulting heterozygotes were bred to produce MARRS +/+ (n=9), +/- (n=9) and -/- (n=12) genotypes.

Mice were fed a standard Teklad S-2335 rodent chow, which contains 3000 IU/kg of vitamin D3.

Pups were toe-clipped at 6 days of age to obtain DNA, and were weaned at 28 days. Mice were housed in rooms exposed to a 12 h light cycles. All procedures were approved by the Animal

Care Committee of the University of Guelph, under Animal Utilization Protocol 3147.

3.3.2 Genotyping

Toe samples were digested in an alkaline lysis buffer for 30 mins at 95°C, then cooled on ice, and neutralized in an acidic buffer. DNA was genotyped using the following primers for

PCR: CGC CAG CCT CTC CAT TTA G (ERp57 forward); CAG AGA TCC TGC CTC TG

(ERp57 reverse); CTG ATC TGA GCT CTG AGT G (Cre forward); and CAT CAC TCG TTG

CAT CGA CC (Cre reverse). Electrophoresis was run using 1.5% w/v agarose gels at 100V for 1 hr, and stained with a carcinogen-free DNA stain (Abcam). A positive Cre result would show a single band at ~250bp, a MARRS +/+ mouse would show a single band ~222bp, MARRS-/-

would show a single band at ~410bp, and a MARRS +/- mouse would have two bands, one each at 222bp and 410bp.

3.3.3 Necropsy and Tissue Collection

At 4 wks of age, the estrous stage of the female was determined by performing a vaginal lavage with sterile phosphate buffered saline (PBS), and vaginal epithelial cells were used to stage the estrous cycle, described previously [144]. Mice were sacrificed if they were in any stage of their estrous cycle other than diestrus. At time of death, whole body weight was measured. Left abdominal mammary glands were collected for whole mounting, right abdominal glands were prepped for immunohistochemistry, and inguinal glands were prepped for Western blot.

3.3.4 Whole mounting Mammary Glands

Whole mounting was performed as described previously [145]. Left abdominal glands were collected and fixed in formalin overnight. They were then submerged in acetone for 48 hrs, then rehydrated in decreasing concentrations of ethanol, and stained in carmine alum overnight.

The next day glands were dehydrated, cleared in xylenes, and stored in methyl salicylate within a sealed plastic pouch. Glands were then analyzed using a light microscope for number of TEBs, ductal extension past the lymph node, and total ductal coverage of the fat pad. Ductal extension was measured using a 1mm square grid placed underneath the gland and manually counting the length of the ducts beyond the lymph node.

3.3.5 Western Blotting

At time of death, inguinal mammary glands were collected and digested in DMEM-F12 media with 2µg/mL of collagenase (Sigma Aldrich) at 37 °C for 2hrs, shaking every 20-30 mins.

The cell suspension was then washed with DMEM-F12 until media was clear and free of adipocytes to leave the epithelial cell isolate. Epithelial cells were frozen at -80°C until Western blot analysis, when they would be thawed, and placed in RIPA buffer with 0.2% w/v SDS for 30 mins. Cell lysates were then vortexed, sonicated, and centrifuged at 16,000 xg for 20 mins at

4°C. The supernatant was analyzed for protein concentration via a DC protein assay (Bio-Rad).

20µg of protein from each cell lysate sample were run on Mini-PROTEAN TGX 4-15% pre-cast gels (Bio-Rad) for 1.5 hrs at 100V, then transferred onto nitrocellulose membrane

(Thermo Scientific) using a wet transfer apparatus at 100V for 1 hr. Membranes were left overnight to dry, then blocked in a 1% milk solution in Tris buffered saline with 0.05% v/v

Tween 20 (TBS-T) for 1 hr at room temperature. Blots were then incubated with 1:2000 MARRS

(Thermo Scientific) and 1:10,000 beta-Actin (GeneTex) antibodies. Incubation periods for

MARRS and beta-Actin were both 1 hr at room temperature. The secondary for the 1,25D3-

MARRS antibody was goat anti-rabbit conjugated to horseradish peroxidase (New England

Biolabs) in a 1:50,000 dilution in 1% w/v milk/TBS-T solution. The secondary for actin was goat anti-mouse conjugated to horseradish peroxidase (Santa Cruz) in a 1:10,000 dilution in 1% w/v milk/TBS-T solution. Both secondaries were also used for 1 hr at room temperature. After several rinses in TBS-T, protein bands were visualized with the Clarity chemiluminesence kit

(Bio-Rad).

Densitometry analysis was performed using Alpha-View FluorChem software to compare density of bands (n=4 minimum per group). One internal control sample was loaded onto every gel to allow comparison between blots. Band density was background-corrected, and normalized to the control antibody.

3.3.6 Immunohistochemistry

Right abdominal mammary glands were fixed overnight in formalin, embedded in wax, and sliced into 5µm sections. After deparaffinizing slides in xylenes, sections were quenched in

3% hydrogen peroxide for 20 mins to prevent endogenous peroxidase activity, and then treated with sodium citrate (pH 6.0) for 20 mins at 95°C for antigen retrieval. After cooling, and several washes in PBS, sections were blocked in 3% normal goat serum for 1hr, then incubated in the same MARRS antibody used in western blots at a dilution of 1:4000 overnight at 4°C. Sections were rinsed with PBS, then incubated in Biotin-SP-affiniPure goat anti-rabbit IgG and

Peroxidase-Streptavidin (Cedarlane) for 1 hr each at 1:200 dilutions. Protein expression was visualized using a 3,3’-Diaminobenzidine (DAB) kit (Vector) for 90s, and then observed under a light microscope.

3.3.7 Statistical Analysis

Data was tested for normal distribution. Differences between means of normally distributed data were analyzed by one-way Analysis of Variance (ANOVA). If there was a significant main effect, the Tukey’s post-hoc multiple comparison test was used to determine differences between means. Differences between non-normally distributed data medians were analyzed using a Kruskal-Wallis Test with pairwise comparisons to indicate differences between

groups. Differences were considered significant at p < 0.05. All tests were conducted using SPSS

Statistics (version 24, IBM, Armonk, N.Y., USA).

3.4 Results

3.4.1 Phenotype Characteristics

By visual inspection of offspring, it was not possible to discriminate between wildtype and homozygous or heterozygous MARRS knockout animals. Gross body weight did not significantly differ between genotypes at 4wks of age. MARRS -/- females were found to be fertile and able to nurse litters, with no significant differences amongst litter sizes or pup survival between groups (data not shown, gathered from 51 litters, through 4 generations).

3.4.2 Mammary gland characteristics

The number of TEBs differed significantly between MARRS +/+ and mice with altered

MARRS expression. Wildtype mice had significantly more TEBs than both MARRS +/- mice

(p=0.026) and MARRS -/- (p=0.019). (Figure 3.1A). There were also significant differences regarding ductal extension in MARRS modified mice compared to wildtype controls (Figure

3.1B). Both MARRS +/- and MARRS -/- had reduced branch lengthening (p=0.04 and p=0.018 respectively). There were no significant differences between groups regarding ductal coverage of the fat pad. Representative wholemount images of all genotypes show the differences in duct growth between groups (Figure 3.2).

3.4.3 Western immunoblots

Western blot analysis showed successful reduction in MARRS protein expression in mammary gland epithelial cells (Figure 3.3A). Densitometry results reveal wildtype mice to have the most MARRS expression within epithelial cells, and there is a reduction in heterozygous and homozygous cells. However, epithelial cells heterozygous for MARRS had varying degrees of

MARRS protein expression, and MARRS was not completely knocked out in all homozygous mice (Figure 3.3B).

3.4.4 Immunohistochemistry

Tissue analysis confirmed reduced MARRS expression between +/+ and +/- animals

(Figure 3.4A and 3.4B). Qualitative observations between staining intensity showed +/- and -/- animals had less protein expression within epithelial cells. Immunohistochemistry also shows unchanged MARRS expression in adipocytes, confirming the specificity of the Cre/lox system reducing MARRS in the epithelial cells only. Immunohistochemical analysis showed concentrated MARRS expression within the apical cells of ducts (Figure 3.4A), as well as within the TEBs, often on the leading edge facing the direction of growth (Figure 3.4C).

A 9 8 7 6 * * 5 4 3

2 1 Average Number of TEBs of Number Average 0 (+/+) (+/-) (-/-) Genotype

B 6.2 6 5.8 5.6

5.4 5.2 * * 5

Extension (mm) 4.8 4.6 Median Length Median of Ductal 4.4 (+/+) (+/-) (-/-) Genotype

Figure 3.1. Mean number of TEBs (A), and median ductal extension (B) of mammary glands in MARRS +/+ (n=9), +/- (n=9) and -/- (n=12) mice. Significance was set to p=0.05 and noted with a *. There was a significant difference between number of TEBs between +/+ and both +/- and - /- mice (p=0.026 and 0.019 respectively, SD=2.88). There was also a significant difference between ductal extension in MARRS +/+ and both +/- and -/- animals (p=0.04 and 0.018 respectively). Maximum/minimum values for ductal extension: +/+ (8/4,) +/- (6/3), -/- (6/4).

A B

Figure 3.2. Representative whole mounts of mammary glands showing terminal end buds, ductal extension and coverage for MARRS +/+ (A), +/- (B) and -/- (C) animals at 4 weeks of age. The entire purple area is the mammary fat pad. Lymph nodes are noted with a * and examples of TEBs noted with an arrow. Figures (B) and (C) show the reduced length and area coverage of the ducts compared to wildtype in (A). There are more TEBs in +/+ animals than in +/- or -/- animals.

B 2.5

a 2

1.5

ab 1

b Relative MARRS expression MARRS Relative expression 0.5

0 (+/+) (+/-) (-/-) Genotype

Figure 3.3. Western blot (A) and densitometry analysis results (B) in MARRS and β-Actin expression in mammary epithelial cells in MARRS wildtype (+/+), heterozygote (+/-) and knockout (-/-) mice. Western blot results show one animal per lane, while densitometry results show average expression of 4 animals per genotype. Western blot shows that MARRS was not completely knocked out in -/- epithelial cells, but was reduced approximately 50% in +/- animals, and a further 50% in -/- animals.

A B

C D

Figure 3.4. Immunohistochemistry of mammary glands for a +/+ animal (A) and a +/- animal (B). MARRS is depicted with brown staining and epithelial cells in blue. Figures (A) and (B) show MARRS localized to the apical side of ductal epithelial cells. Figure (C) highlights a TEB with MARRS localized to the leading edge, in the same direction as duct growth through the fat pad. Cross sections of ductal branching are noted with arrows, and TEBs noted with a *. Figure (D) depicts an MARRS +/+ animal stained without primary MARRS antibody to show the minimal non-specific binding of this antibody.

3.5 Discussion

In the current study we have confirmed the presence of MARRS expression in mammary gland epithelial cells by both western blot and immunohistochemistry, and successfully reduced the level of expressed protein using the Cre/loxp system. However, there were variances in

MARRS expression between mice with the same genotype. Some MARRS +/- mice had similar levels of MARRS protein expression as wildtype mice. This suggests that the remaining allele may be able to compensate for a reduction in allele dosage. MARRS -/- mice had visible and quantifiably reduced levels of MARRS, but to varying degrees. Therefore, knocking out

MARRS in epithelial cells using the Cre/loxp system does not produce the same reduction in protein expression between individual mice. However, it is also possible that additional MARRS signal is coming from adjacent cells, or the isolation of epithelial from other cell types when preparing western blot samples was not 100% pure and contained non-epithelial cells.

Overall, a decrease in MARRS expression altered the growth trajectory of mammary gland development, manifesting in fewer TEBs and stunted ductal extension through the fat pad.

This suggests that MARRS may play a role in stimulating cellular proliferation of the TEB and is a necessary part of normal mammary gland development.

Interestingly, although TEB development and ductal extension altered significantly between wildtype and MARRS +/- mice, there were no significant differences between MARRS

+/- and MARRS -/-. This suggests that removal of only one MARRS allele is sufficient to alter development trajectory, and inhibiting both alleles does not further alter the tissue.

As depicted in the IHC figures, MARRS is present throughout the mammary gland in both adipocyte and epithelial cell fractions. However, expression throughout epithelial cells is

not uniform, as shown in Figures 3.4A and 3.4B. MARRS is localized to the apical side of epithelial duct cells and within TEBs, which would correspond to MARRS being highly expressed within the body cells of the TEB [8]. Since the role of these cells is to rapidly divide and provide the framework of the functional part of the tissue, MARRS may be participating in controlling the rate of cellular division.

However, MARRS is also expressed in the cap cells of the TEB on the leading edge corresponding to the direction of growth (Figure 3.4C). The cap cells of a TEB are known to interact with the surrounding stroma, but ultimately become the basal lamina of the fully formed duct [142]. However, since MARRS is not expressed in the basal lamina of ducts, this may suggest that MARRS is only required in the relatively undifferentiated cap cell. This, together with the TEB results, further suggests MARRS has a role in cell signaling during mammary gland development. Due to its placement within the TEB, MARRS may play a part in deciding the speed or direction of terminal end bud growth. Since MARRS +/- and -/- mice had decreased ductal expansion and TEB formation compared to controls, this suggests a role for MARRS in

TEB cell proliferation.

Possible signaling pathways involved include phospholipase C, IP3 and DAG, PKC and

ERK/MAPK as suggested earlier [141,146]. It has been previously shown in chondrocytes that

MARRS can induce release of arachidonic acid and activation of PKC, PKA [141] and MAPK

[147] resulting in upregulation of cell cycle genes.

There has been other compelling evidence showing MARRS’s ability to interact with several other cell signaling molecules and transcription factors that alter cell growth. Cell culture studies in human melanoma and lymphoma cell lines showed MARRS binds Ref-1 to form a

stable complex [148] with the ability to activate downstream transcription factors. Ref-1 has a long list of recognized abilities including activating NFκB [149], which could explain a lack of cell growth in our MARRS knockouts. MARRS can also directly bind STAT3 [150], which has already been implicated in mammary gland transformation processes [151] as an important signaling molecule in cell growth. Future studies will need to determine whether MARRS is manipulating any of these signaling pathways during mammary gland development.

The present study showed that MARRS plays a role in mammary gland development, and inhibiting MARRS protein expression results in fewer TEBs and overall slower development at the 4wk time point. However, functionality of the fully developed mammary gland does not seem to be inhibited, as MARRS -/- mice did not show any signs of inability to nurse a healthy litter. Therefore, it can be speculated that MARRS -/- mice may benefit from fewer TEBs, while still maintaining full functionality of the tissue.

A receptor with the ability to alter or slow down TEB proliferation could provide new insights into breast cancer treatments and prevention strategies. Although TEBs are a normal and necessary part of mammary gland development, their vulnerability to carcinogens and growth factors has implicated them as possible targets for tumorigenic events [152]. The human equivalent of a TEB, the terminal ductal lobular unit, is the structure from which many breast tumors originate [112], and more TEBs than normal could increase cancer risk when exposed to a carcinogen [152]. This theory has been tested in multiple animal studies, and correlations between fewer TEBs and a decreased risk of breast cancer were found [115,117,143]. Inhibiting the MARRS receptor in these cells may provide a novel avenue for prevention in individuals at high risk of breast cancer.

Previous studies examining murine mammary gland morphology have found that the classic VDR has a role in controlling mammary growth and attenuating tumorigenesis [66,108].

Other studies acknowledging the presence of a non-genomic 1,25D3 receptor have argued for the ability of the nuclear VDR to translocate to the plasma membrane and perform rapid non- genomic actions. However, the results of the present study provide further evidence for the rapid actions of 1,25D3 being carried out by a completely separate receptor from the nuclear VDR, and that MARRS is capable of providing a different function from the VDR. MARRS +/- and -/- mice in the present study showed a delay, or decrease in development compared to wildtype mice, which is the opposite from effects seen in VDR knockout mice [66]. Had the MARRS receptor been the same, or provide the same function as VDR, then we would expect the number of TEBs to be much higher in MARRS knockout mice compared to wildtype.

Limitations of this study include the vitamin D content of the chow. 3000 IU/kg is three times the recommended amount for mice [153] and may impact the growth trajectory of the mammary gland. From what is known from VDR knockout studies, increased vitamin D intake as well as reduced MARRS receptors could enhance the anti-proliferative effects of 1,25D3, and may overestimate the effect MARRS has on growth. Furthermore, this particular cross-strain of mouse has not been tested in regard to mammary growth before, and is therefore difficult to compare these results to other mouse models.

3.6 Conclusion

The results of this study provide evidence that MARRS has a role in mammary gland development, possibly in determining speed or direction of TEB growth. The result of knocking out the MARRS receptor in mammary epithelial cells is opposite to that of knocking out VDR,

thus providing evidence that MARRS is a separate molecule from VDR, and provides a different role. However, this study did not directly investigate MARRS as a 1,25D3 receptor, therefore future studies should focus on the relationship between MARRS and 1,25D3 within the mammary gland and how these signals affect growth.

Chapter 4: Differential Effects of the 1,25D3-MARRS Receptor (ERp57/PDIA3) on Murine Mammary Gland Development Depend on Vitamin D3 Dose

4.1 Abstract

1,25 dihydroxyvitamin D3 (1,25D3) is the most potent biologically active form of vitamin D3 and its actions on the mammary gland include cell growth inhibition and anti-cancer effects. Most of the actions of 1,25D3 continue to be attributed to the classic nuclear vitamin D receptor despite the identification of the membrane-associated rapid response steroid-binding receptor (MARRS). The purpose of this study was to explore the role of MARRS in the mammary gland using a tissue-specific knockout mouse model and a vitamin D3 dietary intervention. Three genotype groups were created using the Cre/loxp system to knock-down and knockout the MARRS receptor in epithelial cells of mouse mammary glands (MG). Abdominal

MGs were collected from 6-week old female mice (n=94) on diets of 10,000 IU/kg (replete),

1,000 IU/kg (sufficient) or 0 IU/kg (deficient) of D3. The growth of MGs was measured by counting the number of terminal end buds (TEBs) of alveolar branches, the length of the longest ductal extension, and total MG area covered by ducts. Differences between means of normally distributed data were analyzed using two-way Analysis of Variance (ANOVA) tests. Differences were considered significant at p < 0.05. There was a significant interaction between genotype and diet regarding TEBs (p=0.001) and ductal coverage (p=0.03). Knockout mice on the sufficient diet had significantly fewer TEBs (p=0.001) compared to wildtypes on the same diet, but the opposite effect was seen in mice on the replete diet. There were no effects of genotype on

TEBs when animals were vitamin D3 deficient. Visual differences were also observed in mice on deficient diets: epithelial ductal structures were thinner and had less secondary branching compared to other diet groups. These results suggest that there is an effect of MARRS on mammary gland development that is dependent on 25(OH)D status, specifically, altering the

number of highly proliferative TEBs. Increased numbers of TEBs have been correlated with increased breast cancer risk later in life. Therefore the results of this study may warrant further examination of 25(OH)D status and recommendations in adolescent humans to reduce dietary effects on future breast cancer risk.

4.2 Introduction

Vitamin D3 (cholecalciferol) can be obtained either through limited options in the diet, or converted from 7-dehydrocholesterol in the skin by UVB rays. Once absorbed into the circulation, vitamin D3 is bound to vitamin D binding protein (DBP) and carried to the liver for hydroxylation by 25-hydroxylase to form calcidiol (25(OH)D) [52]. The 25(OH)D form is inactive, and is measured to estimate vitamin D status due to its long half-life [52]. 25(OH)D is finally converted to its active form calcitriol (1,25D3) by the 1-α-hydroxylase (CYP27B1) commonly described in the kidney, but has been found in other tissues around the body such as the skin [154], bone [155], and the mammary gland [156].

1,25D3 is essential in maintaining the body’s calcium and phosphate homeostasis.

However, it has become commonly accepted that 1,25D3 exhibits a multitude of extra-bone effects including manipulation of the cell growth cycle and initiation of cell differentiation, subject to the target tissue [157,158]. These effects of 1,25D3 are often credited to action through the classic nuclear vitamin D receptor (VDR), which, upon activation, will dimerize with the retinoid X receptor and bind to vitamin D response elements on DNA to alter transcription of target genes [159].

Although actions of 1,25D3 continue to be ascribed to the VDR, there is evidence for a separate, membrane-bound receptor with distinct non-genomic effects [70,160]. This receptor

has been named the membrane associated rapid response steroid-binding receptor (MARRS).

Other names include ERP57, Grp58, and Pdia3 [161], and current research has focused on examining the role of MARRS in tissues with known VDR effects. Evidence for an independent receptor includes studies performed in the basolateral membrane of intestinal epithelium in chicks showing the rapid uptake of calcium and phosphate was eradicated after tissue-specific knockout of the MARRS receptor [132]. MARRS has been shown to have a role in cell differentiation of bone marrow cell lines [71]. MARRS has also been studied in relation to cell proliferation in MCF-7 cells. In these cell culture studies, MCF-7 cells with reduced MARRS expression exhibited increased sensitivity to 1,25D3 and the anti-growth actions of the metabolite were enhanced, resulting in reduced cell growth [110].

These previous studies suggest that in the absence of MARRS, actions of 1,25D3 are mediated solely through the VDR. Further evidence for this statement has been presented through researching the development of post-natal mammary glands. Studies using VDR knockout mice have confirmed an anti-proliferative role for 1,25D3 and the VDR within the mammary gland. In the absence of a functional VDR, mammary gland growth is explosive, resulting in high numbers of terminal end buds (TEBs), elongated ducts, and widespread secondary branching [66]. The VDR knockout was also bred into a mouse model for breast cancer to study effects of morphology on tumorigenesis. MMTV-neu mice with a whole-body

VDR knockout produced histologically atypical ducts, earlier detection of tumorigenesis and reduced survival [109].

Conversely, a tissue-specific knockout study in animals has shown opposite effects when knocking out MARRS, resulting in decreased numbers of TEBs and reduced branching compared to wildtype animals [162]. Therefore, the aim of this study was to expand on the

results from the tissue-specific MARRS knockout studies and test the effects of 25(OH)D status on mammary gland development and MARRS. In the present study, we generated a tissue- specific knockout mouse model with reduced expression of MARRS in the epithelial cells and monitored several variables of development under diets of differing vitamin D3 doses. It is important to note that whole-body MARRS knockout is not possible as this is embryonic lethal and animals do not survive to birth.

4.3 Methods

4.3.1 Animals

C57Bl/6-Pdia mice with the 1,25D3-MARRS gene floxed were obtained from Utah State

University (Dr. Ilka Nemere, with permission from Dr. Gunter Hammerling) and B6129F1-

Tg(MMTV-cre)4Mam/J mice were purchased from Jackson Laboratories. MARRS- floxed mice were crossed with MMTV Cre+ mice, and pups were backcrossed to MMTV Cre+ mice 7 times to reduce C57Bl/6 background. Mice were housed in rooms exposed to a 12 h dark/light cycles.

4.3.2 Diets

At a minimum of 3 mo of age, MARRS heterozygote animals were weaned onto AIN-

93G semi-purified diets with modified vitamin D3 doses: replete (DR - 10,000 IU/kg), sufficient

(DS - 1,000 IU/kg), and deficient (DD - 0 IU/kg). Diets were formulated by Research Diets, Inc.

(NJ, USA). Table 1 contains information regarding the nutrient composition of the diets. Dams were fed the vitamin D3 diet for a minimum of 1 mo to alter baseline plasma 25(OH)D levels before being set up to breed. Heterozygotes were bred to produce MARRS +/+ (WT), +/-

(knockdown) and -/- (knockout) genotypes. Pups were weaned at 3 wk. and placed on the same diet as their dam.

Table 4.1 Nutrient composition of modified animal diets

Component AIN-93G semi purified diet kcal%

Protein 20

Carbohydrate 64

Fat 16

g/kg diet

Casein 200

L-Cystine 3

Corn Starch 397.486

Sucrose 100

Cellulose 50

Soybean Oil 70 t- 0.014 Butylhydroquinone Mineral Mix 35 S10022G

Dietary composition of AIN-93G semi-purified diets from Research Diets Inc. (New Brunswick, USA). All 3 diets (replete, sufficient, deficient) are identical in nutrients listed in the table. The only difference between diets is the content of Vitamin D3. Macronutrients are listed as kcal%, whereas remaining components are listed as # of grams of component per kg of diet.

4.3.3 Genotyping

Pups were toe-clipped at 6 days of age to obtain DNA. Toe samples were digested in an alkaline lysis buffer for 30 min at 95°C, then cooled on ice, and neutralized in an acidic buffer.

DNA was genotyped using the following primers for PCR: CGC CAG CCT CTC CAT TTA G

(MARRS forward); CAG AGA TCC TGC CTC TG (MARRS reverse); CTG ATC TGA GCT

CTG AGT G (Cre forward); and CAT CAC TCG TTG CAT CGA CC (Cre reverse). Cre

Primers are the suggested sequences from Jackson Laboratories, and MARRS primers were made using PubMed Gene databases. Electrophoresis was run using 1.5% w/v agarose gels at

100 V for 1h, and stained with a carcinogen-free DNA stain (SafeView Classic, ABM). A positive Cre result would show a single band at ~250bp, a MARRS WT mouse would show a single band ~222bp, MARRS knockout would show a single band at ~410bp, and a MARRS knockdown mouse would have two bands, one each at 222bp and 410bp.

4.3.4 Necropsy and Tissue Collection

At 6 wk of age, the estrous stage of the female was determined by performing a vaginal lavage with sterile phosphate buffered saline (PBS), and vaginal epithelial cells were used to stage the estrous cycle, described previously [144]. Mice were sacrificed by CO2 if they were in any stage of their estrous cycle other than diestrus (n= 7 minimum per group). At time of death, whole body weight was measured. Left abdominal mammary glands were collected for whole mounting, right abdominal glands were prepped for immunohistochemistry, and inguinal glands were prepped for Western blot. Blood was collected at time of sacrifice, spun at 1500g for 10 min and heparinized plasma was stored at -80°C until analysis. Plasma samples were sent to

Heartland Assays (IO, USA) for 25(OHD) analysis via radioimmunno assay, and total calcium detection.

4.3.5 Whole mounting Mammary Glands

Whole mounting was performed as described previously [163]. Briefly, left abdominal glands were collected and fixed in formalin overnight. They were then submerged in acetone for

48 hr, then rehydrated in decreasing concentrations of ethanol, and stained in carmine alum overnight. The next day glands were dehydrated, cleared in xylenes, and stored in methyl salicylate within a sealed plastic pouch. Glands were then analyzed using a light microscope for number of TEBs, ductal extension past the lymph node, and branching ductal area of the fat pad as a percent of total fat pad area. Ductal extension was measured using a 1mm square grid placed underneath the gland and manually counting the length of the ducts beyond the lymph node.

4.3.6 Western Blotting

At time of termination, inguinal mammary glands were collected and digested in DMEM-

F12 media with 2µg/mL of collagenase (Sigma Aldrich) at 37 °C for 2 h, shaking every 20-30 min. The cell suspension was then washed with DMEM-F12 until media was clear and free of adipocytes to leave the epithelial cell isolate. Epithelial cells were frozen at -80°C until Western blot analysis, when they were thawed, and placed in RIPA buffer with 0.2% w/v SDS for 30 min.

Cell lysates were then vortexed, sonicated, and centrifuged at 16,000g for 20 min at 4°C. The supernatant was analyzed for protein concentration via a DC protein assay (Bio-Rad).

20µg of protein from each cell lysate sample were run on Mini-PROTEAN TGX 4-15% pre-cast gels (Bio-Rad) for 1.5 h at 100 V, then transferred onto nitrocellulose membrane

(Thermo Scientific) using a wet transfer apparatus at 100 V for 1 hr. Membranes were left

overnight to dry, then blocked in a 1% w/v milk solution in Tris buffered saline with 0.05% v/v

Tween 20 (TBS-T) for 1 h at room temperature. Blots were then incubated with 1:2000 Anti-

ERP57 (MARRS, from Pierce, ThermoFisher Scientific) and 1:10,000 β-Actin (GeneTex, CA,

USA) antibodies. Incubation periods for ERP57 and β-Actin were both 1 h at room temperature.

The secondary for the MARRS antibody was goat anti-rabbit conjugated to horseradish peroxidase (New England Biolabs) in a 1:50,000 dilution in 1% milk/TBS-T solution. The secondary for actin was goat anti-mouse conjugated to horseradish peroxidase (Santa Cruz) in a

1:10,000 dilution in 1% milk/TBS-T solution. Both secondaries were also used for 1 h at room temperature. After several rinses in TBS-T, protein bands were visualized with the Clarity chemiluminesence kit (Bio-Rad).

Blots were then analyzed via densitometry to compare brightness of bands (n=3-7 per group). One internal control sample was loaded onto every gel to allow comparison between blots.

4.3.7 Immunohistochemistry

Right abdominal mammary glands were fixed overnight in formalin, embedded in wax, and sliced into 5µm sections. After deparaffinizing slides in xylenes, sections were quenched in

3% hydrogen peroxide for 20 min to prevent endogenous peroxidase activity, and then treated with sodium citrate (pH 6.0) for 20 min at 95°C for antigen retrieval. After cooling, and several washes in PBS, sections were blocked in 3% normal goat serum for 1h, then incubated in the same MARRS antibody used in Western blots at a dilution of 1:8000 overnight at 4°C. Sections were rinsed with PBS, then incubated in Biotin-SP-affiniPure goat anti-rabbit IgG and

Peroxidase-Streptavidin (Cedarlane) for 1 h each at 1:400 dilutions. Protein expression was

visualized using a 3,3’-Diaminobenzidine (DAB) kit (Vector, Cedarlane) for 60 s, and then observed under a light microscope.

4.3.8 Statistical Analysis

Data were tested for normal distribution. Any non-normally distributed data were transformed until normality was achieved. Differences between means of normally distributed data were analyzed using two-way Analysis of Variance (ANOVA) tests. For a significant test, differences between means were found pairwise by the Least Significant Difference post-hoc test. Differences were considered significant at p < 0.05. All tests were conducted using SPSS

Statistics (version 25, IBM, Armonk, N.Y., USA).

4.4 Results

4.4.1 Phenotype characteristics

There were no significant differences amongst 6 wk females’ whole body weight at time of sacrifice. However, deficient mice did exhibit signs of vitamin D deficiency from 4-6 weeks of age as hind legs were slightly contorted (Figure 4.1) compared to replete or sufficient mice of any genotype. This alteration in foot position was not seen in adult mice, and no other physical manifestations of rickets (such as alopecia) were observed.

4.4.2 Blood tests

Genotype did not have any significant effect on 25(OH)D or calcium plasma levels. A vitamin D3 intervention of 10,000 IU/kg in the diet was enough to significantly raise plasma

25(OH)D levels of all mice, regardless of genotype, above those on a sufficient diet (p=0.001)

(Figure 4.2). 25(OH)D levels for deficient diet mice were undetectable by the assay

(<2.5ng/mL).

Total plasma calcium levels were not significantly different between any groups, regardless of genotype (results not shown). There was no difference between plasma calcium of mice on sufficient vs. replete diets. The average plasma calcium level across all groups was

10.83 mg/dL.

4.4.3 Mammary gland growth variables

There was a significant overall interaction between diet and genotype concerning TEBs

(p=0.001) (Figure 4.3A & 4.4). There were no observable differences between genotypes when mice were on a deficient diet, but there were significant differences between genotypes in both the sufficient and replete diets. Within the sufficient diet group, knockdowns had significantly more TEBs than knockouts (p=0.001). And within the replete diet group, knockout animals had significantly more TEBs than both wildtypes (p=0.041) and knockdowns (p=0.005) (Figure

4.3A).

There was also a significant overall interaction between diet and genotype when comparing branching ductal area coverage (p=0.03) (Figure 4.3B). Overall, knockout mice on deficient diets had on average the greatest coverage of fat pad by ductal growth (significantly different from knockouts on the sufficient diet, p=0.004). There were no significant differences between groups concerning length of ductal extension.

Finally, there were visual, qualitative differences noticed between groups. Although mice on the deficient diet were quantitatively similar to other diet groups concerning the length and

area coverage of ductal branching, the epithelial structures were thinner and had less secondary branching compared to animals on sufficient or replete diets (Figure 4.4E).

4.4.4 Western immunoblots

Western blot (Figures 4.5A, 4.5B, 4.5C) and densitometry (Figure 4.5D) analysis of mice on deficient diet showed successful reduction of MARRS expression in epithelial cells regardless of diet. Knockdowns showed an approximately 50% average reduction compared to WT, and knockouts showed a further reduction compared to knockdowns, although there was some individual variation between mice within the same genotype. Knockout animals did not always show a complete elimination of MARRS expression. WT mice on deficient diets seemed to express 50% more MARRS protein than WT mice on other diets. However, due to the variation in expression across all mice, this increase was not significant.

4.4.5 Immunohistochemistry

Immunohistochemistry results also showed successful reduction in MARRS expression in knockdown and knockout animals (Figure 4.6). MARRS is depicted with brown staining. All

WT animals across all vitamin D diets had similar levels of MARRS expression, depicted by the dark brown staining around epithelial cells.

Knockdown animals on sufficient and replete diets had similar MARRS expression, and both groups showed reduced expression compared to WT. Knockdowns on the deficient diet had the least MARRS expression compared to other knockdown animals.

Knockouts on sufficient and deficient diets also had similar staining patterns with minimal MARRS expression throughout mammary epithelial cells. However, knockouts on the replete diet still showed MARRS expression in some, but not all, epithelial cells.

Overall, MARRS expression seems to be most highly expressed in the luminal side of epithelial-lined branches, and on the inner-most cells within a TEB (Figure 4.6).

Figure 4.1. Hind foot abnormalities between deficient (A) and replete (B) mice at 6wks of age. Hind feet of deficient mice were bent outwards regardless of genotype. Deficient mice did not show any other clear signs of rickets such as alopecia. Hind foot abnormalities were not seen in adult mice of any diet.

Figure 4.2. Mean plasma 25(OH)D levels of mice at 6 wks of age. Legend describes mice on deficient (DD), sufficient (DS) and replete (DR) diets. MARRS genotype (x-axis) did not have a significant effect on 25(OH)D levels. Mice on replete diets had significantly higher 25(OH)D levels than mice on sufficient diets, regardless of genotype (p=0.001). Levels for deficient mice were undetectable by the assay (<2.5 ng/mL), labeled NA.

Figure 4.3. Mammary gland growth variables. Average number of TEBs (A) and area of ductal branching coverage (% of total fat pad) (B) of whole mounts of mammary glands at 6 wks of age. There was a significant interaction between genotype and diet regarding both TEBs (p=0.001) and branching coverage (p=0.03) noted by an asterisk. The data suggest in regards to TEBs that there must be sufficient vitamin D3 in the diet to see an effect from MARRS, as shown by the similar TEB numbers for DD animals. The reduced number of TEBs in knockouts on the sufficient diet is restored with a diet replete in vitamin D3.

Figure 4.4. Whole mount representations of left abdominal mammary glands of mice at 6 wks of age. Legends on figures describe the diet and genotype of each mouse: sufficient (DS), replete (DR), deficient (DD), knockdown (+/-) and knockout (-/-). Knockouts on sufficient diets (B) had significantly fewer TEBs, and reduced overall growth compared to knockdowns (A). When mice were on a replete diet, the phenotype was reversed, and TEB numbers and overall growth were restored in the knockout animal (D). (E) is a representative mammary gland of a mouse on the deficient diet. Although all animals on DD were quantitatively similar in measurements, there

were qualitative visual differences compared to animals on sufficient or replete diets including thinner branching structures, and fewer secondary branches.

A B C

MARRS β-Actin

+/+ +/- -/- +/+ +/- -/- +/+ +/- -/- Deficient (DD) Sufficient Replete (DR) (DS) D

2.5

1.5 (+/+) 1 (+/-)

0.5 (-/-) Relative MARRS expression MARRS Relative expression 0 DD DS DR Diet

Figure 4.5. MARRS protein expression in inguinal mammary gland epithelial cells. Representative western blots (A, B, C) and mean densitometry analysis (D) of MARRS protein expression in mammary gland epithelial cells. n=2-7 per group. Overall, western blots show successful reduction of MARRS protein in epithelial cells regardless of diet. Variations between animals within the same genotype exist.

Figure 4.6. Immunohistochemistry results for right abdominal mammary glands stained with Anti-MARRS and counter-stained in hematoxylin. MARRS expression is depicted by brown colouring. (A) represents a DR +/- animal; (B) represents a DR -/- animal. Alveolar branches are noted with an arrow while TEBs are noted with an asterisk. Reduced MARRS expression is shown in epithelial cells, with similar expression in surrounding adipocytes. Slides also show that MARRS is not eliminated in all epithelial cells, which would account for some variation between animals and shows the imperfect deletion of the protein expression in knockouts. MARRS is localized to the apical side of epithelial cells within the ductal branch. (C) represents tissue with no primary antibody.

4.5 Discussion

Although mice were only on the vitamin D3 intervention diet for 3 wk post-weaning, this was sufficient time to alter their 25(OH)D status (Figure 4.2). 1,000 IU/kg is the standard vitamin D3 dose for mice [153] and 10,000 IU/kg was well-tolerated by all mice. 1,000 IU/kg resulted in all mice on the sufficient diet achieving an average plasma 25(OH)D level of 27.62 ng/ml, which is considered vitamin D sufficient for humans according to the Institute of

Medicine [164]. Mice on the replete diet achieved an average plasma 25(OH)D level of

58.11ng/mL, which is well below the threshold of 150ng/mL, at which clinical toxicity symptoms usually begin to arise in humans [165].

Total plasma calcium levels were not significantly altered between groups, and the average plasma calcium level was similar to previous studies [108,166]. Importantly, there was no difference between plasma calcium in mice on sufficient vs. replete diets, thus showing the high intake of vitamin D3 did not cause hypercalcemia.

Currently, studies examining vitamin D3 interventions do not use standardized doses.

What is considered a high dose in some studies (4000-5000 IU/kg range [167]) would merely be an intermediate dose in others (where 10,000-20,000 IU/kg is labeled as a high dose [168]). The recommended dose of 1000 IU/kg of diet as set by the National Research Council [153] was not formally studied, and only set by personal communication and observations that this amount did not result in deficiencies or rickets [169]. However, in some strains of mice this dose results in

25(OH)D plasma levels that exceed optimal concentrations for humans [166]. Furthermore, commercially available non-purified diets often contain upwards of 3000IU/kg of vitamin D3.

Although there are a few studies that study the dose-response for vitamin D3 in mice for several

outcomes (plasma calcium, plasma 25(OH)D, bone health) [166], mice of differing genetic backgrounds could respond differently, thus changing the outcomes of studies. It is clear there is a gap in the literature concerning optimal vitamin D3 doses for mice, and researchers should carefully consider doses used in feeding trials.

One limitation of this study is that 1,25D3 concentrations were not directly measured at the tissue level, so we do not know for certain that there was increased 1,25D3 available for

MARRS within the mammary gland. However, the 25(OH)D-1α-hydroxylase enzyme that converts circulating 25(OH)D into the active 1,25D3 is present in mouse mammary glands [170] as well as normal human mammary tissue [57]. Further, CYP27B1 RNA expression has been shown to increase in mouse mammary organ cultures after treatment with 25(OH)D [171], so it is reasonable to expect that active 1,25D3 would be available at the mammary gland with increased circulating 25(OH)D.

Previous reports studying the effects of VDR on mammary gland growth and development demonstrate a growth-controlling effect. VDR null mice showed significant bursts of growth at 6 wk of age, including increased numbers of TEBs [66] and increased ductal extension [172] compared to wildtype controls. VDR knockout mice also have increased circulating estrogen levels, as well as an increased sensitivity to hormone-related growth [66]. In vitro organ studies have also showed direct incubation with either 25(OH)D or 1,25D3 inhibited mammary growth [11, 20]. In contrast, a previous study with MARRS knockouts [162] showed that a lack of MARRS protein expressed in mammary epithelial cells resulted in decreased number of terminal end buds and overall reduced growth of the gland during puberty. These studies provide further evidence that MARRS is a separate receptor from VDR, whose actions in mammary gland growth result in different, and opposing effects.

TEBs are one of the most-studied structures within the developing mammary gland as they are a key component to the laying of the ductal framework [173]. It is clear that there is an interaction between MARRS expression and 25(OH)D status concerning TEB growth, and this has manifested in two ways in our study. First, there were no differences in TEB numbers in mice on the deficient diet, regardless of genotype. This would imply that any role MARRS may play in mammary gland development is dependent on vitamin D3 being adequately available in the diet, thus suggesting that MARRS is activated by 1,25D3 in this tissue.

Second, the differential effects that were seen when comparing knockdown and knockout animals on sufficient and replete diets show both MARRS expression and 25(OH)D status are important. The mice on sufficient diets show similar results to a previous study [162] in that knockouts have fewer TEBs and overall less growth development than wildtypes. However, when mice are supplemented with an abundant supply of vitamin D3 in their diet, the reverse effect is seen, now with knockout animals having a high number of TEBs, and the same ductal coverage and extension as controls. Thus, the reduction in growth caused by a loss of MARRS expression was rescued with a high vitamin D3 dose.

However, the significantly higher (p <0.05) number of TEBs in knockout mice on the replete diet was unexpected, and implies an interaction between MARRS and other signaling pathways. As stated, previous research concerning vitamin D and mammary gland growth would suggest that having both a MARRS knockout and excess 25(OH)D or 1,25D3 would result in very little gland growth. Presumably, without MARRS, VDR would be the only activated receptor, and the anti-growth properties of 1,25D3 and VDR would manifest in limited branching and ductal extension. However, our results suggest there may be a possible interaction between MARRS and VDR, and the loss of one receptor does not simply result in a complete

shunt to the remaining pathway. Previous studies have shown that different 1,25D3 analogs are able to selectively activate either MARRS or VDR [70]. The use of either of these analogs may help to determine the pathways responsible for these unexpected growth results by being certain that only MARRS or VDR is being activated. Further, real-time PCR analysis of changes in

RNA expression and sensitivity analysis would also be helpful to understand the signaling changes at the cellular level. Some important signaling modulators with possible roles here would include estradiol, TGFβ, and EGFR which have been previously implicated in MARRS signaling pathways ([111,174], unpublished Meckling).

The results of this study may also have implications for humans regarding 25(OH)D status during puberty. Although mouse research does not directly translate into the same conclusions for humans, mammary gland development is comparable between these species [6].

The TEB equivalent in a human, the terminal ductal lobular unit, is an epithelial structure from which ductal carcinomas arise [112]. It has also been shown that more invasive carcinomas tend to originate from less specialized structures within the mammary gland, [112] and an increased number of TEBs is correlated with an increased risk of breast cancer in animal models

[115,117,143]. With an increase in the number of vitamin D deficient adolescents [175], and a diet-gene interaction manifesting in changes to organ development, it is worth pursuing further research into optimal doses for adolescents.

4.6 Conclusions

Overall, the results of this study build upon previous studies’ conclusions that there is a role for MARRS in the development of the mammary gland by altering the number of TEBs and overall growth rate of the tissue during puberty. However, these morphological changes seem to

be dependent on the availability of 25(OH)D in the body, with differential effects depending on if the subject is deficient or replete in 25(OH)D. These results may have consequences for adolescents with 25(OH)D deficiencies, and the results of this study warrant further investigation into 25(OH)D status at puberty and correlations with breast cancer incidence. This study also provides further evidence of MARRS’ interaction with vitamin D, and shows changes to mammary gland development regardless of VDR status. Vitamin D’s actions are more complicated than previously thought, and the impact of vitamin D can no longer be viewed solely through the VDR as there are likely other signaling pathways being activated.

Chapter 5: Effects of knocking out both the 1,25D3-MARRS Receptor and Vitamin D Receptor in Murine Pubertal Mammary Gland Development

5.1 Abstract

1,25 dihydroxyvitamin D3 (1,25D3) has been extensively studied for its anti-proliferative and anti-cancer effects in a multitude of tissues. The role of the vitamin D receptor (VDR) has been elucidated in the murine pubertal mammary gland as a factor in regulating growth.

However, another receptor for 1,25D3 – the membrane associated rapid response steroid receptor

(MARRS) – has seemingly the opposite role, in being required for optimal growth. This study produced a novel double knockout, altering the expression of both the MARRS and VDR receptors to study the individual efforts and possible signaling interactions between both receptors. VDR knockout (-/-) mice were crossed with MARRS -/- mice, and resulting heterozygotes were crossed to produce double knockouts. Mammary glands were collected from

6 week old mice (n=38) for mammary gland growth variables, MARRS and VDR protein expression, and RNA analysis of MARRS, VDR, estrogen receptor alpha, and transforming growth factor beta. There was a significant overall interaction between groups (p=0.004) for mean mammary gland area covered by branching. The MARRS+/VDR- animals had the least ductal coverage and were significantly different from all other groups (p=0.035). Reducing

MARRS expression also had an effect on ductal coverage and MARRS-/VDR+ animals had reduced ductal coverage compared to MARRS+/VDR+ animals. Overall these results do not concur with those of previous studies and suggest that 1,25D3 signaling is more complicated than previously thought.

5.2 Introduction

Vitamin D (or 1,25D3 when in the active metabolite form) has roles in many tissues throughout the body. In the classic capacity, activation of the vitamin D nuclear receptor (VDR) results in transcriptional activation of many genes leading to, most notably, upregulation of calcium and phosphate absorption at the intestine [176]. VDR has been found in a great number of human tissues and has implications in a multitude of cellular actions, including arresting the cell cycle [177], regulating apoptosis [178], and boosting immune function [179].

Recent research into the actions of 1,25D3 has found its ability to bind to a second receptor, noted ERP57, or the membrane-associated rapid response steroid-binding receptor

(MARRS). Originally discovered as a chaperone protein [135], MARRS can also be activated by

1,25D3 resulting in rapid cellular signaling [157]. VDR-independent, non-genomic actions of

1,25D3 have been observed in chick intestinal cells [132], mouse osteoblasts [180], and mouse chondrocytes [80].

The mammary gland is a unique tissue in that the majority of its development happens post-natally. A rudimentary epithelial structure is formed within the fat pad in utero, and remains relatively unchanged until the onset of puberty when ovarian and pituitary hormones trigger epithelial cells to rapidly divide and form a structure termed the Terminal End Bud (TEB) [6].

TEBs are club-shaped epithelial configurations that bifurcate and push through the fat pad to lay the alveolar ductwork required for milk production and secretion during pregnancy [8]. Due to their proliferative and undifferentiated nature, TEBs are highly sensitive to the effects of growth hormones and carcinogens [114], making them both a common measurement of tissue growth, and a risk factor for mutagenic events [115].

In relation to the murine mammary gland, vitamin D has been studied for both its role in mammary gland growth and breast cancer protection and treatment. Previous studies have observed that VDR knockout mice experience enhanced growth during puberty [66,67] noted by higher numbers of TEBs and increased ductal growth. VDR knockout mice also exhibit increased sensitivity to carcinogen-induced tumorigenesis [108]. These studies highlight VDR’s growth-controlling abilities. However, MARRS knockout mice have exhibited delayed or hindered growth compared to wildtypes [162].

A recent development in our lab has found that the actions MARRS differ depending on the 25(OH)D status of the animal during puberty (unpublished Meckling). These differential results have led our lab to study the relationship between MARRS and VDR. The purpose of this study was to develop a novel double knockout to further explore the individual efforts and possible signaling interactions between 1,25D3 receptors.

5.3 Methods

5.3.1 Animals and Diets

B6.129S4-Vdr/J VDR knockout (-/-) mice were obtained from Jackson

Laboratories and placed on a recovery diet with 2200 IU vitamin D3, 2% Ca2+, 1.25% phosphorus and 20% lactose (Teklad.94112, Envigo) as in previous studies using this mouse model [181,182]. The same MARRS -/- mice from Chapters 3 and 4 were crossed to VDR -/- mice.

MARRS -/- and VDR -/- mice were bred and first generation heterozygous progeny were crossed again to MARRS -/- mice. Pups resulting from this cross were subsequently bred to

VDR -/- mice to produce double knockouts. Mice used in this study were either MARRS wildtype (WT) or MARRS -/-, and VDR wildtype or -/-. Thus, the four groups in this study are noted as MARRS+/VDR+, MARRS-/VDR+, MARRS+/VDR-, or MARRS-/VDR-.

Mice were housed in rooms exposed to a 12 h light/dark cycles. All procedures were approved by the Animal Care Committee of the University of Guelph, under Animal Utilization

Protocol 3147.

5.3.2 Genotyping

Pups were toe-clipped at 6 days of age to obtain DNA. Toe samples were digested in an alkaline lysis buffer for 30 min at 95°C, then cooled on ice, and neutralized in an acidic buffer.

Primers used are in Table 5.1; VDR primers are the suggested sequence from Jackson

Laboratories. Electrophoresis was run using 1.5% w/v agarose gels at 100 V for 1h, and stained with a carcinogen-free DNA stain (Safeview Classic, Abcam). A MARRS WT mouse would show a single band ~222bp, and a MARRS knockout would show a single band at ~410bp.

5.3.3 Necropsy and Tissue Collection

At 6 wks of age, vaginal epithelial cells were collected to determine the estrous stage of the mouse, described previously [144]. To control for cycling hormones, tissue was not collected when mice were in diestrus. At time of death, whole body weight was measured. Left abdominal mammary glands were collected for whole mounting, right abdominal glands were collected for immunohistochemistry, and inguinal glands were collected for Western blot and RNA analysis

(n= 7-12 per group).

5.3.3 Whole mounting Mammary Glands

At time of termination, left abdominal glands were removed and fixed in formalin overnight. They were then immersed in acetone for 48 h, then rehydrated in 100%, 90%, and

75% ethanol over a period of 3h, immersed in water for 1 h, and finally stained in carmine alum overnight. The next day glands were dehydrated using the same dilutions of ethanol as above, cleared in xylenes, and stored in methyl salicylate. Glands were then manually analyzed using a light microscope for number of TEBs, ductal extension past the lymph node (in mm), and total ductal coverage of the fat pad (in %).

5.3.4 Western immunoblotting

At time of death, inguinal mammary glands were collected and prepared for Western blot exactly as previously described [162].

Blots were incubated with 1:2000 Anti-ERP57 (Pierce), 1:1500 Anti-VDR 9A7

(ThermoFisher) and 1:10,000 beta-Actin (GeneTex) antibodies. Incubation periods for ERP57 and beta-Actin were both 1 h at room temperature in a 1% w/v milk/TBS-T solution. Anti-VDR

9A7 was incubated with the blots overnight at 4°C in a 1% w/v BSA/TBS-T solution. The secondary for both the ERP57/MARRS and VDR antibodies was goat anti-rabbit conjugated to horseradish peroxidase (New England Biolabs) in a 1:50,000 dilution in 1% milk/TBS-T solution. The secondary for actin was goat anti-mouse conjugated to horseradish peroxidase

(Pierce) in a 1:10,000 dilution in the buffer recommended by the manufacturer. All secondaries were also used for 1 h at room temperature. Protein bands were visualized with the Clarity chemiluminesence kit (Bio-Rad).

Blots were then analyzed via densitometry to compare brightness of bands (n=4 minimum per group). One internal control sample was loaded onto every gel to allow comparison between blots.

5.3.5 Immunohistochemistry

At time of death, right abdominal mammary glands were fixed overnight in formalin, embedded in paraffin, and sliced into 5µm sections onto charged slides. Slides were dried at

37°C overnight. To prepare for staining, slides were deparaffinized in xylenes, quenched in 3% hydrogen peroxide for 20 min, and then treated with sodium citrate (pH 6.0) for 20 min at 95°C.

After cooling, slides were blocked in 3% normal goat serum for 1h, then incubated in the same

ERP57 antibody used in western blots at a dilution of 1:8000 overnight at 4°C. Anti-VDR 9A7

(ThermoFisher) was used at a dilution of 1:1000, and also incubated overnight at 4°C. Sections were rinsed with PBS, then incubated in Biotin-SP-affiniPure goat anti-rabbit IgG and

Peroxidase-Streptavidin (Cedarlane) for 1 h each diluted at 1:1000. Protein expression was visualized using a 3,3’-Diaminobenzidine (DAB) kit (Vector) for 60s, and then observed under a light microscope.

5.3.6 RNA analysis

Mammary tissue was digested in collagenase as described for Western blotting. RNA extraction was performed on the resulting cell pellet using the Quiagen RNeasy Lipid Kit

(Germany). RNA was tested for purity and quality using a Nanodrop (ThermoFisher), and cDNA was transcribed using TaqMan Reverse Transcription Reagents (Applied Biosystems). Primers are listed in Table 5.2. cDNA products were run on a 1.5% w/v agarose gel at 80 V for 1.5 h and viewed under UV light using the Safeview Classic DNA stain (Abcam).

5.3.7 Statistics

Data was tested for normal distribution. Differences between means of normally distributed data were analyzed using factorial two-way Analysis of Variance (ANOVA) tests. If the overall model was significant, differences between groups were found pairwise using the

Least Significant Difference post-hoc test. Differences were considered significant at p < 0.05.

All tests were conducted using SPSS Statistics (Version 25, IBM, Armonk, N.Y., USA).

Table 5.1. List of primers used in genotyping.

Gene Primer

MARRS forward CGC CAG CCT CTC CAT TTA G

MARRS reverse CAG AGA TCC TGC CTC TG

VDR wildtype CTC CAT CCC CAT GTG TCT TT

VDR common TTC TTC AGT GGC CAG CTC TT

VDR mutant CAC GAG ACT AGT GAG ACG TG

Table 5.2. List of primers used in RNA analysis.

Gene Forward Primer Reverse Primer Product Size (Base Pairs) VDR CCT GTG AAG GCT TGT TGC TCC TCA 243 GCA AGG GTT T GAC AGC TTG G MARRS CTG AGC TAG CCG GTC TCG GAA GGG 403 CAA AAC C TAG TGC TGFβ TGA CGT CAC TGG GCA CCA TCC ATG 170 AGT TGT ACG G ACA TGA ACC ERα GTT GGA TGC TGA TTG TGC CCC TCT ATG 664 ACC GCC AT ACC TGC TC GAPDH TGC ACC ACC AAC GAT GAC CTT GCC 209 TGC TTA GCC CAC AGC CTT

GAPDH primers were adapted from Hugel et al. [183]

ERα primers from Carley et al. [184]

TGFβ primers adapted from Yiakouvaki et al. [185]

MARRS primers from Richard et al. [110]

5.4 Results

5.4.1 Physical Characteristics and Breeding

B6.129S4-Vdr/J VDR knockout animals began to develop alopecia by week

5. These mice lost almost all fur by 8 wks, and females were not healthy enough to carry a healthy litter to full term. Therefore, to obtain double knockout animals, male VDR-/MARRS- were bred to VDR heterozygote/MARRS- females.

Mice tolerated the Teklad.94112 diet well. There was no overall significant interaction between genotypes and body weight; however, there were significant differences amongst VDR animals (Figure 5.1). The effect of VDR was significant (p=0.021), notably; VDR- animals had a

6-7% decrease in body weight compared to the MARRS+/VDR+ animals. MARRS status does not appear to affect body weight during puberty (Figure 5.1).

5.4.2 Mammary gland growth variables

There was a significant interaction between genotypes (p=0.004) for mean % mammary gland area covered by branching (Figures 5.2, 5.4). The MARRS+/VDR- animals had the least ductal coverage and were significantly different from all other groups (p=0.035). Reducing

MARRS expression also had an effect on ductal coverage and MARRS-/VDR+ animals had significantly (p=0.035) reduced ductal coverage compared to MARRS+/VDR+ animals.

There were quantitative (although non-significant) differences in ductal elongation past the lymph node (Figures 5.3, 5.4). MARRS+/VDR+ animals had the longest ductal extension, while any animal with either gene knockout showed shorter branches. Although the overall effect

of both MARRS and VDR genotypes were not significant, the effect of VDR alone was significant (p=0.002); VDR- animals (regardless of MARRS status) had the shortest ductal extension.

There were no significant interactions between groups regarding TEB numbers; however there is a tendency for MARRS+/VDR+ animals to have fewer TEBs, and double knockout animals to have the most (Figure 5.5).

5.4.3 Western immunoblots

Western blots and densitometry analysis (Figure 5.6) confirmed genotyping by demonstrating a successful reduction in both MARRS and VDR protein expression in the epithelial cells of the mammary gland (Figure 5.6). MARRS expression was reduced in knockout animals by greater than 50% compared to their wildtype counterparts (a reduction that approached significance, p=0.052), although MARRS expression was still detected in some knockout animals. VDR expression was also reduced by about 50% between wildtypes and VDR knockout animals, although due to much variation in expression between individual animals, this reduction was not significant.

5.4.4 Immunohistochemistry

Immunohistochemistry slides showed no gross differences in mammary gland structure at the cellular level between genotype groups. MARRS proteins seem to stain uniformly across adipocytes regardless of MARRS genotype (Figure 5.7). MARRS knockouts also clearly show a reduction in MARRS protein expression within epithelial cells. Within epithelial structures,

MARRS seems to be expressed on the apical side of ductal tubes.

We were unable to show a difference in VDR protein expression using VDR-9A7, with both wildtype and knockout animals showing similar staining patterns.

5.4.5 RNA expression

RNA expression of various genes showed some differences in gene expression other than the targets MARRS and VDR. MARRS-/VDR+ mice showed reductions in TGFβ gene expression compared to all other groups (Figure 5.8). ERα was also tested, and RNA expression was similar in all genotypes.

18.5 18 17.5 17 16.5 16 15.5 Mean Body Weight (g) Weight Body Mean 15 14.5

Genotype

Figure 5.1. Mean body weight (g) of 6wk old mice. No significant overall interaction between genotypes, but the effect of VDR was significant (p=0.021). VDR-ablated mice weighed less than VDR wildtype mice.

80 c b 70 b c 60 a 50

Mean % Area Coverage Coverage Area % Mean 10

Genotype

Figure 5.2. Mean % area of branching coverage throughout the mammary gland. There was a significant overall interaction (p=0.004), with significant differences between the MARRS+/VDR- group and all other groups (p=0.035), as well as between MARRS+/VDR+ and both MARRS-/VDR+ and MARRS+/VDR- (p=0.035).

2 Mean Ductal Extension Mean

Genotype

Figure 5.3. Mean ductal extension of alveolar branching through mammary glands. There was no significant overall interaction, but the effect of VDR was significant (p=0.002). VDR knockout mice had decreased ductal extension compared to other groups.

Figure 5.4. Whole mounts of female mammary glands at 6wks of age stained in carmine alum. Figures represent mammary glands from MARRS+/VDR+ (A), MARRS+/VDR- (B), MARRS- /VDR+ (C) and MARRS-/VDR- (D). Lymph nodes are noted with a star; TEBs are pointed out with an arrow. There were significant differences in area of ductal coverage with MARRS+/VDR+ branching covering significantly more area than MARRS+/VDR- or MARRS- /VDR+ animals (p=0.035). There were no significant differences between groups concerning number of TEBs.

Mean number of TEBs of number Mean 2

Genotype

Figure 5.5. Mean number of TEBs per genotype at 6 weeks of age. No significant differences between groups.

A B 0.3 0.25 0.2 0.15 MARRS 0.1 β-Actin 0.05

(corrected to Actin) to (corrected 0 +/+ +- -/- (+/+) (+/-) (-/-) Relative MARRS expression MARRS Relative expression Genotype

C D 0.15

0.1 VDR 0.05 β-Actin 0 +/+ +/- -/- Actin) to (corrected (+/+) (+/-) (-/-) Relative VDR expression VDR expression Relative Genotype

Figure 5.6. Western blot (A and C) and densitometry analysis (B and D) of MARRS (top row) and VDR (bottom row) in mammary gland epithelial cells. Westerns show successful reduction of both MARRS and VDR between wildtype (+/+) animals and knockouts respective of protein.

Figure 5.7. Immunohistochemistry results. Top row shows differences in MARRS staining (1:1000 dilution) between a MARRS+/VDR- (A) and a MARRS-/VDR+ animal (B). MARRS staining is concentrated within epithelial cells on the apical side of the alveolar tube. Middle row shows Anti-VDR staining between a MARRS-/VDR+ (C) and a MARRS-/VDR- animal (D). VDR staining is similar between animals, regardless of VDR genotype status. Bottom row is a sample with no primary antibody staining (E).

A B

GAPDH ERα

+ - + - MARRS + + - - + + - - MARRS - + - + VDR + - + - + - + - VDR

Figure 5.8. A) Reverse-transcribed RNA analysis of TGFβ and MARRS for a MARRS+/VDR- and MARRS-/VDR+ animal. The TGFβ band is 170bp and expression is reduced in MARRS- /VDR+ animals. Gel also shows a successful reduction of MARRS RNA in a MARRS- animal. B) Reverse-transcribed RNA analysis of ERα and GAPDH. The ERα is at 664bp and expression is similar between all genotypes White arrows indicate bands of interest. GAPDH included as housekeeping gene.

5.5 Discussion

MARRS is only now being more widely accepted as an alternate receptor for 1,25D3 and has only recently begun to be included in studies exploring 1,25D3 at the molecular level.

Furthermore, very few studies have investigated the relationship between MARRS and VDR and studies exploring any possible crosstalk between the two pathways are limited.

Previous studies exploring the roles of both VDR and MARRS suggest a potential interaction in the downstream signaling of both pathways [186]. In a recent study re-examining pathways involved in photoprotection, it was discovered using co-immunoprecipitation assays that VDR and MARRS physically interact with one another in nonnuclear extracts of human fibroblasts [186]. MARRS has also been found to interact with NFκB [71] as well as chaperone proteins MHC class I [187] and calnexin [135]. Therefore the purpose of this study was to attempt to elucidate any signaling pathway intersections between MARRS and VDR that manifest in differences at the tissue level using a double knockout mouse model.

Previous studies investigating the roles of either MARRS or VDR separately on murine mammary gland development have found opposing effects. While VDR knockout mice have increased numbers of TEBs, longer ductal extension, and increased number of alveolar branch points [66,188], MARRS knockout animals experience the opposite effect of fewer TEBs and minimal ductal extension through the fat pad [162]. In this study we attempted to determine if one vitamin D receptor’s signaling pathway supersedes the other in murine morphogenesis. We found that number of TEBs is consistent with the VDR knockout literature, with VDR knockout mice developing more TEBs than MARRS knockout or wildtype animals. However, these

differences were marginal, and did not reach statistical significance. These results suggest that neither MARRS nor VDR is essential for TEB development.

Interestingly, both ductal extension and area of branch coverage were lowest in

MARRS+/VDR- and MARRS-/VDR- groups, and highest in MARRS+/VDR+ animals. These results are the opposite from most other studies [66,188] where VDR ablation causes much more growth at this age. One possible explanation for the difference in apparent growth is that these mice are a combination of two different background strains previously not studied. Background strain has been found to elicit differential effects in mammary gland development when mice are exposed to both industrial toxins [189], n-3 polyunsaturated fatty acids [190], and high fat diets

[191], therefore genetic background may have a significant role in growth at this time that was not explored in this study. Also, the mean body weights of mice at 6 weeks also followed the same trend as ductal extension and branching coverage, in that VDR knockouts were also the smallest animals. This may have contributed to their mammary growth, and mammary development was consistent relative to their gross size.

The Western blots do not demonstrate a true knockout for either MARRS or VDR protein expression in mammary gland epithelial cells. However, the amount of protein reduction is consistent with previous studies in our lab [9, unpublished Meckling]. VDR expression is very low in epithelial cells, even in wildtype mice, which is in accordance with previous studies [172].

Also, there were no discernable differences in MARRS protein quantity between mice with differing VDR statuses (or vice versa, data not shown), thus suggesting there is no obvious connection or cross-talk regarding protein expression between these two receptors i.e. The reduction of one receptor did not result in the upregulation of the other.

The RNA reverse transcription results offer some insight into possible hormonal and growth factor changes in the mammary gland when either MARRS or VDR is knocked out.

There was a decrease in TGFβ gene expression in only the MARRS-/VDR+ group while all other groups had similar TGFβ expression. TGFβ is an important growth factor in mammary gland development being one of the few inhibitory influences on proliferation to control disorderly growth during puberty [8,45]. It has been previously shown in rat intestinal epithelial cells that TGFβ acts as a transcriptional regulator for MARRS [174], Rat IEC-6 cells were exposed to TGFβ-1, and MARRS transcript and protein expression increased compared to vehicle controls. Furthermore, in IEC-6 cells with a deleted smad 3 response element within the

MARRS receptor promoter, TGFβ-1 response was eliminated. This smad 3 element is also present in mice and humans and is clearly an important proponent of TGFβ -MARRS signaling

[174]. Although this interaction has not been extensively studied, there is also evidence that when MARRS is inhibited in platelet cells, the effects of TGFβ -1 are inhibited [192].

There were overall no differences in ERα expression in any of the genotype groups.

Estrogen is one of the most important regulators of mammary gland development, acting on the

Estrogen receptor α to cause an upregulation of amphiregulin, which in turn upregulates growth factors within the mammary epithelium to stimulate TEB formation and ductal growth [8]. ERα knockout mice experience impaired ductal proliferation and differentiation [18]. Although all animals in this study had similar ERα RNA expression, it is possible the sensitivity of the cells to estradiol has changed. A study in MARRS knockdown MCF-7 cells indicated an increased sensitivity to the growth effects of estradiol with no change in ERα protein expression

(unpublished Meckling). An increased sensitivity to estradiol in the epithelial cells of the

mammary gland in the mice in this study would ensure TEB development even amongst any growth-inhibition effects caused by a MARRS knockout.

Ultimately there were few significant differences in mammary gland growth between the double knockout animals and wildtypes, suggesting that in the absence of both VDR and

MARRS there is a more powerful growth regulator that ensures proper development. This regulator is likely hormonal, as estrogen and growth hormone have been shown to be essential in mammary gland growth. Estrogen is required for the formation of TEBs and outward ductal growth during puberty [8,18] and mice lacking mammary estrogen receptors do not form ductal structures [21,193]. Growth hormone, through insulin-like growth factor 1, is also essential for

TEB formation [194], and can rescue ductal growth in oophorectomized rats [31].

5.6 Conclusions

Overall the results of this study do not fully coincide with those of previous vitamin D- related mammary gland development studies. The roles of both MARRS and VDR are still not clear, and it is possible that background strain in animal studies may alter results. Moving forward, studies using multiple background strains will give a fuller picture of the effects of a treatment that will not be confounded by background genetics.

Future studies will need to further explore double knockouts at the cellular level, and determine more information about the availability of estrogen, growth hormone, and IGF-1.

Perhaps using a tissue-specific VDR knockout mouse instead of a whole-body knockout will focus the results on the mammary gland and account for any whole-body changes interfering with growth.

Chapter 6: Integrated Discussion

6.1 General Summary

The purpose of this thesis was to elucidate a role for MARRS in a tissue with known

VDR/1,25D3 actions to both provide further evidence for a separate 1,25D3 receptor, as well as to help define its actions in the developing pubertal mammary gland, as this is currently unknown. By using tissue-specific knockout animals, we were able to contain the deletion of

MARRS to the epithelial cells of the mammary gland, leaving expression unchanged in adipocytes. By utilizing a localized gene disruption that would potentially contain disrupted signalling to the mammary gland, this would minimize disruption to whole-body signaling that would confound results. By knocking out MARRS in these cells and comparing the results to wildtype littermates, we were able to study how a lack of MARRS protein alters mammary gland morphology, and therefore hypothesize on its actions.

6.1.1 Study 1

The results of the first study provide evidence that MARRS does indeed play a role in mammary gland morphology, in that its absence results in slower, possibly hindered growth at the 6-week time point. Therefore, MARRS exerts a growth-positive signal on mammary glands.

This is important information as the reduction in number of TEBs and shorter ducts at this time point is the exact opposite phenomenon from VDR knockouts; VDR is known to control and slow down growth. Hence, we hypothesized that any 1,25D3 in circulation would bind to VDR in the absence of MARRS and exert anti-growth effects. However, the reduction in TEBs and overall slower ductal lengthening did not seem to hinder the function of the gland, as knockout pregnant females were able to nurse healthy litters. Therefore, it is not necessarily that absence

of MARRS has a negative impact on mammary gland growth; it may, in fact, have a benefit, in that the same functional tissue can be achieved by adulthood without increasing the number of

TEBs and subsequent risks associated with them.

Once we had ensured that MARRS played a significant role in mammary gland development, we sought to evaluate the impact of MARRS abolition in different vitamin D environments.

6.1.2 Study 2

The rationale behind Study 2 was based on previous cell experiments showing reduction of MARRS protein in breast cancer cells enhanced their sensitivity to the anti-growth effects of

1,25D3. Therefore, we wanted to translate this hypothesis into an in vivo model, as well as provide evidence that MARRS is capable of interacting with, or responding to 1,25D3. Since we had established that in the absence of MARRS, mammary glands experience slower growth, we hypothesized that by introducing an abundance of vitamin D3 and presumably, 1,25D3, the anti- growth effects of the VDR would be enhanced. However, the results were not so straightforward.

It is important to note that while plasma 25(OH)D levels correlated with vitamin D3 doses in the diets, this study design did not directly measure active 1,25D3 at the tissue level, nor the direct interaction between 1,25D3 and MARRS. We assumed active 1,25D3 would be increased/decreased at the tissue relative to D3 dose due to previous studies finding CYP27B1

RNA expression in mouse mammary glands [57]. However, it would have been beneficial to know if the amount of active metabolite was affected by either the changes in MARRS expression or the availability of circulating 25(OH)D.

An important conclusion from this study is that MARRS indeed responds to changes in

25(OH)D status in mice, thus providing further indirect evidence of MARRS as a 1,25D3 receptor, which is highly controversial in the literature. The fact that there were limited changes in TEBs and mammary morphology when mice were on a deficient diet shows MARRS requires an adequate 25(OH)D status for at least part of its actions in this tissue, and all 1,25D3 actions cannot be assigned to VDR alone.

Some of the results from this study were unexpected compared to the known actions of

1,25D3 and MARRS in this tissue. We expected growth to be limited in groups with both reduced MARRS expression and increased 25(OH)D. This was true for MARRS knockdowns on replete diets, but not true for the knockouts, where growth was suddenly increased despite the abundance of 25(OH)D in circulation. It is unknown why, with both reduced MARRS expression and excess 25(OH)D – both growth-slowing treatments – would suddenly cause a burst of TEB growth. These results led us to believe that when considering 1,25D3 signaling through MARRS or VDR pathways, taking away one pathway does not simply result in a shunt to the other. There may be potential cross-talk between the pathways, or interactions with other signaling molecules such as IGF-1, retinol, estrogen, progesterone, etc.

6.1.3 Study 3

In light of the results of study 2, we believed the downstream signaling pathways of

MARRS and VDR were possibly interacting, and we decided to study both knockouts simultaneously. By developing a combination of mouse models differing in MARRS and VDR receptor status, we hoped to determine if the actions of one pathway supersedes the other when

determining the mammary physiology. This is the first study, to our knowledge, that uses a double knockout of both MARRS and VDR in mice.

We hypothesized that the number of TEBs would continue to follow the patterns as seen previously i.e. MARRS knockouts would have fewer TEBs, and VDR knockouts would have more. However, this was not the case, and all groups had similar numbers of TEBs. This would suggest a much more important factor in TEB formation is at play to ensure proper development.

In regard to other growth variables, while MARRS knockouts had shortened ducts and slightly less branching coverage as usual, the VDR knockouts had even further reductions compared to double wildtypes, which is uncharacteristic of this genotype. However, it is possible that overall body size is playing the most important role here, as the MARRS+/VDR- group had the shortest ducts and least amount of coverage and was also significantly smaller in total body size than other groups. It may have been beneficial to use a tissue-specific knockout model for the VDR knockouts instead of a whole-body knockout. The whole-body knockout seemed to have an effect on overall growth of the animals, which in turn affected the growth of the tissue.

The results of studies 2 and 3 do not provide a clear-cut interaction between MARRS,

VDR and 1,25D3, and suggest that the signaling within the gland, at this time point, is much more complicated than 1 ligand and 2 receptors providing opposing actions. These studies would have benefited from further molecular analysis of signaling molecules within the gland including estrogen and other hormones/growth factors involved in development. Further details can be found in the Future Directions section below.

One limitation across all study methodologies pertains to the method of separating epithelial cell from adipocytes. As MARRS signaling was localized to epithelial cells, while

adipocytes remained untouched, it was important when collecting cells for Western blotting and

RNA analysis that adipocytes be discarded from the sample. The method we used for separating cells reduced the fraction of adipocytes significantly, however, it is possible that the remaining sample was still contaminated by adipocytes and possibly other matrix cells (fibroblasts etc.), which may explain the unclear RNA results in study 3, and the imperfect knockout results in all

Western blots. Especially due to the complicated stromal-epithelial signaling within this tissue,

FACS separation of epithelial cells from adipocytes would have been a valuable next step to provide a pure sample for further analysis.

6.2 Future Directions

Now that we know MARRS plays a role in mammary gland development, and it has a relationship with 1,25D3 in this tissue, the next step is to approach the MARRS knockouts from a more molecular angle to elucidate the pathways and signaling molecules at play. Firstly, estrogen signaling would be an obvious molecule to investigate. Its role in mammary development is essential, but there are reports of 1,25D3 affecting estrogen production [195] as well as VDR knockout mice having increased sensitivity to hormones [66]. If one of the ways

VDR controls growth is by inhibiting estrogen, it is possible MARRS contributes the opposite effect.

Also, it is known that MARRS can interact directly with TGFβ signaling [174] and TGFβ is a primary anti-growth signal during puberty. It is also possible MARRS is interfering in TGFβ signaling to produce a growth response. Study 3 aimed to investigate both changes in ERα receptor and TGFβ expression, but the results were not conclusive. Perhaps a more direct and dependable way of measuring expression of these molecules would be to dissect mammary

organoids from pubertal mice and treat them ex vivo to examine the direct effects of 1,25D3 in the knockout model.

Another avenue to explore would be the differences between stromal and epithelial signaling regarding MARRS and 1,25D3. Although known for decades that stromal signaling affects the expressed receptors and subsequent growth of alveolar ducts, it was important to first ensure a role for MARRS in mammary glands before investigating inter-compartment signaling.

Now that it is known MARRS can affect epithelial growth, we can also study its effects on adipocytes, and how they interact with epithelial cells. By knocking out MARRS in the adipocyte fraction we will be able to see if MARRS is essential in either compartment.

Currently there is not a lot of research defining a causal relationship between number of

TEBs and future risk of breast cancer. Now that we have established a role for MARRS in mammary gland development as well as its differing sensitivity to available 1,25D3 in circulation, the next step would be to translate this MARRS knockout into a mouse model of breast cancer. By using either MMTV-neu mice, or DMBA-induced tumorigenesis, it would be very interesting to see if knocking out MARRS in vivo will alter the sensitivity of TEBs to carcinogen-induced tumorigenesis during puberty and how that affects breast cancer rates later in life. Also it would be interesting to know if the MARRS knockout will enhance breast tumor sensitivities to 1,25D3 treatment, as it has in previous cell culture studies [110]. Considering previous studies using 1,25D3 to treat breast cancer requires doses that result in hypercalcemia

[196], finding ways to reduce the 1,25D3 dose while maintaining the effectiveness of anti-tumor properties would be extremely beneficial. These experiments will enhance our understanding of

1,25D3 and its anti-cancer properties; it is possible that 1,25D3 treatment of breast cancer could be more effective in conjunction with locally acting MARRS-blocking antibodies.

6.3 Conclusions

This thesis provides the first evidence of MARRS’s role in mammary gland development and provides a new avenue to further understand mammary gland morphology and the origins of breast cancer risk. These results also provide further evidence that MARRS’s actions are affected by changes in 25(OH)D status, and can have consequences in mammary gland morphology and possibly breast cancer development and treatment. It is now important, moving forward, that studies investigating any roles of 1,25D3 or VDR also take into account the presence and actions of MARRS, and not to assume all actions of 1,25D3 act through the VDR. Further appreciation of the actions of both receptors and their possible intertwining signaling pathways will enhance our ability to understand and treat disease.

REFERENCES

[1] H. Macias, L. Hinck, Mammary Gland Development, Wiley Interdiscip. Rev. Dev. Biol. 1 (2013) 533–557. doi:10.1002/wdev.35.Mammary. [2] L. Hennighausen, G.W. Robinson, Signaling Pathways in Mammary Gland Development, Dev. Cell. 1 (2001) 467–475. doi:10.1016/S1534-5807(01)00064-8. [3] H. Parmar, G.R. Cunha, Epithelial-stromal interactions in the mouse and human mammary gland in vivo, Endocr. Relat. Cancer. 11 (2004) 437–458. doi:10.1677/erc.1.00659. [4] L.A. Jerome-Majewska, F. Zindy, V.E. Papaioannou, C.J. Sherr, E. Ernstoff, G.P. Jenkins, Tbx3 , the ulnar-mammary syndrome gene, and Tbx2 interact in mammary gland development through a p19 Arf /p53-independent pathway, Dev. Dyn. 234 (2005) 922– 933. doi:10.1002/dvdy.20575. [5] H. Dürnberger, K. Kratochwil, Specificity of tissue interaction and origin of mesenchymal cells in the androgen response of the embryonic mammary gland, Cell. 19 (1980) 465– 471. doi:10.1016/0092-8674(80)90521-8. [6] G. Musumeci, P. Castrogiovanni, M.A. Szychlinska, F.C. Aiello, G.M. Vecchio, L. Salvatorelli, G. Magro, R. Imbesi, Mammary gland: From embryogenesis to adult life, Acta Histochem. 117 (2015) 379–385. doi:10.1016/j.acthis.2015.02.013. [7] S. Mcnally, T. Stein, Overview of Mammary Gland Development: A Comparison of Mouse and Human, in: F. Martin, T. Stein, J. Howlin (Eds.), Methods Mol. Biol., Humana Press, New York NY, 2017. doi:10.1007/978-1-4939-6475-8. [8] I.S. Paine, M.T. Lewis, The Terminal End Bud: the Little Engine that Could, J. Mammary Gland Biol. Neoplasia. 22 (2017) 93–108. doi:10.1007/s10911-017-9372-0. [9] L. Bai, L.R. Rohrschneider, s-SHIP promoter expression marks activated stem cells in developing mouse mammary tissue, Genes Dev. 24 (2010) 1882–1892. doi:10.1101/gad.1932810. [10] M. Shackleton, F. Vaillant, K.J. Simpson, J. Stingl, G.K. Smyth, M.L. Asselin-Labat, L. Wu, G.J. Lindeman, J.E. Visvader, Generation of a functional mammary gland from a single stem cell, Nature. 439 (2006) 84–88. doi:10.1038/nature04372. [11] A.J. Ewald, A. Brenot, M. Duong, B.S. Chan, Z. Werb, Collective Epithelial Migration and Cell Rearrangements Drive Mammary Branching Morphogenesis, Dev. Cell. 14 (2008) 570–581. doi:10.1016/j.devcel.2008.03.003. [12] A.J. Ewald, R.J. Huebner, H. Palsdottir, J.K. Lee, M.J. Perez, D.M. Jorgens, A.N. Tauscher, K.J. Cheung, Z. Werb, M. Auer, Mammary collective cell migration involves transient loss of epithelial features and individual cell migration within the epithelium, J. Cell Sci. 125 (2012) 2638–2654. doi:10.1242/jcs.096875.

[13] M. Nassour, Y. Idoux-Gillet, A. Selmi, C. Côme, M.L.M. Faraldo, M.A. Deugnier, P. Savagner, Slug Controls Stem/Progenitor Cell Growth Dynamics during Mammary Gland

Morphogenesis, PLoS One. 7 (2012). doi:10.1371/journal.pone.0053498. [14] L. Bakiri, S. MacHo-Maschler, I. Custic, J. Niemiec, A. Guió-Carrión, S.C. Hasenfuss, A. Eger, M. Müller, H. Beug, E.F. Wagner, Fra-1/AP-1 induces EMT in mammary epithelial cells by modulating Zeb1/2 and TGFβ expression, Cell Death Differ. 22 (2015) 336–350. doi:10.1038/cdd.2014.157. [15] B.S. Wiseman, M.D. Sternlicht, L.R. Lund, C.M. Alexander, J. Mott, M.J. Bissell, P. Soloway, S. Itohara, Z. Werb, Site-specific inductive and inhibitory activities of MMP-2 and MMP-3 orchestrate mammary gland branching morphogenesis, J. Cell Biol. 162 (2003) 1123–1133. doi:10.1083/jcb.200302090. [16] T.Y. Feinberg, R.G. Rowe, T.L. Saunders, S.J. Weiss, Functional roles of MMP14 and MMP15 in early postnatal mammary gland development, Development. 143 (2016) 3956– 3968. doi:10.1242/dev.136259. [17] F.F. Bolander, Y.J. Topper, Loss of Differentiative Potential of the Mammary Gland in Ovariectomized Mice: Prevention and Reversibility of the Defect, Endocrinology. 107 (1980) 1281=1285. [18] S. Mallepell, A. Krust, P. Chambon, C. Brisken, Paracrine signaling through the epithelial estrogen receptor-alpha is required for proliferation and morphogenesis in the mammary gland, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 2196–2201. [19] C.W. Daniel, G.B. Silberstein, P. Strickland, Direct action of 17β-estradiol on mouse mammary ducts analyzed by sustained release implants and steroid autoradiography, Cancer Res. 47 (1987) 6052–6057. [20] G.R. Cunha, P. Young, Y.K. Hom, P.S. Cooke, J.A. Taylor, D.B. Lubahn, Elucidation of a role for stromal steroid hormone receptors in mammary gland growth and development using tissue recombinants, J. Mammary Gland Biol. Neoplasia. 2 (1997) 393–402. doi:10.1023/A:1026303630843. [21] S.O. Mueller, J.A. Clark, P.H. Myers, K.S. Korach, Mammary gland development in adult mice requires epithelial and stromal estrogen receptor α, Endocrinology. 143 (2002) 2357–2365. doi:10.1210/endo.143.6.8836. [22] N. Zeps, J.M. Bentel, J.M. Papadimitriou, M.F. D’Antuono, H.J.S. Dawkins, Estrogen receptor-negative epithelial cells in mouse mammary gland development and growth, Differentiation. 62 (1998) 221–226. doi:10.1046/j.1432-0436.1998.6250221.x. [23] L. Ciarloni, S. Mallepell, C. Brisken, Amphiregulin is an essential mediator of estrogen receptor-alpha function in mammary gland development, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 5455–5460. [24] M.D. Sternlicht, S.W. Sunnarborg, H. Kouros-Mehr, Y. Yu, D.C. Lee, Z. Werb, Mammary ductal morphogenesis requires paracrine activation of stromal EGFR via ADAM17-dependent shedding of epithelial amphiregulin, Development. 132 (2005) 3923–3933. doi:10.1242/dev.02314.

[25] J.F. Wiesen, P. Young, Z. Werb, G.R. Cunha, Signaling through the stromal epidermal growth factor receptor is necessary for mammary ductal development., Development. 126 (1999) 335–44. http://www.ncbi.nlm.nih.gov/pubmed/9847247. [26] K.L. Hull, S. Harvey, Growth Hormone and Reproduction: A Review of Endocrine and Autocrine/Paracrine Interactions, Int. J. Endocrinol. 2014 (2014) 1–24. doi:10.1155/2014/234014. [27] S. Mukhina, D.X. Liu, K. Guo, M. Raccurt, S. Borges-Bendris, H.C. Mertani, P.E. Lobie, Autocrine growth hormone prevents lactogenic differentiation of mouse mammary epithelial cells, Endocrinology. 147 (2006) 1819–1829. doi:10.1210/en.2005-1082. [28] D.L. Kleinberg, T.L. Wood, P.A. Furth, A. V. Lee, Growth hormone and insulin-like growth factor-I in the transition from normal mammary development to preneoplastic mammary lesions, Endocr. Rev. 30 (2009) 51–74. doi:10.1210/er.2008-0022. [29] P.D. Walden, W. Ruan, M. Feldman, D.L. Kleinberg, Evidence That the Mammary Fat Pad Mediates the Action of Growth Hormone in Mammary Gland Development, Endocrinology. 139 (1998) 659–662. [30] W. Ruan, D.L. Kleinberg, Insulin-Like Growth Factor I Is Essential for Terminal End Bud Formation and Ductal Morphogenesis during Mammary Development, Endocrinology. 140 (1999) 5075–5081. doi:10.1210/en.140.11.5075. [31] W. Ruan, V. Catanese, R. Wieczorek, M. Feldman, D.L. Kleinberg, Estradiol enhances the stimulatory effect of insulin-like growth factor-I (IGF-I) on mammary development and growth hormone-induced IGF-I messenger ribonucleic acid, Endocrinology. 136 (1995) 1296–1302. [32] W. Ruan, M.E. Monaco, D.L. Kleinberg, Progesterone stimulates mammary gland ductal morphogenesis by synergizing with and enhancing insulin-like growth factor-I action, Endocrinology. 146 (2005) 1170–1178. doi:10.1210/en.2004-1360. [33] L.M. Arendt, C. Kuperwasser, Form and function: how estrogen and progesterone regulate the mammary epithelial hierarchy, J Mammary Gland Biol Neoplasia. 20 (2015) 9–25. doi:10.1007/s10911-015-9337-0.Form. [34] C.S. Atwood, R.C. Hovey, J.P. Glover, G. Chepko, E. Ginsburg, J. Robison, B.K. Vonderhaar, Progesterone induces side-branching of the ductal epithelium in the mammary glands of peripubertal mice, J. Endocrinol. 167 (2000) 39–52. doi:10.1677/joe.0.1670039. [35] J.P. Lydon, F.J. DeMayo, C.R. Funk, S.K. Mani, A.R. Hughes, C.A. Montgomery, G. Shyamala, O.M. Conneely, B.W. O’Malley, Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities, Genes Dev. 9 (1995) 2266–2278. doi:10.1101/gad.9.18.2266. [36] C. Brisken, S. Park, T. Vass, J.P. Lydon, B.W. O’Malley, R.A. Weinberg, A paracrine role for the epithelial progesterone receptor in mammary gland development, Proc. Natl.

Acad. Sci. 95 (1998) 5076–5081. [37] S.Z. Haslam, G. Ghyamala, Relative distribution of estrogen and progesterone receptors among the epithelial, adipose, and connective tissue coponents of the normal mammary gland, Endocrinology. 108 (1981) 825–830. [38] M. Beleut, R.D. Rajaram, M. Caikovski, A. Ayyanan, D. Germano, Y. Choi, P. Schneider, C. Brisken, Two distinct mechanisms underlie progesterone-induced proliferation in the mammary gland, Proc. Natl. Acad. Sci. 107 (2010) 2989–2994. doi:10.1073/pnas.0915148107. [39] T.K. Said, O.M. Conneely, D. Medina, B.W. O’Malley, J.P. Lydon, Progesterone, in addition to estrogen, induces cyclin D1 expression in the murine mammary epithelial cell, in vivo, Endocrinology. 138 (1997) 3933–3939. doi:10.1210/endo.138.9.5436. [40] R. Fernandez-Valdivia, A. Mukherjee, Y. Ying, J. Li, M. Paquet, F.J. DeMayo, J.P. Lydon, The RANKL signaling axis is sufficient to elicit ductal side-branching and alveologenesis in the mammary gland of the virgin mouse, Dev. Biol. 328 (2009) 127– 139. doi:10.1016/j.ydbio.2009.01.019. [41] L.M. Wakefield, Y.A. Yang, O. Dukhanina, Transforming growth factor-β and breast cancer: Lessons learned from genetically altered mouse models, Breast Cancer Res. 2 (2000) 100–106. doi:10.1186/bcr41. [42] G.B. Silberstein, C.W. Daniel, Reversible inhibition of mammary gland growth by transforming growth factor-beta, Acad. OneFile. 237 (1987) 291–294. [43] W. V. Ingman, S.A. Robertson, Mammary Gland Development in Transforming Growth Factor Beta1 Null Mutant Mice: Systemic and Epithelial Effects, Biol. Reprod. 79 (2008) 711–717. doi:10.1095/biolreprod.107.067272. [44] D.F. Pierce, M.D. Johnson, Y. Matsui, S.D. Robinson, L.I. Gold, A.F. Purchio, C.W. Daniel, B.L.M. Hogan, H.L. Moses, Inhibition of mammary duct development but not alveolar outgrowth during pregnancy in transgenic mice expressing active TGF-β1, Genes Dev. 7 (1993) 2308–2317. doi:10.1101/gad.7.12a.2308. [45] K.B. Ewan, G. Shyamala, S.A. Ravani, Y. Tang, R. Akhurst, L. Wakefield, M.H. Barcellos-hoff, Latent Transforming Growth Factor-beta Activation in Mammary Gland - Regulation by Ovarian Hormones Affects Ductal and Alveolar, Am. J. Pathol. 160 (2002) 2081–2093. [46] E.A. Kohn, Z. Du, M. Sato, C.M.H. Van Schyndle, M.A. Welsh, Y. an Yang, C.H. Stuelten, B. Tang, W. Ju, E.P. Bottinger, L.M. Wakefield, A novel approach for the generation of genetically modified mammary epithelial cell cultures yields new insights into TGFβ signaling in the mammary gland, Breast Cancer Res. 12 (2010) R83. doi:10.1186/bcr2728. [47] Y. Yang, B. Tang, G. Robinson, L. Hennighausen, B. SG, C. Deng, L. Wakefield, Smad3 in the mammary epithelium has a nonredundant role in the induction of apoptosis, but not

in the regulation of proliferation or differentiation by transforming growth factor-beta, Cell Growth Differ. 13 (2002) 123–130. [48] K.B.R. Ewan, H.A. Oketch-Rabah, S.A. Ravani, G. Shyamala, H.L. Moses, M.H. Barcellos-Hoff, Proliferation of estrogen receptor-α-positive mammary epithelial cells is restrained by transforming growth factor-β1 in adult mice, Am. J. Pathol. 167 (2005) 409– 417. doi:10.1016/S0002-9440(10)62985-9. [49] J. Massagué, TGFb in cancer, Cell. 134 (2008) 215–230. doi:10.1016/j.cell.2008.07.001.TGF. [50] D. Llobet-Navas, R. Rodriguez-Barrueco, J. de la Iglesia-Vicente, M. Olivan, V. Castro, L. Saucedo-Cuevas, N. Marshall, P. Putcha, M. Castillo-Martin, E. Bardot, E. Ezhkova, A. Iavarone, C. Cordon-Cardo, J.M. Silva, The MicroRNA 424/503 Cluster Reduces CDC25A Expression during Cell Cycle Arrest Imposed by Transforming Growth Factor in Mammary Epithelial Cells, Mol. Cell. Biol. 34 (2014) 4216–4231. doi:10.1128/mcb.00611-14. [51] H. Moses, M.H. Barcellos-hoff, B.F. Sloane, G.W. Robinson, M.A. Labarge, J.M. Rosen, D. Piwnica-worms, TGF-beta Biology in Mammary Development and Breast Cancer, Cold Spring Harb. Perspect. Biol. 3 (2011). [52] M.F. Holick, Resurrection of vitamin D and rickets, J. Clin. Invest. 116 (2006) 2062– 2072. doi:10.1172/JCI29449.2062. [53] Vitamin D and Calcium: Updated Dietary Reference Intakes, (2019). https://www.canada.ca/en/health-canada/services/food-nutrition/healthy-eating/vitamins- minerals/vitamin-calcium-updated-dietary-reference-intakes-nutrition.html. [54] C.S. Trump, Donald L., Deeb, Kristen., Johnson, Vitamin D: Considerations in the Continued Development as an Agent for Cancer Prevention and Therapy, Cancer J. 16 (2010) 1–9. doi:10.1097/PPO.0b013e3181c51ee6.Vitamin. [55] S. Christakos, P. Dhawan, A. Verstuyf, L. Verlinden, G. Carmeliet, Vitamin D: Metabolism, Molecular Mechanism of Action, and Pleiotropic Effects, Physiol. Rev. 96 (2016) 365–408. doi:10.1152/physrev.00014.2015. [56] G. Bivona, L. Agnello, M. Ciaccio, The immunological implication of the new Vitamin D metabolism, Cent. Eur. J. Immunol. 43 (2018) 331–334. doi:10.5114/ceji.2018.80053. [57] N. Zhalehjoo, Y. Shakiba, M. Panjehpour, Gene expression profiles of CYP24A1 and CYP27B1 in malignant and normal breast tissues, Mol. Med. Rep. 15 (2017) 467–473. doi:10.3892/mmr.2016.5992. [58] R.A. Replogle, Q. Li, L. Wang, M. Zhang, J.C. Fleet, Gene-by-diet interactions influence calcium absorption and bone density in mice, J. Bone Miner. Res. 29 (2014) 657–665. doi:10.1002/jbmr.2065. [59] J.W. Pike, S. Christakos, Biology and Mechanisms of Action of the Vitamin D Hormone, Endocrinol. Metab. Clin. North Am. 46 (2017) 815–843. doi:10.1016/j.ecl.2017.07.001.

100

[60] S. Akhter, G.D. Kutuzova, S. Christakos, H.F. DeLuca, Calbindin D9k is not required for 1,25-dihydroxyvitamin D3-mediated Ca2+ absorption in small intestine, Arch. Biochem. Biophys. 460 (2007) 227–232. doi:10.1016/j.abb.2006.12.005. [61] B.S. Benn, D. Ajibade, A. Porta, P. Dhawan, M. Hediger, J. Bin Peng, Y. Jiang, T.O. Goo, E.B. Jeung, L. Lieben, R. Bouillon, G. Carmeliet, S. Christakos, Active intestinal calcium transport in the absence of transient receptor potential vanilloid type 6 and calbindin-D 9k, Endocrinology. 149 (2008) 3196–3205. doi:10.1210/en.2007-1655. [62] S. Kim, M. Yamazaki, L.A. Zella, N.K. Shevde, J.W. Pike, Activation of Receptor Activator of NF- B Ligand Gene Expression by 1,25-Dihydroxyvitamin D3 Is Mediated through Multiple Long-Range Enhancers, Mol. Cell. Biol. 26 (2006) 6469–6486. doi:10.1128/mcb.00353-06. [63] N. Takahashi, N. Udagawa, T. Suda, Vitamin D endocrine system and osteoclasts, Bonekey Rep. 3 (2014) 1–9. doi:10.1038/bonekey.2013.229. [64] Y.C. Li, A.E. Pirro, M. Amling, G. Delling, R. Baron, R. Bronson, M.B. Demay, Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia., Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 9831–5. doi:10.1073/pnas.94.18.9831. [65] D. Yang, P.H. Anderson, A.R. Wijenayaka, K.R. Barratt, R. Triliana, C.J.M. Stapledon, H. Zhou, D.M. Findlay, H.A. Morris, G.J. Atkins, Both ligand and VDR expression levels critically determine the effect of 1α,25-dihydroxyvitamin-D 3 on osteoblast differentiation, J. Steroid Biochem. Mol. Biol. 177 (2018) 83–90. doi:10.1016/j.jsbmb.2017.09.005. [66] G. Zinser, K. Packman, J. Welsh, Vitamin D(3) receptor ablation alters mammary gland morphogenesis., Development. 129 (2002) 3067–76. http://www.ncbi.nlm.nih.gov/pubmed/12070083. [67] A.L. Johnson, G.M. Zinser, S.E. Waltz, Loss of vitamin D receptor signaling from the mammary epithelium or adipose tissue alters pubertal glandular development, Am. J. Physiol. - Endocrinol. Metab. 307 (2014) 674–685. doi:10.1152/ajpendo.00200.2014. [68] I. Nemere, C.M. Szego, Early actions of parathyroid hormone and 1,25- dihydroxycholecalciferol on isolated epithelial cells from rat instestine: 1. Limited lysosomal enzyme release and calcium uptake, Endocrinology. 108 (1981) 1450–1462. [69] I. Nemere, M.C. Farach-Carson, Membrane receptors for steroid hormones: A case for specific cell surface binding sites for vitamin D metabolites and estrogens, Biochem. Biophys. Res. Commun. 248 (1998) 443–449. doi:10.1006/bbrc.1998.8492. [70] I. Nemere, M.C. Dormanen, M.W. Hammond, W.H. Okamura, A.W. Norman, Identification of a Specific Binding Protein for la,25-Dihydroxyvitamin D, in Basal- Lateral Membranes of Chick Intestinal Epithelium and Relationship to Transcaltachia*, J. Biol. Chem. 269 (1994) 23750–23756. http://www.jbc.org/content/269/38/23750.full.pdf#page=1&view=FitH.

101

[71] W. Wu, G. Beilhartz, Y. Roy, C.L. Richard, M. Curtin, L. Brown, D. Cadieux, M. Coppolino, M.C. Farach-Carson, I. Nemere, K.A. Meckling, Nuclear translocation of the 1,25D3-MARRS (membrane associated rapid response to steroids) receptor protein and NFkB in differentiating NB4 leukemia cells, Exp. Cell Res. 316 (2010) 1101–1108. doi:10.1016/j.yexcr.2010.01.010. [72] I. Nemere, R. Ray, W. McManus, Immunochemical studies on the putative plasmalemmal receptor for 1,25(OH)-D3 . I. Chick intestine, Am. J. Physiol. Metab. 278 (2000) E1104– E1114. doi:10.1152/ajpendo.2000.278.6.e1104. [73] I. Nemere, K. Campbell, Immunochemical studies on the putative plasmalemmal receptor for 1,25-dihydroxyvitamin D3. Vitamin D status, Steroids. 65 (2000) 451–457. [74] I. Nemere, N. Garbi, G.J. Hämmerling, R.C. Khanal, Intestinal cell calcium uptake and the targeted knockout of the 1,25D 3-MARRS (Membrane-associated, Rapid Response Steroid-binding) receptor/PDIA3/Erp57, J. Biol. Chem. 285 (2010) 31859–31866. doi:10.1074/jbc.M110.116954. [75] I. Nemere, N. Garcia-Garbi, G.J. Hämmerling, Q. Winger, Intestinal cell phosphate uptake and the targeted knockout of the 1,25D 3-MARRS receptor/PDIA3/ERp57, Endocrinology. 153 (2012) 1609–1615. doi:10.1210/en.2011-1850. [76] R. Khanal, I. Nemere, The ERp57/GRp58/1,25D3-MARRS Receptor: Multiple Functional Roles in Diverse Cell Systems, Curr. Med. Chem. 14 (2007) 1087–1093. doi:10.2174/092986707780362871. [77] R.C. Khanal, T.M. Sterling Peters, N.M. Smith, I. Nemere, Membrane receptor-initiated signaling in 1,25(OH)2D 3-stimulated calcium uptake in intestinal epithelial cells, J. Cell. Biochem. 105 (2008) 1109–1116. doi:10.1002/jcb.21913. [78] M. Mesbah, I. Nemere, P. Papagerakis, J. Nefussi, S. Orestes-cardoso, C. Nessmann, A. Berdal, Expression of a 1,25-dihydroxyvitamin D3 Membrane-Associated Rapid- Response Steroid Binding Protein During Human Tooth and Bone Development and Biomineralization, J. Bone Miner. Res. 17 (2002) 1588–1596. [79] A. Berdal, M. Mesbah, P. Papagerakis, I. Nemere, Putative Membrane Receptor for 1,25(OH) 2 Vitamin D 3 in Human Mineralized Tissues During Prenatal Development, Connect. Tissue Res. 44 (2003) 136–140. doi:10.1080/03008200390152223. [80] B.D. Boyan, V.L. Sylvia, N. McKinney, Z. Schwartz, Membrane Actions of Vitamin D Metabolites 1α,25(OH)2D 3 and 24R,25(OH)2D3 Are Retained in Growth Plate Cartilage Cells from Vitamin D Receptor Knockout Mice, J. Cell. Biochem. 90 (2003) 1207–1223. doi:10.1002/jcb.10716. [81] J. Meszaros, N. Karin, M.C. Farach-Carson, Voltage-sensitive calcium channels in osteoblasts: mediators of plasma membrane signalling events, Connect. Tissue Res. 35 (1996) 107–111.

[82] B.D. Boyan, L. Wang, K.L. Wong, H. Jo, Z. Schwartz, Plasma membrane requirements

102

for 1α,25(OH)2D3 dependent PKC signaling in chondrocytes and osteoblasts, Steroids. 71 (2006) 286–290. doi:10.1016/j.steroids.2005.09.018. [83] P.J. Hughes, A. Steinmeyer, R.A.S. Chandraratna, G. Brown, 1α,25-dihydroxyvitamin D3 stimulates steroid sulphatase activity in HL60 and NB4 acute myeloid leukaemia cell lines by different receptor-mediated mechanisms, J. Cell. Biochem. 94 (2005) 1175–1189. doi:10.1002/jcb.20377. [84] P.J. Hughes, J.S. Lee, N.E. Reiner, G. Brown, The vitamin D receptor-mediated activation of phosphatidylinositol 3-kinase (PI3Kα) plays a role in the 1α,25-dihydroxyvitamin D 3- stimulated increase in steroid sulphatase activity in myeloid leukaemic cell lines, J. Cell. Biochem. 103 (2008) 1551–1572. doi:10.1002/jcb.21545. [85] B.J. Willcox, D.C. Willcox, Caloric Restriction, CR Mimetics, and Healthy Aging in Okinawa: Controversies and Clinical Implications, Curr Opin Clin Nutr Metab Care. 17 (2014) 51–58. doi:10.1037/emo0000122.Do. [86] I. Nemere, N. Garbi, Q. Winger, The 1,25D3-MARRS receptor/PDIA3/ERp57 and lifespan, J. Cell. Biochem. 116 (2015) 380–385. doi:10.1002/jcb.24986. [87] Canadian Cancer Statistics 2018, Toronto, ON, 2018. cancer.ca/Canadian-Cancer- Statistics-2018-EN (accessed May 3, 2019). [88] E.T. Jacobs, L.N. Kohler, A.G. Kunihiro, P.W. Jurutka, Vitamin D and colorectal, breast, and prostate cancers: A review of the epidemiological evidence, J. Cancer. 7 (2016) 232– 240. doi:10.7150/jca.13403. [89] K.A.Y.W. Colston, U.T.A. Berger, R.C. Coombes, Possible role for vitamin D in controlling breast cancer cell proliferation, Lancet. (1989) 188–191. [90] A. Vrieling, P. Seibold, T.S. Johnson, J. Heinz, N. Obi, R. Kaaks, D. Flesch-Janys, J. Chang-Claude, Circulating 25-hydroxyvitamin D and postmenopausal breast cancer survival: Influence of tumor characteristics and lifestyle factors?, Int. J. Cancer. 134 (2014) 2972–2983. doi:10.1002/ijc.28628. [91] M.-H. Shin, M.D. Holmes, S.E. Hankinson, K. Wu, G.A. Colditz, W.C. Willett, Intake of Dairy Products, Calcium, and Vitamin D and Risk of Breast Cancer, J. Natl. Cancer Inst. 94 (2002) 1301–1310. doi:10.1093/jnci/94.17.1301. [92] J.L. Tommie, S.M. Pinney, L.A. Nommsen-Rivers, Serum Vitamin D Status and Breast Cancer Risk by Receptor Status: A Systematic Review, Nutr. Cancer. 70 (2018) 804–820. doi:10.1080/01635581.2018.1470653. [93] R.T. Chlebowski, K.C. Johnson, C. Kooperberg, M. Pettinger, J. Wactawski-Wende, T. Rohan, J. Rossouw, D. Lane, M.J. O’Sullivan, S. Yasmeen, R.A. Hiatt, J.M. Shikany, M. Vitolins, J. Khandekar, F.A. Hubbell, Calcium plus vitamin D supplementation and the risk of breast cancer, J. Natl. Cancer Inst. 100 (2008) 1581–1591. doi:10.1093/jnci/djn360. [94] B. Dawson-Hughes, R.P. Heaney, M.F. Holick, P. Lips, P.J. Meunier, R. Vieth, Estimates of optimal vitamin D status, Osteoporos. Int. 16 (2005) 713–716. doi:10.1007/s00198-

103

005-1867-7. [95] R. Vieth, A. Zittermann, S.J. Whiting, J.J. McGrath, C.F. Garland, R.P. Heaney, B.W. Hollis, C. Lamberg-Allardt, M.F. Holick, R. Scragg, B.J. Boucher, A.W. Norman, W.C. Willett, B. Dawson-Hughes, H. Bischoff-Ferrari, The urgent need to recommend an intake of vitamin D that is effective, Am. J. Clin. Nutr. 85 (2007) 649–650. doi:10.1093/ajcn/85.3.649. [96] F.C. Garland, C.F. Garland, E.D. Gorham, J.F. Young, Geographic variation in breast cancer mortality in the United States: A hypothesis involving exposure to solar radiation, Prev. Med. (Baltim). 19 (1990) 614–622. doi:10.1016/0091-7435(90)90058-R. [97] R.D. Zamoiski, D.M. Freedman, M.S. Linet, C.M. Kitahara, W. Liu, E.K. Cahoon, Prospective study of ultraviolet radiation exposure and risk of breast cancer in the United States, Environ. Res. 151 (2016) 419–427. doi:10.1016/j.envres.2016.08.013. [98] M. Moukayed, W.B. Grant, The roles of UVB and vitamin D in reducing risk of cancer incidence and mortality: A review of the epidemiology, clinical trials, and mechanisms, Rev. Endocr. Metab. Disord. 18 (2017) 167–182. doi:10.1007/s11154-017-9415-2. [99] C. Chouvet, E. Vicard, M. Devenoec, S. Saez, 1,25-Dihydroxyvitamin D3 inhibitory effect on the growth of two human breast cancer cell lines (MCF-7, BT-20), J. Steroid Biochem. Mol. Biol. 24 (1986) 373–376. [100] S.P. Xie, G. Pirianov, K.W. Colston, Vitamin D analogues suppress IGF-I signalling and promote apoptosis in breast cancer cells, Eur. J. Cancer. 35 (1999) 1717–1723. doi:10.1016/S0959-8049(99)00200-2. [101] M. Simboli-Campbell, C.J. Narvaez, K. van Weelden, M. Tenniswood, J. Welsh, 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 (1997) 31–41. http://www.ncbi.nlm.nih.gov/pubmed/9116316. [102] Q.M. Wang, J.B. Jones, G.P. Studzinski, Cyclin-dependent kinase inhibitor p27 as a mediator of the G1-S phase block induced by 1,25-dihydroxyvitamin D3 in HL60 cells, Cancer Res. 56 (1996) 264–267. [103] C.M. Kemmis, S.M. Salvador, K.M. Smith, J. Welsh, 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 (2006) 887–892. doi:136/4/887 [pii] ET - 2006/03/22. [104] C. Kemmis, J. Welsh, Mammary epithelial cell transformation is associated with deregulation of the vitamin D pathway, J. Cell. Biochem. 105 (2008) 980–988. doi:10.1002/jcb.21896. [105] G.M. Zinser, K. McEleney, J. Welsh, Characterization of mammary tumor cell lines from wild type and vitamin D3 receptor knockout mice, Mol. Cell. Endocrinol. 200 (2003) 67– 80. doi:S0303720702004161 [pii].

104

[106] M.E. Valrance, A.H. Brunet, A. Acosta, J.E. Welsh, Dissociation of growth arrest and CYP24 induction by VDR ligands in mammary tumor cells, J. Cell. Biochem. 101 (2007) 1505–1519. doi:10.1002/jcb.21263. [107] M.E. Valrance, J. Welsh, 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 (2004) 221–225. doi:10.1016/j.jsbmb.2004.03.082. [108] G.M. Zinser, M. Suckow, J.E. Welsh, Vitamin D receptor (VDR) ablation alters carcinogen-induced tumorigenesis in mammary gland, epidermis and lymphoid tissues, J. Steroid Biochem. Mol. Biol. 97 (2005) 153–164. doi:10.1016/j.jsbmb.2005.06.024. [109] G.M. Zinser, J. Welsch, Vitamin D receptor status alters mammary gland morphology and tumorigenesis in MMTV-neu mice, Carcinogenesis. 25 (2004) 2361–2372. doi:10.1093/carcin/bgh271. [110] C.L. Richard, M.C. Farach-Carson, B. Rohe, I. Nemere, K.A. Meckling, Involvement of 1,25D3-MARRS (membrane associated, rapid response steroid-binding), a novel vitamin D receptor, in growth inhibition of breast cancer cells, Exp. Cell Res. 316 (2010) 695– 703. doi:10.1016/j.yexcr.2009.12.015. [111] E. Gaucci, F. Altieri, C. Turano, S. Chichiarelli, The protein ERp57 contributes to EGF receptor signaling and internalization in MDA-MB-468 breast cancer cells, J. Cell. Biochem. 114 (2013) 2461–2470. doi:10.1002/jcb.24590. [112] J. Russo, Y.-F. Hu, X. Yang, I.H. Russo, Chapter 1: Developmental, Cellular, and Molecular Basis of Human Breast Cancer, J. Natl. Cancer Inst. Monogr. 2000 (2000) 17– 37. doi:10.1093/oxfordjournals.jncimonographs.a024241. [113] I.H. Russo, J. Russo, Developmental Stage of the Rat Mammary Gland as Determinant of Its Susceptibility to 7, 12-Dimethylbenzanthracene, JNCI J. Natl. Cancer Inst. 61 (1978) 1439–1449. doi:10.1093/jnci/61.6.1439. [114] J. Russo, L.K. Tay, D.R. Ciocca, I.H. Russo, Molecular and cellular basis of the mammary gland susceptibility to carcinogenesis., Environ. Health Perspect. 49 (1983) 185–199. doi:10.1289/ehp.8349185. [115] S.E. Olivo, L. Hilakivi-Clarke, Opposing effects of prepubertal low- and high-fat n-3 polyunsaturated fatty acid diets on rat mammary tumorigenesis, Carcinogenesis. 26 (2005) 1563–1572. doi:10.1093/carcin/bgi118. [116] F. de Oliveira Andrade, C.C. Fontelles, M.P. Rosim, T.F. de Oliveira, A.P. de Melo Loureiro, J. Mancini-Filho, M.M. Rogero, F.S. Moreno, S. de Assis, L.F. Barbisan, L. Hilakivi-Clarke, T.P. Ong, Exposure to lard-based high-fat diet during fetal and lactation periods modifies breast cancer susceptibility in adulthood in rats, J. Nutr. Biochem. 25 (2014) 613–622. doi:10.1016/j.jnutbio.2014.02.002. [117] L. Hilakivi-Clarke, R. Clarke, I. Onojafe, M. Raygada, E. Cho, M. Lippman, A maternal diet high in n - 6 polyunsaturated fats alters mammary gland development, puberty onset,

105

and breast cancer risk among female rat offspring, Proc. Natl. Acad. Sci. 94 (1997) 9372– 9377. doi:10.1073/pnas.94.17.9372. [118] T. Muto, S. Wakui, N. Imano, K. Nakaaki, H. Takahashi, H. Hano, M. Furusato, T. Masaoka, Mammary gland differentiation in female rats after prenatal exposure to 3,3′,4,4′,5-pentachlorobiphenyl, Toxicology. 177 (2002) 197–205. doi:10.1016/S0300- 483X(02)00224-X. [119] L.A. Korde, A.H. Wu, T. Fears, A.M.Y. Nomura, D.W. West, L.N. Kolonel, M.C. Pike, R.N. Hoover, R.G. Ziegler, Childhood soy intake and breast cancer risk in Asian American women, Cancer Epidemiol. Biomarkers Prev. 18 (2009) 1050–1059. doi:10.1158/1055-9965.EPI-08-0405. [120] M. Messina, L. Hilakivi-Clarke, Early Intake Appears to Be the Key to the Proposed Protective Effects of Soy Intake Against Breast Cancer, Nutr. Cancer. 61 (2009) 792–798. doi:10.1080/01635580903285015. [121] W.B. Murrill, N.M. Brown, J.-X. Zhang, P.A. Manzolillo, S. Barnes, C.A. Lamartiniere, Prepubertal genistein exposure suppresses mammary cancer and enhances gland differentiation in rats, Carcinogenesis. 17 (1996) 1451–1457. doi:10.1093/carcin/17.7.1451. [122] X. Su, G.A. Colditz, L.C. Collins, H.J. Baer, L.A. Sampson, W.C. Willett, C.S. Berkey, S.J. Schnitt, J.L. Connolly, B.A. Rosner, R.M. Tamimi, Adolescent Intakes of Vitamin D and Calcium and Incidence of Proliferative Benign Breast Disease, Breast Cancer Res. Treat. 134 (2012) 783–791. doi:10.1016/j.freeradbiomed.2008.10.025.The. [123] C.E. Boeke, R.M. Tamimi, C.S. Berkey, G.A. Colditz, E. Giovannucci, S. Malspeis, W.C. Willett, A.L. Frazier, Adolescent Dietary Vitamin D and Sun Exposure in Relation to Benign Breast Disease, Cancer Causes Control. 26 (2015) 1181–1187. doi:10.1183/09031936.00063810.The. [124] L.N. Anderson, M. Cotterchio, V.A. Kirsh, J.A. Knight, Ultraviolet sunlight exposure during adolescence and adulthood and breast cancer risk: A population-based case-control study among Ontario women, Am. J. Epidemiol. 174 (2011) 293–304. doi:10.1093/aje/kwr091. [125] J.A. Knight, M. Lesosky, H. Barnett, J.M. Raboud, R. Vieth, Vitamin D and reduced risk of breast cancer: A population-based case-control study, Cancer Epidemiol. Biomarkers Prev. 16 (2007) 422–429. doi:10.1158/1055-9965.EPI-06-0865. [126] M.C. Pike, M.D. Krailo, B.E. Henderson, J.T. Casagrande, D.G. Hoel, ‘Hormonal’ risk factors, ‘breast tissue age’ and the age-incidence of breast cancer, Nature. 303 (1983) 767–770. doi:10.1038/303767a0. [127] A. Colditz, Models of Breast Cancer Show That Risk Is Set by Events of Early Life: Prevention Efforts Must Shift Focus, Cancer Epidemiol. Biomarkers Prev. 4 (1995) 567– 571.

106

[128] G.A. Colditz, K. Bohlke, C.S. Berkey, Breast cancer risk accumulation starts early: Prevention must also, Breast Cancer Res. Treat. 145 (2014) 567–579. doi:10.1007/s10549-014-2993-8. [129] R.M. Tamimi, D. Spiegelman, S.A. Smith-Warner, M. Wang, M. Pazaris, W.C. Willett, A.H. Eliassen, D.J. Hunter, Population attributable risk of modifiable and nonmodifiable breast cancer risk factors in postmenopausal breast cancer, Am. J. Epidemiol. 184 (2016) 884–893. doi:10.1093/aje/kww145. [130] C.I. Li, K.E. Malone, J.R. Daling, J.D. Potter, L. Bernstein, P.A. Marchbanks, B.L. Strom, M.S. Simon, M.F. Press, G. Ursin, R.T. Burkman, S.G. Folger, S. Norman, J.A. McDonald, R. Spirtas, Timing of menarche and first full-term birth in relation to breast cancer risk, Am. J. Epidemiol. 167 (2008) 230–239. doi:10.1093/aje/kwm271. [131] Y. Liu, G.A. Colditz, B. Rosner, C.S. Berkey, L.C. Collins, S.J. Schnitt, J.L. Connolly, W.Y. Chen, W.C. Willett, R.M. Tamimi, Alcohol intake between menarche and first pregnancy: A prospective study of breast cancer risk, J. Natl. Cancer Inst. 105 (2013) 1571–1578. doi:10.1093/jnci/djt213. [132] I. Nemere, N. Garbi, G. Hammerling, K.J. Hintze, Role of the 1,25D3-MARRS receptor in the 1,25(OH)2D3-stimulated uptake of calcium and phosphate in intestinal cells., Steroids. 77 (2012) 897–902. doi:10.1016/j.steroids.2012.04.002. [133] A. Hettinghouse, R. Liu, C. ju Liu, Multifunctional molecule ERp57: From cancer to neurodegenerative diseases, Pharmacol. Ther. (2017). doi:10.1016/j.pharmthera.2017.07.011. [134] N. Garbi, S. Tanaka, F. Momburg, G.J. Hämmerling, Impaired assembly of the major histocompatibility complex class I peptide-loading complex in mice deficient in the oxidoreductase ERp57, Nat. Immunol. 7 (2006) 93–102. doi:10.1038/ni1288. [135] J.D. Oliver, H.L. Roderick, D.H. Llewellyn, S. High, ERp57 Functions as a Subunit of Specific Complexes Formed with the ER Lectins Calreticulin and Calnexin, Mol. Biol. Cell. 10 (1999) 2573–2582. doi:10.1091/mbc.10.8.2573. [136] P.A. Wearsch, P. Cresswell, Selective loading of high-affinity peptides onto major histocompatibility complex class I molecules by the tapasin-ERp57 heterodimer, Nat. Immunol. 8 (2007) 873–881. doi:10.1038/ni1485. [137] K.M. Dong, Gang; Wearsch, Pamela A; Peaper, David R; Cresswell, Peter; Reinisch, Insights into MHC class I peptide loading from the structure of the tapasin/ERp57 heterodimer, Immunity. 30 (2009) 21–32. doi:10.1002/nbm.3066.Non-invasive. [138] L. Zhu, N.C. Santos, K.H. Kim, Disulfide isomerase glucose-regulated protein 58 is required for the nuclear localization and degradation of retinoic acid receptor ??, Reproduction. 139 (2010) 717–731. doi:10.1530/REP-09-0527. [139] M.H. Choe, J.W. Min, H.B. Jeon, D.-H. Cho, J.S. Oh, H.G. Lee, S.-G. Hwang, S. An, Y.- H. Han, J.-S. Kim, ERp57 modulates STAT3 activity in radioresistant laryngeal cancer

107

cells and serves as a prognostic marker for laryngeal cancer., Oncotarget. 6 (2015) 2654– 66. doi:10.18632/oncotarget.3042. [140] C.M. Leys, S. Nomura, B.J. LaFleur, S. Ferrone, M. Kaminishi, E. Montgomery, J.R. Goldenring, Expression and prognostic significance of prothymosin-α and ERp57 in human gastric cancer, Surgery. 141 (2007) 41–50. doi:10.1016/j.surg.2006.05.009. [141] B.D. Boyan, J. Chen, Z. Schwartz, Mechanism of Pdia3-dependent 1,25-dihydroxy vitamin D3 signaling in musculoskeletal cells, Steroids. 77 (2012) 892–896. doi:10.1016/j.steroids.2012.04.018. [142] J.L. Inman, C. Robertson, J.D. Mott, M.J. Bissell, Mammary gland development: cell fate specification, stem cells and the microenvironment, Development. 142 (2015) 1028–1042. doi:10.1242/dev.087643. [143] L. Hilakivi-Clarke, I. Onojafe, M. Raygada, E. Cho, T. Skaar, I. Russo, R. Clarke, Prepubertal exposure to zearalenone or genistein reduces mammary tumorigenesis, Br. J. Cancer. 80 (1999) 1682–1688. doi:10.1038/sj.bjc.6690584. [144] A.C. McLean, N. Valenzuela, S. Fai, S.A.L. Bennett, Performing Vaginal Lavage, Crystal Violet Staining, and Vaginal Cytological Evaluation for Mouse Estrous Cycle Staging Identification, J. Vis. Exp. (2012) 4–9. doi:10.3791/4389. [145] B.M. Anderson, M.B. MacLennan, L.M. Hillyer, D.W.L. Ma, Lifelong exposure to n-3 PUFA affects pubertal mammary gland development, Appl. Physiol. Nutr. Metab. 39 (2014) 699–706. doi:10.1139/apnm-2013-0365. [146] J.C. Fleet, Rapid, Membrane-Initiated Actions of 1,25 Dihydroxyvitamin D: What Are They and What Do They Mean?, J. Nutr. 25 (2004) 3215–3218. http://www.ncbi.nlm.nih.gov/pubmed/15570014. [147] J.L. Costa, P.P. Eijk, M. a van de Wiel, D. ten Berge, F. Schmitt, C.J. Narvaez, J. Welsh, B. Ylstra, Anti-proliferative action of vitamin D in MCF7 is still active after siRNA-VDR knock-down., BMC Genomics. 10 (2009) 499. doi:10.1186/1471-2164-10-499. [148] C. Grillo, C. D'Ambrosio, A. Scaloni, M. Maceroni, S. Merluzzi, C. Turano, F. Altieri, Cooperative activity of Ref-1/APE and ERp57 in reductive activation of transcription factors, Free Radic. Biol. Med. 41 (2006) 1113–1123. doi:10.1016/j.freeradbiomed.2006.06.016. [149] S. Xanthoudakis, G. Miao, F. Wang, Y.C. Pan, T. Curran, Redox activation of Fos-Jun DNA binding activity is mediated by a DNA repair enzyme., EMBO J. 11 (1992) 3323– 35. [150] S.S. Shishkin, L.S. Eremina, L.I. Kovalev, M.A. Kovaleva, AGR2, ERp57/GRP58, and Some Other Human Protein Disulfide Isomerases., Biochem. Biokhimi͡ia. 78 (2013) 1415– 30. doi:10.1134/S000629791313004X. [151] R.S. Chapman, P. Lourenco, E. Tonner, D. Flint, S. Selbert, K. Takeda, S. Akira, A.R. Clarke, C.J. Watson, The Role of Stat3 in Apoptosis and Mammary Gland Involution,

108

Springer, Boston, MA, 2002. doi:10.1007/0-306-46832-8_16. [152] Leena Hilakivi-Clarke, Nutritional Modulation of Terminal End Buds: Its Relevance to Breast Cancer Prevention, Curr. Cancer Drug Targets. 7 (2007) 465–474. doi:10.2174/156800907781386641. [153] Reeves, F.H. Nielsen, G.C. Fahey, AIN-93 Purified Diets for Laboratory Rodents: Final Report of the American Institute of Nutrition Ad Hoc Writing Committee on the Reformulation of the AIN-76A Rodent Diet., J. Nutr. 123 (1993) 1939–1951. doi:10.1017/CBO9781107415324.004. [154] V.N. Sumantran, P. Mishra, R. Bera, N. Sudhakar, Microarray analysis of differentially- expressed genes encoding CYP450 and Phase II drug metabolizing enzymes in psoriasis and melanoma, Pharmaceutics. 8 (2016) 1–15. doi:10.3390/pharmaceutics8010004. [155] P.H. Anderson, S. Iida, J.H.T. Tyson, A.G. Turner, H.A. Morris, Bone CYP27B1 gene expression is increased with high dietary calcium and in mineralising osteoblasts, J. Steroid Biochem. Mol. Biol. 121 (2010) 71–75. doi:10.1016/j.jsbmb.2010.03.021. [156] U. Segersten, P.K. Holm, P. Björklund, O. Hessman, H. Nordgren, L. Binderup, G. Åkerström, P. Hellman, G. Westin, 25-Hydroxyvitamin D3-1α-hydroxylase expression in breast cancer and use of non-1α-hydroxylated vitamin D analogue, Breast Cancer Res. 7 (2005) R980. doi:10.1186/bcr1332. [157] K.K. Deeb, D.L. Trump, C.S. Johnson, Vitamin D signalling pathways in cancer: potential for anticancer therapeutics, Nat. Rev. Cancer. 7 (2007) 684–700. doi:10.1038/nrc2196. [158] N. Lopes, J. Paredes, J.L. Costa, B. Ylstra, F. Schmitt, Vitamin D and the mammary gland: a review on its role in normal development and breast cancer, Breast Cancer Res. 14 (2012) 211. doi:10.1186/bcr3178. [159] S. a Lamprecht, M. Lipkin, Chemoprevention of colon cancer by calcium, vitamin D and folate: molecular mechanisms., Nat. Rev. Cancer. 3 (2003) 601–14. doi:10.1038/nrc1144. [160] A.W. Norman, X. Song, L. Zanello, C. Bula, W.H. Okamura, Rapid and genomic biological responses are mediated by different shapes of the agonist steroid hormone , 1, 25 ( OH ) 2 vitamin D 3, Steroids. 64 (1999) 120–128. [161] J. Chen, K.S. Lobachev, B.J. Grindel, M.C. Farach-Carson, S.L. Hyzy, K.B. El-Baradie, R. Olivares-Navarrete, M. Doroudi, B.D. Boyan, Z. Schwartz, Chaperone properties of pdia3 participate in rapid membrane actions of 1α,25-dihydroxyvitamin d3., Mol. Endocrinol. 27 (2013) 1065–77. doi:10.1210/me.2012-1277. [162] A.M. Wilkin, A. Harnett, M. Underschultz, C. Cragg, K.A. Meckling, Role of the ERp57 protein (1,25D3-MARRS receptor) in murine mammary gland growth and development, Steroids. (2018) 1–6. doi:10.1016/j.steroids.2018.02.006. [163] B.M. Anderson, M.B. MacLennan, L.M. Hillyer, D.W.L. Ma, Lifelong exposure to n-3 PUFA affects pubertal mammary gland development 1, Appl. Physiol. Nutr. Metab. 39 (2014) 699–706. doi:10.1139/apnm-2013-0365.

109

[164] IOM (Institute of Medicine), Dietary Reference Intakes for Calcium and Vitamin D, 2011. doi:10.17226/13050. [165] M.F. Holick, N.C. Binkley, H.A. Bischoff-Ferrari, C.M. Gordon, D.A. Hanley, R.P. Heaney, M.H. Murad, C.M. Weaver, Evaluation, treatment, and prevention of vitamin D deficiency: An endocrine society clinical practice guideline, J. Clin. Endocrinol. Metab. 96 (2011) 1911–1930. doi:10.1210/jc.2011-0385. [166] J.C. Fleet, C. Gliniak, Z. Zhang, Y. Xue, K.B. Smith, R. McCreedy, S. a Adedokun, Serum metabolite profiles and target tissue gene expression define the effect of cholecalciferol intake on calcium metabolism in rats and mice., J. Nutr. 138 (2008) 1114– 1120. doi:10.1038/jid.2014.371. [167] K.L. Seldeen, M. Pang, M. Rodríguez-Gonzalez, M. Hernandez, Z. Sheridan, P. Yu, B.R. Troen, A mouse model of vitamin D insufficiency: is there a relationship between 25(OH) vitamin D levels and obesity?, Nutr. Metab. (Lond). 14 (2017) 26. doi:10.1186/s12986- 017-0174-6. [168] L. Williamson, A. Hayes, E.D. Hanson, P. Pivonka, N.A. Sims, J.H. Gooi, High dose dietary vitamin D3increases bone mass and strength in mice, Bone Reports. 6 (2017) 44– 50. doi:10.1016/j.bonr.2017.02.001. [169] National Research Council, Nutrient Requirements of Laboratory Animals; Fourth Revised Edition, The National Academies Press, Washington, DC, 1995. [170] G.M. Zinser, J. Welsh, Accelerated Mammary Gland Development during Pregnancy and Delayed Postlactational Involution in Vitamin D 3 Receptor Null Mice, Mol. Endocrinol. 18 (2004) 2208–2223. doi:10.1210/me.2003-0469. [171] X. Peng, M. Hawthorne, A. Vaishnav, R. St-Arnaud, R.G. Mehta, 25-Hydroxyvitamin D3 is a natural chemopreventive agent against carcinogen induced precancerous lesions in mouse mammary gland organ culture, Breast Cancer Res. Treat. 113 (2009) 31–41. doi:10.1007/s10549-008-9900-0. [172] A.L. Johnson, G.M. Zinser, S.E. Waltz, Loss of vitamin D receptor signaling from the mammary epithelium or adipose tissue alters pubertal glandular development., Am. J. Physiol. - Endocrinol. Metab. 307 (2014) E674–E685. doi:doi:10.1152/ajpendo.00200.2014. [173] L. Hinck, G.B. Silberstein, Key stages in mammary gland development: The mammary end bud as a motile organ, Breast Cancer Res. 7 (2005) 245–251. doi:10.1186/bcr1331. [174] B. Rohe, S.E. Safford, I. Nemere, M.C. Farach-Carson, Regulation of expression of 1,25D3-MARRS/ERp57/PDIA3 in rat IEC-6 cells by TGFβ and 1,25(OH)2D3, Steroids. 72 (2007) 144–150. doi:10.1016/j.steroids.2006.11.013. [175] L.L. Munasinghe, Y. Yuan, N.D. Willows, E.L. Faught, J.P. Ekwaru, P.J. Veugelers, Vitamin D deficiency and sufficiency among Canadian children residing at high latitude following the revision of the RDA of vitamin D intake in 2010., Br. J. Nutr. 117 (2017)

110

457–465. doi:10.1017/S0007114517000320. [176] J.W. Ryan, P.H. Anderson, H.A. Morris, Pleiotropic activities of vitamin D receptors - adequate activation for multiple health outcomes, Clin. Biochem. Rev. 36 (2015) 53–61. [177] S.S. Jensen, M.W. Madsen, J. Lukas, J. Bartek, L. Binderup, Inhibitory Effects of 1α,25- Dihydroxyvitamin D 3 on the G 1 –S Phase-Controlling Machinery, Mol. Endocrinol. 15 (2014) 1370–1380. doi:10.1210/mend.15.8.0673. [178] C.J. Narvaez, J.E. Welsh, Role of Mitochondria and Caspases in Vitamin D-mediated Apoptosis of MCF-7 Breast Cancer Cells, J. Biol. Chem. 276 (2001) 9101–9107. doi:10.1074/jbc.M006876200. [179] B. Prietl, G. Treiber, T.R. Pieber, K. Amrein, Vitamin D and Immune Function, Nutrients. 5 (2013) 2502–2521. doi:10.3390/nu5072502. [180] R.K. Wali, J. Kong, M.D. Sitrin, M. Bissonnette, Y.C. Li, Vitamin D receptor is not required for the rapid actions of 1,25-dihydroxyvitamin D3 to increase intracellular calcium and activate protein kinase C in mouse osteoblasts, J. Cell. Biochem. 88 (2003) 794–801. doi:10.1002/jcb.10432. [181] J. Welsh, J.A. Wietzke, G.M. Zinser, S. Smyczek, S. Romu, E. Tribble, J.C. Welsh, B. Byrne, C.J. Narvaez, Impact of the Vitamin D3 receptor on growth-regulatory pathways in mammary gland and breast cancer, J. Steroid Biochem. Mol. Biol. 83 (2002) 85–92. doi:10.1016/S0960-0760(02)00277-7. [182] S. Ching, S. Kashinkunti, M.D. Niehaus, G.M. Zinser, Mammary Adipocytes Bioactivate 25-hydroxyvitamin D3 and Signal via Vitamin D3 Receptor, Modulating Mammary Epithelial Cell Growth, J. Cell. Biochem. 112 (2011) 3393–3405. doi:10.1002/jcb.23273. [183] U. Hügel, L. Weber, J. Reichrath, O. Mehls, G. Klaus, Rat growth plate chondrocytes express low levels of 25-hydroxy-1a-hydroxylase, J. Steroid Biochem. Mol. Biol. 90 (2004) 143–147. doi:10.1016/j.jsbmb.2004.03.114. [184] M.E. Carley, D.J. Rickard, J.B. Gebhart, M.J. Webb, K.C. Podratz, T.C. Spelsberg, Distribution of estrogen receptors a and b mRNA in mouse urogenital tissues and their expression after oophorectomy and estrogen replacement, Int. Urogynceology J. 14 (2003) 141–145. doi:10.1007/s00192-002-1020-5. [185] A. Yiakouvaki, M. Dimitriou, I. Karakasiliotis, C. Eftychi, S. Theocharis, D.L. Kontoyiannis, Myeloid cell expression of the RNA-binding protein HuR protects mice from pathologic inflammation and colorectal carcinogenesis, J. Clin. Invest. 122 (2012) 48–61. doi:10.1172/JCI45021.48. [186] V.B. Sequeira, M.S. Rybchyn, W. Tongkao-on, C. Gordon-Thomson, P.J. Malloy, I. Nemere, A.W. Norman, V.E. Reeve, G.M. Halliday, D. Feldman, R.S. Mason, The Role of the Vitamin D Receptor and ERp57 in Photoprotection by 1α,25-Dihydroxyvitamin D 3, Mol. Endocrinol. 26 (2012) 574–582. doi:10.1210/me.2011-1161. [187] N.A. Morrice, S.J. Powis, A role for the thiol-dependent reductase ERp57 in the assembly

111

of MHC class I molecules, Curr. Biol. 8 (1998) 713–716. [188] D.. Matthew, J. D’Angelo, J. Drelich, J. Welsh, Adipose-specific VDR deletion alters body fat and enhances mammary epiehtlial density, J. Steroid Biochem. Mol. Biol. 164 (2016) 299–308. doi:10.1016/j.jsbmb.2015.09.035. [189] C. Yang, Y.S. Tan, J.R. Harkema, S.Z. Haslam, Differential effects of peripubertal exposure to perfluorooctanoic acid on mammary gland development in C57Bl/6 and Balb/c mouse strains, Reprod. Toxicol. 27 (2009) 299–306. doi:10.1016/j.reprotox.2008.10.003. [190] M.B. Maclennan, B.M. Anderson, D.W.L. Ma, Differential mammary gland development in FVB and C57Bl/6 mice: Implications for breast cancer research, Nutrients. 3 (2011) 929–936. doi:10.3390/nu3110929. [191] L.K. Olson, Y. Tan, Y. Zhao, M.D. Aupperlee, S.Z. Haslam, Pubertal exposure to high fat diet causes mouse strain-dependent alterations in mammary gland development and estrogen responsiveness, Int. J. Obes. 34 (2010) 1415–1426. doi:10.1038/ijo.2010.51. [192] T.M. Brophy, B.S. Coller, J. Ahamed, Identification of the Thiol Isomerase-binding Peptide , Mastoparan , as a Novel Inhibitor of Shear-induced Transforming Growth Factor-beta 1 Activation, J. Biol. Chem. 288 (2013) 10628–10639. doi:10.1074/jbc.M112.439034. [193] Y. Feng, D. Manka, K. Wagner, S.A. Khan, Estrogen receptor-alpha expression in the mammary epithelium is required for ductal and alveolar morphogenesis in mice, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 14718–14723. [194] D.L. Kleinberg, M. Feldman, W. Ruan, IGF-I: An essential factor in terminal end bud formation and ductal morphogenesis, J. Mammary Gland Biol. Neoplasia. 5 (2000) 7–17. doi:10.1023/A:1009507030633. [195] C. Lee, J. Wang, K. Chou, M. Hsu, 1,25-Dihydroxyvitamin D3 Increases Testosterone- induced 17beta-estradiol Secretion and Reverses Testosterone-Reduced Connexin 43 in Rat Granulosa Cells, Reprod. Biol. Endocrinol. 12 (2014) 1–12. [196] R.G. Mehta, X. Peng, F. Alimirah, G. Murillo, R. Mehta, Vitamin D and breast cancer: Emerging concepts, Cancer Lett. 334 (2013) 95–100. doi:10.1016/j.canlet.2012.10.034.

112

113