THE CELLULAR AND FUNCTIONAL ROLE OF MELANOTRANSFERRIN IN MELANOMAS

by

Mel Mel Tian

B.Sc., University of British Columbia, 2001

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

in

THE FACULTY OF GRADUATE STUDIES

(Microbiology and Immunology)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

September 2009

© Mei Mci Tian, 2009 Abstract

Melanoma, one of the deadliest forms of skin cancers, is a malignant tumor derived from abnormal proliferation of epidermal melanocytes, and its formation involves a series of events leading to altered cellular properties and rapid proliferation. High cellular iron content has been connected to the development of cancers in human. One way iron may affect cancer development and progression is by altering cell growth and proliferation. Iron has been proposed to promote progression from 1G to S phase of the cell cycle through activation of ribonucleotide reductase during DNA synthesis. Iron is normally absorbed by the enterocyte of the duodenum where it is transported throughout the body via serum transferrin. However, the existence of transferrin-transferrin receptor independent iron transport pathways has been shown. Such pathways are thought to play critical roles in non-transferrin bound iron overload in diseases such as hemochromatosis and Alzheimer’s disease, though the molecular details of these pathways remain obscure. Many genes are up-regulated to aid the abnormal proliferation of cancer cells and subsequent invasion of tissues. In the case of melanoma, the expression of melanotransferrin (p97), an iron binding molecule, is elevated. In this thesis, the cellular and functional role of glycosyiphosphatidylinositol (GPI)-anchored p97 in melanoma was investigated. The first aim of this thesis was to perform detailed investigation of the internalization pathway of GPI-anchored p97 in melanoma cells. Although many GPI anchored proteins have been studied previously, there is still an on-going debate whether caveolae-dependent or clathrin-dependent pathways are involved. Using confocal immunofluorescence microscopy and quantitative immunoelectron microscopy, iron bound GPI-anchored p97 was shown to be internalized via a caveolae vesicles dependent endocytotic pathway. In addition, endosomal disruption studies demonstrated that the intracellular trafficking of the GPI-anchored p97 in melanoma cells is endosomal dependent. The studies performed in this thesis demonstrate that GPI-anchored p97 protein can mediate the iron uptake in melanoma cells. As GPI-anchored p97 is highly expressed in melanoma cells, this thesis also examined the functional role of GPI anchored p97 in melanoma cells and tumors. The expression of GPI-anchored p97 in III melanoma cells through over-expression and down-regulation techniques. The results demonstrated that the GPI-anchored p97 promote melanoma cell proliferation, melanin production and secretion in vitro and melanoma tumor growth in vivo. Taken together, this thesis sheds new light on the relationship between GPI-anchored p97, melanoma development and cellular iron uptake. iv

Table of Contents

Abstract ii Table of Contents iv List of Tables vii List of Figures viii List of Abbreviations x Acknowledgments xiii Dedication xv

Chapter 1 : General introduction 1

1.1 Melanoma 1 1.2 Iron and melanoma 4 1.2.1 Iron metabolism 4 1.2.1.1 Brief history of iron 4 1.2.1.2 Distribution and function of iron 4 1.2.2 Role of iron in cancer 9 1.3 Melanotransferrin and iron 10 1.3.1 Melanotransferrin and transferrin family of proteins 10 1.3.1.1 Transferrin 10 1.3.1.2 Other transferrin family member 16 1.3.1.3 Melanotransferrin 16 1.3.2 Melanotransferrin and iron uptake 18 1.3.2.1 Iron binding and melanotransferrin 18 . 1.3.2.2 Transferrin-transferrin receptor independent iron transport and melanotransferrin . ,.21 1.4 Possible function of melanotransferrin 21 1.5 Project rationale and general approach 22 Chapter 2 : Materials and methods 24 2.1 Tissue culture, cells and antibodies 24 2.2 Constructs 27 2.2.1 p97 suppression construct: pSUPERIOR.neo.GFP-p97 27 2.2.2 Histidine tagged p97 construct: pMXpie-His-p97 30 2.2.3 Non-tagged p97 construct: pCMV-Tag3-p97 30 2.3 Stable transfection 34 2.4 Isothermal titration calorimetry 34 2.5 RNA isolation 35 2.6 RT-PCR 2.7 Real-time PCR 2.8 Western blots 2.9 Immunoprecipitation V

2.10 Detection of surface GPI-anchored murine p97 expression on B16F1O melanoma cells 38 2.11 Fab fragment production for immunofluorescence and confocal microscopy 38 2.12 Immunofluorescence and confocal microscopy 39 2.12.1 Quantification of confocal results 39 2.13 Cryoimmunoelectron microscopy 40 2.14 Proliferation assay 40 2.15 Melanin assay. ill 2.16 In vivotumor growth assays 41 2.16.1 Animals 41 2.16.2 Tumor studies 41 2.16.2.1 B16F1Omouse melanoma cells over-expressing p97 41 2.16.2.2 JBMS mouse melanoma cells with knock-down of p97 expression 42 2.17 Statistical analysis. ‘I,

Chapter3 : Iron uptake in human melanoma cells mediated by GPI-anchoredp97 — 43 3.1 Background 43 3.1.1 Introduction 43 3.1.2 Rationale 45 3.2 Results 46 3.2.1 Binding affinity of soluble p97 to iron 46 3.2.2 GPI-anchored p97 is endocytosed through caveolae dependent pathway 49 3.2.2.1 Immunofluorescence microscopy studies of the localization of GPI-anchored p97 in SK-MEL 28 human melanoma cell 49 3.2.2.2 Cryo-immunoelectron microscopy on the localization of GPI-anchored p97 in SK MEL 28 human melanoma cells 56 3.2.3 GPI-anchored p97 internalize iron utilizing an endosomal pathway in SK-MEL 28 melanoma cell 59 3.2.3.1 GPI-anchored p97 internalize to endosomes, but trafficks more slowly than TfR 59 3.2.3.2 Endosomes are essential in the internalization of GPI-anchored p97 63 3.2.3.3 Endosomes are essential in the trafficks of iron to ferritin mediated by GPI-anchored p97 in SK-MEL 28 melanoma cells 63 3.3 Discussion 69 Chapter 4: The role of GPI-anchoredp97 expression in melanoma malignancy 72 4.1 luuaai 72 4.2 Results .73 4.2.1 Selection of mouse melanoma cells for over-expression and down-regulation of GPI anchored p97 cell models 73 4.2.2 GPI-anchored p97 in over-expression B16F1Omouse melanoma cell models 76 4.2.3 Characterization of GPI-anchored p97 over-expression in mouse melanoma cells in vitro and in vivo 80 4.2.3.1 Over-expression of GPI-anchored p97 in B16F10 mouse melanoma cells promote cell proliferation 80 4.2.3.2 Over-expression of GPI-anchored p97 promotes melanin production and secretion in BI6F1O mouse melanoma cells 82 vi

4.2.3.3 Over-expression of GPI-anchored p97 in B16F1O mouse melanoma cells promote tumor growth in vivo 85 4.2.4 Characterization of GPI-anchored p97 down-regulation in mouse melanoma cells in vitro and in vivo 87 4.2.4.1 Down-regulation of GPI-anchored p97 expression in JBMS mouse melanoma cell... 87 4.2.4.2 Down-regulation of GPI-anchored p97 decrease cell proliferation in JBMS mouse melanoma cells 89 4.2.4.3 Down-regulation of GPI-anchored p97 in JBMS mouse melanoma cells reduce tumor growth in vivo 91 4.3 Chapter 5: Conclusion and future perspectives 98 References 103 Appendix i: Crystallization of p97 114 Appendix ii: Cellular regulatioin of p97 expression 117 VII

List of Tables

Table 1.1 Examples of dietary iron food sources and factors which increase and reduce iron absorption 6

Table 1.2 Iron binding lobes of transferrin and melanotransferrin 20

Table 2.la Primary antibodies used in the experiments presented in this thesis 25

Table 2.lb Secondary antibodies used in the experiments presented in this thesis .... 26

Table 2.2 PCR primers used in this thesis 33

Table 3.1 Apparent binding constant of iron to p97 as determined using differential scanning calorimetry in the presence of bicarbonate at 25°C 47 VIII

List of Figures

Figure 1.1 Progression of melanocytes to malignant melanoma 3

Figure 1.2 Absorption of dietary iron by enterocytes in the duodenum 7

Figure 1.3 Sequence alignments of selective members of the transferrin family of protein 12

Figure 1.4 Iron uptake mediated by transferrin-transferrin receptor complex 14

Figure 1.5 Evolution of selected members of transferrin family proteins 17

Figure 2.1 pSUPERIOR.neo.GFP and murine short interfering RNA 29

Figure 2.2 Nucleotide sequence of the modified mouse GPI-anchored p97 used in the construction of histidine tagged GPI-anchored p97 32

Figure 3.1 Calorimetry thermogram of iron binding to p97 at various ratios 48

Figure 3.2 Localization of TfR and GPI-anchored p97 to either caveolae or clathrin coated pits in SK-MEL 28 melanoma cells 51

Figure 3.3 GPI-anchored p97 co-localize with caveolin, not clathrin in human SK MEL 28 melanoma cells 54

Figure 3.4 GPI-anchored p97 localize in the same vesicles as caveolin- 1, while TfR are localized in the clathrin coated vesicles in human SK-MEL 28 melanoma cells 57

Figure 3.5 Localization of TfR and GPI-anchored p97 to early endosomes in human SK-MEL 28 melanoma cells 60

Figure 3.6 Endosomes are essential for GPI-anchored p97 internalization in human SK MEL 28 cells 65

Figure 3.7 Functional endosomes are essential for the GPI-anchored p97 mediated trafficking of iron to ferritin in SK-MEL 28 melanoma cells 68

Figure 3.8 The proposed model of iron uptake mediated by GPI-anchored p97 71

Figure 4.1 GPI-anchored p97 mRNA expression level in melanoma cell lines 75

Figure 4.2 GPI-anchored p97 is up-regulated at both mRNA and protein levels in GPI anchored p97 over-expressing transfectants 78 ix

Figure 4.3 Iron binding ability of GPI-anchored p97 over-expressing transfectants.... 79

Figure 4.4 The effect of GPI-anchored p97 over-expression on cellular proliferation of B16F1Omouse melanoma cells 81

Figure 4.5 The effect of GPI-anchored p97 over-expression on melanin production and secretion 83

Figure 4.6 GPI-anchored p97over-expression promotes the growth of subcutaneous B16F1O melanoma tumors 86

Figure 4.7 GPI-anchored p97 is down-regulated in pSUPERIOR.neo.GFP-p97 transfected JBMS cells 88

Figure 4.8 The effect of GPI-anchored p97 down-regulation on cellular proliferation

of JBMS mouse melanoma cells 90

Figure 4.9 GPI-anchored p97 down-regulation promotes the growth of subcutaneous JBMS mouse melanoma tumors 92

Figure 4.10 Schematic pathways of melanogenesis 97 x

List of Abbreviations

AD Alzheimer’ s disease ANOVA analysis of variance

AP- 1 activator protein-i

ATP adenosine triphosphate BSA bovine serum albumin

CDNA complementary deoxyribonucleic acid

DHICA 5,6-dihydroxyindole-2-carboxylic acid

DMEM Dulbecco’s modified eagle’s medium

DMT 1 divalent metal transporter 1

DNA deoxyribonucleic acid DTT dithiothreitol

ER endoplasmic reticulum

FACS fluorescent-activated cell sorting

FBS fetal bovine serum 2Fe ferrous irons 3Fe ferric irons Fe-NTA Fe-nitrilotriacetate

FWD forward (for PCR primer) GFP green fluorescent protein GPI glycosylphosphatidylinositol

HEPES 4-(2-hydroxyethyl) piperazine- 1-ethanesulfonic acid His histidine

IL interleukin

IIVIDM Iscove’s Modified Dulbecco Media xi

IRE iron regulatory element

kb kilobase kDa kilodalton LTf lactotransferrin

mAb monoclonal antibody

MAVID multiple sequence alignment program

mRNA messenger ribonucleic acid MTf melanotransferrin

NAPS nucleic acid protein service unit NP-40 nonidet-P40

NTA nitrilotriacetic acid

NTB 1 non-Tf bound iron

•OH hydroxide radical OTf ovotransferrin

P97 melanotransferrin

PAGE polyacrylamide gel electrophoresis

PBS phosphate-buffered saline

PCR polymerase chain reaction

PT-PLC phosphoinositol-phospholipase C

REV reverse (for PCR primer)

RGP radial growth phase

RNA ribonucleic acid

RNAi ribonucleic acid interference

ROS reactive oxygen species

RPMT Roswell Park Memorial Institute defined cell culture media rRNA ribosomal ribonucleic acid

RT-PCR Reverse Transcription - Polymerase Chain Reaction XII

S15 small ribosomal subunit rRNA used as a housekeeping gene control

S.C. Subcutaneous

S.D. Standard deviation

SDS sodium dodecyl sulfate

SEM standard error of the mean

Sig signal siRNA small interference ribonucleic acid

SK-MEL-28 the human melanoma cell line Tf transferrin

TfR 1 transferrin receptor 1

TfR transferrin receptor

TNFcL tumor necrosis factor alpha

UVR ultra violet radiation VGP vertical growth phase

WT wild-type XIII

Acknowledgments

In completing this thesis, I would like to recognize the encouragement and supports from my family, friends and colleagues. First of all, I would like to thank my supervisor, Dr. Wilfred A. Jefferies, for taking me in and giving me the opportunity to work on this exciting and challenging project. His guidance through the good and bad times has taught me to face difficulties head on, challenge myself and never give up. I am grateful to him in mentoring me for the past seven years and turning me into a Ph.D. I would also like to thank my supervisory committee members: Dr. Ninan Abraham, Dr. Pauline Johnson, Dr. Fumio Takei, for guidance, support and ideas through my Ph.D. journey. In addition, I would like to give many thanks to my present and past lab mates and colleagues, including Dr. Jacqueline Tiong, Kyla Omilusik, Kaan Biron, Dr. Jason Grant, Dr. Cheryl Pfeifer, Dr. Dara Dickstein, Dr. Maya Kotturi, Dr. Tim Vitalis, Bing Cai, Rayshad Gopaul, Dr. Genc Basha, Dr. Francesca Setiadi, for creating a wonderful work environment and helping me get through the ups and the downs of my Ph.D. I would like to thank Kyla especially for taking time out of her schedule to help me proofread this thesis. I would also like to give my thanks to Biomedical Research centre, Michael Smith Laboratories and Department of Microbiology and Immunology for providing me with the facility in conducting all the experiments in this thesis. The study of iron uptake through GPI-anchored p97 was originally started by Dr. Jacqueline Tiong (Microbiology and Immunology, University of British Columbia, Vancouver, BC) and me. Jacqueline Tiong studied the binding affinity of p97 to different metal ions and started the initial work on localization of GPI-anchored p97 in melanoma cells. In Chapter 3, the experiments to generate figure 3.1, 3.2 and 3.5 were performed by both Dr. Jacqueline Tiong and I. The binding affinity of soluble p97 to iron in Table 3.1 was done in collaboration with Dr. Louise Creagh (Michael Smith Laboratories, University of British Columbia, Vancouver, BC). The Cryo-immunoelectron micrograph in figure 3.4 was done with assistance by Dr. Elaine Humphries and Garnet Martin (UBC EM lab, University of British Columbia, Vancouver, BC) and the data analysis and statistics were performed with the assistance by Dr. Timothy Vitalis (Michael Smith Laboratories, University of British Columbia, Vancouver, BC). Subcutaneous injection of melanoma xiv cells in mice in Chapter 4 were performed with assistance by Rayshad Gopaul (Department of Zoology & Michael Smith Laboratories, University of British Columbia, Vancouver, BC). I would also like to acknowledge John Turner’s award for the financial support. Most of all, I would like to thanks my father Jia Xu Tian, my mother Isabel Xiao, my husband-to-be Jason Morris, and my sisters May Tian and Estella Tian, for putting up with me over the last seven years through both the happy and not so happy moments. Without them I would not be where I am today. Thus at the end of my Ph.D. journey, I would like to give all of them a big hug and kiss to thank for their unconditional love, support, a shoulder to cry on when I needed it and listening to me vent when I was frustrated. Through these years I have learned many new things both professionally and personally all thanks to everyone mentioned above, without them I would not be the same person I am today. xv

Dedication

I would like to dedicate this work to my parents, Jia Xu Tian and Isabel Xiao, for working hard all their lives for giving my sisters and I better lives in Canada. Without their sacrifice I would not be who I am and where I am today. Thanks for all your love and support, I am eternally grateful! 1

Chapter 1: Introduction

1.1 Melanoma

Melanoma is one of the deadliest cancers, it accounts for less than 5% of all skin cancer cases but about 80% of skin cancer related deaths (Weinstock, 2001) in America. It is the second most common cancer amongst American women between the age of 20 and 35, and is the leading cause of cancer related deaths in woman ages 25 to 30 (SEER Surveillance, 2009). It was estimated that in 2008 there were 62,480 new cases of melanoma and about 8,420 melanoma caused deaths in the United States (Society, 2009). Locally, it was estimated that there were 4,600 new cases and 910 deaths in Canada in 2008 with the provinces of Ontario and British Columbia having the highest number of new cases (Canada, 2008).

Melanoma has garnered increasing attention in the recent years. Melanoma can be traced back as early as 2,400 years ago in the Peruvian Inca mummies which showed apparent melanotic masses in the skin (Urteaga and Pack, 1966). The first reported case of metastatic melanoma was reported by John Hunter in 1787 from a 35 year old male. The melanoma was originally described as a “cancerous fungous excrescence”, which was later confirmed to be a malignant melanoma using more advanced microscopic examination (Davis, 1980). Melanoma was first described as a disease by Dr. René Laennec during a lecture in 1804 and later published in 1806 (Roguin, 2006). The term melanose, Greek for the color black, was coined by Dr. Laennec to describe the tumors (Roguin, 2006).

Melanoma, malignant tumor derived from epidermal melanocytes, can occur in any tissue where melanocytes are present. However, the most common form is cutaneous melanoma (Cummins et al., 2006). Melanocytic nevi, UV exposures, genetic and phenotypic characteristics are known risk factors of the disease (Cummins et al., 2006). Significant numbers of melanoma cases arise from pre-existing nevi; however, in recent years, there are increasing evidences linking UVR exposure and the development of 2 melanoma through DNA damage (Cummins et al., 2006). Melanoma develops through a series of transformations (Figure 1). In normal skin tissues, melanocytes are found within the basal layer of the epidermis. Melanoma cells may develop either from melanocytic nevi (Figure 1; black arrows) or directly from normal tissues (Cumniins et al., 2006)

(Figure 1; red arrows). A series of events take place during the melanomagenesis that lead to alterations to the cellular properties of the melanocytes, such as cell proliferation, differentiation and death (Swetter, 2009). Melanoma progression begin with cells proliferating locally within the epidermis in a lateral direction, known as the Radial Growth Phase (RGP), followed by the Vertical Growth Phase (VGP) characterized by an invasion into the dermis and associated vascularization (Swetter, 2009). As melanoma cells proliferate, they remain clustered together. Once the mass of cells reach the critical size of 1 nmi in diameter, diffusion is no longer sufficient to supply nutrients to the cells thus pro-angiogenic factors are secreted to establish new vascularization (Rolz-Cruz and Kim, 2008). The establishment of this vascularization thus facilitate the growth of tumor mass. 3

Normal Nevus RGP VGP Metastatie

Stratumcomeum Granularlayer Spinouslayer • :. ; 0

Basallayer oc 0

Melanomacell N Directtransformation/

Bloodvessel

Figure 1.1: Melanocyte progression to malignant melanoma. Melanoma cells can arise either from a nevus (black arrow) or from melanocyte directly. The initial stage of the disease is characterized by the lateral cell proliferation within the epidermis (RGP, Radial Growth Phase), and progress into the metastatic form which involve cell invasion into the dermis (VGP, Vertical Growth Phase) and vascularisation of the lesion. 4

1.2 Iron and melanoma

1.2.1 Iron metabolism

1.2.1.1 Brief history of iron

lion (Fe), one of the most abundant elements in the universe, it has played an important role in human history. Following the lion Age, aside from helping mankind build the world as we know it, it also plays an influential role in the history of medicine. For example, in early t116 century iron was used to treat variety of disorders from minor acne and diarrhea to more serious tuberculosis (Beutler, 2002; Beutler et al., 1963; Fairbanks et al., 1971). The first demonstration of iron in blood was demonstrated by Lemery and Geoffroy in 1713, however, it was not until 1925 that the existence of non-hemoglobin iron was documented by Fontés and Thivolle (Beutler, 2002). A pivotal moment in the understanding of iron took place in 1937 when ferritin was crystallized from horse spleen by Laufberger. This lead to the discover of the different chemical forms of iron and gradual understanding of iron homeostasis (Beutler, 2002).

1.2.1.2 Distribution and function of iron

lion is an essential element that plays a crucial role in cellular metabolism; it is required by almost all organisms except for a few species of bacteria (Chua et al., 2007; Neilands, 1974; Stubbe, 1990). lion is present in relatively higher amounts comparing to the other metals in human body (Lindley, 1996). The variation of iron levels in the body depending on a variety of factors, such as age, gender, nutrition, and health status. For example, iron levels are limited at birth; however they increase gradually over time until adulthood (Chua et al., 2007). Adult men usually contain between 35 to 45 mg of iron per kilogram of body weight, while premenopausal women possess much less due to the periodic blood loss during menstruation (Andrews, 1999). lion imbalance can lead to diseases such as iron deficiency or anemia which affects an estimated 3 billion people worldwide, and 5 hemochromatosis which is caused by iron overload (Andrews, 1999). Iron supply in the body is affected by the absorption of dietary iron from foods in the duodenum, and factors that either promote or reduce the absorption such as alcohol consumption, pregnancy and calcium level (Table 1.1).

The absorption of dietary iron in the duodenum occurs via the passing of iron from gut lumen through enterocytes. In biological systems, iron functions through the conversion between the ferrous )2(Fe and ferric j3(Fe form (Beinert, 1976). In nature, iron is commonly found in the form of ferric )3(Fe and is relatively insoluble at neutral pH. There are two forms of iron obtained through normal healthy diets: heme iron from meat products and non-heme iron from plant and dairy products (Chua et al., 2007). As depicted by Chua et al, the non-heme iron )3(Fe is enzymaticlly reduced to ferrous )2(Fe form by brush border ferrireductase, and subsequently transported into the enterocyte through divalent metal transporter 1 (DMT1, also known as Nramp2). On the other hand, the heme iron is taken up as part of metalloporphyrin which binds to a heme receptor, HCP 1, on the apical membrane of the duodenal enterocytes. Once the iron is transported into the gut cell, the iron j2(Fe is released from metalloporphyrin by heme oxygenase, an enzyme that liberates iron by breaking down heme. The iron is either stored inside the enterocyte as ferritin or transported and released into the blood circulation via iron exporter on the basolateral surface. Hephestin oxidize the iron from 2Fe back to the 3Fe form during the export into the blood stream, where it is distributed throughout the body via transferrin (TO (Chua et al., 2007) (Figure 1.2). Majority of the iron bound to circulating plasma Tf is used during erythropoiesis (Finch, 1982). Tf is also believed to be responsible for the delivery of iron to most cells in the body, where diferric loaded Tf binds to the transferrin receptor (TfR) on cell surface and internalize iron via receptor mediated endocytosis (Klausner et al., 1983). 6

Table 1.1 Examples of dietary iron food sources and factors which increase and reduce iron absorption (Natural Health Information Centre, 2009).

Source mg I lOOg Source mg I lOOg

Kelp 100 Pigs kidneys 6.4

Curry powder 29.6 Almonds 4.7

Breakfast cereals 16.7 Prunes 3.9

Molasses 16 Tomato puree 3.5

Pumpkin seeds 11.2 Brazil nuts 3.4

Soy flour 8.0 Eggs 2.0

Lambs liver 7.5 Beef 1.9

Sunflower seeds 7.1 Red wine 0.5

Factors (t iron absorption)

Alcohol Pregnancy

Malic acid (fruits) Fruits

Fructose Tartaric acid (baking powder, tart fruits)

Lactic acid (milk, yoghurt) Vitamin C

Factors (.1.iron absorption)

Tannins (tea, red wine) Phosphates (in soft drinks)

Antacid use Lignin(in fibrous vegetables)

Phytates (cereals) Low stomach acid

Calcium Manganese 7

Figure 1.2: Absorption of dietary iron by enterocytes in the duodenum. Non-heme iron is transported into the enterocyte via divalent metal transporter 1 (DMT1), while heme iron is taken up as part of metalloporphyrin which binds to a heme receptor, HCP 1, on the surface of the enterocyte. Iron taken up into the cell is either stored as ferritin or transported out through basolateral surface into the blood circulation. Figure is adapted from Chua et al. (2007). Gut Lumen Enterocyte Blood

FPN

Fe2 - Fe Fe2 -1--> Non-heme A iron Fe3 Hephaestin Fe3 Transit Iron Pool ITt 4 7 Fe2Tf Fe2 1 Ferritin Heme

— Oxygenase Heme I >Heme iron

00 9

Through evolution, biological systems have evolved to utilize iron in a variety of processes. Iron containing proteins such as cytochrome C reductase I oxidase and NADH-ubiquinone reductase are involved in the transfer of protons within mitochondria, which leads to the generation of an electrochemical gradient and subsequent ATP production (Rich, 1984). Iron is also an essential in cell proliferation as it is vital to the activity of an enzyme important in DNA synthesis (Simonart et al., 2000). The cellular requirement of iron is a tightly regulated process. Although iron is essential for life, excess deposits in a cell can also be detrimental leading to the production of the highly toxic hydroxide radical (.OH)(Lindley, 1996). The hydroxide radical, a reactive oxygen species (ROS), is highly reactive due to unpaired electrons in its valence shell. It is generated through the Haber-Weiss reaction utilizing iron as a catalyst (Lindley, 1996).

First step: 3Fe + •02 —f Fe + 02 Second step: 2Fe + 20H —* 3Fe + OH + .OH Net reaction: •02 + 20H —* OH + .OH +02 This redox potential is capable of causing damage to fatty acids, nucleic acids and proteins in the cell (Aruoma et al., 1989; Chua et al., 2007; Halliwell and Gutteridge, 1986).

1.2.2 Role of iron in melanoma

The detailed role of iron in melanoma development and progression remains unclear. However, several studies have connected cellular iron to the development of cancers (Sussman, 1992; Weinberg, 1996). Risk assessments have found positive correlations between body iron levels and risk of developing cancer (Stevens et al., 1994), as well as between hemochromatosis and an enhanced probability of developing liver cancer and other malignancies (Weinberg, 1996). Cancer cells are generally characterized by abnormal proliferation of neoplastic cells, which is aided

by an up-regulation of transferrin receptor 1 (TfR 1) (Chitambar et al., 1983; Larrick and Cresswell, 1979; Trowbridge and Omary, 1981). In addition, it has also been 10 demonstrated that cancer cells such as melanoma and hepatoma cells take up iron at a much higher rate (Kwok and Richardson, 2002; Richardson and Baker, 1992; Richardson and Baker, 1990; Trinder et al., 1996). Simonart et al. summarized five possible pathways in which iron could affect cancer: (1) promote the formation of mutagenic reactive oxygen species, (2) inhibit the activity of immune cells to minimize host’s defences, (3) promote production of viral nucleic acids, (4) induce anti-apoptotic signals, and lastly (5) promote the growth of neoplastic cells (Simonart et al., 2000). Iron has been shown to directly affect the growth of neoplastic cells, specifically the progression from 1G to S phase of the cell cycle (Kwok and Richardson, 2002; Trowbridge and Omary, 1981), by affecting the activation of ribonucleotide reductase that is responsible for the reduction of ribonucleotide to deoxyribonucleotide during DNA synthesis (Cazzola et al., 1990; Simonart et al., 1999; Simonart et al., 2000).

1.3 Melanotransferrin and iron

1.3.1 Melanotransferrin and transferrin family of proteins

The transferrin family of proteins are a group of single chain, glycosylated proteins that function to transport or help regulate iron levels in both vertebrates and invertebrates (Lambert et al., 2005b). Family members include serum transferrin (Tf), melanotransferrin (MTf or p97), lactotransferrin (LTf), and ovotransferrin (OTf). All members possess two homologous ferric iron binding lobes (N- and C- lobes) linked by a short hinge region (Lambert et al., 2005b). Each lobe is capable of binding one atom of ferric iron together with a synergistic anion such as bicarbonate anion. Protein sequence alignment of the family members show great conservation in the region involved in coordinating iron binding (Figure 1.3).

1.3.1.1 Transferrin

Human serum transferrin was first isolated from the non-hemoglobin fraction of blood plasma in 1946 (Schade and Caroline, 1946). The gene is approximately 33.5 kb in length, and is made up of 17 exons coding for an 80 kDa molecular 11 weight (Lambert et al., 2005b; Schaeffer et al., 1987). It has two iron binding lobes (N-terminal and C-terminal) that are globular structures made up of iron binding subunits and anion binding inter-subdomain cleft. The cleft switches between the open and closed form during iron release and binding respectively (Baker and Lindley, 1992). Serum Tf is widely accepted as the main iron transport and regulatory molecule that plays a key role during erythropoiesis and active cell division (Macedo and de Sousa, 2008). Tf function via binding to transferrin receptor (TfR), which is a transmembrane protein made up of disulfide-linked homodimer of two 760 amino acids glycoproteins, with each monomer capable of binding a molecule of Tf (Cheng et al., 2004). The intracellular tail of TfR, more specifically the Tyr-Thr-Arg-Phe sequence, functions to signal for endocytic internalization (Collawn et al., 1993).

The iron uptake by Tf is a receptor mediated process (Figure 1.4). Briefly, serum Tf is capable of binding an atom of ferric iron per each of its lobe. At physiological pH, the iron bound Tf docks with the TfR on the cell surface and is endocytosed into the cell via the formation of clathrin-coated vesicles. Subsequently, these vesicles lose their clathrin-coat and fuse with each other to become endosomes. The early endosomes are located near the cell periphery and are the first vacuolar compartments along the endocytic pathway. Eventually, early endosomes fuse and mature into late endosomes resulting in a gradual decrease in pH within the endosomes. This leads to a reduction in the affinity of iron for the Tf-TfR complex. The ferric iron is consequently released within the endosomes and is converted into ferrous iron by a ferrireductase before released into the cytoplasm of the cell via endosomal divalent metal transporter 1 (DMT1) to be either used immediately or stored as ferritin (Bali et al., 1991; De Domenico et al., 2008; Sipe and Murphy, 1991). 12

Human Melanotransferrin MRGPSGALWLLLALRTVL—-GGME -VRWCATSDPEQHKCGNMSEAFREA- 46 Mouse Melanotransferrin MRLLSVTFWLLLSLRTVV- -CVME-VQWCTISDAEQQICCKDMSEAFQGA- 46 Human Lactotransferrin MKLVFLVLLFLGALGLCLAGRRRS -VQWCAVSQPEATKCFQWQRMIRKV- 48 Human Transferrin MRLAVGALLVCAVLGLCLAVPDKT-VRWCAVSEHEATKCQSFRDHMKSVI 49 Mouse Transferrin MRLTVGALLACAALGLCLAVPDKT -VKWCAVSEHENTKCISFRDHMKTVL 49 Chicken Ovotransferrin MKLILCTVLSLGIAAVCFAAPPKSVIRWCTISSPEEKKGNNLRDLTQQ- - 48

Human Melanotransferrin - GIQPSLLCVRGTSADHCVQLIAAQEADAITLDGGAIYEAG - KEHGLKPV 94 Mouse Melanotransferrin - GIRPSLLCVQGNSADHCVQLIKEQKADAITLDGGAIYEAG - KEHGLKPV 94 Human Lactotransferrin - -RGPPVSCIKRDSPIQCIQAIAENRADAVTLDGGFIYEAGLAPYKLRPV 96 Human Transferrin PSDGPSVACVKKASYLDçIRAIAAN VTLDAGLVYDAYLAPNNLKPV 99 Mouse Transferrin PPDGPRLACVKKTSYPDIKAI SAS TLDGGWVYDAGLTPNNLICPV 99 Chicken Ovotransferrin - -ERISLTCVQKATYLDCIICZ4ANN ISLDGGQVFEAGLAPYKLKPI 96

Human Melanot rans f errin VGEVYD- -QEVGTSYYAV RSSHVTIDTLKGVKSCHTGINRTVGWNV 142 Mouse Melanot rans f errin VGEVYD- -QDIGTSYYAVA RNSNVTINTLKGVKSCHTGINRTVGWNV 142 Human Lactotransferrin AAEVYGTERQPRTHYYAVAVVICICGGSFQLNELQGLKSCHTGLRRTAGWNV 146 Human Transferrin VAEFYGSKEDPQTFYYAVAVVKIcDSGFQMN RGKKSCHTGLGRSAGWNI 149 Mouse Transferrin AAEFYGSVEHPQTYYYAVAVVICICGTDFQLN GKKSCHTGLGRSAGWVI 149 Chicken Ovotransferrin AAE IYEHTEGSTTSYYAVAVVICKGTEFTVND QGKNSCHTGLGRSAGWNI 146

Human Melanotransferrin PVGYLVESGRLSVMGC- - -DVLICAVSDYFGG9VPGAGETSYSESLCRLC 189 Mouse Melanotransferrin PVGYLVESGHLSVMGC - - - DVLKAVGDYFGGSCVPGTGETSHSESLCRLC 189 Human Lactotransferrin PIGTLRPFLNWTGP- - - PEPIEAAVARFFSASCVPGADKGQF- PNLCRLC 192 Human Transferrin PIGLL- -YCDLPEP- - -RKPLEKAVANFFSGSCAPCADGTDF- PQLCQLC 193 Mouse Transferrin PIGLL- - FCKLSEP- - - RSPLEKAVSSFFSGSCVPCADPVAF- PKLCQLC 193 Chicken Ovotransferrin PIGTLLHWGAIEWEGIESGSVEQAVAKFFSASCVPGATI - - -EQKLCRQC 193

Human Melanotransferrin RGDSSGEGVCDKSPLERYYDYSGAFRCLAEGAGDVAFVKHSTVLENTDGK 239 Mouse Melanotransferrin RGDSSGHNVCDKSPLERYYDYSGAFRCLAEGAGDVAFVKHSTVLENTDGN 239 Human Lactotransferrin AGT- -GENKCAFSSQEPYFSYSGAFKCLRDGAGDVAFIRESTVFEDLSDE 240 Human Transferrin PG CGCSTLNQYFGYSQAFKCLKDGAGDVAFVKHSTIFENLANK 236 Mouse Transferrin PG CGCSSTQPFFGVVKCLKDGGGDVAFVKHTTI FEVLPEK 236 Chicken Ovotransferrin KGD- - PKTK- - CARNAPYSGYSGAFHCLKDGKGDVAFVKHTTVNENAPD 238

Human Melanotransferrin TLPSWGQALLSQDFELLCRDGSRADVTEWR VPAHAVVVRADTDG 289 Mouse Melanotransferrin TLPSWGKSLMSEDFQLLCRDGSRADITEWR PAHAVVVRGDMDG 289 Human Lactotransferrin A ERDEYELLCPDNTRKPVDKFKD VPSHAVVARSVNGK 282 Human Transferrin A DRDQYELLCLDNTRKPVDEYKD QVPSHTVVARSMGGK 278 Mouse Transferrin A DRDQYELLCLDNTRKPVDQYED IPSHAVVARKNNGK 278 Chicken Ovotransferrin LNDEYELLCLDGSRQPVDNYKTGNWARVAAHAVVARDDNKV 279

Human Melanotransferrin -GLIFRLLNEGQRLFSHEGS-SFQMFSSEAYG- - -QKDLLFKDSTSELVP 334 Mouse Melanotransferrin -GLIFQLLNEGQLLFSHEDS-SFQMFSSKAYS- -QKNLLFKDSTLELVP 334 Human Lactotransferrin EDAIWNLLRQAQEKFGKDKSPKFQLFGSPSG- - - -QKDLLFKDSAIGFSR 328 Human Transferrin EDLIWELLNQAQEHFGKDKSKEFQLFSSPHG XDLLFKDSAHGFLK 323 Mouse Transferrin EDLIWEILKVAQEHFGKGKSKDFQLFSSPLG KDLLFKDSAFGLLR 323 Chicken Ovotransferrin - EDIWSFLSKAQSDFGVDTKSDFHLFGPPGKKDPVLKDFLFIcDSAIMLKR 328

Human Melanotransferrin IATQ-TYEAWLGHEYLIThNKGLL CDPNRLPPYLRWCVLS 372 Mouse Melanotransferrin IATQ-NYEAWLGQEYLQAMKGLL CDPNRLPHYLRWCVLS 372 Human Lactotransferrin VPPRIDSGLYLGSGYFTAIQNLRKSEEEVAA-RR ARVVWCAVG 370 Human Transferrin VPPRNDAKMYLGYEYVTAIRNLREGTCPEAPTDEC KPVKWCALS 367 Mouse Transferrin VPPRMDYRLYLGHNYVTAIRNQQEGVCPEGSIDN SPVKWCALS 366 Chicken Ovotransferrin VPSLMDSQLYLGFEYYSAIQSMRKDQLTPSP-RE NRIQWCAVG 370

Human Melanotransferrin TPEIQKCGDMAVAFRRQRLKPE IQCVSAKSPQHCMERIQAEQVDAVTLSG 422 Mouse Melanotransferrin APEIQKCGDMAVAFSRQNLKPE IQCVSAESPEHCMEQIQAGHTDAVTLRG 422 Human Lactotransferrin EQELRKCNQWSGLS EGSVTCSSASTTEDCIALVLKGEADN4SLDG 415 Human Transferrin HHERLKCDEWS VMS VGKIECVSAETTEDCIAKIMMGEADANSLDG 412 Mouse Transferrin HLERTICCDEWSIIS EGKIECESAETTEDCIEKIVNGEADAMTLDG 411 Chicken Ovotransferrin KDEICSKCDRWSVVS NGDVECTVVDETKDCI IKIMKGEADAVALDG 415 Human Melanotransferrin EDIYTAGKICYGLVPAAGEHYAPED - S SNSYYVVAVVRR 459 13

Mouse Melanotransferrin EDIYRAGKVYGLVPAAGELYAEED-R SNSYFVVAVARR 459 Human Lactotransferrin GYVYTAGK- CGLVPVLAENYKSQQ- SSDPDPNCVDRPV-EGYLAVAVVRR 462 Human Transferrin GFVYIAGK-CGLVPVLAENYNKSDN CEDTPE-AGYFAVAVVKK 453 Mouse Transferrin GHAYIAGQ- CGLVPVMAEYYESS -NCAIPSQQGI - - F- PKGYYAVAVVKA 456 Chicken Ovotransferrin GLVYTAGV- CGLVPVMAERYDDESQCSKTDE H- PASYFAVAVARK 458

Human Melanotransferrin DSSHAFTLDELRGKRSCHAGFGSPAGWDVPVGALIQRGFIRPKDCDVLTA 509 Mouse Melanotransferrin DSSYSFTLDELRGKRSCHPYLGSPAGWEVPIGSLIQRGFIRPKDCDvLTA 509 Human Lactotransferrin - SDTSLTNNSVKGKKSCHTAVDRTAQWNIPMGLLFNQ- - - -TGSCK- - - - 503 Human Transferrin - SASDLTWDNLKGKKSCHTAVGRTAGWNIPMGLLYNK- - - - INHCR- - - - 494 Mouse Transferrin - SDTSITWNNLKGKKSCHTGVDRTAGWNIPMGMLYNR- - - - INHCK- - - - 497 Chicken Ovotransferrin DS- -NVNWNNLKGKKSCHTAVGRTAGWVIPMGLIHNR- - - -TGTCN- - - - 498

Human Melanotransferrin VSEFFHASCVPVNNPKNYPSSLCALCVGDEQG- - -RNKCVGMSQERYYGY 556 Mouse Melanotransferrin VSQFFNASCVPVNNPKNYPSALCALCVGDEKG- - - RNKCVGSSQERYYGY 556 Human Lactotransferrin FDEYFSQSCAPGSDPR- - - SNLCALCIGDEQG- - - ENKCVPNSNERYYGY 547 Human Transferrin FDEFFSEGCAPGSKKD- - - SSLCKLCMGS - -G- - - LNLCEPNNKEGYYGY 536 Mouse Transferrin FDEFFSQGCAPGYEKN- - - STLCDLCIG PLKCAPNNKEEYNGY 537 Chicken Ovotransferrin FDEYFSEGCAPGSPPN- - - SRLCQLCQGS- -GGIPPEKCVASSHEKYFGY 543

Human Melanotransferrin RGAFRCLVENAQDVAVRHTTVFDNTNGHNSEPWAAELRSEDYELLCPNG 606 Mouse Melanotransferrin SGAFRCLVEHAGDVAFVKHTTVFENTNGHNPEPWASHLRWQDYELLCPNG 606 Human Lactotransferrin TGAFRCLAENAGDVAFVKDVTVLQNTDGNNNEAWAKDLKLADFALLCLDG 597 Human Transferrin TGAFRCLVEK-GDVAF’VKHQTVPQNTGGKNPDPWAKNLNEKDYELLCLDG 585 Mouse Transferrin TGAFRCLVEIC- GDVAFVKHQTVLDNTEGICNPAEWAKNLKQEDFELLCPDG 586 Chicken Ovotransferrin TGALRCLVEK- GDVAFIQHSTVEENTGGKNKADWAKNLQMDDFELLCTDG 592

Human Melanotransferrin ARAEVSQFAACNLAQIPPHAVMVRPDTNIFTVYGLLDKAQDLFGDD - -HN 654 Mouse Melanotransferrin ARAEVDQFQACNLAQMPSH PDTNIFTVYGLLDKAQDLFGDD - -HN 654 Human Lactotransferrin KRKPVTEARSCHLAMAPNHA S DKV-ERLKQVLLHQQAICFGRNGSDC 646 Human Transferrin TRKPVEEYANCHLARAP T KDKE -ACVHKILRQQQHLFGSNVTDC 634 Mouse Transferrin TRKPVKDFASCHLAQAP S KEKA-ARVKAVLTSQETLFGGS- -DC 633 Chicken Ovotransferrin RRANVMDYRECNLAEV2T PEKA-NKIRDLLERQEKRFGVNGSE - 640

Human Melanotransferrin KNGFKMFDSSNYHGQD ATVRAVPVGEKTTYRGWLGLDYVAALEGM 704 Mouse Melanotransferrin KNGFQMFDSSKYHSQD ATVRAVPVREKTTYLDWLGPDYVVALEGM 704 Human Lactotransferrin PDKFCLFQSE - - - T NTECLARLHGKTTYEKYLGPQYVAGITNL 693 Human Transferrin SGNFCLFRSE- - -TKDLL DTVCLAKLHDRNTYEKYLGEEYVKAVGNL 681 Mouse Transferrin TGNFCLFKST- - -TKDLLFR DTKCFVKLPEGTTPEKYLGAEYMQSVGNN 680 Chicken Ovotransferrin KSKFMMFESQ- - -NKDLLFILTKCLFKVREGTTYKEFLGDKFYTVI5NL 687

Human Melanotransferrin SSQQCSGAAAPAPGAPLLPLLLPALAARLLPPA L 738 Mouse Melanotransferrin LSQQCSGAGAAVQRVPLLALLLLTLAAGLLPRV L 738 Human Lact ot rans f e rrin K--KCST SPLLEACEFLR-K 710 Human Transferrin R--KCST SSLLEACTFRR-P 698 Mouse Transferrin R--KCST SRLLEACTFHK-H 697 Chicken Ovotransferrin K--TCNP SDILQMCSFLEGK 705

Figure 1.3: Sequence alignment of selective members of the transferrin family of proteins including human and mouse melanotransferrin (p97), human and mouse transferrin (TO, human lactotransferrin (LEFt),and chicken ovotransferrin (OTt). Conserved amino acids are highlighted in yellow, while red colour amino acids denote residues involved in iron binding. Sequence alignment was constructed using MAVID multiple alignment server version 2.0.4 (Bray and Pachter, 2004). 14

Figure 1.4: Iron uptake mediated by transferrin-transferrin receptor complex. Each transferrin is capable of binding a maximum of two molecules of ferric iron. At optimum condition, two circulating holo-transferrin binds to one cell surface transferrin receptor at physiological pH, and the complex is endocytosed through formation of the clathrin-coated pit. As these vesicles mature, the clathrin coat is shed and these vesicles mature into early endosomes. As the early endosomes gradually matures into late endosomes in the endocytic pathway, the decrease in intravascular pH facilitate the disassociation of iron from transferrin. The iron is transported out of the late endosomes through divalent metal transporter, DMT1, and is either incorporated into ferritin for storage or transferred to other organelles for cellular usage. Extracellular Fe2-Tf 0 0 QApo-TI Space (PH 7L4) 0 øj TfR Cytoplasm 0

Clathrin ‘ 4. DMT1 PH<5.5 Clathrin coated pit Late endosome

Early Endosome St Ferritin

I-a 0’ 16

1.3.1.2 Other transferrin family members

Lactotransferrin (LTf) has been proposed to be a relatively recent addition to the transferrin family (Lambert et al., 2005b). Similar to Tf, it is an 80 kDa glycoprotein with an iron binding ability. LTf can be found in mammalian milk (Campbell et al., 1992), tears (Janssen and van Bijsterveld, 1983), and other secretions such as pancreatic fluid (Colomb et a!., 1976) and bile (Saito and Nakanuma, 1992). It has been shown to bind iron at a higher affinity than serum Tf (Baker and Lindley, 1992). However, LTf knockout mice show no obvious disruption to overall iron metabolism (Ward et a!., 2003). LTf has been proposed to function mainly in the immune response. LTf seems to be released during inflammation from neutrophils (Crouch et al., 1992), and promote the migration of neutrophils from blood to the inflammation site (Asako et al., 1994; Kurose et al., 1994). LTf has also been shown to reduce production of IL-i, 1L6, and TNFa by monocytes in response to lipopolysaccaride (Crouch et al., 1992; Otsuki et al., 1998).

Ovotransferrin (OTf) is the serum Tf of birds and can be found in both serum and egg whites (Jeltsch et al., 1987; Lambert et al., 2005b). Its expression varies depending on time, location, and degree of glycosylation (Lambert et al., 2005b; Williams et a!., 1982). OTf is believed to have both an iron transport and anti-microbial role in egg whites (Lambert et al., 2005a).

1.3.1.3 Melanotransferrin

Melanotransferrin, also known as melanoma associated tumor antigen p97, is a 97 kDa protein first identified as a cell surface marker for human skin cancer (Brown et al., 1981b). It has about 40% sequence homology with chicken transferrin, human serum transferrin and lactoferrin (Brown et al., 1982). It is highly expressed in neoplastic cells, especially in malignant melanoma cells, while mostly non-detectable in normal tissues (Brown et al., 1981b; Woodbury et a!., 1980). Subsequent studies have found it to be expressed at various levels in the liver, intestine, sweat gland, placenta, umbilical cord, and more recently in human brain endothelium (Alemany et al., 1993; Barresi and Tuccari, 1994; Rothenberger et a!., 1996; Sciot et a!., 1989). Increased levels of p97 have been observed in both the sera of Alzheimer’s disease (AD) patients and the reactive 17 microglia associated with the amyloid plaque in the brain of AD patients (Jefferies et al., 1996; Kennard et aL, 1996).

Using multi-sequence alignments and neighbour-joining trees of 71 transferrin family sequences from 51 different species, Lambert et al. proposed that p97 is one of the oldest member of the transferrin family, dating back to more than 670 million years ago (Lambert et al., 2005b). P97 may have diverged from serum Tf soon after duplication of the iron binding lobes (Lambert et al., 2005b) (Figure 1.5).

Figure 1.5 Evolution of selected members of mammalian transferrin family showing the N-and C-lobes of p97 (MTF and CMTF respectively) are more similar to the C-lobe of serum transferrin (CTF) than the N-lobe (TF). The numbers on the branches represent the percentage of 2,000 bootstrap samples supporting that branch. This linearized bootstrap NJ tree was adapted from Lambert et al. (2005) with permission (Lambert et al., 2005b). LTF = lactoferrin, ICA = inhibitor of carbonic anhydrase. 18

1.3.2 Melanotransferrin and iron uptake

1.3.2.1 Iron binding and melanotransferrin

In order to localize the chromosomal location of the p97 gene, Plowman et al. characterized somatic cell hybrids for p97 expression and their presence in human chromosomes, and found that similar to Tf and TfR genes, p97 is located on chromosome

3 (Plowman et al., 1983). Like other transferrin family members, p97 possess two metal binding lobes (Table 1.2). Unlike its family members, p97 is found in two different forms, which are thought to be the product of alternative splicing of the p97 mRNA (Food and Des Richardson, 2002; McNagny et al., 1996). It can exist either as a membrane protein attached to the cell surface via a glycosyiphosphatidylinositol (GPI) anchor or as a free soluble form in the serum (Alemany et al., 1993; Brown and Rose, 1992; Food and Des Richardson, 2002). Protein sequence alignment of p97 with Tf has shown that amino acids involved in binding of the iron atom in the N-lobe are totally conserved (Rose et al., 1986)(Table 1.2). In fact, iron binding studies have shown that the binding affinity of iron to p97 is intermediate between the N- and C-lobes of the serum Tf (Creagh et al., 2005). Hhowever there are at least four amino acid substitutions in the C-lobe of p97 which render its iron binding ability nonfunctional: (1) The substitution of aspartic acid by serine residue are position 421 is thought to affect the hydrogen bonding across the iron binding cleft, therefore resulting in the inability to lock iron atom in place when it is bound. (2) The substitution of threonine by alanine at position 478 and (3) the substitution of arginine by serine at position 482 are both thought to affect the proper binding of carbonate anion at the binding site. And lastly, (4) the substitution of threonine by proline at position 483 is thought to disrupt the hydrogen bonding with aspartic acid at position 421 (Bailey et al., 1988; Baker and Lindley, 1992; Sekyere and Richardson, 2000). These modifications to the iron binding pocket are likely to weaken the iron binding ability of the C-lobe as compared to the totally conserved N lobe.

Most proteins involved in iron metabolism are regulated through the iron response element-iron regulatory protein system, such as TfR and ferritin. The iron response 19

element is a conserved RNA stem-loop structure located on the untranslated regions of the mRNA. When iron is in short supply, the iron regulatory protein binds to the iron regulatory element and leads to translational regulation. However, when iron is in abundant supply, the binding between iron regulatory protein and iron response element does not take place. In contrast, there has been no iron response element identified for p97 and its expression does not seem to be regulated by iron levels (Kennard et al., 1995; Richardson, 2000). P97 does have a regulatory element located within the enhancer region, deletion of which severely impairs the p97 expression (Duchange et al., 1992).

Also included within the enhancer region are two activator protein-i (AP-1) binding sites.

AP- 1 is a well known transcription factor that is involved in regulation of gene expression in response to various stimuli, and controls cellular processes such as proliferation, differentiation and apoptosis (Eferl and Wagner, 2003). Tnfact, AP- 1 has been shown to be up-regulated after UV irradiation, which may be associated with the high expression level of p97 in melanoma cells (Devary et al., 1991). Furthermore, it has been previous shown in our laboratory that the presence of GPI-anchored p97 doubled the iron uptake in CHO cell line defective in the TfR and transfected with GPI-anchored p97 (Kennard et al., 1995). The treatment of these transfected CHO cells with either phosphatidylinositol-phospholipase C or monoclonal antibody against p97 resulted in a more than 50% reduction and a 47% increase in the iron uptake respectively (Kennard et al., 1995). These results implied that GPI-anchored p97 is involved in the transferrin independent iron uptake in mammals. 20

Table 1.2 Iron binding lobes of transferrin (TO and melanotransferrin (p97). Iron binding is mediated by the amino acid residues aspartic acid (D), tyrosine (Y) and histidine (H). Aspartic acid residue in the C-lobe of p97 is changed to serine (S).

Amino acid residues involved in Iron binding lobes iron binding

N-Lobe DYYH Transferrin C-Lobe DYYH

N-Lobe DYYH melanotransferrin C-Lobe SYYH 21

1.3.2.2 Transferrin-transferrin receptor independent iron transport and melanotransferrin

Tf-TfR is considered the main iron uptake and transport mechanism. However, evidence also exists for iron transport via Tf-TfR independent system. Such as the case for a rare genetic disorder, atransferrinemia, where patients possess none to extremely low concentrations of plasma transferrin (Huggenvik et al., 1989). Although these patients lack plasma transferrin, there is still evidence of dietary iron (non-transferrin bound form) absorption and iron transport to organs such as heart, liver, kidney and pancreas (Beutler et al., 2000; Hamill et al., 1991; Huggenvik et al., 1989). However, non-transferrin bound iron cannot cure the disorder due their inability to be used for erythropoiesis (de Silva et al., 1996). In addition, hypotransferrinemia mice show dramatic iron overload in all but hematopoietic tissues despite lower serum Tf levels (Bernstein, 1987; Craven et al., 1987). It was also demonstrated that the iron burden in the liver of these mice was about 100 fold higher than the normal wild type mice (Trenor et al., 2000), thus offering strong evidence for Tf-TfR independent iron uptake. Furthermore, it was previously demonstrated that by transfecting GPI-anchored p97 into CHO cells, which are defective in its TfR, GPI-anchored p97 is able to transport iron resulting in an increased iron content within the cells (Kennard et al., 1995). Since these CHO cells are defective in TfR, the uptake of iron must be mediated by GPI-anchored p97, independent of Tf-TfR iron transport pathway. In addition, more recent structural studies by Cheng et al. have substantially revised our understanding of this process. The binding of Tf to TfR results in a conformational change in Tf, hindering the iron release from its iron binding N-lobe, thus suggesting that only one atom of iron per each Tf is efficiently released during each cycle of internalization of the Tf-TfR complex (Cheng et al., 2004).

1.4 Possible function of melanotransferrin

Various candidate functions have been proposed for p97 over the years. Truncated serum p97 have been proposed to inhibit angiogenesis (Michaud-Levesque et 22

al., 2007; Sala et al., 2002). P97 has been shown to possibly act as a carrier to shuttle therapeutic drugs across blood brain barrier. It is also linked to Alzheimer’s disease due to its elevated expression in the cerebral spinal fluid and serum of these patients (Kennard et al., 1996; Kim et al., 2001), as well as in the reactive microglia associated with senile plaques (Jefferies et al., 1996). Finally, GPI-anchored p97 has been shown to bind iron and is highly expressed in melanoma cells so further investigations are needed to understand its involvement in melanoma development.

1.5 Project rationale and objectives

The focus of this thesis is to study the relationship between p97 and melanoma. The first part of the thesis will examine the iron binding and transport ability of GPI anchored p97 and the intracellular pathway by which melanoma cells utilize p97 bound iron. Some features relating to the internalization of GPI-anchored proteins have been established (Anderson, 1994; Verkade et al., 2000) but debate still exists on whether the internalization is mediated by a specialized subset of lipid raft called caveolae vesicles or by clathrin-coated vesicles (Rijnboutt et al., 1996; Shyng et al., 1994; Smart et al., 1996). There are also contradictory reports on the normal distribution of GPI-anchored proteins in the cell. While some report a preferential localization to caveolae (Ying et al., 1992), others have shown that under normal conditions, GPI-anchored proteins are diffusely distributed over the cell surface and only upon cross-linking with multiple secondary antibodies do they move into caveolae (Mayor et aL, 1994). However, the majority of the data seem to indicate that GPI-anchored proteins tend to cluster in or near caveolae (Anderson, 1998). Thus, it was hypothesized that GPI-anchored p97 on the surface of melanoma cells, bind and uptake iron through a Tf-TfR independent and clathrin independent endocytosis pathway. In the second part of the thesis, in vitro and in vivo experiments were performed to examine the role of GPI-anchored p97 expression in relation to melanoma formation and progression. It was hypothesized that the high expression of p97 may cause an increase in iron uptake by melanoma cells, and subsequently increases the metabolic activities of the cell. As result p97 may be associated with the rapid division and growth that is commonly observed with melanoma 23 cells and tumor. Taken together, these studies provide new insight into Tf-TfR independent iron uptake, as well as the cellular and functional role of p97 in melanoma. 24

Chapter 2: Materials and Methods

2.1 Tissue culture, cells and antibodies

The human melanoma cell line, SK-MEL-28, and murine melanoma cell line, B16F1O which expresses high levels of p97 were obtained from American Tissue Culture Collection (ATCC, Manassas, VA). These cell lines were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen Life Technologies, Burlington, ON) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS) (Invitrogen Life Technologies, Burlington, ON), 2 mM glutamine and 20 mM Hepes. JBIMS, a murine melanoma cell line which also expresses high levels of p97, was provided by Dr. Vincent J. Hearing (National Institute of Health, Bathesda, MD). The cell line was maintained in RPMI-1640 (Invitrogen Life Technologies, Burlington, ON) supplemented with 10% (vlv) FBS, 2 mM glutamine and 20 mM Hepes. The murine microglia cell line, BV-2, provided by Dr. Michael McKinney (Mayo Clinic, Jacksonville, FL). This cell line exhibit morphological and functional similarities to primary microglia (Blasi et al., 1990). It has been previously reported these cells can undergo phenotype changes after multiple passages (Li et al., 2004). To prevent this, the BV-2 cells used were passaged no more than three times. BV-2 cells were cultured in DMEM supplemented with 10% (v/v) FBS, 2 mM glutamine and 20 mM Hepes. All cells were maintained at 37°C in a 5% 2CO humidified incubator.

Antibodies used in this study are listed in Table 2.1 a) and b). 25

Table 2.1 a: Primary antibodies used in the experiments presented in this thesis.

Antibody Antben Host Concentration Source

ATCC, Manassas, OKT-9 HumanTfR Mouse 12 mg/mi VA

ATCC, Manassas, L235 Human p97 Mouse 15 mg/mi VA

Human ferritin L- DAKO, Carpinteria Ferritin Rabbit 1.3 mg/mi chain CA.

BD Transduction

Caveolin Human caveolin 1 Rabbit 250 Ig/m1 Laboratories, Franklin Lakes, NJ.

Carboxyl terminal Santa Cruz Clathrin HC of human clathrin Rabbit 200 tg/m1 Biotechnology Inc., heavy chain Santa Cruz, CA.

BD Transduction Adaptin- Human 3-adaptin Mouse 250 ig/m1 Laboratories, Franklin Lakes, NJ.

Santa Cruz Amino terminal of EEA- 1 Goat 200 pig/mi Biotechnology Inc., human EEA-1 Santa Cruz, CA.

Clontech Laboratories Histidme tagged 6xHis Mouse 500 fig/mi Inc., Mountain View, protein CA. 26

TabIe2.1 b: Secondary antibodies used in the experiments presented in this thesis

Antibody Concentration Working dilution Source

Alexa 488 nm Molecular Probes conjugated Goat 2 mg/mi 1:500 Inc., Eugene, OR anti-mouse IgG

Alexa 568 nm Molecular Probes 2mg/mi 1:500 conjugated Goat Inc., Eugene, OR rabbit-mouse IgG

Alexa 4568 nm Molecular Probes conjugated Rabbit 2 mg/mi 1:500 Inc., Eugene, OR anti-mouse IgG

Alexa 568 nm Molecular Probes conjugated Rabbit 2 mg/mi 1:500 Inc., Eugene, OR anti-goat IgG

Alexa 568 Molecular Probes conjugated Goat 2 mg/mi 1:500 Inc., Eugene, OR anti-mouse IgG 27

2.2 Constructs. 2.2.1 P97 suppression construct: pSUPERIOR.neo.GFP-p97. A pair of custom oligonucleotides that contain a unique 19 nucleotide sequence (Figure 2.ib) from the mRNA transcript of full length mouse p97 (according to GenBank accession AB024336 from the National Centre for Biotechnology Information database) was generated using OligoEngine RNAi Design Tool (Oligoengine). The oligonucleotides were annealed in 100 m]V1NaCl and 50 mM Hepes PH 7.4 under following conditions: 94°C 4 minutes, 85°C 4 minutes, 80°C 4 minutes and then slowly cooled to 10°C. The annealed oligos are ligated to the pSUPERIOR vector using T4 DNA ligase (Invitrogen). The recombinant pSUPERIOR vector (Figure 2.la) was transformed into DH5a competent cells (Invitrogen), and plated on LB ampicillin agar plates. The bacteria colonies containing the pSUPERIOR vector were isolated using mini-prep kit (Qiagen) and sequenced at the NAPS Facility (Nucleic Acid Protein Service Unit, University of British Columbia, Vancouver BC, Canada). Because the suppression construct is tetracycline regulated, thus the cells were co-transfected with pSUPERIOR.neo.GFP-p97 vector and the pcDNA6/TR TetR-expressing regulator vector using Lipofectamine 2000 (Invitrogen) (Figure 2.lc). PcDNA6/TR expresses a Tet repressor under the control of a human CMV promoter, which binds to Tet operator sequences in the pSUPERIOR vetor to repress its expression in the absence of tetracycline. The transfectants were selected in media containing neomycin (Invitrogen) at concentration of 700 pUg/miand 10 pg/ml blasticidin (Calbiochem) for approximately four weeks and then sorted for GFP expression using the FACSDiva (BD Pharmingen, San Jose, CA) into single cell clones. The clonal cell lines were then grown until confluency under selective media. 28

Figure 2.1: pSUPERIOR.neo.GFP and murine specific synthetic short interfering RNA (siRNA). This vector uses a RNA polymerase-lil Hi promoter to drive the expression of a short hairpin RNA. The oligo insert targeting the p97 gene (b) consist of a unique 19 nucleotides in both sense and antisense orientation, separated by a 9 nucleotides spacer sequence that together form the hairpin structure. Once inside the cell, the recombinant vector constitutively expresses the hairpin RNA and process into siRNA to induce suppression of the p97. This vector is tetracycline inducible, whereupon the addition of tetracycline leads to binding with Tet repressor at high affinity. The tetracycline inducibility is accomplished through the co-transfection of this vector along with a pcDNA6/TR TetR-expressing regulator vector (c) (Invitrogen). This causes a conformational change in the repressor that prevents the binding to Tet operator within the Hi promoter sequence, and thus allow for the transcription of RNA hairpin precursor of the siRNA duplex. Neo-GFP fusion protein expressed under the control of separate PGK promoter allows stably transfected cells to be selected using G418 and visually by GFP expression (Adapted from OligoEngine). 29

XmnI Dralil (a) Scal

BsaI IBamHI AhdI Smal XmaI

pSUPERIOR.neo+GFP TthlllI MscI

EGFP Afill BsrGI

BStAPI Agel N[iel BsrnI AatII PpuMI EcoRI * Stul * Bpul 1021 BsmBI

( ) Targquenc: sense (Hair Pfr)) Tagesequence; antir * cAcccrTrcA AQAGACAOGTrTCCATcAGTA1rn rA -3 3 C C-C)CACTkCCTTTCTC c-c CAAGTTCTC1CTCCCACAAACTAAAAAAATTCCJ-5 (BgI II) (HindIIIJ

(c)

pcDNA6ITR© 6662 bp

SV4OpP 30

2.2.2 Histidine tagged p97 construct: pMXpie-His-p97 Genomic DNA from murine melanoma cells was used to construct p97 overexpression vectors. For histidine tagged p97, the full length p97 minus the N- terminus signal peptide was amplified by PCR with Taq polymerase (Invitrogen) using the sense p97FWD primer and antisense p97REV primer (see Table 2 for primer sequences) under following conditions: 95°C 3 minutes, 35 cycles of 95°C, 60°C, 72°C with a final elongation step at 72°C for 10 minutes. A pair of oligonucleotides (SigFWD and SigREV) was synthesized by sigma-Genosys to cover the p97 N-terminus signal peptide and followed by the added 6X histidine tag. The ligation product of the His tagged signal peptide and the p97 gene (Figure 2.2) was subcloned into pBluescript sk+ vector (Stratagene, La Jolla, CA), transformed into DH5a bacteria (Invitrogen) and plated on ampicillin LB agar plates in the presence of X-gal to allow for blue-white colony selection. The white colonies, which contain the pBluescript sk+ vector and the histidine tagged p97 insert, were screened by DNA sequencing. The clone with the correct orientation of signal sequence, histidine tag and p97 was then subcloned into the pMXpie vector using restriction enzymes BamH I and Not I (Invitrogen). The final histidine tagged p97 construct (pMXpie- His-p97) was sequenced again to ensure sequence integrity and correct orientation within the vector.

2.2.3 Non-tagged p97 construct: pCMV-Tag3-p97 The non-tagged p97 overexpression vector was constructed by amplifying the full length p97 using PCR with the p97Full FWD and p97Full REV primers (Table 2.2). The

PCR conditions used were 95°C for 3 minutes, 35 cycles of 95°C 30 seconds, 60°C 30 seconds, 72°C 2.5 minutes, followed by 72°C extension step for 10 minutes. The full length p97 obtained by PCR was then subcloned into pCR2.1-TOPO cloning vector (Invitrogen Life Technologies Burlington, ON), transformed into TOP 10 E. coli bacteria (Invitrogen) and plated on selective ampicillin LB agar plates. Bacteria colonies were screened for full length p97 insert by DNA sequencing. The clone with correct insertion was then subcloned into pCMV-Tag3 vector (Stratagene, La Jolla, CA) using restriction enzymes BamH I and Hind III (Invitrogen, Burlington, ON). The resulting pCMV-Tag3- 31 p97 construct was sequenced with T3 (5’-AAflAACCCTCACTAAAGGG-3’) and T7 primers (5’-TAATACGACTCACTATAGGG-3’). 1111111111111111 I I 111111111 I 11111111 I I 1111 III I I III III 11111111111111111 111111 III I III III

Signal Peptide 6X His tag p97 N-terminus

I I II 1111 1111111111111111111 I III 11111 III 1111111111 111111 1111111111111111 11111111 111111111 I I

1111111 I I I III I I I 11111111 III III 11111111 I I 111111111111111 II 11111111111 III I 11111 11111111

11111 11111 111111 III 111111 III 111111 11111 1111111111 I liii I III 11111 I I III III I 11111111111

p97 C-terminus

Figure 2.2: Nucleotide sequence of the modified mouse GPI-anchored p97 used in the construction of histidine tagged GPI-anchored p97. GPI-anchored p97 precursor cDNA is normally consist of a N-terminus signal peptide, which directs the translocation of nascent chain of p97 into ER, followed by p97 sequence and a C-terminus signal peptide that responsible for the attachment of the GPI-anchor. Upon the processing and attachment of the anchor, both N- and C-terminus signal peptides are removed and thus making tagging the protein difficult. In order to solve this problem, a 6X histidine tag (red) is inserted between the N-terminus signal peptide and the precursor sequence corresponding to the N-terminus of mature p97.

Cu 33

Table 2.2: PCR primers used in this thesis (all synthesized by sigma-Genosys).

PCR Primer name Primer sequence product size (bp) p97FWD 5’- GAGCTCGTGATGGAGGTGCAGTGGTGT-3’ 2160

P97REV 5’- GCGGCCGCTCAGAGAACGCGAGGAAGGA-3’

5’- Sig FWD GATCCGCCGCCATGAGGCTCCTGAGCGTGACTTTT 78 TGGCTACTCCTGTCCCTGCGCACTGTCGTCTGTGC ACATCATCATCATCATCATG-3’

5,- Sig REV AGCTCATGATGATGATGATGATGTGCACAGACGA CAGTGCGCAGGGACAGGAGTAGCCAAAAAGTCAC GCTCAGGAGCCTCATGGCGGCG-3’

P97Fu11FWD 5’-ATGAGGCTCCTGAGCGTGACTfTTfGG-3’ 2217

P97Fu11REV 5’-CAGGCCTCCTTCCTCGCGTFCTCTGA-3’

S15 FWD 5’-TTCCGCAAGHCACCTACC-3’ 361

S15 REV 5’-CGGGCCGGCCATGCmACG-3’

c-JUN FWD 5’-CGTTCTATGACTGCAAAGATGGA-3’ 1027 c-JUN REV 5’-CCCTGACAGTCTGTTCTCAAAA-3’

c-FOS FWD 5’-TCAACACACAGGACTTTfGCG-3’ 880

c-FOS REV 5’-TAAGTAGTGCAGCCCGGAGTA-3’ 34

2.3 Stable transfection Cells at approximately 70% confluency were transfected using TMLipofectamine 2000 reagent (Invitrogen) with 5 g of vector DNA under serum free condition. SK MEL-28 cells were transfected with the pEGFP-Rab5-S34N construct (a generous gift from Dr. Robert E. Lodge, National Institutes of Health) and selected in media containing 500 ig/m1 G418 (Invitrogen). This construct has a pEGFP-C1 (Clontech) backbone. B16F1O cells were transfected either with pMXpie-His-p97 construct, pCMV-Tag3-p97 construct or empty vectors alone as control. Cells containing the pMXpie-His-p97 and pCMV-Tag3-p97 constructs were selected and maintained in puromycin (Calbiochem) at 2 ig/m1 and G418 (Invitrogen) at 700 igIm1, respectively. JBMS cells were co transfected with the p97 suppression construct (pSUPERIOR.neo.GFP-p97) and the pcDNA6/TR TetR-expressing regulator vector (Invitrogen) at a ratio of 6:1 using Lipofectamine 2000 (Invitrogen). The transfectants were selected in media containing 700 ig/m1 G418 (Invitrogen) and 10 ig/m1 blasticidin (Calbiochem) for approximately four weeks and then sorted for GFP expression using the FACSDiva (BD Pharmingen, San Jose, CA) into single cell clones. The clonal cell lines were then grown until confluency under selective media. Since the pSUPERIOR suppression vector is tetracycline inducible, GFP positive clonal cell lines were treated with 1 g/m1 of tetracycline (Sigma) and analyzed for p97 expression using the real-time PCR. The clone with the greatest amount of p97 suppression was selected and used in further experiments.

2.4 Isothermal titration calorimetry Cells that produce the soluble form of p97 were cultured in DMEM supplemented with 10% FBS, 20 mM Hepes, 2 mM L-glutamine and 500 M methotrexate (Faulding). The culture supernatant was collected and passed through a column of L235 mAb immobilized on AfiGellO (Bio-Rad). The soluble p97 was eluted from the column using

0.1 M citric acid at a pH of 2.5, and neutralized with 1 M Hepes. The eluant was dialyzed for two days in 0.2 M 24POKH 1 mM EDTA and 1 mM sodium nitrilotriacetic acid (NTA) in a 10,000 MWCO, dialysis cassette (Pierce) at 4°C accompanied with constant stifling. The apo-p97 was recovered from the dialysis and further dialysed in 50 mlvi Hepes overnight at 4°C. The Hepes dialysis buffer was previously treated with 35 chelex- 100 (Bio-Rad) to remove any traces of metals. Thereafter, the apo-p97 was concentrated in a Centriprep-30 concentrator (Amicon) and its concentration was determined via spectrophotometer at 280 nm using an extinction coefficient of 1.214 (mglml)’ .1cm The isothermal titration calorimetry study was carried out on a MicroCal MCS-ITC (MicroCal Inc., Northampton, MA). All experiments were carried out in 100 mM Hepes, 25 mM 3NaHCO at a pH of 7.5.

2.5 RNA isolation Total RNA from B16F1O and JBMS cells was isolated using Trizol (Invitrogen) followed by DNase-1 digested (Fermentas, Burlington, ON) according to the manufacturer’s instructions. RNA concentrations were determined using NanoDrop ND 1000 (Thermo Scientific, Wilmington DE) at wavelength of 260 and 280.

2.6 RT-PCR Reverse Transcriptase and Polymerase Chain Reaction (RT-PCR) was performed with primer pairs synthesized by Sigma Genosis (Table 2). 1 tg of RNA was converted to cDNA with the Superscript II kit and Oligo dT primers (Invitrogen) as per manufacturer’s instructions, followed by RNase H treatment (Invitrogen). PCR amplification was performed at 95°C 3 mm, 35 cycles at 95°C, 60°C, 72°C and a final elongation step at 72°C for 10 mm using Taq polymerase (Invitrogen). As a control, templates were also amplified using above conditions with S15 primers. S15 rRNA is a well used “housekeeping” gene or loading control for RT-PCR. This is due to its relative constant expression level across cell lines and tissues, and is often unaffected by drug treatments. Amplification products were then separated on a 1% agarose gel and visualized with SyberSafe (Invitrogen) on an Aiphalmager (Alpha Innotech, San Leandro, CA).

2.7 Real-time PCR Semi-quantitative evaluation of RT-PCR results were performed using real-time

PCR. For each sample, 1 ig of RNA was reverse transcribed using Superscript II kit (Invitrogen). Subsequently, real-time PCR was performed using the Roch Light Cycler. 36

Briefly, equal amount of cDNA per sample and gene specific primers were added to SYBR®Green Taq Ready TMMix (Sigma) and PCR amplified. The amplification protocol was as follow: denaturation at 95°C 5 minutes, amplification and quantification at 95°C 5 seconds, 60°C 10 seconds, 72°C 30 seconds for 40 cycles, melting curve at 65-95°C with a heating rate of 0.2°C per second, and final cooling step to 40°C. Simultaneously, all templates were also evaluated with S15 primers as internal control. The transcripts were quantified using Delta-delta CTmethod. Specifically, the cycle at which the fluorescence passed the fixed threshold of a sample was used in calculating the fold of induction of the gene of interest. The formulas used in this calculation are listed below.

ACT= CT(gene of interest) - CT(S15, endogenous control)

Az\CT= ACT(sample x) - ACT(control sample) Fold induction (gene of interest) 2T

2.8 Western blots

B16 cells and B16 cells transfected with the pMXpie-His-p97 were washed two times with cold PBS. Equal number of cells were collected in PBS and centrifuged for 5 minutes at 4°C at 1200 rpm. The pellets were lysed on ice in 1 ml of 1% (vlv) Triton X 100 in 140 mM KC1, 10 mM Tris-Cl for 30 minutes. 20 tl of each sample lysate was mixed with 10 pi of 3X non-reducing sample buffer containing 250 mM Tris-Ci, 3% (vlv) glycerol, 0.16% (wlv) SDS, 0.002% (wlv) bromophenol blue and 1% (w/v) dithiothreitol

(DTT) (Roche). The resulting mixtures were boiled for 5 minutes and cooled to room temperature before 5 iil of 0.5 M iodoacetamide (Sigma) was added to each. Equal amount of samples were ran on 10% (wlv) SDS-PAGE gel and proteins were separated at

120 V for 1 hour and transferred to a nitrocellulose membrane (Biorad). The membrane was first blocked for non-specific binding with 5% skim milk in PBS containing 0.1% (vlv) Tween-20 detergent (Biorad). The membrane was then incubated overnight at 4°C in 5% skim milk / 0.1% Tween-20 in PBS containing primary antibody against mouse histidine tagged protein (1:5000 dilution). After three washes of 30 minutes each in 5% skim milk / 0.1% Tween-20 in PBS, peroxidise conjugated goat anti-mouse antibody (1:10000 dilution) in 5% skim milk / 0.1% Tween-20 in PBS was used to detect the 37 primary antibody (Pierce). The histidine tagged p97 proteins were detected using the chemiluminescence ECL Western blotting detection system (Amersham Biosciences, Piscataway, NJ).

2.9 Immunoprecipitation SK-MEL-28 cells stably transfected with pEGFP-Rab5-S34N construct and pEGFP-C1 vector alone (Clontech) were grown separately to 100% confluency in 60 x 15 mm tissue culture dishes. For each transfected cell line, the cells were washed three times in Iscove’s modified Dulbecco media (IMDM) (Invitrogen) for 1 hour each time at 37°C to remove endogenous iron and serum transferrin. The cells were then incubated with 1.5 ml of 1 M of 55Fe-nitrilotriacetate (Fe-NTA) for 90 minutes at 37°C. This was followed by ice cold PBS wash before the cells were lysed in 1 ml of 1% (vlv) Nonidet P40 (NP-.40)(Sigma) in lysing buffer containing 20 mM Tris-Ci pH 7.4, 150 mM NaC1,2 mM EDTA and 1 tablet of protease inhibitor cocktail (Roche) for 30 minutes on ice. The lysate was collected and centrifuged at 11000 rpm for 15 minutes at 4°C. The supernatant was collected and pre-cleared with 3 pl of normal rabbit serum for an hour and another subsequent hour with 30 il of protein G sepharose bead slurry (Amersham Biosciences, Arlington Height, IL) under agitation. The lysate was immunoprecipitated with antibody against ferritin for two hours at 4°C, and followed by the addition of pre washed protein G sepharose bead slurry (30 !Il) for an hour at 4°C. The sample was washed twice in buffer B (0.2% NP-40, 10 mM Tris-Cl pH 7.5, 150 mM NaC1, 2 mM EDTA), once in buffer C (0.2% NP-40, 10 mM Tris-Ci pH 7.5, 500 mM NaC1, 2 mM EDTA) and once in buffer D (10 mM Tris-Cl pH 7.5). Finally, the sample was resuspended in 1 ml of PBS before transferred to scintillation vials along with the Ready Safe scintillation counting fluid (Beckman Coulter Inc., Fullerton, CA). The radioactivity associated with the immunoprecipitate was measured using Beckman LS6000IS Liquid Scintillation counter. 38

2.10 Detection of surface GPI-anchored murine p97 expression on B16F1O melanoma cells Equal number of B16F1O cells stably transfected with pMXpie-His-p97, pCMV Tag3-p97 construct or empty vector control were plated separately in 60 x 15 mm tissue culture dishes. For each transfected cell line, the cells were washed three times in

Iscove’s modified Dulbecco media (IMDM) (Invitrogen) for 1 hour each time at 37°C to remove endogenous iron and serum transferring. The cells were then incubated with 1.5 ml of 1 M of 55Fe-nitrilotriacetate (Fe-NTA) for 90 minutes at 4°C. The cells were washed three times with cold PBS before being collected in 1 ml of PBS and transferred to scintillation vials along with Ready Safe scintillation counting fluid (Beckman Coulter Inc., Fullerton, CA). The radioactivity was measured using Beckman LS6000IS Liquid Scintillation counter.

2.11 Fab fragment production for immunofluorenscence and confocal microscopy Anti-human p97 monoclonal antibody (mAb), L235, was digested by the method from Parham et al.(Parham et al., 1982). Briefly, L235 mAb was dialyzed overnight in 0.1 mol!L sodium acetate pH 5.5. Papain (Sigma) was activated in 0.1 molfL sodium acetate pH 5.5 with 1 mmol/L EDTA, 50 mmolIL cysteine (Invitrogen) and 1 mmol]L DTI’ for 30 minutes at 37°C. Excess DIT was removed by Sephadex G-25 gel filtration (Pharmacia). The activated papain was added to L235 mAb at a ratio of 1:20 (wlw) and rotating for 5 hours at 37°C. The papain digestion of L235 mAb was stopped by addition of 75 mmolfL iodoacetamide (Fluca). The digest was applied to a Protein A Sepharose column (Amersham) and the column was washed in binding buffer (20 mM sodium phosphate pH 7.0). The Fab fragments were eluted from the column with 0.1 M glycine buffer pH 2.5 and neutralized with 1 M Tris-HC1 pH 9.0. The resulting L235 Fab fragments were conjugated with FITC by BRC antibody facility (University of British Columbia, Vancouver BC, Canada). 39

2.12 Immunofluorescence and confocal microscopy

SK-MEL-28 cells were grown on sterile No.1 18 mm coverslips (Fisher Scientific Inc.) in 10cm plates overnight. At 70% confluency, the coverslips were washed twice in cold PBS and blocked in 1% (wlv) BSA/PBS for an hour at 4°C. The coverslips were stained with 200 i1 of anti-human p97 (L235), FITC conjugated anti-human p97 Fab fragment (L235-Fab), Alexa 488 conjugated transferrin or anti-human transferring receptor (OKT9) antibodies in 1% (wlv) BSA/PBS for 30 minutes on ice. The coverslips were then washed five times for 10 minutes in 1% BSAJPBS before transferring to pre warmed 1% BSAIPBS and incubated at 37°C for various time points to allow endocytosis to take place. Following incubation, cells were fixed in 4% paraformaldehyde and 2% glutaldehyde (Sigma) in PBS for 30 minutes at room temperature. Cells were permeabilized with 0.1% saponin in 1% BSAJPBS for 20 minutes at room temperature. Cells were also stained with a second set of antibodies for various vesicles for 30 minutes at room temperature: rabbit anti-human caveolin (1:50 dilution), goat anti-human clathrin

(1:40 dilution), and goat anti-human early endosome antigen 1 (EEA1) (1:40 dilution). Alexa Fluors (Molecular Probes) were used to visualize the staining pattern (Table I b). The secondary Alexa Fluor antibodies were diluted in 1% BSA/PBS at 1:500 dilutions. Cells were washed five times in 0.1% saponin, 1% BSA/PBS before incubated 10 minutes in Slow Fade (Molecular Probes) equilibration buffer and mounted in Slow Fade glycerol solution. The coverslips were sealed on the slides with nail polish. The BioRad Radiance Plus confocal laser scanning microscope and Nikon TE2000 inverted microscope with EZ-C 1 software version 3.0 were used to capture the fluorescent images.

Data analysis was performed on mid-section single plane images with ImageJ 1.37C and Adobe Photoshop CS version 8.0.

2.12.1 Quantification of confocal results The number of co-localized pixels relative to total pixels was counted per cell with ImageJ 1.37c in merged images. Bar graphs show the means +1-SEM. 40

2.13 Cryo-immunoelectron microscopy

SK-MEL-28 cells grown on coverslips were labelled with 5 nm gold conjugated primary antibody against either p97 or transferrin receptor at 37°C for 30 minutes to allow for internalization. Excess antibodies were removed and the cells were fixed in 4% paraformaldehye and 2% glutaldehyde for 20 minutes. The cells were then permeabilized with 0.1% saponin in PBS before labelled with 10 nm gold conjugated anti-caveolin or lOnm gold conjugated anti-clathrin antibody for 30 minutes at 4°C. Excess antibody was removed before the cells were fixed with 1% osmium tetroxide in 0.1M phosphate buffer

(Canenmco) in the microwave (Microwave fixation: 2 mm on 100W under vacuum, 2 mm off under vacuum and 2 mm on 100W under vacuum). This was repeated and then the samples were rinsed twice in distilled water. The cells were removed from the coverslips before the dehydration protocol in the microwave. The dehydration protocol in the microwave is as followed. Sequential dehydrations in ethanol (50%, 70%, 95%, 100%, 100%, and 100%) were carried out at 200W for 40 seconds, with a gentle centrifugation between each step. The cells were pelleted and infiltrated with 1:1 acetone: resin in the microwave at 300W under vacuum. The resin was a 1:1 mixture of EPON and sTSpurr (soft mixture). The cells were infiltrated three times with 100% resin (same as above, fresh resin each infiltration) and baked at 65°C overnight. The blocks were cut using a diatome diamond knife and a Leica Ultracut T ultramicrotome and stained with 2% uranyl acetate and finally with Reynold’s lead citrate for 30 and 15 minutes respectively. The sections were viewed on a Hitachi H7600 TEM and digital pictures taken using an AMT camera built into the H7600.

2.14 Proliferation assay

B16F10 and JBMS mouse melanoma transfected cells were plated at a density of 8000 cells per well in 96 well flat bottom culture plates. Cell proliferation was measured by the alarmBlue® dye reduction assay (Invitrogen). The assay utilizes a nontoxic, non fluorescent indicator dye, resazurin, which is converted to highly red fluorescent resorufin via metabolic reactions within the living cells. The amount of fluorescence produced by the assay is proportional to the number of living cells. Briefly, 20 hours after initial plating, 10% final concentration (vlv) of alarmBlue® dye was added to each 41 well and returned to incubation for four hours at 37°C 5% .2CO For the next 7 hours following the incubation, the plates were removed at each hour and absorbance readings were measured at 570 nm excitation and 600 nm emission. Dye reduction was measured in a SpectraMax® 190 spectrophotometer (Molecular Devices, Sunnyvale CA). Cell proliferation was calculated as the fold of increase compared with the value at t = 1 hour.

2.15 Melanin assay

Melanin in the cultured B16F10 and B16F10 transfected cells was measured according to Lee et al. (Lee et al., 2005). Briefly, equal number of cells from each cell line was seeded in separate culture dishes. Four days after initial plating, cells were harvested with 0.25% trypsin solution. The cells were collected and resuspended in PBS.

Total cell number in each culture was determined using hemocytometer. 5x10 cells from each cell line were pelleted by centrifugation, and the pellets were solubilized in 1

M NaOH then heated to 100°C for 10 minutes. Melanin content was determined with a SpectraMax® 190 spectrophotometer at 400 nm, and calculated from a standard curve using synthetic melanin (Sigma).

Melanin secretion by B16F10 cells and B16F10 transfected cells was measured by collecting the culture media four days after initial plating. Melanin in the culture media was determined with a SpectraMax® 190 spectrophotometer at 400 nm, and quantified using a standard curve created with synthetic melanin (Sigma).

2.16 In vivo tumor growth assays 2.16.1 Animals The C57BL/6 strain of mice was maintained at the Biotechnology Breeding Facility (University of British Columbia, Vancouver BC, Canada). The mice were maintained according to the regulations of the Canadian Council on Animal Care, and all of the animal procedures were approved by University of British Columbia Animal Ethics Committee. The mice used were between 7 and 12 weeks of age. 42

2.16.2 Tumor studies 2.16.2.1 B16F1O mouse melanoma cells over-expressing p97 For each trial, mice (n = 15) were injected subcutaneously in the hindquarter with 51.5x10 B16F1O cells stably transfected with His-p97-pMXpie, p97-pCMV-Tag3, or empty vector controls. Mice were sacrificed 15 days post injection and tumor masses were removed and measured. Three trials were conducted.

2.16.2.2 JBMS mouse melanoma cells with knock-down of p97 expression For each trial, mice (n = 20) were injected with lxi of either JBMS cells or JBMS cells stably transfected with p97-pSUPERIOR.neo+GFP construct. For mice injected with each cell line, half were treated with 1 mg/mi tetracycline (Sigma) plus 5% sucrose (Sigma) and half were treated with only 5% sucrose in their drinking water. Sucrose was added to mask the bitter taste of the tetracycline in drinking water. Drinking water was protected from light and changed every three days. Mice were sacrificed 21 days post injection and tumor masses were measured. Three trials were conducted.

2.17 Statistical analysis Statistical analysis for real-time PCR, melanin assay and tumor weight was performed with GraphPad Prism 5 software for Windows (GraphPad Software, San Diego CA, USA) using one-way ANOVA with Tukey’s post test. The Tukey’s test was used for multiple comparisons to determine the differential effects of the treatment. Error bars represent the standard error of mean. Data were considered significantly different if the p

GraphPad Prism 5 software for Windows (GraphPad Software, San Diego CA, USA) using linear regression test. Error bars represent the standard deviation. 43

Chapter 3: Iron uptake in human melanoma cells mediated by GPI-anchored p97

3.1 Background 3.1.1 Introduction Endocytosis:

Endocytosis is an essential process that the cell utilizes to control the intake of extracellular nutrients and membrane proteins. Classically, endocytosis is divided into two general types based on size: phagocytosis and pinocytosis. Phagocytosis refers to the internalization of particles larger than 200 nm, whereas pinocytosis refers to the internalization of extracellular fluid and small particles less than 200 nm (Liu and Shapiro, 2003). Phagocytosis is often used by single cell organisms to obtain nourishment. However, specialized cells in multi-cellular organisms have also adapted to utilize this mechanism, such as macrophages and neutrophils (Liu and Shapiro, 2003). Tn mammalian cells, pinocytosis is the main mechanism used for cellular import of particles, and can be further divided into clathrin-dependent and clathrin-independent endocytosis. One of the clathrin-independent endocytotic pathways is proposed to be caveolae mediated endocytosis. Caveolae are 50-80 nm flask-shaped, non-coated plasma membrane invaginations (ilconenet al., 2004). They are found in a variety of cell types and tissues (Gabella, 1976; Gil, 1983; Napolitano, 1963; Thomas and Smart, 2008), as well as different cell types possess different number of caveolae on the membrane surface (Guillot et al., 1990; Thomas and Smart, 2008; Thorn et al., 2003), notably abundant in endothelial (Bruns and Palade, 1968) and muscle cells (Ikonen et al., 2004; Ishikawa, 1968). Caveolae are often considered a specialized lipid raft due to their similar lipid composition properties (Razani et al., 2002), which enable them to be highly detergent insoluble with low buoyant density (Ikonen et al., 2004). Evidence has also shown that caveolae contain more cholesterol relative to sphingolipids comparing to other type of detergent resistant 44 cellular membranes (Ikonen et al., 2004; Iwabuchi et al., 1998; Pike, 2002). However, caveolae are distinguished from other lipid rafts by the presence of caveolin, cholesterol- binding protein associated with the filarnentous coat of caveolae (Rothberg et al., 1992). In mammalian cells, there are three caveolin genes, caveolin- 1, caveolin-2, and caveolin

3 (Scherer et aL, 1996; Tang et al., 1996; Way and Parton, 1995). Caveolin-l and -2 are often found co-expressed in adipocytes, endothelial cells, pneumocytes, and fibroblasts (Scherer et al., 1996), however, only caveolin-1 is essential for the formation of caveolae (Parolini et al., 1999; Razani et al., 2002). On the other hand, caveolin-3 expression is limited exclusively to the muscle cells (Song et a!., 1996). Caveolae endocytosis have been shown to be highly mobile, able to be rapidly internalized and cleared from the plasma membrane (Echarri et al., 2007; Oh et al., 2007; Parton and Richards, 2003). It has been proposed that caveolae-mediated endocytosis involves switching from cycles of “kiss-and-run” events with the plasma membrane to microtubule-dependent transport cycles, resulting in internalization of caveolae to destined intracellular organelles

(Pelkmans and Zerial, 2005). Recently, caveolin- 1 has been shown to activate Rab5, a small GTPase that regulates vesicle docking and fusion during endosome formation, via direct binding and resulting in enhanced endocytosis (Hagiwara et al., 2009). Transferrin independent iron uptake: Iron plays many crucial roles in cellular metabolism. There are many advances in the iron field, however the best studied iron uptake is still the clathrin-dependent endocytosis through the classical Tf-TfR system (Figure 1.5). However, there are evidences exist for Tf-TfR independent iron uptake system (Bernstein, 1987; Beutler et al., 2000; Craven et al., 1987; Hamill et al., 1991; Huggenvik et al., 1989; Kennard et al., 1995). One such candidate is GPI-anchored p97 protein. GPI-anchored proteins consist of a large functionally diverse group of proteins that has the ability to attach to the membrane solely dependent on the GPI-anchor (Chatterjee and Mayor, 2001). The synthesis of the GPI-anchor portion of the protein take place mainly in the endoplasmic reticulum (ER) membrane system and is attached to the designated protein in a series of enzymatic events (Chatterjee and Mayor, 2001; Ferguson, 1999; Udenfriend and Kodukula, 1995). The mature GPI-anchor precursor is added to the carboxyl-terminus of the protein during post-translational lipid modification. 45

GPI-anchored proteins have two signal sequences prior to anchor attachment. The amino-terminal signal sequence direct the protein to the ER membrane, whereas the carboxyl-terminal signal sequence responsible for its own cleavage and the attachment of GPI-anchor (Kinoshita et al., 1997; Udenfriend and Kodukula, 1995). Subsequently, the GPI-anchored protein is transported to the cell surface.

3.1.2 Rationale Various GPI-anchored proteins have been studied, however, the precise mechanism involved in the endocytotic pathway of GPI-anchored protein remain a hotly debated subject. One of the controversies revolves around how GPI-anchored proteins are internalized from the surface of the cell. A considerable amount of information about the GPI-anchored proteins came from the studies of GPI-anchored folate receptor (GPI FR). Rothberg et al. found that GPI-anchored folate receptors were clustered in the caveolae vesicles and proposed that GPI-FR are sequestered into the cell via closure of the caveolae vesicles from the plasma membrane (Rothberg et al., 1992; Rothberg et al., 1990). However, this proposal was questioned when it was shown that the localization of GPI-anchored proteins in caveolae vesicles only took place upon cross-linking with secondary antibody (Mayor et al., 1994). In addition, there have been other findings supporting the clustering of GPI-anchored proteins in clathrin-coasted vesicles, such as prion protein (Shyng et al., 1994) and placental alkaline phosphatise (Makiya et al., 1992). Since then there are also increasing evidence supporting the importance of caveolae vesicles in internalization of GPI-anchored proteins such as prion proteins (Harmey et al., 1995; Peters et al., 2003) and urokinase plasminogen activator receptor (Cavallo-Medved et al., 2003; Stahl and Mueller, 1995). Thus, the on-going debate over the caveolae vesicles involvement in the endocytosis of GPI-anchored proteins continues. To determine the role of GPI-anchored p97 with respect to iron uptake, the cellular internalization pathway of GPI-anchored p97 is studied in depth within this chapter. 46

3.2 Results 3.2.1 Binding affinity of soluble p97 to iron In collaboration with Dr. Louise Creagh, the binding ability of soluble p97 to iron was investigated using Isothermal Titration Calorimetry (1TC). Calorimetry has been used extensively to measure the energy involved during the interaction between macromolecules and their ligands (Brandts and Lin, 1990). The TIC is used to measure the binding interactions between the ligand and the macromolecule at a constant temperature. Figure 3.1 shows the thermogram of apo-p97 in the presence of iron at concentrations up to 100 times the p97 concentration. Results demonstrate that the binding of iron to the functioning N-lobe of p97 increases the melting temperature of the N lobe by more than 30 °C, resulting in the appearance of a thermal transition at 82.4 °C and the concomitant disappearance of the thermal transition observed for the apo-N lobe at 52.1 °C. The results show that with iron to p97 binding ratio of 1:1 at pH 7.5 and 25°C, the apparent association constant was found to be 4.4 x 1017M’ in the presence of bicarbonate (Table 3.1). This affinity is found to be intermediate between those reported for N-lobe and C-lobe of Tf by Lin et al. (Lin et al., 1994) (Table 3.1). This data offer support to the notion of p97 may play significant role in Tf-TIR independent iron transport. 47

Table 3.1: Apparent binding constant (Kapp) of Fe to p97 as determined using differential scanning calorimetry in the presence of bicarbonate at 25°C. Data is compared with Lin et al. (1994) results for human Tf performed under the same conditions. This data suggest that the N-lobe of p97 can be just as efficient as Tf in binding iron.

Kapp in the presence of bicarbonate at 25°C

FeNTA to p97 (N-lobe) 4.4 x 7iO’ 1W

FeNTA to N-lobe hTf (Lin et al., 1994) 1.1 x 1022 M’

FeNTA to C-lobe hTf (Lin et al., 1994) 8.0 x 3iO’ 1M 48

300 Fe:7 I - • 250

100 - 05

10 20 30 40 50 60 70 80 90 Temperature, °c

Figure 3.1: Calorimetry thermogram of p97 binding to iron at various ratios, as indicated, in 500 mM Hepes, 25 mlvi 3NaHCO and pH 7.5. This is a representative of three experiments. , 49

3.2.2 GPI-anchored p97 is endocytosed through caveolae dependent pathway 3.2.2.1 Immunofluorescence microscopy studies of the localization of GPI-anchored p97 in SK-MEL 28 human melanoma cells To determine whether GPI-anchored p97 co-localizes with caveolae or clathrin coated pits, we employed a modified internalization assay previously used in the measurement of endocytosis of other membrane receptors (Chu et al., 1997). Briefly, surface TfR or GPI-anchored p97 were pre-labelled with anti-TfR and anti-p97 antibody at 4°C (a temperature that blocks endocytosis). Cells were returned to 37°C for various time periods to allow endocytosis to resume and internalization of the labelled surface proteins was visualized using confocal microscopy. Antibodies against caveolin and clatbrin were used to stain for the respective vesicles. In addition, to prevent the proposed sequestration of GPI-anchored proteins in caveolae triggered by antibody cross- linking, we employed stronger fixing reagents. Mayor et al. proposed that although GPI anchored proteins tend to redistribute under normal fixing conditions, and this redistribution can be prevented by using 0.3% to 0.5% glutaraldehyde along with 3% paraformaldehyde (Mayor et al., 1994). In this thesis, fixing reagent containing 2% glutaraldehyde and 4% paraformaldehyde was used to prevent the redistribution due to antibody cross-linking. Results demonstrate that the co-localization pattern was only observed between TfR and clathrin, as well as between GPI-anchored p97 and caveolin (Figure 3.2), but not vice versa. In addition, majority of the co-localization were targeted to the cell surface at 0 minute. At 20 minutes post-endocytosis, the majority of the GPI anchored p97 and TfR molecules were no longer located on the cell surface. The GPI anchored p97 molecules were observed inside the caveolae vesicles (Figure 3.2a) and TfR molecules were observed inside the clathrin-coated vesicles (Figure 3.2b)

In conjunction, a FITC-labelled Fab fragment of antibody against GPI-anchored p97, which completely eliminates the antibody cross-linking problem, was also used to confirm the co-localization between caveolin and p97. The Fab fragment was made by papain digestion of monoclonal antibody again human p97 according to the method based on Parham et al. (Parham et al., 1982). Papain is a cysteine hydrolase enzyme from 50 papaya, and is known to cleave the Fc portion of the antibody from the antigen binding Fab portion. Since the Fc portion is what binds the secondary antibodies and responsible for the cross-linking effect, its removal would eliminate the proposed redistribution of

GPI-anchored protein in caveolae due to antibody cross-linking. (Mayor Ct al., 1994). The F1TC-labelled Fab fragment was subsequently used to study the localization of GPI anchored p97. Results confirm those from previous experiment using the full antibody (L235), and showed that GPI-anchored p97 co-localize with caveolin defined vesicles, not clathrin (Figure 3.3). 51

Figure 3.2: Localization of TfR and GPI-anchored p97 to either caveolae or clathrin coated pits in human SK-MEL 28 melanoma cells using confocal microscopy. Labeled TfR (OKT9) or GPI-anchored p97 (L235) were visualized with Alexa 488-conjugated goat anti-mouse and labeled caveolae or clathrin vesicles were visualized with Alexa 568-conjugated goat anti-rabbit antibodies. Merged images are shown on the right. Cells were fixed in 4% paraformaldehye and 2% glutaldehye to prevent redistribution of GPI anchored protein trigger by antibody cross-linking. Co-localization is identified by the yellow color. (a) At 0 minute, co-localization between TfR and clathrin was only observed on the cell surface, while 20 minutes after endocytosis, the co-localization pattern is observed throughout the cell. There was no co-localization observed for GPI anchored p97 and clathrin. (b) GPI-anchored p97 show co-localize with caveolae vesicles on the cell surface at 0 minutes and in the interior of the cell at 20 minutes post endocytosis. TfR did not show any co-localization with caveolae vesicles at either 0 or 20 minutes after endocytosis. The scale bar represents 10 rim. Data shown were compiled from at least 20 individual cells from three separate individual preparations. Data shown represent a single plane at mid-section of the cell. Co-localizations were quantified using ImageJ 1.37c. 52

(a)

o miii

20 miii

p97 Clathrun Merged

o mm

20 miii

Transferrin receptor Clathrin Merged

p97iTfR colocalization with clathrin at 20 minutes 15

10• . 5.

I I p97 TfR 53

(b)

o mm

20 mm

Caveolin Merged

0 mm

20 mm

Transferrin receptor Caveolin Merged

p97ITfR colocalization with caveolin at 20 minutes 15 0

10

5

I I - p97 TfR 54

Figure 3.3: GPI-anchored p97 co-localize (pink) with caveolin, but not clathrin in human SK-MEL 28 melanoma cells. GPI-anchored p97 (blue) was labelled with FITC-Fab fragment made from monoclonal antibody against p97 (L235). Caveolae vesicles (red) were identified using caveolin- 1, while clathrin vesicles (red) were identified using antibody against clathrin heavy chain. Cells were stained directly with FITC conjugated anti-human p97 Fab fragment (L235-Fab), for 30 minutes on ice. Endocytosis was allowed to take place for 20 minutes at 37°C, and followed by fixation in 4% paraformaldehyde and 2% glutaldehyde (Sigma) in PBS for 30 minutes at room temperature. Cells were permeabilized and stained for caveolae I clathrin vesicles at room temperature. Both caveolae and clathrin vesicles were visualized with Alexa 568- conjugated goat anti-rabbit antibodies. Merged images are shown on the right. The scale bar represents 10im. Data shown were compiled from at least 20 individual cells from two to three separate individual preparations. Data shown represent a single plane at mid-section of the cell. Co-localizations were quantified using ImageJ 1.37c. 55

clathrin merge

caveolin merge

p97 colocalization with clathr inlcaveolin at 20 minutes

0C

0 0 0 4- C I0 04) caveolin 56

3.2.2.2 Cryo-immunoelectron microscopy on the localization of GPI-anchored p97 in SK-MEL 28 human melanoma cells To further confirm the localization of GPI-anchored p97 in caveolae vesicles, SK

MEL 28 melanoma cells were labelled with 5 nm gold particles conjugated directly to antibodies against either GPI-anchored p97 or TfR and 10 nm gold particles directly conjugated to antibodies against either caveolin-1 or clathrin. SK-MEL 28 cells grew on coverslips were labelled with either anti-p97 or anti-TfR antibodies (5 nm gold conjugate) at 37°C for 30 minutes to allow for internalization. Afterwards the cells were washed and fixed in 4% paraformaldehyde and 2% glutaldehyde followed by treatment with 0.1% saponin in PBS. The cells were then labelled with either anti-caveolin or anti-clathrin antibodies (10 nm gold conjugate). After further washing cells were fixed with phosphate buffered osmium tetroxide. The cells were then removed from the coverslips and processed and sectioned (80 nm) for TEM. After 30 minutes, vesicles were examined for the presence of 5 and 10 nm gold particles corresponding to either GPI-anchored p97 or TfR and either caveolin or clathrin. The number of vesicles showing both 5 and 10 nm particles as a proportion of the total number of vesicles showing either GPI-anchored p97 or TfR specific labelling was quantified. The results show GPI-anchored p97 localized in the same vesicle as caveolin- 1(Figure 3.4a) and TfR localized with clathrin (Figure 3.4b). The z statistic comparing two proportions was calculated to determine the statistical significance of the prevalence of co-localization. There was a significantly greater prevalence of caveolin rather than clathrin in p97 labelled vesicles (p

Figure 3.4: GPI-anchored p97 localize in the same vesicle as caveolin- 1, while TfR are localized in the clathrin coated vesicles in human SK-MEL 28 melanoma cells. (a) Cryo immunoelectron micrograph demonstrating the co-localization of caveolin- 1 and GPI anchored p97 within the same vesicle (arrow). (b) Immunoelectron micrograph demonstrating the co-localization of clathrin and transferrin receptor. Caveolin- 1and Clathrin are identified by the lOnm gold particles whereas GPI-anchored p97 and TfR are directly conjugated with 5nm gold particles. The sections were viewed on a Hitachi H7600 ThM and digital pictures taken using an AMT camera built into the H7600. (C) The statistical prevalence of p97 co-localization with caveolae vesicles. ___

58 (a)

(b)

* S a • . •

SO. 4.

lOOrvW

(C)

60

k 50

40

II30 Ct V 20

C. 10

CbthnnltfR CaveohnlTfR CfathrnIp97 Caveolrnip97 59

3.2.3 GPI-anchored p97 internalize iron utilizing an endosomal pathway in SK-MEL 28 melanoma cells 3.2.3.1 GPI-anchored p97 internalize to endosomes, but trafficks more slowly than TfR The fate of GPI-anchored proteins after they are internalized is currently under debate (Anderson, 1994; Anderson, 1998). To determine whether vesicles carrying GPI anchored p97 mature into endosomes, SK-MEL 28 melanoma cells were triple labelled for TfR, p97 and the early endosome marker, early endosome antigen-i (EEA-1). The SK-MEL 28 cells were starved for serum Tf before they were stained for Alexa-488 conjugated Tf, anti-p97 (L235) and anti-EEA1 antibodies. The cells were incubated at 37°C for various time periods before they were fixed and labelled with secondary antibodies at 4°C (Figure 3.5a). At 0 minute, both p97 and TfR were present at the cell surface of the SK-MEL 28 cells, and EEA-1 exhibited a typical punctuate endosomal staining. After 10 minutes, most of the TfR were endocytosed into the cells and co localized with endosomes, whereas most of the GPI-anchored p97 molecules were still present on the cell surface. A small number of GPI-anchored p97 molecules were seen in the endosomes containing TfR, represented by the white punctuate staining when the three immunofluorescence dyes Alexa-488, Alexa-568 and Cy5 co-localized together (Figure 3.5b). This implied that majority of the GPI-anchored p97 traffic more slowly to the endosomes as compared to the TfR. 20 minutes after endocytosis, the majority of TfRs were not present in the endosomes and there is an increase in the number of p97 molecules associated with endosomes as shown by the pink color (when Alexa-568 co localizes with Cy5). After 30 minutes, there were a significant number of endosomal vesicles containing p97 (Figure 3.5b). The results show that both GPI-anchored p97 and TfR trafficks to the endosomal compartments, however, TfR enters the endosomal compartments at earlier time points compared to GPI-anchored p97. 60

Figure 3.5: Localization of TfR and GPI-anchored p97 to early endosomes (EEA1) in human SK-MEL 28 melanoma cells. Cells were labeled with Alexa 488-conjugated Tf (green) and antibodies against GPI-anchored p97 at 4°C before the initiation of internalization for various periods of time at 37°C. The cells were fixed, permeabilized before stained for antibodies against EEA1. (a) Labeled GPI-anchored p97 were visualized with Cy5-conjugated rabbit anti-mouse (blue) and labeled EEA1 were visualized with Alexa 568-conjugated rabbit anti-goat (red) antibodies respectively. Merged images are shown on the right. Co-localization of GPI-anchored p97 and early endosomes is indicated by the pink color in the merged panels. The co-localization of TfR and early endosomes is visualized as yellow color in the merged panels. Co localization of TfR, GPI-anchored p97 and early endosomes is visualized as white color in the merged panels. (b) The enlarged images of cells from the 10 minute and 30 minute time point merged panels. The arrows point to the co-localization of TfR, GPI-anchored p97 and early endosomes. The scale bar represents 10 tm. Data shown represent a single plane at mid-section of the cell. Co-localizations were quantified using Imagel 1.37c. 61 (a)

p97ITfR colocalization with EEA-1 at various time points

5 1Omn . 20mm . 30mm 8 62

(b) 63

3.2.3.2 Endosomes are essential in the internalization of GPI anchored p97

To further confirm the involvement of endosomal pathway in internalization of GPI-anchored p97, studies were conducted utilizing endosome disruption through dominant negative Rab5. Rab5 is a small GTPase that regulates the fusion between endocytic vesicles and early endosome, as well as the homotypic fusion among the early endosomes (Olkkonen and Stenmark, 1997). SK-MEL 28 melanoma cells were transfected with pEGFP-S34N rab5 construct and stable clones were isolated. The S34N mutation in Rab5 is a dominant negative mutation that disrupts the GTPase function, thus render the endosomal pathway non-functional. The SK-MEL 28 cells were stained for anti-TfR (OKT9), anti-p97 (L235) and anti-EEA1 antibodies. The cells were incubated at 37°C for 20 minutes before they were fixed and labelled with secondary antibodies at 4°C. The confocal results show that the dominant negative rab5 restricted the localization of GPI-anchored p97 to the periphery of the cell with no significant co-localization with

EEA- 1 inside the cells, whereas in the cells with functioning rab5, there is a greater co localization between EEA- 1 and p97 inside the cell (Figure 3.6b). Similar result was also observed for TfR in these cells (Figure 3.6a). These data suggest that functioning endosomes are essential in the intracellular trafficking of GPI-anchored p97 in melanoma cells.

3.2.3.3 Endosomes are essential in the trafficking of iron to ferritin mediated by GPI-anchored p97 in SK-MEL 28 melanoma cells

To demonstrate the importance of endosomal pathway, SK-Mel 28 melanoma cells transfected with dominant negative rab5 mutation were used to investigate the trafficking of iron within the cells. The S34N rab5 transfected SK-MEL 28 cells were incubated with either 55Fe-NTA or 55Fe-Tf and the cell lysates were collected. Ferritin was immunoprecipitated and the radioactivity associated with each protein was determined. To differentiate the iron uptake between Tf-TfR mediated and GPI-anchored p97 mediated pathways, different media type and iron complex were used. To evaluate iron uptake through solely GPI-anchored p9’7, IMDM medium was used to eliminate 64 serum Tf which effectively prevented any iron uptake through the Tf-TfR pathway. On the other hand, normal DMEM medium and 55Fe bound to Tf were used to evaluate the iron uptake through Tf-TfR pathway. Furthermore, cells were also treated with P1-PLC to cleave off the GPI-anchored p97 from cell surface, thus ensuring that iron is only uptake through Tf-TfR pathway. Figure 3.7 shows that very little radioactive iron associated with the ferritin in the cells with disrupted endosomes. These cells were not able to mediate the transport of iron from either the Tf/TfR or the GPI-anchored p97 pathways compare to the vector control. The results show that disruption of endosomes effectively prevents the trafficking of iron to ferritin, and thus confirming that endosomes are required for the iron transport mediated by GPI-anchored p97 to ferritin. 65

Figure 3.6: Endosomes are essential for GPI-anchored p97 internalization in human SK MEL 28 melanoma cells. SK-MEL 28 cells were transfected with either vector control or S34N rab5 dominant negative vector. GPI-anchored p97 was labeled with monoclonal antibody against p97 (L235), TfR was labeled with OKT9 antibody and early endosomes were identified using EEA- 1. (a) Transfectants showing co-localization between TfR and

EEA- 1. (b) Transfectants showing co-localization between p97 and EEA- 1. Merged images are shown on the right. The scale bar represents lOm. Data shown were compiled from at least 20 individual cells from two to three separate individual preparations. Data shown represent a single plane at mid-section of the cell. Co localizations were quantified using JmageJ 1.37c. 66

(a)

S34N rab5

EEA-1 TfR Merge

TfR colocalization with EEA-1

0

0 0 0 •1-’ C C) 0) 0. ::::::::::::::•:•:•:•:•::•:•:•:•::: vector control S34N rab5 67

(b) Vector control

S34N rab5

EEA-1 p97 Merge

p97 colocahzation with EEA-1 8 0

Cu 0C.) 0 C.)

I. Ivector control S34N rab5 ______

68

Vector alone S34N Rab5

400 350 I 300 250 2O0 150 100

0 I 1.1 NRS Ferritin(TfR) Ferritin(p97)

Figure 3.7: Trafficking of iron to ferritin mediated by Tf-TfR or GPI-anchored P97 in SK-MEL 28 melanoma cell. Melanoma cells transfected with dominant negative rab5 were pre-starved for iron for one hour and incubated with 1iM of 55FeNTA, 55Fe-Tf or NRS (negative control) for 90 minutes. The lysed cells were immunoprecipitated for ferritin, and the associated radioactivity was determined. Results indicate that the expression of S34N Rab5 inhibited the uptake of iron by either p97 or TfR to ferritin, thus confirming that endosomes are essential for the uptake of iron by GPI-anchored p97. Each set of columns in the figure represent an individual experiment. 69

3.3 Discussion

if Tf-TfR is an efficient pathway for iron to transport into cells, why do organisms need an alternate pathway for iron uptake? While there are many examples of specific uptake of non-Tf bound iron (NTBI) into cell types (Wessling-Resnick, 2000), little is known about the molecules involved in these process. In addition, these alternate iron uptake processes are energy dependent and are not regulated by the level of iron in the cell (Gutierrez et al., 1998). In the case for a rare genetic disorder, atransferrinemia, where patients possess no plasma transferrin or at extremely low concentrations, although these patients lack plasma transferrin, there are still evidence of absorption of dietary iron (non-transferrin bound form) and transport to organs such as heart, liver, kidney and pancreas (Beutler et al., 2000; Hamill et al., 1991; Huggenvik et al., 1989). However, these non-transferrin bound iron cannot cure the disorder due to the inability of these iron to be used for erythropoiesis (de Silva et al., 1996). Furthermore, hypotransferrinemia mice show dramatic iron overload in all but hematopoietic tissues despite lower serum Tf levels (Bernstein, 1987; Craven et al., 1987). The results here show that iron internalized via GPI-anchored p97 is one of the mechanisms that the cell employs to internalize NTBI. This pathway delivers iron to the cells independent of Tf. Due to the high expression of GPI-anchored p97 in melanoma cells, it is also possible that GPI-anchored p97 act as an additional iron uptake system in providing the essential iron for rapid cell growth / proliferation during melanogenesis. In collaboration with Dr. Louise Creagh, calorimetric study results show that the binding affinity of p97 to iron is intermediate between the high affinity and low affinity lobes of Tf (Creagh et al., 2005) (Table 3). In addition, it was recently demonstrated that upon binding of Tf to TfR, a conformation change occurs in Tf that hinders the release of iron from N-lobe of Tf suggesting perhaps only one atom of iron is released within the cell during each round of internalization (Cheng et al., 2004). Given that mammalian cells in culture may express approximately l.Ox106molecules of p97/cell (Kennard et al., 1995), the efficiency of iron uptake via p97 may thus rival that of the classical Tf-TfR pathway. In the immunofluorescence experiments, results demonstrated that GPI-anchored p97 enters the SK-MEL 28 cell through caveolae vesicles (Figure 3.2b), not clathrin coated vesicles. There are studies suggesting that GPI-anchored proteins by nature do not 70 constitutively concentrate in caveolae, however, use of polyclonal secondary antibody triggers a cross-linking event that redistribute GPI-anchored proteins into caveolae (Mayor et al., 1994). In the same study, Mayor et al. stated that the redistribution of the GPI-anchored proteins could be prevented by using stronger fixing reagents, namely, 0.3 to 0.5% glutaraldehyde along with 3% paraformaldehyde (Mayor et al., 1994). To solve this problem, we employed much stronger fixing reagents than suggested by Mayor et al., as well as using fluorescent labeled Fab fragments made from monoclonal antibody against p97 to eliminate the cross-linking. Furthermore, hnmunoelectron microscopy experiments were conducted using direct gold conjugated antibodies. The results show that GPI-anchored p97 co-localized with caveolin in the same vesicular compartment, and the various techniques used in the experiments strongly confirm that the localization is a real phenomenon (Figure 3.2-3.4). The internalization of GPI-anchored p97 and TfR to the endosomal compartment is demonstrated through a series of time course experiments. Immunofluorescence microscopy results show that P97 trafficks through the endosomal compartment slower than the Tf-TfR complex. Specifically, most of the Tf moved into endosomes after 10 minutes at 37CC,whereas the bulk of GPI-anchored p97 trafficked to the endosomes after 20 minutes (Figure 3.5). In addition, the endosome disruption through dominant negative Rab5 prevented internalization of GPI-anchored p97 (Figure 3.6) and led to a decrease in iron load (Figure 3.7). These data suggests the fusion and maturation of endosomal vesicles are required for the transport of iron by GPI-anchored p97 to the iron storage protein. Recently, the function of p97 as an iron transporter has been questioned because p97 does not have an iron regulatory element (IRE) (Richardson, 2000). Therefore, its expression is likely not regulated by the level of intracellular iron, similar to the transferrin receptor 2 and Tf itself, which do not possess IRE (Kawabata et al., 2000). These data clearly indicate that p97 is able to uptake iron and donate to ferritin in melanoma cells and support the proposal that iron bound to GPI-anchored p97 is internalized through a caveolae dependent endosomal pathway as illustrated in Figure 3.8. Furthermore, these results establish GPI-anchored p97 as an alternative iron uptake mechanism in melanoma cells. 0 Apo-Tf Fe3 Basal-lateral surface GPI-p97 F -TI - if TIR Apical surface 0 f DM11 Clathrin 40 + Caveolin+ — DMT1 Clatflrin coated + pit + Late endosome

Caveolae + + Ar /!

(. Early endosome Early endosome Ferntin $‘% Late endosome

Figure 3.8: The proposed model of iron uptake mediated by GPI-anchored p97. GPI-anchored p97 bound with ferric iron is internalized via caveolae vesicles (red arrows). As a comparison, T mediated iron-Tf uptake is shown. HoloTf bound to TfR is endocytosed through clathrin-coated pits (yellow arrows). The caveolae and the clathrin-coated vesicles both proceed to the endosomal pathway where iron is released from either GPI-anchored p97 or Tf due to the low acidity within the endosome. Iron then exits the endosome and incorporates into ferritin in the cytoplasm for storage. 72

Chapter 4: The role of GPI-anchored p97 expression in melanoma malignancy

4.1 Rationale P97 is highly expressed in neoplastic cells, especially in malignant melanoma cells, while mostly non-detectable in normal tissues (Brown et al., 198ib; Woodbury et aL, 1980). Subsequent studies have found it to be expressed at various levels in the liver, intestine, sweat gland, placenta, umbilical cord and more recently in human brain endothelium (Alemany et al., 1993; Barresi and Tuccari, 1994; Rothenberger et al., 1996; Sciot et al., 1989). Being a member of the Tf family of proteins, GPI-anchored p97 possess the ability to bind iron. In fact, its apparent binding affinity is intermediate between the high and low affinity lobes of Tf (Creagh et al., 2005). As investigated in chapter 3 of this thesis, GPI-anchored p97 is also capable of effectively uptake iron into the melanoma cell through a Tf-TfR independent endosomal pathway involving caveolae. Taken together with the fact that melanoma cells express the highest levels of GPI anchored p97, it may be hypothesized that GPI-anchored p97 aid the rapid proliferation in these cells by offering an additional mechanism of iron uptake. However, other studies have suggested that p97 does not play any essential role in iron metabolism (Dunn et al., 2006). Phenotypic characterization of p97 knock-out mice showed no apparent effect on the growth, development, behavior or metabolism of the mice comparing to the wild-type littermates (Dunn et aL, 2006). It should be noted that these mice are not available to the general scientific community and indeed have apparently been ‘retired’ despite their key importance in more fully understanding the role of p97. In the same study, it was proposed that the down-regulation of p97 in human melanoma cells impaired cellular proliferation in vitro, and the subcutaneous injection of these cells in immunodeficient nude mice resulted in depressed tumor growth (Dunn et al., 2006). To further understand the role of p97 in melanoma, modulation of p97 expression in mouse melanoma cells were examined through either over-expression or suppression. The effects of these on tumourigenesis in vivo were also examined in this chapter. 73

4.2 Results 4.2.1 Selection of mouse melanoma cells for over-expression and down-regulation of GPI-anchored p97 cell models GPI-anchored p97 is expressed in various type of tissues ranging from neonatal tissues, such as melanocytes and liver, to normal adult tissues, such as salivary glands, sweet glands, brain endothelial cells and smooth muscle (Smith et al., 2006). Normal adult cells display the lowest surface expression of GPI-anchored p97 at about <4,000 molecules per cell, while the highest surface expression is found in smooth muscle at about 8,000 molecules per cell (Woodbury et al., 1981). However, these are in no way comparable to the level displayed in most melanoma cells. It has been reported that most melanoma cell expresses between 50,000 to 500,000 molecules of GPI-anchored p97 on their cell surface (Woodbury et al., 1981), and SKMEL-28 human melanoma cell has been calculated to express at least 400,000 molecules of GPI-anchored p97 on its cell surface (Brown et al., 1981a). This significant up-regulation of GPI-anchored p97 expression in melanoma cells implies possible important role in melanoma formation and development. In order to study the relevance of GPI-anchored p97 in melanoma in both in vitro and in vivo setting, two mouse melanoma cell lines were selected to be used in the expression modification experiments. B16F1O mouse melanoma cell line is the subline derived from the parental B16 melanoma that arose spontaneously in C57BL/6 mouse in 1954 (Berkelhammer et al., 1982). B16F1Osubline was created by selecting the parental B16 line for their ability to form lung colonies in vivo and subsequent 10 cycles of lung colony formation in vitro (Poste et al., 1982). B16F1O cells are known for its efficiency in metastasizing around the subcutaneous (S.C.) injection site (Fidler, 1978; Poste et al., 1982; Poste et al., 1980). JBMS is another mouse melanoma cell line, although unlike B16F1O, it did not arise spontaneously in mice. Since spontaneous development of melanoma in mice is quiet rare (Berkelhammer et al., 1982), studies have been done to create induced melanoma models to facilitate the understanding of the biological, immunological and biochemical parameters that affect the development of malignant melanomas. One of these models is JBMS mouse melanoma induced by single application of 7, 12-dimethylbenz(a)anthracene to the scapular region of four day old C57BL/6 mice (Berkelhammer et al., 1982). These melanomas maintained their 74 melanotic appearance after transplantation to normal C57BL/6 mice via s.c. and metastasized spontaneously in the transplant recipients (Berkelhammer et al., 1982). In the last 20 years, JBMS melanoma cell line has proven to be an excellent comparative system for the B16 melanoma. With respect to the GPI-anchored p97 expression in these two cell lines, mRNA level was measured using real-time PCR (Figure 4.1). Results show that although both cell lines expression GPI-anchored p9’7,JBMS mouse melanoma cells express a significantly higher level (- 16 fold) of endogenous GPI-anchored p97 than B 16F10. Therefore, B16F10 melanoma cell line was chosen for the GPI-anchored p97 over-expression experiment, where the effect of GPI-anchored p97 up-regulation in B16F1O cells both in vitro and in vivo will be investigated. On the other hand, since JBMS mouse melanoma cells expression relatively high level of endogenous GPI anchored p9’7, it will be used in the study of the in vitro and in vivo effect of GPI anchored p97 down-regulation. _____

75

25’ V II.0 g20. 15 0) 0. ia ;C)

Cu

C.) 5.

•E — B16 JBMS

p

Figure 4.1: GPI-anchored p97 mRNA expression level in two well known melanoma cell lines were evaluated using real-time RT-PCR. The JBMS mouse melanoma cell line was found to express p97 at —16fold higher level than the B16F1O mouse melanoma cell line.

Significance of the data was analyzed using one-way ANOVA with Tukey’s post test. Error bar represent the standard error of mean. 76

4.2.2 GPI-anchored p97 in over-expression B16F1O mouse melanoma cell models Due to the lack of an available antibody to mouse p97, study of GPI-anchored p97 in the mouse system has proven to be a challenge. To overcome this challenge, a 6X His tag was added to GPI-anchored mouse p97. Briefly, GPI-anchor p97 protein possesses two signal peptides. An amino-terminal that directs the translocation of the nascent chain of p97 into the ER, and the carboxyl-terminal signal peptide that is responsible for the attachment of the GPI-anchor. Since both signal peptides will be cleaved in the GPI anchor attachment process, it is not possible to place a tag on either ends of p97. To solve this problem, the amino-terminal peptide was identified and a 6X His tag was inserted between the signal peptide and the rest of the p97 sequence. The resulting His tagged mouse p97 was subcloned into pMX-pie vector (pMX-pie-His-p97). As a control, a non-tagged p97 expression vector (pCMV-Tag3-p97) was also constructed to ensure that the His tag is not affecting the function of the resulting tagged protein. B16F10 cells were stably transfected with the pMX-pie-His-p97 (Bi6hisp97) and pCMV-Tag3-p97

(Bi6p9’7). As a control, B16F10 cells were transfected with vector alone. The antibiotic- resistant clones were tested for mouse GPI-anchored p97 expression by real-time PCR and Western blotting (Figure 4.2). Real-time PCR results demonstrated that both Bl6hisp97 and B16p97 transfectants had relatively similar level of up-regulation in GPI anchored p97 at mRNA level, approximately 180 fold higher (p

(a)

f1\ — O) 3 5 300

0 U,

200

I

a, 100 (U

C.) a) >

CU a) •0 I 6’

1p

VC Bl6hisp97 his control

Figure 4.2: GPI-anchored p97 is over-expressed at both mRNA and protein level. B16F1O mouse melanoma cells were transfected with pMX-pie-His-p97 (Bl6hisp97), pCMV-tag3-p97 (B16p97) constructs and vector control for each (pMX-pie and pCMV tag3, respectively). Stable transfectants were analyzed for their GPI-anchored p97 mRNA expression using real-time PCR (a). Bl6hisp97 and pMX-pie vector control stable transfectants was used to analyze the protein expression of GPI-anchored p97 by western blotting for the histidine tag (b). Significance of the data was analyzed using

one-way ANOVA with Tukey’s post test. Error bar represent the standard error of mean. 79

I’

P

Figure 4.3: fron binding ability of GPI-anchored p97 over-expressing transfectants. GPI anchored p97 over-expressing transfectants (Bl6hisp97 and B16p97) and untransfected B16F1O cells were incubated with 55Fe-nitrilotriacetate 55Fe-NTA) at 4°C. The radioactivity count associated with the cells was measured. (The error bars represent the standard deviations of three separate experiments. Significance of the data was analyzed using one-way ANOVA with Tukey’s post test. Error bar represent the standard error of mean. 80

4.2.3 Characterization of GPI-anchored p97 over-expression in mouse melanoma cells in vitro and in vivo

4.2.3.1 Over-expression of GPI-anchored p97 in B16F1Omouse melanoma cells promote cell proliferation In order to determine the cellular role of GPI-anchored p97 in melanoma cells, the effect of p97 over-expression on cellular growth of melanoma cells was investigated. The same number of Bl6Hisp97, B16p97, and vector control (pMX-pie and pCMV-tag3, respectively) stable transfectants were seeded in 96 well culture plates and grown for 24 hours. On day 2, cell proliferation was measured using alamarBlue® dye reduction assay (Invitrogen). During proliferation, the intracellular environment undergoes reduction due to increase in the ratios of NADP/NADPH, FAD/FADH, FMNIFMNH and NADINADH (Akahoshi et al., 2006). The non-toxic alamarBlue® dye is able to be reduced by these metabolic intermediates from a non-fluorescent indicator dye resazurin into a highly red fluorescent resorufin, and is thus a useful tool in quantitating cell proliferation (Akahoshi et al., 2006). Cell proliferation was calculated as fold of change compared to the value at t=lhour for each cell line. The results show that during the seven hours incubation and growth period, GPI-anchored p97 over-expressing B16F1O cells (B16h1sp97 and B16p97) proliferated significantly faster than the vector control transfected B16F10 cells, pMX-pie and pCMV-tag3, respectively (Figure 4.4). There was no difference observed between the proliferation of B16F10 cells transfected with either pMX-pie or pCMV-tag3 vector control. The B16hisp97 proliferated slightly more slowly than the Bl6p97 transfectants, however, the difference is not significant. On the other hand, both GPI-anchored p97 over-expressing B16F1Otransfectants (Bl6hisp97 and B16p97) show significantly higher proliferation compare to the vector controls (pMX-pie and pCMV-tag3, respectively).

These results show that up-regulation of GPI-anchored p97 in B16 melanoma cells promotes cell growth I proliferation, likely associated with the iron uptake ability of GPI anchored p97. 81

10• -.— pMX-pie -.- pCMV-Tag3 5 —.— B16p97 Bl6his-p97

I 6

4-

r2-

U- I I I I I 0 1 2 3 4 5 6 7 8 Time (hour)

P

Figure 4.4: The effect of GPI-anchored p97 over-expression on cellular proliferation of B16F1O mouse melanoma cells. Cellular proliferation of Bl6hisp97, B16p97 and vector control stable transfectants were assessed by alamarBlue® dye reduction assay. During the 7 hour period, both GPI-anchored p97 over-expression transfectants had significantly higher proliferation when compared to the vector control cells. B16p97 transfectants proliferated at a slightly higher rate compare to the B16hisp97 transfectants, however the difference is not significant. Results are expressed as mean ± SEM. Data were analyzed using linear regression test. 82

4.2.3.2 Over-expression of GPI-anchored p97 promotes melanin production and secretion in B16F1Omouse melanoma cells The over-expression of GPI-anchored p97 in B16F1O mouse melanoma cells lead to no obvious phenotypic changes, with an exception of melanin production. Melanin is a pigmented polymer synthesized in melanocytes and stored in specialized endosome-like organelles, melanosomes (Raposo and Marks, 2002; Sarangarajan and Apte, 2006). Melanin plays various important protective roles, such as protecting tissues from UV damage and acting as an antioxidant by binding to metal ions and other toxic chemical substances (Hu, 2008). Melanoma arise due to the uncontrolled proliferation of melanocytes, thus is often characterized by melanin production. It was observed during the initial culturing stage of Bl6hisp97 and B16p97 transfectants that the culturing media often turn from bright red to black color. Since B16 melanoma cells are known to produce and secrete melanin, the effect of GPI-anchored p97 up-regulation on melanin production and secretion was investigated. Equal number of cells from Bl6hisp97, B16p97 and B16 vector control transfectants were seeded in separate culture dishes. Four days after initial plating, cells were harvested and collected.

5x1 cells from each cell line were pelleted and solubilized in 1 M NaOH heated to 100°C for 10 minutes. Melanin content was determined with a SpectraMax® 190 spectrophotometer at 400 nm, and calculated from a standard curve using synthetic melanin (Sigma). Figure 4.5 show that the expression of GPI-anchored p97 in either over-expression cell lines (B16hisp97 and Bl6p97) correlated with a significant increase in total cellular melanin content (Figure 4.5a). Furthermore, B 16hisp97 and Bl6p97 cell lines also show significant secretion of melanin into the culture media after four days in comparison to B16 vector controls (pMX-pie and pCMV-tag3 respectively) (Figure 4.5b). After 4 days of culturing, neither the GPI-anchored p97 over-expressing cells nor the vector control cells show any significant difference in cell death, thus the melanin in the culture media is attributed to cellular secretion and not a result of release after cell death. Taken together, these results indicate that GPI-anchored p97 might be connected to the melanin synthesis in B16F1Omelanoma cells. 83

Figure 4.5: The effect of GPI-anchored p97 over-expression on melanin production and secretion. At the in vitro level, there are no obvious morphological differences between the GPI-anchored p97 over-expressing transfectants (Bl6bisp97 and B16p97) compared to the vector control cells. However, dramatic increases in the number of melanosomes present on Bl6hisp97 and B16p97 transfectants were observed. Melanin assays were performed on these cells as well as their culture supernatant to evaluate melanin production within the cell (a) and secretion after 4 days of culturing (b) by the B16hisp97, B16p97 and vector controls. Bl6hisp97 and B16p97 produced significant higher level of cellular melanin comparing to the vector controls. This was also true for melanin secreted by the cells. Both B16hp97 and B16p97 transfectants secreted significant levels of melanin, while the vector control transfectants did not show any detectable melanin secretion after 4 days of culture. Results in (a) and (b) are normalized to cell number. Significance difference between the over-expression and vector control was analyzed using one-way ANOVA with Tukey’s post test. Error bar represent the standard error of mean. melanin secretion (jig/i 06 cells) 4 melanin content (pg/b6 cells) 4 s,. 4-

4 ‘4 4

-v -v A A p 0 0 b 0 0 I-

00 85

4.2.3.3 Over-expression of GPI-anchored p97 in B16F1O mouse melanoma cells promote tumor growth in vivo Since the in vitro data from previous section show GPI-anchored p97 up-

regulation promote cell proliferation and melanin production and secretion in B16F10 cells, its effect in an in vivo setting was evaluated next. In order to see whether the up- regulation of GPI-anchored p97 has any relevance to melanoma tumor formation and development in vivo, cells from each of the GPI-anchored p97 over-expression cell line (Bl6hisp97 and B16p97) and vector control (pMX-pie and pCMV-tag3) were injected

subcutaneously into C57BL16 mice. Tumor mass was harvested and measured 15 days post-injection. Figure 4.6 shows that tumors arose from the Bl6hisp97 and B16p97 transfectants were significantly larger than those from vector control transfectants. There was no difference in size between the tumors arose from Bl6hisp97 vs. B16p97. This is reasonable since there was no significant difference between these two GPI-anchored p97 over-expressing cell lines in terms of mRNA level. In addition, although Bl6hisp97 proliferated at a slightly slower rate than B16p97, however it was not statistically significant enough to cause any major difference between the two cell lines. There was also no difference in tumor size observed between the two vector controls, implying that the vector backbone is not contributing to the increase in the tumor size. Taken together, these results indicate that the presence of excess GPI-anchored p97 alone is able to promote B16 melanoma tumor growth, perhaps by providing extra iron supply to aid the rapidly dividing tumor cells. ______

86

2.0-

a —— — 0) aI —— 0) == 10- 05-

0.0-- — — — A 4JLP c .3- 6’

p

Figure 4.6: GPI-anchored p97 over-expression promotes the growth of subcutaneous B16F1O melanoma tumors. C57/BL6 mice (5 mice per group) were injected subcutaneously with 5l.5x10 cells (Bl6hisp97, Bl6p97 or vector controls) in the hindquarter. Tumor mass were measured after 15 days. One way ANOVA with Tukey’s post test showed that both B16hisp97 and Bl6p97 transfectants with increase GPI anchored p97 expression promoted tumor growth. This experiment is a representative of three separate experiments. 87

4.2.4 Characterization of GPI-anchored p97 down-regulation in mouse melanoma cells in vitro and in vivo Complimentary to the experiments in section 4.2.2, a model of GPI-anchored p97 down-regulation in melanoma cells was also established. Since results from the previous sections demonstrate that GPI-anchored p97 alone is able to promote cell proliferation, the goal of this section is to see if down-regulation of endogenous GPI-anchored p97 could cause an opposite effect, reduce cell proliferation. Melanin production and secretion was not investigated for JBMS mouse melanoma cells, because these cells cultured under the culture media used does not show any significant measureable cellular melanin production or secretion.

4.2.4.1 Down-regulation of GPI-anchored p97 decreases cell proliferation in JBMS mouse melanoma cells Due to the relatively high endogenous level of GPI-anchored p97 compare to B16F1O cells (Figure 4.1), JBMS mouse melanoma cell line was chosen for the study of GPI-anchored p97 down-regulation in melanoma cells. The down-regulation of GPI anchored p97 was achieved through the use of tetracycline inducible siRNA vector from TMOligoengine Cells were stably transfected with tetracycline inducible pSUPERIOR.neo.GFP-p97 siRNA vector. Single cell clones were isolated based on the highest. GFP expression and GPI-anchored p97 down-regulation in the presence of 1 ig/ml tetracycline. The clone with the greatest relative suppression of endogenous GPI anchored p97 expression was selected and used for subsequent studies. As mentioned earlier, due to the lack of antibody for mouse p97, the expression level was evaluated solely based on the mRNA expression. The real-time PCR result (Figure 4.7) show that cells derived from the chosen clone yield a relative 50 fold suppression of GPI-anchored p97 expression when treated with 1 pg/rnl tetracycline comparing to no tetracycline treatment. In addition, there was no difference in the expression of GPI-anchored p97 expression observed between the untransfected JBMS cells grown in presence of tetracycline and JBMS-pSUPERIOR.neo.GFP-p97 grown in absence of tetracycline. This implies that tetracycline at concentration used does not affect the expression of p97. ______

88

20

•0 2 JBMS-Psuperior-no tet JBMS-tet JBMS-Psuperior-tet

20• p I —

o -40 a) >

-60

P

Figure 4.7: GPI-anchored p97 is down-regulated in pSUPERIOR.neo.GFP-p97 transfected JBMS cells. JBMS mouse melanoma cells were transfected with tetracycline inducible pSUPEROIOR.neo.GFP-p97. Untransfected JBMS cells and transfected cells grown in absence of tetracycline were used as controls. Stable transfectants were analyzed for their GFP expression, and the cells with highest GFP expression were subjected to single cell cloning. Each clone was grown in the presence of 1 ig/ml tetracycline, and the amount of down-regulation of endogenous GPI-anchored p97 was analyzed using real-time PCR. Significance difference between the down-regulation and vector control was analyzed using one-way ANOVA with Tukey’s post test. Error bar represent the standard error of mean. 89

4.2.4.2 Down-regulation of GPI-anchored p97 decrease cell proliferation in JBMS mouse melanoma cells To examine the role of GPI-anchored p97 in melanoma cells, its expression was down-regulated via siRNA as mentioned previously. The clone that elicited -5O fold down-regulation of endogenous GPI-anchored p97 was expanded and grown in the presence of 1 ig/m1 tetracycline to activate the siRNA knock-down effect. As a control, the clone was also grown in the absence of tetracycline to act as a vector control. Since siRNA knock-down effect is only induced by tetracycline, the transfected JBMS cells grown in the absence of tetracycline should maintain normal level of endogenous GPI anchored p97 expression. Untransfected JBMS cells grown in presence of tetracycline were also used as control to ensure that tetracycline itself is not affecting proliferation of cells. Equal number of cells from each of the cell lines was grown in 96 well plates for 24 hours. On Day 2, alamarBlue® dye was added and cell proliferation was measured at each hour for 7 hours. Results show that the JBMS transfectants with induced knock down effect by siRNA proliferated at a significantly slower rate than the control cells (transfectants grow in the absence of tetracycline, and untransfected JBMS cells) (Fig

4.8). Tetracycline at the concentration of 1 tg/m1 shows no apparent effect on the proliferation of the cells, as there is no difference between untransfected JBMS cells grown in the presence of tetracycline and JBMS transfectants grown in the absence of tetracycline. In addition, it should be noted that there is no growth difference observed in untransfected JBMS cells grown in presence or absence of tetracycline. The result from this experiment demonstrate that decrease in endogenous GPI-anchored p97 expression can lead to decrease in cell proliferation, thus suggesting that the expression of GPI anchored p97 alone is able to significantly affect the proliferation of JBMS mouse melanoma cells, possibly related to the iron uptake ability of GPI-anchored p97 in melanoma cells. 90

10• —— JBMS(1ig/mItet) —— JBMS-pSUPERIOR-p97(no tet) 8 —— JBMS-pSUPERIOR-p97(1 jtg/mI tet) 5 I 6 . 4.

2 I a’ I I 0 2 4 6 8 Time (hour)

P

Figure 4.8: The effect of GPI-anchored p97 down-regulation on cellular proliferation of JBMS mouse melanoma cells. Cellular proliferation of JBMS-pSUPERIOR-p97 transfectants grown in the presence of tetracycline and controls (untransfected JBMS cells grown in presence of tetracycline and JBMS-pSIJPERIOR-p97 transfectants grown in absence of tetracycline) were assessed by alamarBlue® dye reduction assay. JBMS transfectants with induced siRNA knock-down effect show significantly lower proliferation when compared to the control cells (JBMS transfectants without induction of siRNA knock-down effect and untransfected JBMS cells). The control cells show no difference in cellular proliferation between them. Results were analyzed using linear regression test. 91

4.2.4.3 Down-regulation of GPI-anchored p97 in JBMS mouse melanoma cells reduce tumor growth in vivo To extend our finding on the effect of GPI-anchored p97 down-regulation in JBMS mouse melanoma cells to an in vivo setting, melanoma tumors were established in C57BL/6 mice by subcutaneous injection of stably transfected JBMS/pSUPERIOR-p97 single cell clone. In order to induce the siRNA effect, half of the injected mice were fed with tetracycline orally through their drinking water at the concentration of 1 mg/mi. As control, mice were also subcutaneously injected with untransfected JBMS cells, and half were fed tetracycline water while the other half were fed normal drinking water. Tumor mass was measured after 21 days. The tumors that grew from the tetracycline induced JBMS/pSUPERIOR-p97 transfectants were significantly smaller than those from tetracycline treated or untreated untransfected JBMS and non-induced

JBMS/pSUPERIOR-p97 transfectants (Figure 4.9). Tnabsence of tetracycline, there was no difference in size between tumors derived from JBMS untransfected cells and JBMS/pSUPERIOR-p97 transfectants. However, tumors derived from tetracycline treated untransfected JBMS cells were about 50% smaller than those derived from JBMS/pSUPERIOR-p97 transfectants with no tetracycline treatment and JBMS untransfected cells alone, indicating that tetracycline alone is able to reduced tumor growth. In the presence of tetracycline, the size of tumors derived from JBMS/pSUPERIOR-p97 transfectants were significantly reduced compare to those derived from untransfected JBMS cells, indicating that the reduction in the endogenous GPI-anchored p97 level alone is able to affect the growth of tumor in vivo. 92

2.O

1.5

O.5

0.0 — — I

0

P

JBMS-pSUPERIOR--p97 transfectants were fed 1 mg/mi tetracycline through drinking water, and the other half were fed regular drinking water with no addition of tetracycline. The same feeding pattern was also applied to mice injected with untransfected JBMS cells. One way ANOVA with Tukey’s post test showed that tetracycline induced siRNA knock-down of GPI-anchored p97 in JBMS transfectants retarded tumor growth. This experiment is a representative of three separate experiments. 93

4.3 Discussion GPI-anchored p97 can be found in a variety of tissues ranging from normal tissues, such as salivary glands, sweat glands, smooth muscle, brain endothelial cells and liver, to neoplastic cell and fetal tissues (Alemany et al., 1993; Brown et al., 1981b; Richardson, 2000; Sciot et al., 1989; Woodbury et al., 1981). Study have shown that normal adult tissues display surface expression of GPI-anchored p97 ranging from less than 4,000 molecules per cell to about 8,000 molecules per cell in smooth muscle (Woodbury et al., 1981). However, this is not the case in melanoma cells. Most melanoma cells have been found to express between 50,000 to 500,000 molecules of GPI-anchored p97 per cell (Woodbury et al., 1981), with SKMEL-28 human melanoma cells proposed to express at least 400,000 molecules per cell (Brown et al., 198la). Since GPI-anchored p97 possess a functional ferric iron binding site identical to Tf, thus the proposal of a role in iron metabolism particularly in melanoma cells is plausible. In an attempt to understand the relationship between the up-regulation of GPI-anchored p97 expression and melanoma cells, two models of expression modification of GPI-anchored p97 in mouse melanoma cells were developed via over-expression and down-regulation. To further investigate the role of GPI-anchored p97 in melanoma cells, expression modification studies were conducted in this chapter of the thesis. Two mouse melanoma cell lines, B16F10 and JBMS, were used in order to study both the in vitro and in vivo effects. Since JBMS melanoma cells express relatively high level of endogenous GPI anchored p97 compare to B16F1O melanoma cell, it was used in the GPI-anchored p97 down-regulation model, while B16F1O cells were used in the over-expression model. In the over-expression model, two constructs were created: pMX-pie-his-p97 and pCMV tag3-p97. pMX-pie-his-p97 construct was created to enable identification of the over- expressed GPI-anchored p97 via histidine tag, since there is no antibody available for the mouse version of GPI-anchored p97. GPI-anchored proteins are often difficult to tag since they possess signal peptide at both N- and C-terminus. The N-terminus signal peptide directs the translocation of the nascent chain of the protein to the endoplasmic reticulum, and is cleaved upon arrival. The C-terminus is composed of a signal peptide which is responsible for the attachment of the GPI-anchor to the mature protein, and is cleaved off during the attachment of the GPI-anchor. Since both signal peptides are 94 cleaved, thus attachment of any tag on either ends are not useful. To solve this problem, a 6x histidine tag was inserted in between the N-terminus signal peptide and the p97 sequence. As a result, this enables the identification of over-expressed GPI-anchored p97 at protein level via the histidine tag. In addition, to make sure that the histidine tag is not affecting the overall function of over-expressed GPI-anchored p97, a non-tagged GPI anchored over-expression construct (pCMV-tag3-p97) was also used. The down- regulation of GPI-anchored p97 was achieved through tetracycline inducible siRNA. Since the only difference between the transfectants and the control cells is the amount of GPI-anchored p97, thus any effects are solely the attributes of GPI-anchored p97. In vitro results demonstrated that when GPI-anchored p97 is over-expressed in B16F1O mouse melanoma cells the proliferation was significantly faster than the vector control transfected cells (Figure 4.4), while down-regulation of GPI-anchored p97 in JBMS mouse melanoma cells led to significant slower proliferation compare to the control cells (Figure 4.8). Therefore, the in vitro cell proliferation data showed that modification in GPI-anchored p97 expression alone is able to cause change in the proliferation of melanoma cells, thus implying that GPI-anchored p97 play important role in melanoma cell proliferation. Consistent with the in vitro data, in vivo tumor results demonstrated that over-expression of GPI-anchored p97 expression in B16F1O melanoma cells was able to more than double the size of the tumors compare to the control (Figure 4.6), while the down-regulation of GPI-anchored p97 in JBMS mouse melanoma cells resulted in significant reduction in tumor size compare to the controls (Figure 4.9). Taken together, these results suggest that the presence of GPI-anchored p97 is able to promote melanoma cell proliferation, as results the increased proliferation is translated into higher tumor growth in vivo. These results are consistent with recent findings by Dunn et at. where GPI anchored p97 was down-regulated in SKMEL-28 human melanoma cells (Dunn et al., 2006). It was found that GPI-anchored p97 down-regulation in SKMEL-28 human melanoma cells decrease the proliferation of these cells in vitro, and in xenotransplant into nude immunodeficient mice these cells gave rise to tumors about 20 to 22% in size of that found for SKIVIEL-28control cells (Dunn et al., 2006). Because the nude mice are immune deficient, it is not a good representative of physiological melanoma development. 95

On the other hand, the expression of mouse GPI-anchored p97 in healthy mice demonstrates a more realistic development of melanoma tumors. My studies strengthen the proposal that GPI-anchored p97 plays an important role in melanoma cell proliferation and tumourigenesis. Dunn et al. also stated that addition of iron, apo- and holo-soluble p97 was not able to rescue the depressed proliferation in SKIvIEL-28cells with down-regulated GPI anchored p97(Dunn et al., 2006). However, the mechanism through which soluble p97 enters the cell remain a mystery, thus adding apo- or holo-soluble p97 to GPI-anchored p97 down-regulated SKMEL-28 cells should not rescue the depressed proliferation. Furthermore, since iron has to be bound to GPI-anchored p97 in order to be internalized into the cell, the amount of iron internalized is based on the number of GPI-anchored p97 molecules present, not the amount of iron available. The same researcher has also previously proposed that GPI-anchored p97 does not play an important role in iron uptake in melanoma cells (Richardson and Baker, 1991; Richardson, 2000). The results from chapter 3 clearly indicated that GPI-anchored p97 is able to bind iron and internalize through caveolae fractions, endosomes and subsequently donated to ferritin (Figure 3.1 — 3.6). In addition, iron was shown to accumulate in ferritin when human GPI-anchored p97 was transfected in transferrin receptor deficient CHO cells (Kennard et al., 1995). When these cells were treated with PT-PLC, the amount of iron donated to ferritin decreased proportionally to the amount of decrease in iron uptake due to the PT-PLC removal of GPI-anchored p97 (Kennard et al., 1995). An interesting phenotypic characteristic was observed in mouse melanoma cells over-expressing GPI-anchored p97 (Bl6hisp97 and B16p97). It was found that the higher expression of GPI-anchored p97 in both Bl6hisp97 and B16p97 promoted higher production and secretion of melanin compare to control cells (Figure 4.5). After 4 days of culture, Bl6hisp97 and B16p97 showed around 50% more cellular melanin than the control cells (Figure 4.5a), and secreted large amount of melanin out of the cell whereas the control cells did not show any significant measurable amount of melanin (Figure 4.5b). These results suggest that the GPI-anchored p97 are somehow connected to the melanin synthesis in the melanoma cells. This is most like through the deposition of excess iron by GPI-anchored p97 in the Bl6hisp97 and B16p97 transfectants. 96

Presumably, the over-expressed GPI-anchored p97 transport extra iron into the melanoma cells and some of these iron are used to facilitate cell proliferation, while others are used in melanin synthesis. Iron is important in melanin synthesis due to its role in conversion of dopachrome to 5,6-dihydroxyindole-2-carboxylic acid (DHICA). More specifically, iron in its ferric state is important in the function of dopachrome tautomerase, a melanocyte-specific enzyme that is responsible for the conversion of dopachrome to DHICA (Orlow et al., 1992). DHICA is one of the intermediate products involved in melanogenesis pathway (Figure 4.10), and its oxidation products forms eumelanin which is the most common form of melanin (Blarzino et al., 1999). The pigments derived from DHICA are considered more abundant than any other forms (Ito, 1986). Taken together, it is plausible that the additional iron internalized by GPI-anchored p97 in Bl6hisp97 and Bhip97 transfectants were utilized by the dopachrome tautomerase in the synthesis of eumelanin. In summary, the role of GPI-anchored p97 in melanoma cells was investigated by expression modification. From both in vitro and in vivo experiments, it was concluded that GPI-anchored p97 is important role in melanoma cell growth I proliferation and tumorigenesis, likely due to its iron uptake ability in melanoma cells. In addition, it also suggests that GPI-anchored p97 is able to affect melanin synthesis in melanoma cells possibly through interaction with dopachrome tautomerase. These results establish GPI anchored p97 as an important player in the biology of melanomas. — HO’D”z :x,—° Tyrosire Dopaquinone Dopa

O_r__i GOGH A Dopchrome Fe3

:o?cooH DHI DHICA P/H2o\

EUMELANINS

Figure 4.10: Schematic pathway of melanogenesis. Eumelanins are produced from either tyrosine or dopa through a series of intermediates including dopaquinone, dopachrome and DHI I DHICA. T = tyrosinase, P = peroxidase. This figure is taken from Blarzion et al. (1999) (Blarzino et al., 1999) with permission. 98

Chapter 5 Conclusion and future perspectives Since the first reported case of melanoma over 200 years ago, it has become one of the deadliest forms of skin cancer. With the rise in the popularity of outdoor activities, the overexposure to ultraviolet (UV) radiation has become a leading cause of melanoma in recent years. Through years of research, many aspects with regard to melanoma formation and development have been deciphered, yet there are still so much not known and a cure is still not within our reach. One of the unique characteristics of melanoma is its high expression of .p97 P97 is not exclusively expressed in melanomas, however no other cell type contains comparable levels of p97. It begs the question why is p97 so highly expressed in melanoma cells and melanomas? As a member of the transferrin family of proteins, p97 possesses two conserved iron binding sites, however the amino acid substitutions in one of the iron binding site renders it nonfunctional. Since its discovery almost 30 years ago, various studies have been done on the protein to gain further understanding. GPI-anchored p97 has been shown to internalize iron in p97 transfected transferrin deficient CR0 cells (Kennard et al., 1995). There has also been data linking GPI-anchored p97 to Alzheimer’s disease (AD). In non-AD brain, the distribution of GPI-anchored p97 mRNA and protein has been shown to restrict to the endothelium, while the expression pattern extent to include reactive microglia associated with amyloid plaques (Jefferies et al., 1996; Rothenberger et al., 1996; Yamada et al., 1999). Recently, it was shown that the binding affinity between iron and p97 is intermediate between the high and low affinity lobes of transferrin (Creagh et al., 2005), thus proving the iron binding ability of p97. Rowever, the debate over what exactly happens after the binding remains a hot topic. Iron is essential for cell growth and survival, especially in the malignant neoplastic cells such as melanomas where uncontrolled growth of melanocytes is the root of the problem. The most well characterized and accepted model of cellular iron uptake is facilitated through Tf-TfR complex mediated internalization. The dietary ferric iron binds to the circulating apo-Tf and subsequently binds to the TfR on the cell surface. The iron bound Tf-TfR complex internalize via clathrin-coated pits dependent endosomal pathway. Within the endosomes, the progressive acidification prompts the release of iron from the Tf-TfR complex, and the dissociated iron is transported across the membrane 99 into the cytoplasm by DMT1. Some of the iron is used for cellular metabolism, while the unused iron is sequestered by ferritin, the iron storage protein, for future usage. The rate of iron uptake via Tf-TfR complex is determined by the level of cell surface expression of TfR (Casey et al., 1989; Dautry-Varsat et al., 1983; Klausner et aL, 1983). On the other hand, the level of TfR biosynthesis is also regulated by the cellular iron level (Casey et al., 1989; Mattia et al., 1984; Ward et al., 1982) through the iron responsive element (IRE) located at the 3’ untranslated region of the Tifi mRNA (Casey et al., 1988; Casey et al., 1989). TfR are found in great abundance during erythropoiesis with a gradual increase to a maximum expression of about 800,000 sites per cell on the intermediate normoblast (Busfield et al., 1997; Ponka and Lok, 1999; Sawyer and Krantz, 1986), however during the maturation to erythrocytes, the expression of TfR are lost (Jandi et al., 1959; Pan et al., 1983; Ponka and Lok, 1999). In rapidly proliferation cells, TfR expression have been found to vary between 10,000 to 100,000 sites per cell found on carcinoma cells (Inoue et al., 1993; Qian et al., 2002). Tf-TfR dependent iron transport is by no way the exclusive mechanism for efficient iron uptake. In cases where the expression of Tf is low to none existent, such as atransferrineniia and hypotransferrinemia, significant amount of iron have been shown to efficiently uptake by vital organs (Bernstein, 1987; Beutler et al., 2000; Craven et al., 1987; Hamill et al., 1991; Huggenvik et al., 1989; Trenor et al., 2000). In addition, non Tf bound iron in hereditary hemochromatosis has been found to circulate at relatively high levels (Batey et al., 1980; Gutteridge et al., 1985; McNamara et al., 1999). There growing evidence prompt the researchers to identify the possible mechanisms behind Tf TfR independent iron transport. One such candidate is GPI-anchored p97. GPI-anchored p97 has been shown to transport iron efficiently in TfR deficient Chinese hamster ovary cells transfected with GPI-anchored p97, and the process was both temperature dependent and saturable (Kennard et al., 1995). Recent our studies using calorimetry have demonstrated that the N-lobe of GPI-anchored p97 is able to tightly bind to ferric iron at 4.4x 170’M’, with is intermediate between the high affinity N-lobe and low affinity C-lobe of Tf (1.1M221x10 and 31M8x10’ respectively) (Creagh et al., 2005). Even though the iron binding ability of GPI-anchored is well accepted, however the fate of the iron and GPI-anchored p97 subsequently after binding remain a hotly 100

debated subject. Since GPI-anchored p97 is expressed at the highest levels in melanomas, the iron uptake and internalization ability mediated by GPI-anchored p97 was studied. Being a GPI-anchored protein, p97 represents a good opportunity to shed light on the on-going debate over the internalization of ligand bound GPI-anchored proteins. One of the main issues in the debate is over the involvement of caveolae vesicles. Caveolae vesicles are flask shaped invaginations formed from the plasma membrane and are found in various cell types, such as endothelial cells and adipocytes. Caveolae vesicles are distinct from clathrin-coated vesicles and are characterized by their rich lipid content (Anderson, 1998). While some reports show GPI-anchored proteins are preferentially localized in the caveolae (Anderson, 1998; Ying et al., 1992), others argue the distribution pattern is only a result of cross-linking with multiple secondary antibodies (Mayor et al., 1994). In this thesis, the study of iron trafficking by GPI-anchored p97 mediated through caveolae vesicles dependent endosomal pathway in melanoma cells represents a novel iron uptake pathway separate from the Tf-TfR system, and sheds new light on the on-going debate over GPI-anchored protein internalization. In addition, the efficient transport of iron by GPI-anchored p97 to ferritin offers clue to the reason for its high expression in melanoma cells especially. To date, the highest expression of GPI-anchored p97 is found in malignant melanomas at about 50,000 to 500,000 molecules per cell, with the expression level in normal cells and tissues found to range between less than 4,000 to at maximum 8,000 molecules per cell in smooth muscle (Woodbury et al., 1981). In comparison, rapid proliferating cells such as some carcinomas express TfR up to 100,000 molecules per cell (Inoue et al., 1993; Qian et al., 2002). Tf-TfR complex is known to be very efficient in transport of iron to cells. However, the existence of iron uptake independent of Tf-TfR system questions the universal role of iron transport assigned to Tf-TfR. The dramatic difference in the expression of GPI-anchored p97 between normal cells and malignant neoplastic cells such as melanomas infers that the heightened expression of GPI-anchored p97 may be necessary in delivering excess iron to meet the demand of uncontrolled proliferation of melanocytes in malignant melanomas. In this thesis, the studies on expression modification of GPI-anchored p97 in mouse melanoma cells shed light on the mystery relationship between GPI-anchored p97 and melanomas. Along with the iron 101

uptake data, the fact that GPI-anchored p97 alone is able to positively change the cell proliferation and tumor growth suggests that it is effectively binding and transporting iron into the cell due to the demands for iron during cell proliferation. This is further supported by similar study done in human SKMEL-28 melanoma cells (Dunn et aL, 2006; Richardson and Baker, 1991). In conjunction, the surprising finding on the increased melanin production and secretion in mouse melanoma cells over expressing GPI anchored p97 further supports the model of GPI-anchored p97 acting as an alternative iron uptake pathway in melanoma cells. If GPI-anchored p97 play negligible role in cellular iron uptake in melanoma cells as proposed by Richardson group (Dunn et al., 2006; Richardson and Baker, 1991), one would not expect any change in melanin production by melanoma cells. Since iron plays an important role in melanin synthesis, the increase in the amount of melanin per cell in melanoma cells over-expressing GPI anchored p97 clearly support the role of GPI-anchored p97 as an iron uptake molecule in melanoma cells. The studies in this thesis have opened doors for various potential future projections. Although the process of crystallizing p97 has already been started (see Appendix I), the crystal structure of p97 has yet to be resolved. The identification of crystal structure of p97 will enable the visualization of the binding between p97 and iron. In addition, studies examining the factors, such as hypoxia and transcription factor

activator protein 1 (AP- 1), associated with or responsible for the up-regulation of GPI anchored p97 expression are underway (see Appendix II). Preliminary data shows that

both hypoxic condition and hypoxia-inducible factor 1 (HIF-1) are able to up-regulate

GPI-anchored p97 at mRNA level. Since p97 possess two AP- 1 binding sites in its enhancer region, the ability of AP-1 to up-regulate GPI-anchored p97 expression has been investigated. Thus far, results show that AP- 1 transcription factor is able to effectively up-regulate endogenous GPI-anchored p97. However, additional studies are needed to understand the relationship between HIF-1, AP-l and GPI-anchored expression. Furthermore, it has been shown that soluble p97 can carry iron and drugs (unpublished data) across the BBB and function as a pro-angiogenic factor (Sala et al., 2002) by binding to the TfR1 and possibly via the low-density lipoprotein receptor protein (LRP) (Demeule et al., 2003). Human recombinant s-p97 has been shown to form 102 a complex with plasminogen activator and thus modify endothelial tumourigenesis (Michaud-Levesque et al., 2005). These data further suggest that p97 may be important for iron transport, inflammation and angiogenesis, all of which are important features in AD.

In conclusion, the results in this thesis suggest that GPI-anchored p97 can serve as an alternative iron transport protein to effectively deliver excess iron supply needed by the rapidly proliferating melanocytes in melanoma cells. 103

References

Akahoshi, E., S. Yoshimura, and M. Ishihara-Sugano. 2006. Over-expression of AhR (aryl hydrocarbon receptor) induces neural differentiation of Neuro2a cells: neurotoxicology study. Environ Health. 5:24. Alemany, R., M.R. Vila, C. Franci, G. Egea, F.X. Real, and T.M. Thomson. 1993. Glycosyl phosphatidylinositol membrane anchoring of melanotransferrin (p97): apical compartmentalization in intestinal epithelial cells. J Cell Sci. 104 ( Pt 4): 1155-62. Anderson, R.G. 1994. Functional specialization of the glycosylphosphatidylinositol membrane anchor. Semin Immunol. 6:89-95. Anderson, R.G. 1998. The caveolae membrane system. Annu Rev Biochem. 67:199-225. Andrews, N.C. 1999. Disorders of iron metabolism. NEnglJMed. 341:1986-95. Aruoma, 0.1., B. Halliwell, M.J. Laughton, G.J. Quinlan, and J.M. Gutteridge. 1989. The mechanism of initiation of lipid peroxidation. Evidence against a requirement for an iron(II)-iron(III) complex. Biochem J. 258:617-20. Asako, H., I. Kurose, R.E. Wolf, and D.N. Granger. 1994. Mechanisms of lactoferrin induced leukocyte-endothelial cell adhesion in postcapillary venules. Microcirculation. 1:27-34. Bailey, S., R.W. Evans, R.C. Garratt, B. Gorinsky, S. Hasnain, C. Horsburgh, H. Jhoti, P.F. Lindley, A. Mydin, R. Sarra, and et al. 1988. Molecular structure of serum transferrin at 3.3-A resolution. Biochemistry. 27:5804-12. Baker, E.N., and P.F. Lindley. 1992. New perspectives on the structure and function of transferrins. J Inorg Biochem. 47:147-60. Bali, P.K., 0. Zak, and P. Aisen. 1991. A new role for the transferrin receptor in the release of iron from transferrin. Biochemistry. 30:324-8. Barresi, G., and G. Tuccari. 1994. Immunocytochemical demonstration of melanotransferrin (p97) in thyroid tumors of follicular cell origin. Pathology. 26:127-9. Batey, R.G., P. Lai Chung Fong, S. Shamir, and S. Sherlock. 1980. A non-transferrin bound serum iron in idiopathic hemochromatosis. Dig Dis Sci. 25:340-6. Beinert, H. 1976. fron-sulfur proteins, the most numerous and most diversified components of the mitochondrial electron transfer system. Adv Exp Med Biol. 74:137-49. Berkeihammer, J., R.W. Oxenhandler, R.R. Hook, Jr., and J.M. Hennessy. 1982. Development of a new melanoma model in C57BL/6 mice. Cancer Res. 42:3 157- 63. Bernstein, S.E. 1987. Hereditary hypotransferrinemia with hemosiderosis, a murine disorder resembling human atransferrinemia. J Lab Clin Med. 110:690-705. Beutler, E. 2002. History of iron in medicine. Blood Cells Mol Dis. 29:297-308. Beutler, E., V.F. Fairbanks, and J.L. Fahey. 1963. Clinical disorders of iron metabolism. Grune & Stratton, New York. iv, 267 p. pp. Beutler, E., T. Gelbart, P. Lee, R. Trevino, M.A. Fernandez, and V.F. Fairbanks. 2000. Molecular characterization of a case of atransferrinemia. Blood. 96:4071-4. Blarzino, C., L. Mosca, C. Foppoli, R. Coccia, C. De Marco, and M.A. Rosei. 1999, Lipoxygenase/H202-catalyzed oxidation of dihdroxyindoles: synthesis of 104

melanin pigments and study of their antioxidant properties. Free Radic Biol Med. 26:446-53. Blasi, E., R. Barluzzi, V. Bocchini, R. Mazzolla, and F. Bistoni. 1990. Immortalization of murine microglial cells by a v-raf/v-myc carrying retrovirus. J Neuroimmunol. 27:229-37. Brandts, J.F., and L.N. Lin. 1990. Study of strong to ultratight protein interactions using differential scanning calorimetry. Biochemistry. 29:6927-40. Bray, N., and L. Pachter. 2004. MAVID: constrained ancestral alignment of multiple sequences. Genome Res. 14:693-9. Brown, D.A., and J.K. Rose. 1992. Sorting of GPI-anchored proteins to glycolipid enriched membrane subdomains during transport to the apical cell surface. Cell. 68:533-44. Brown, J.P., R.M. Hewick, I. Hellstrom, K.E. Hellstrom, R.F. Doolittle, and W.J. Dreyer. 1982. Human melanoma-associated antigen p97 is structurally and functionally related to transferrin. Nature. 296:171-3. Brown, J.P., K. Nishiyama, I. Helistrom, and K.E. Helistrom. 1981a. Structural characterization of human melanoma-associated antigen p97 with monoclonal antibodies. J Immunol. 127:539-46. Brown, J.P., R.G. Woodbury, C.E. Hart, I. Helistrom, and K.E. Helistrom. 198lb. Quantitative analysis of melanoma-associated antigen p97 in normal and neoplastic tissues. Proc Nati Acad Sci US A. 78:539-43. Bruns, R.R., and G.E. Palade. 1968. Studies on blood capillaries. II. Transport of ferritin molecules across the wall of muscle capillaries. J Cell Biol. 37:277-99. Busfield, S.J., P.A. Tilbrook, B.A. Callus, A. Spadaccini, L. Kuhn, and S.P. Kdinken. 1997. Complex regulation of transferrin receptors during erythropoietin-induced differentiation of J2E erythroid cells--elevated transcription and mRNA stabilisation produce only a modest rise in protein content. Eur J Biochem. 249:77-84. Campbell, T., R.A. Skilton, R.C. Coombes, S. Shousha, M.D. Graham, and Y.A. Luqmani. 1992. Isolation of a lactoferrin cDNA clone and its expression in human breast cancer. Br J Cancer. 65:19-26. Canada, C.C.S.N.C.I.o. 2008. Canadian Cancer Statistics. Vol. 2009, Toronto. Casey, J.L., M.W. Hentze, D.M. Koeller, S.W. Caughman, T.A. Rouault, R.D. Klausner, and J.B. Harford. 1988. lion-responsive elements: regulatory RNA sequences that control mRNA levels and translation. Science. 240:924-8. Casey, J.L., D.M. Koeller, V.C. Ramin, R.D. Klausner, and J.B. Harford. 1989. lion regulation of transferrin receptor mRNA levels requires iron-responsive elements and a rapid turnover determinant in the 3’ untranslated region of the mRNA. EMBO J. 8:3693-9. Cavallo-Medved, D., J. Dosescu, B.E. Linebaugh, M. Sameni, D. Rudy, and B.F. Sloane. 2003. Mutant K-ras regulates cathepsin B localization on the surface of human colorectal carcinoma cells. Neoplasia. 5:507-19. Cazzola, M., G. Bergamaschi, L. Dezza, and P. Arosio. 1990. Manipulations of cellular iron metabolism for modulating normal and malignant cell proliferation: achievements and prospects. Blood. 75:1903-19. 105

Chatterjee, S., and S. Mayor. 2001. The GPI-anchor and protein sorting. Cell Mol Life Sci. 58:1969-87. Cheng, Y., 0. Zak, P. Aisen, S.C. Harrison, and T. Walz. 2004. Structure of the human transferrin receptor-transferrin complex. Cell. 116:565-76. Chitambar, C.R., E.J. Massey, and P.A. Seligman. 1983. Regulation of transferrin receptor expression on human leukemic cells during proliferation and induction of differentiation. Effects of gallium and dimethylsulfoxide. J Clin Invest. 72:1314- 25. Chu, P., 5. Murray, D. Lissin, and M. von Zastrow. 1997. Delta and kappa opioid receptors are differentially regulated by dynamin-dependent endocytosis when activated by the same alkaloid agonist. J Biol Chem. 272:27124-30. Chua, A.C., R.M. Graham, D. Trinder, and J.K. Olynyk. 2007. The regulation of cellular iron metabolism. Crit Rev Clin Lab Sci. 44:413-59. Collawn, J.F., A. Lai, D. Domingo, M. Fitch, S. Hatton, and I.S. Trowbridge. 1993. YTRF is the conserved internalization signal of the transferrin receptor, and a second YTRF signal at position 31-34 enhances endocytosis. J Biol Chem. 268:21686-92. Colomb, E., C. Pianetta, J.P. Estevenon, 0. Guy, C. Figarella, and H. Sarles. 1976. Lactoferrin in human pancreas. Immunohistological localization in normal and pathological pancreatic tissues. Digestion. 14:242-9. Craven, C.M., 3. Alexander, M. Eldridge, J.P. Kushner, S. Bernstein, and 3. Kaplan. 1987. Tissue distribution and clearance kinetics of non-transferrin-bound iron in the hypotransferrinemic mouse: a rodent model for hemochromatosis. Proc Natl Acad Sci USA. 84:3457-61. Creagh, A.L., J.W. Tiong, M.M. Tian, C.A. Haynes, and W.A. Jefferies. 2005. Calorimetric studies of melanotransferrin (p97) and its interaction with iron. I Biol Chem. 280:15735-41. Crouch, S.P., K.J. Slater, and J. Fletcher. 1992. Regulation of cytokine release from mononuclear cells by the iron-binding protein lactoferrin. Blood. 80:235-40. Cummins, D.L., J.M. Cummins, H. Pantle, M.A. Silverman, A.L. Leonard, and A. Chanmugam. 2006. Cutaneous malignant melanoma. Mayo Clin Proc. 81:500-7. Dautry-Varsat, A., A. Ciechanover, and H.F. Lodish. 1983. pH and the recycling of transferrin during receptor-mediated endocytosis. Proc NatI Acad Sci U S A. 80:2258-62. Davis, N.C. 1980. William Norris, M.D.: a pioneer in the study of melanoma. Med fAust. 1:52-4. De Domenico, I., D. McVey Ward, and J. Kaplan. 2008. Regulation of iron acquisition and storage: consequences for iron-linked disorders. Nat Rev Mol Cell Biol. 9:72- 81. de Silva, D.M., C.C. Askwith, and J. Kaplan. 1996. Molecular mechanisms of iron uptake in eukaryotes. Physiol Rev. 76:31-47. Demeule, M., Y. Bertrand, 3. Michaud-Levesque, 3. Jodoin, Y. Rolland, R. Gabathuler, and R. Beliveau. 2003. Regulation of plasminogen activation: a role for melanotransferrin (p9’7)in cell migration. Blood. 102:1723-31. 106

Devary, Y., R.A. Gottlieb, L.F. Lau, and M. Karin. 1991. Rapid and preferential activation of the c-jun gene during the mammalian UV response. Mol Cell Biol. 11:2804-11. Duchange, N., A. Ochoa, GD. Plowman, A. Roze, M. Amdjadi, and M.M. Zakin. 1992. Identification of an enhancer involved in the melanoma-specific expression of the tumor antigen melanotransferrin gene. Nucleic Acids Res. 20:2853-9. Dunn, L.L., E.O. Sekyere, Y.S. Rahmanto, and D.R. Richardson. 2006. The function of melanotransferrin: a role in melanoma cell proliferation and tumorigenesis. Carcino genesis. 27:2157-69. Echarri, A., 0. Muriel, and M.A. Del Pozo. 2007. Intracellular trafficking of raftlcaveolae domains: insights from integrin signaling. Semin Cell Dev Biol. 18:627-37. Eferl, R., and E.F. Wagner. 2003. AP-1: a double-edged sword in tumorigenesis. Nat Rev Cancer. 3:859-68. Fairbanks, V.F., E. Beutler, and J.L. Fahey. 1971. Clinical disorders of iron metabolism. Grune & Stratton, New York,. ix, 486 p. pp. Ferguson, M.A. 1999. The structure, biosynthesis and functions of glycosyiphosphatidylinositol anchors, and the contributions of trypanosome research. J Cell Sci. 112 ( Pt 17):2799-809. Fidler, I.J. 1978. Tumor heterogeneity and the biology of cancer invasion and metastasis. Cancer Res. 38:2651-60. Finch, C.A. 1982. Erythropoiesis, erythropoietin, and iron. Blood. 60:1241-6. Food, M.R., and R. Des Richardson. 2002. Iron uptake by melanoma cells from the soluble form of the transferrin homologue, melanotransferrin. Redox Rep. 7:279- 82. Gabella, G. 1976. Quantitative morphological study of smooth muscle cells of the guinea- pig taenia coli. Cell Tissue Res. 170:161-86. Gil, J. 1983. Number and distribution of plasmalemmal vesicles in the lung. Fed Proc. 42:2414-8. Guillot, F.L., K.L. Audus, and T.J. Raub. 1990. Fluid-phase endocytosis by primary cultures of bovine brain microvessel endothelial cell monolayers. Microvasc Res. 39:1-14. Gutierrez, J.A., R.S. Inman, T. Akompong, J. Yu, and M. Wessling-Resnick. 1998. Metabolic depletion inhibits the uptake of nontransferrin-bound iron by K562 cells. J Cell Physiol. 177:585-92. Gutteridge, J.M., D.A. Rowley, E. Griffiths, and B. Halliwell. 1985. Low-molecular- weight iron complexes and oxygen radical reactions in idiopathic haemochromatosis. Clin Sci (Lond). 68:463-7. Hagiwara, M., Y. Shirai, R. Nomura, M. Sasaki, K. Kobayashi, T. Tadokoro, and Y. Yamamoto. 2009. Caveolin-1 activates RabS and enhances endocytosis through direct interaction. Biochem Biophys Res Commun. 378:73-8. Halliwell, B., and J.M. Gutteridge. 1986. Oxygen free radicals and iron in relation to biology and medicine: some problems and concepts. Arch Biochem Biophys. 246:501-14. Hamill, R.L., J.C. Woods, and B.A. Cook. 1991. Congenital atransferrinemia. A case report and review of the literature. Am J Clin Pathol. 96:215-8. 107

Harmey, J.H., D. Doyle, V. Brown, and M.S. Rogers. 1995. The cellular isoform of the prion protein, PrPc, is associated with caveolae in mouse neuroblastoma (N2a) cells. Biochem Biophys Res Commun. 210:753-9. Hegedus, D.D., T.A. Pfeifer, D.A. Theilmann, M.L. Kennard, R. Gabathuler, W.A. Jefferies, and T.A. Grigliatti. 1999. Differences in the expression and localization of human melanotransferrin in lepidopteran and dipteran insect cell lines. Protein Expr Purif 15:296-307. Hu, D.N. 2008. Methodology for evaluation of melanin content and production of pigment cells in vitro. Photochem Photobiol. 84:645-9. Huggenvik, J.I., C.M. Craven, R.L. Idzerda, S. Bernstein, J. Kaplan, and G.S. McKnight. 1989. A splicing defect in the mouse transferrin gene leads to congenital atransferrinemia. Blood. 74:482-6. Ikonen, E., S. Heino, and S. Lusa. 2004. Caveolins and membrane cholesterol. Biochem Soc Trans. 32:121-3. Inoue, T., P.G. Cavanaugh, P.A. Steck, N. Brunner, and G.L. Nicolson. 1993. Differences in transferrin response and numbers of transferrin receptors in rat and human mammary carcinoma lines of different metastatic potentials. J Cell Physiol. 156:212-7. Ishikawa, H. 1968. Formation of elaborate networks of T-system tubules in cultured skeletal muscle with special reference to the T-system formation. J Cell Biol. 38:51-66. Ito, S. 1986. Reexamination of the structure of eumelanin. Biochim Biophys Acta. 883:155-61. Iwabuchi, K., K. Handa, and S. Hakomori. 1998. Separation of “glycosphingolipid signaling domain” from caveolin-containing membrane fraction in mouse melanoma B 16 cells and its role in cell adhesion coupled with signaling. J Biol Chem. 273:33766-73. Jandl, J.H., J.K. Inman, R.L. Simmons, and D.W. Allen. 1959. Transfer of iron from serum iron-binding protein to human reticulocytes. J Clin Invest. 38:161-85. Janssen, P.T., and O.P. van Bijsterveld. 1983. Origin and biosynthesis of human tear fluid proteins. Invest Ophthalmol VisSci. 24:623-30. Jefferies, W.A., M.R. Food, R. Gabathuler, S. Rothenberger, T. Yamada, 0. Yasuhara, and P.L. McGeer. 1996. Reactive niicroglia specifically associated with amyloid plaques in Alzheimer’s disease brain tissue express melanotransferrin. Brain Res. 712: 122-6. Jeltsch, J.M., R. Hen, L. Maroteaux, J.M. Gamier, and P. Chambon. 1987. Sequence of the chicken ovotransferrin gene. Nucleic Acids Res. 15:7643-5. Kawabata, H., R.S. Germain, P.T. Vuong, T. Nakamaki, J.W. Said, and H.P. Koeffler. 2000. Transferrin receptor 2-alpha supports cell growth both in iron-chelated cultured cells and in vivo. J Biol Chem. 275:16618-25. Kennard, M.L., H. Feldman, T. Yamada, and W.A. Jefferies. 1996. Serum levels of the iron binding protein p97 are elevated in Alzheimer’s disease. Nat Med. 2:1230-5. Kennard, M.L., D.R. Richardson, R. Gabathuler, P. Ponka, and W.A. Jefferies. 1995. A novel iron uptake mechanism mediated by GPI-anchored human p97. EMBO J. 14:4178-86. 108

Kim, D.K., M.Y. Seo, S.W. Lim, S. Kim, J.W. Kim, B.J. Carroll, D.Y. Kwon, T. Kwon, and S.S. Kang. 2001. Serum melanotransferrin, p97 as a biochemical marker of Alzheimer’s disease. Neuropsychopharmacology. 25:84-90. Kinoshita, T., K. Ohishi, and J. Takeda. 1997. GPI-anchor synthesis in mammalian cells: genes, their products, and a deficiency. J Biochem. 122:251-7. Klausner, R.D., G. Ashwell, J. van Renswoude, J.B. Harford, and K.R. Bridges. 1983. Binding of apotransferrin to K562 cells: explanation of the transferrin cycle. Proc Nati Acad Sci U S A. 80:2263-6. Kurokawa, H., B. Mikami, and M. Hirose. 1995. Crystal structure of diferric hen ovotransferrin at 2.4 A resolution. J Mol Biol. 254:196-207. Kurose, I., T. Yamada, R. Wolf, and D.N. Granger. 1994. P-selectin-dependent leukocyte recruitment and intestinal mucosal injury induced by lactoferrin. J Leukoc Biol. 55:771-7. Kwok, J.C., and D.R. Richardson. 2002. The iron metabolism of neoplastic cells: alterations that facilitate proliferation? Crit Rev Oncol Hematol. 42:65-78. Laderoute, K.R., J.M. Calaoagan, C. Gustafson-Brown, A.M. Knapp, G.C. Li, H.L. Mendonca, H.E. Ryan, Z. Wang, and R.S. Johnson. 2002. The response of c junJAP- 1 to chronic hypoxia is hypoxia-inducible factor 1 alpha dependent. Mol Cell Biol. 22:25 15-23. Lambert, L.A., H. Perri, P.J. Halbrooks, and A.B. Mason. 2005a. Evolution of the transferrin family: conservation of residues associated with iron and anion binding. Comp Biochem Physiol B Biochem Mol Biol. 142:129-41. Lambert, L.A., H. Perri, and T.J. Meehan. 2005b. Evolution of duplications in the transferrin family of proteins. Comp Biochem Physiol B Biochem Mol Biol. 140:11-25. Larrick, J.W., and P. Cresswell. 1979. Modulation of cell surface iron transferrin receptors by cellular density and state of activation. J Supraniol Struct. 11:579-86. Lee, J., E. Jung, J. Park, K. Jung, E. Park, J. Kim, S. Hong, S. Park, S. Lee, and D. Park. 2005. Glycyrrhizin induces melanogenesis by elevating a cAMP level in b16 melanoma cells. Jlnvest Dermatol. 124:405-11. Li, M., K. Pisalyaput, M. Galvan, and A.J. Tenner. 2004. Macrophage colony stimulatory factor and interferon-gamma trigger distinct mechanisms for augmentation of beta-amyloid-induced microglia-mediated neurotoxicity. J Neurochem. 91:623-33. Lin, L.N., A.B. Mason, R.C. Woodworth, and J.F. Brandts. 1994. Calorimetric studies of serum transferrin and ovotransferrin. Estimates of domain interactions, and study of the kinetic complexities of ferric ion binding. Biochemistry. 33:1881-8. Lindley, P.F. 1996. Iron in biology: a structural viewpoint. Rep. Frog. Phys. 59:867-933. Liu, J., and J.I. Shapiro. 2003. Endocytosis and signal transduction: basic science update. Biol Res Nurs. 5:117-28. Macedo, M.F., and M. de Sousa. 2008. Transferrin and the transferrin receptor: of magic bullets and other concerns. Inflamm Allergy Drug Targets. 7:41-52. Makiya, R., L.E. Thornell, and T. Stigbrand. 1992. Placental alkaline phosphatase, a GPI anchored protein, is clustered in clathrin-coated vesicles. Biochem Biophys Res Commun. 183:803-8.

Parton,

Parolini,

Parham,

Pan,

Otsuki,

Olkkonen,

Orlow,

Oh, Neilands,

Natural

Napolitano,

Michaud-Levesque,

Michaud-Levesque,

McNamara,

McNagny, Mayor,

Mattia,

P.,

B.T.,

new

caveolin-2

M.P.

golgi

Biochem

Monoclonal

and

Okamoto,

Rev

the

melanins

R.G., and

caveolae

Zinn,

Cell in

Press,

health

Microscope

phenomenon

plasminogen

P.

angiogenesis

African and melanotransferrin. proteins

K., Biol

regulation

surface

S.J.,

E.,

Health

S.,

P.,

I.,

the

Borgstrom,

J.B.

R.

V.M.,

functional

A.

interleukin-6

maturation

insights

Cytol.

K.

K.M.,

Res.

V.R.

K.G.

Chem.

complex. L.,

M.

Lisanti.

and

M.J.

M.P.

R.

L.

Blostein,

lung.

New

information

Yoda,

Rao,

1974.

dietary

Information

in

J.

A.P.

Baldwin, antigen,

5:113-21. pumping

Sargiacomo,

from

A.A.

1963.

T.

and Gordeuk.

to

176:1-85.

Rothberg,

of

2

F. caveolae

Nat

Androlewicz,

York.

Osber,

259:2689-92.

Study.

and

antibodies:

10:37-47.

D.S.

]kezu,

J.,

by

for

Y.

J.,

the

the

MacPhail,

Rossi,

Microbial

studies H.

H.

1999.

J

Richards. dihydroxyindole-2-carboxylic

activation

and

iron

Biotechnol.

of

M.

a

Toma,

Biol

common

cell

The

interaction

M.

Witkiewicz,

Stenmark.

plasma

human

xv,

Shapiro,

soluble

E0S47,

J.E.

centre. Blood.

targeted

J

and

sheep

P.E.

R.M.

Demeule,

overload.

1999. triggered

Expression Cell

invasion.

Centre.

Chein.

G.

Demeule,

597

of

and

Differentiation

F.

Testa,

Y.

purification, J.M.

class

Scherer,

E.

Smith,

iron

2003.

transferrin

Johnstone.

Biol.

membrane.

p. form

mechanisms.

87:1343-52.

F.M.

Galbiati,

Non-transferrin-bound

reticulocytes Shimizu,

F.R.

H.H.

is

25:327-37.

by

Mandishona,

antibody

274:25718-25.

pp.

in

1997.

2009.

Pawelek.

I and

by

J

metabolism

Y.

and

Exp

a

amniotic

of

18:663-79.

MHC

Lipid

Gastroenterol

Maxfield.

Brodsky,

and

and Sussman,

membrane-bound

cross-linking.

Li,

of

R.

chicken

melanotransferrin.

J.E.

Role

R.

Cell

Iron.

H.

G.

Mora,

receptor

caveolin-1

B.J.

1983.

T.

rapidly

rafts

fragmentation

R.

Retention

antigens.

Beliveau.

Schnitzer.

Saito,

of

Traffic.

Res.

Rizzo,

of

infection.

1992.

Graf.

in

Borgstrom,

Beliveau.

The

P.

and

homologue

and

N.J.

:

White

Rab

Loss

E.

1994.

vitro.

Bloom,

a

308:479-90.

and

by

and

acid

Rodriguez-Boulan,

energy

Hepatol.

caveolae

comprehensive

Synthesis

1996.

J

R.D.

Science.

4:724-38.

F.

GTPases

of

is

2005.

Holmes,

of

desferrioxamine

Immunol

2007.

T.

specifically

iron

Sequestration

Adv

An

and

Adipose

Grignani,

required

caveolin-2

the

2007.

and

A.C.

Yanaihara.

Klausner.

Carcino

A.

melanotransferrin:

The

mineral.

Stimulation

and

Exp

transferrin

immunological

dihydroxyindole.

of

14:126-32.

as

Live

264:1948-51.

Chrastina,

application

in

and

Paterson,

Methods. and

eosinophil-specific

portals

the

In

hepatic

Med

membrane

for

genesis.

Cells.

dynamic

l.A.

across

characterization

at

treatise. vivo

Vol.

J.P.

1984.

oncofetal

1998.

Biol.

of

the

the

in

C.

for

receptor

K.

Engelman,

of

dysfunction

T.A.

Ways. GPI-anchored

53:133-73.

2009.

An

K562

inhibition

Peschle,

to

transport

endothelium

level

Biosynthetic

endocytosis:

28:280-8.

443:267-71.

Iwata,

cell

Lactoferrin

imaging

traffic.

Academic

approach.

structural

Pigment

Electron

Rouault,

a

antigen

surface

cells.

natural

during

of

1982.

K.R.

and

the

key

cell

109

mt

T.

of

of

of

of

in J 110

Pelkmans, L., and M. Zerial. 2005. Kinase-regulated quantal assemblies and kiss-and-run recycling of caveolae. Nature. 436:128-33. Peters, P.J., A. Mironov, Jr., D. Peretz, E. van Donselaar, E. Leclerc, S. Erpel, S.J. DeArmond, D.R. Burton, R.A. Williamson, M. Vey, and S.B. Prusiner. 2003. Trafficking of prion proteins through a caveolae-mediated endosomal pathway. J Cell Biol. 162:703-17. Pike, I.M. 2002. Outpatient endoscopy possibilities for the office. Gastrointest Endosc Clin NAm. 12:245-58. Plowman, G.D., J.P. Brown, C.A. Enns, J. Schroder, B. Nilcinmaa, H.H. Sussman, K.E. Helistrom, and I. Helistrom. 1983. Assignment of the gene for human melanoma- associated antigen p97 to chromosome 3. Nature. 303:70-2. Ponka, P., and C.N. Lok. 1999. The transferrin receptor: role in health and disease. mtJ Biochem Cell Biol. 31:1111-37. Poste, G., J. Doll, A.E. Brown, J. Tzeng, and I. Zeidman. 1982. Comparison of the metastatic properties of B16 melanoma clones isolated from cultured cell lines, subcutaneous tumors, and individual lung metastases. Cancer Res. 42:2770-8. Poste, G., J. Doll, I.R. Hart, and 1.1. Fidler. 1980. In vitro selection of murine B16 melanoma variants with enhanced tissue-invasive properties. Cancer Res. 40:1636-44. Qian, Z.M., H. Li, H. Sun, and K. Ho. 2002. Targeted drug delivery via the transferrin receptor-mediated endocytosis pathway. Pha rmacol Rev. 54:56 1-87. Raposo, G., and M.S. Marks. 2002. The dark side of lysosome-related organelles: specialization of the endocytic pathway for melanosome biogenesis. Traffic. 3:237-48. Razani, B., S.E. Woodman, and M.P. Lisanti. 2002. Caveolae: from cell biology to animal physiology. Pharmacol Rev. 54:431-67. Rich, P.R. 1984. Electron and proton transfers through quinones and cytochrome bc complexes. Biochim Biophys Acta. 768:53-79. Richardson, D., and B. Baker. 1991. The uptake of inorganic iron complexes by human melanoma cells. Biochim Biophys Acta. 1093:20-8. Richardson, D., and E. Baker. 1992. Two mechanisms of iron uptake from transferrin by melanoma cells. The effect of desferrioxamine and ferric ammonium citrate. J Biol Chem. 267:13972-9. Richardson, D.R. 2000. The role of the membrane-bound tumour antigen, melanotransferrin (p9’7),in iron uptake by the human malignant melanoma cell. Eur J Biochem. 267:1290-8. Richardson, D.R., and E. Baker. 1990. The uptake of iron and transferrin by the human malignant melanoma cell. Biochim Biophys Acta. 1053:1-12. Rijnboutt, S., G. Jansen, G. Posthuma, J.B. Hynes, J.H. Schomagel, and G.J. Strous. 1996. Endocytosis of GPI-linked membrane folate receptor-alpha. J Cell Biol. 132:35- 47. Roguin, A. 2006. Rene Theophile Hyacinthe Laennec (1781-1826): the man behind the stethoscope. Clin Med Res. 4:230-5. Rolz-Cruz, G., and C.C. Kim. 2008. Tumor invasion of the skin. Dermatol Clin. 26:89- 102, viii. 111

Rose, T.M., G.D. Plowman, D.B. Teplow, W.J. Dreyer, K.E. Helistrom, and J.P. Brown. 1986. Primary structure of the human melanoma-associated antigen p97 (melanotransferrin) deduced from the mRNA sequence. Proc Nail Acad Sci U S A. 83:1261-5. Rothberg, K.G., J.E. Heuser, W.C. Donzell, Y.S. Ying, J.R. Glenney, and R.G. Anderson. 1992. Caveolin, a protein component of caveolae membrane coats. Cell. 68:673- 82. Rothberg, K.G., Y.S. Ying, J.F. Kolhouse, B.A. Kamen, and R.G. Anderson. 1990. The glycophospholipid-linked folate receptor internalizes folate without entering the clathrin-coated pit endocytic pathway. J Cell Biol. 110:637-49. Rothenberger, S., M.R. Food, R. Gabathuler, M.L. Kennard, T. Yamada, 0. Yasuhara, P.L. McGeer, and W.A. Jefferies. 1996. Coincident expression and distribution of melanotransferrin and transferrin receptor in human brain capillary endothelium. Brain Res. 712:117-21. Saito, K., and Y. Nakanuma. 1992. Lactoferrin and lysozyme in the intrahepatic bile duct of normal livers and hepatolithiasis. An immunohistochemical study. J Hepatol. 15:147-53. Sala, R., W.A. Jefferies, B. Walker, J. Yang, J. Tiong, S.K. Law, M.F. Carlevaro, E. Di Marco, A. Vacca, R. Cancedda, F.D. Cancedda, and D. Ribatti. 2002. The human melanoma associated protein melanotransferrin promotes endothelial cell migration and angiogenesis in vivo. Eur J Cell Biol. 81:599-607. Sarangarajan, R., and S.P. Apte. 2006. The polymerization of melanin: a poorly understood phenomenon with egregious biological implications. Melanoma Res. 16:3-10. Sawyer, S.T., and S.B. Krantz. 1986. Transferrin receptor number, synthesis, and endocytosis during erythropoietin-induced maturation of Friend virus-infected erythroid cells. JBiol Chem. 261:9187-95. Schade, A.L., and L. Caroline. 1946. An lion-binding Component in Human Blood Plasma. Science. 104:340-341. Schaeffer, E., M.A. Lucero, J.M. Jeltsch, M.C. Py, M.J. Levin, P. Chambon, G.N. Cohen, and M.M. Zakin. 1987. Complete structure of the human transferrin gene. Comparison with analogous chicken gene and human pseudogene. Gene. 56:109- 16. Scherer, P.E., T. Okamoto, M. Chun, I. Nishimoto, H.F. Lodish, and M.P. Lisanti. 1996. Identification, sequence, and expression of caveolin-2 defines a caveolin gene family. Proc Natl Acad Sci USA. 93:131-5. Sciot, R., R. de Vos, P. van Eyken, K. van der Steen, P. Moerman, and V.J. Desmet. 1989. In situ localization of melanotransferrin (melanoma-associated antigen P97) in human liver. A light- and electronmicroscopic immunohistochemical study. Liver. 9:110-9. SEER-Surveillance. 2009. Epidemiology and End Results Data. Vol. 2009. National Cancer Institute. Sekyere, E., and D.R. Richardson. 2000. The membrane-bound transferrin homologue melanotransferrin: roles other than iron transport? FEBS Lett. 483:11-6. Shyng, S.L., J.E. Heuser, and D.A. Harris. 1994. A glycolipid-anchored prion protein is endocytosed via clathrin-coated pits. J Cell Biol. 125:1239-50. 112

Simonart, T., J.P. Van Vooren, I. Herbauts, and J.R. Boelaert. 1999. High incidence of Kaposi’s sarcoma in Iceland and the Faroe Islands. Br J Cancer. 79:373. Simonart, T., J.P. Van Vooren, D. Parent, M. Heenen, and J.R. Boelaert. 2000. Role of iron in dermatology. Dermatology. 200:156-9. Sipe, D.M., and R.F. Murphy. 1991. Binding to cellular receptors results in increased iron release from transferrin at mildly acidic pH. JBiol Chem. 266:8002-7. Smart, E.J., C. Mineo, and R.G. Anderson. 1996. Clustered folate receptors deliver 5- methyltetrahydrofolate to cytoplasm of MA1O4cells. J Cell Biol. 134:1169-77. Smith, C.A., B.F. Anderson, H.M. Baker, and E.N. Baker. 1992. Metal substitution in transferrins: the crystal structure of human copper-lactoferrin at 2.1-A resolution. Biochemistry. 31:4527-33. Smith, L.M., A. Nesterova, S.C. Alley, M.Y. Torgov, and P.J. Carter. 2006. Potent cytotoxicity of an auristatin-containing antibody-drug conjugate targeting melanoma cells expressing melanotransferrinlp97. Mol Cancer Ther. 5:1474-82. Society, A.C. 2009. Cancer Reference Information. Vol. 2009. American Cancer Society. Song, K.S., P.E. Scherer, Z. Tang, T. Okamoto, S. Li, M. Chafel, C. Chu, D.S. Kohtz, and M.P. Lisanti. 1996. Expression of caveolin-3 in skeletal, cardiac, and smooth muscle cells. Caveolin-3 is a component of the sarcolemma and co-fractionates with dystrophin and dystrophin-associated glycoproteins. J Biol Chem. 271:15160-5. Stahl, A., and B.M. Mueller. 1995. The urokinase-type plasminogen activator receptor, a GPI-linked protein, is localized in caveolae. J Cell Biol. 129:335-44. Stevens, R.G., B.I. Graubard, M.S. Micozzi, K. Neriishi, and B.S. Blumberg. 1994. Moderate elevation of body iron level and increased risk of cancer occurrence and death. mtj Cancer. 56:364-9. Stubbe, J. 1990. Ribonucleotide reductases. Adv Enzymol Relat Areas Mol Biol. 63:349- 419. Sussman, H.H. 1992. Iron in cancer. Pathobiology. 60:2-9. Swetter, S.M. 2009. Malignant Melanoma. In Dermatology. Vol. 2009. D.M. Elston, editor. emedicine. Tang, Z., P.E. Scherer, T. Okamoto, K. Song, C. Chu, D.S. Kohtz, I. Nishimoto, H.F. Lodish, and M.P. Lisanti. 1996. Molecular cloning of caveolin-3, a novel member of the caveolin gene family expressed predominantly in muscle. J Biol Chem. 271:2255-61. Thomas, C.M., and E.J. Smart. 2008. Caveolae structure and function. J Cell Mol Med. 12:796-809. Thorn, H., K.G. Stenkula, M. Karlsson, U. Ortegren, F.H. Nystrom, J. Gustavsson, and P. Stralfors. 2003. Cell surface orifices of caveolae and localization of caveolin to the necks of caveolae in adipocytes. Mol Biol Cell. 14:3967-76. Trenor, C.C., 3rd, D.R. Campagna, V.M. Sellers, N.C. Andrews, and M.D. Fleming. 2000. The molecular defect in hypotransferrinemic mice. Blood. 96:1113-8. Trinder, D., 0. Zak, and P. Aisen. 1996. Transferrin receptor-independent uptake of differic transferrin by human hepatoma cells with antisense inhibition of receptor expression. Hepatology. 23:1512-20. Trowbridge, I.S., and M.B. Omary. 1981. Human cell surface glycoprotein related to cell proliferation is the receptor for transferrin. Proc NatlAcad Sci USA. 78:3039-43. 113

Udenfriend, S., and K. Kodukula. 1995. How glycosyiphosphatidylinositol-anchored membrane proteins are made. Annu Rev Biochem. 64:563-9 1. Urteaga, 0., and G.T. Pack. 1966. On the antiquity of melanoma. Cancer. 19:607-10. Verkade, P., T. Harder, F. Lafont, and K. Simons. 2000. Induction of caveolae in the apical plasma membrane of Madin-Darby canine kidney cells. J Cell Biol. 148:727-39. Ward, J.H., J.P. Kushner, and J. Kaplan. 1982. Regulation of HeLa cell transferrin receptors. JBiol Chem. 257:10317-23. Ward, P.P., M. Mendoza-Meneses, G.A. Cunningham, and O.M. Conneely. 2003. Iron status in mice carrying a targeted disruption of lactoferrin. Mol Cell Biol. 23:178- 85. Way, M., and R.G. Parton. 1995. M-caveolin, a muscle-specific caveolin-related protein. FEBS Lett. 376:108-12. Weinberg, E.D. 1996. The role of iron in cancer. Eur J Cancer Prey. 5:19-36. Weinstock, M.A. 2001. Epidemiology, etiology, and control of melanoma. Med Health R I. 84:234-6. Wessling-Resnick, M. 2000. Iron transport. Annu Rev Nutr. 20:129-51. Williams, J., T.C. Elleman, I.B. Kingston, A.G. Wilkins, and K.A. Kuhn. 1982. The primary structure of hen ovotransferrin. Eur J Biochem. 122:297-303. Woodbury, R.G., J.P. Brown, S.M. Loop, K.E. Helistrom, and I. Hellstrom. 1981. Analysis of normal neoplastic human tissues for the tumor-associated protein p97. mtJ Cancer. 27:145-9. Woodbury, R.G., J.P. Brown, M.Y. Yeh, I. Hellstrom, and K.E. Helistrom. 1980. Identification of a cell surface protein, p97, in human melanomas and certain other neoplasms. Proc Nati Acad Sci US A. 77:2183-7. Yamada, T., Y. Tsujioka, J. Taguchi, M. Takahashi, Y. Tsuboi, I. Moroo, J. Yang, and W.A. Jefferies. 1999. Melanotransferrin is produced by senile plaque-associated reactive microglia in Alzheimer’s disease. Brain Res. 845:1-5. Ying, Y.S., R.G. Anderson, and K.G. Rothberg. 1992. Each caveola contains multiple glycosyl-phosphatidylinositol-anchored membrane proteins. Cold Spring Harb Symp Quant Biol. 57:593-604. 114

Appendix i: Crystallization of p97

Rationale:

The crystal structures of several members of the Tf family have been determined (Bailey et a!., 1988; Baker and Lindley, 1992; Kurokawa et al., 1995; Smith et al., 1992), however solving the crystal structure of p97 has been a challenge. Since structure of a protein can often convey important information about its function, thus solving the crystal structure of p97 protein will no doubt enable us to visualize the physical interaction between iron and the iron-binding site of p97. Due to the apparent difficulty with crystallizing human soluble p9’7,a recombinant form of human soluble p97 produced in Spodoptera frugiperda cell line (Sf9) were used. Briefly, 2.4 kb from p97 cDNA was subcloned into mammalian expression vector, from which 2.6 kb fragment was then subcloned into a Orgyia pseudotsugata vector. The resulting vector contain not only p97 but also a immediate early 2 (IE-2) gene promoter that direct constitutive expression in insect cell lines, as well as a chimeric IE-2 bacterial synthetic promoter that confers resistance to zeocin (Hegedus et al., 1999). The Sf9 cell line was stable with no overall decline in p97 expression over 12 passages, and secreted soluble p97 at 658ng/10 cells/hr (Hegedus et al., 1999). By using this recombinant form of human soluble p97, microheterogeneity commonly observed in p97 produced by mammalian cell lines was minimized. The resulting soluble p97 is highly uniform and slightly smaller compare to human soluble p97 due to the less complex carbohydrate modifications (Hegedus et a!., 1999). All of which are beneficial to p97 crystallization. The successful isolation of red crystals, due to the preloading of p97 with iron, proves to be encouraging in our attempts to crystallize p97. However, since crystals have to be >0.1 mm in its smallest dimension in order to be detected by X-ray diffraction, we were unable to confirm the exact identities of the crystals. Further crystallization attempts need to be made using preferable more concentrated p97 proteins to ensure maximum growth. 115

Experimental design and results:

The crystallization process was carried out using the most commonly used setup for protein crystals, the hanging drop technique. This method is based on the idea that overtime the protein/precipitant mixture in the drop slowly increases since the concentration of the reservoir solution is comparatively more concentrated. Subsequently, this would facilitate the formation of crystals. Previous graduate student in our laboratory, Jacqueline Tiong, has narrowed down the condition required using two pre made kits (Hampton research Crystal screen 1 and Crystal screen 2 and Emerald

Biostructures Wizard 1 screen and Wizard 2 screen). For this thesis, different ratio of Tf and precipitants were mixed together and placed on a glass cover slip, which is sealed over the reservoir. The conditions that successfully produced Tf crystals were then used in the crystallization for p97 protein. A month after the initial set up of the crystallization plates for Tf at 20°C, small crystals were observed in selective conditions containing 10/20/25% PEG-8000, 250/300mM NaCI, 100mM CAPS pHlO.5. Subsequently, these conditions were used in crystallization of p97 protein. After another couple of months, small needle or rectangular shaped crystals were observed in some of the wells of the crystallization plates. They appeared to have sharp edges and as well as regular shapes. In addition, the p97 crystals were red in color. On average, the crystals measured to about 0.3mm x 0.05mm x 0.04mm (Appendix Figure 1). 116

(a)

(b)

Appendix Figure 1: Unpolarized (a) and polarized (b) image of iron bound soluble p97 crystals. The red color of the crystal in (a) is the indication of the iron binding by p97. 117

Appendix ii: Cellular regulation of p97 expression. Rationale:

As a tumor grows, the number of tumor cells rapidly outgrows the blood supply that is available, thus leaving the cells deprived of oxygen. This often leads to a hypoxic microenvironment. Since it is known that hypoxia induces AP-l expression (Laderoute et aL, 2002) and GPI-anchored P97 possess two AP-l binding sites in its enhancer region, the expression of p97 in a hypoxic microenvironment and what is the relationship between hypoxia, AP- 1 and p97 was examined. Since the melanoma cell lines that are available express a constant level of p97 unless artificially modified through transfection of expression vector, I wanted to find a cell line that is able to have variation in its p97 expression. It has been shown previously that activated microglia cells in the brain of Alzheimer’ patients show elevated level of p97 compare to control patients (Jefferies et al., 1996; Yamada et al., 1999), so BV-2 mouse microglia cell line was chosen to conduct the following experiments. These experiments will hopefully shed some light on the cellular regulation of p97 expression in cells.

Experimental design and results:

To determine the HIF- 1 and p97 expression in normal non-activated BV-2 microglia cells, real-time RT PCR were performed. Endogenous expressions of HIF-l and GPI-anchored p97 in B16 and JBMS mouse melanoma cells were evaluated. Results show that B16F10 and JBMS cells express relatively the same level of HIF- 1, while non- activated BV-2 cells express HIF- 1 at much lower levels in comparison (Appendix Figure 2a). In addition, JBMS cells show 20 fold higher of p97 expression and the B16F10 cells show 10 fold higher of p97 expression than resting BV-2 cells (Appendix Figure 2b). Furthermore, to determine whether hypoxic environment is able to effect the expression of endogenous GPI-anchored p97expression, BV-2 cells were incubated at 1% oxygen hypoxic chamber over a 4 days period. Real time RT PCR was performed to determine the expression level of HIF-1 (Appendix Figure 3a) and p97 (Appendix Figure 3b) at each day. Results show that HIP- 1 expression increased dramatically on day 1 and decrease gradually over the next 3 days, while expression level of p97 increased 118 gradually over the 4 days period. To further investigate the possible relationship between GPI-anchored p97 expression and hypoxia, non-activated BV-2 cells were transiently transfected with a HIF-lexpression vector and real-time RT PCR was performed on the transfected cells for HIF- 1 (Appendix Figure 4a) and p97 expressions (Appendix Figure

4b). Result showed that increased HIF- 1 expression led to over a 6 fold of increase in p97 expression compare to non-transfected BV2 cells. This indicates that H1F-1 is able to induce the expression of p97 in vitro. Since it is known that hypoxia is able to stimulate expression of c-junJAP-1, and GPI-anchored p97 possess two AP-1 binding sites in its enhancer region, non-activated BV-2 microglia cells were transiently transfected with c-jun expression vector. RT-PCR results demonstrate that the increase in c-jun expression (Appendix Figure 5a) is able to significantly increase the endogenous expression level of GPI-anchored p97 (Appendix Figure 5b) in the transfectants compare to the vector control.

Taken together, these preliminary results show that endogenous GPI-anchored p97 expression in BV-2 microglia cells could possibly be up-regulated by hypoxia through transcription factor AP- 1. 119

(a)

400

0 C .2 U) 300 Co 0. 0) x 200 C 0) C Co

C.) 10 0) > Co Bi 6 JBMS BV-2

I I

(b)

0 C 0 0U) 0) 0. x C)

C

C CO

C.) C) >

Co

B16 JBMS BV-2

Appendix figure 2: Endogenous expression of HIF-1 (a) and GPI-anchored p97 (b) in B16F1Oand JBMS mouse melanoma cell lines and BV-2 mouse microglia cell line. ______

120

(a)

25000 0 0 g 20000

15000

ioooo 0)

5000

—.—,...— I 1 2 3 4 Culture Time (Day)

(b)

6’ 10 I 0

4

I:I:!::lT1 Culture time (Day)

Appendix figure 3: The effect of hypoxia on BV-2 mouse microglia cells. Over 4 days of culture, the level of endogenous HIF-l decrease (a), while the endogenous p97 level increased gradually (b). 121

(a)

600

0 C 0 0 00

200 a)

C.) a)

a)

C BV-2 + vector BV-2 ÷ HIF1

(b)

•0 0

C 0 0 0 a) a. a) a) a. C a) 0) C a) U a)

a) a)

BV-2 + vector BV-2 + HIF1

Appendix figure 4: HIP-i expression is able to up-regulate the endogenous expression of p97 in BV-2 mouse microglia cells. Non-activated BV-2 cells were transiently transfected with HIP-i expression vector. The transfectants show significant up- regulation of HIP-i (a), and subsequently led to -4 fold increase in endogenous p97 (b). 122

(a) I$9

Lb SIS

c-jun

(b) 4 -4-k -1; Lv 0Lv S15

p97

Appendix figure 5: c-junIAP-l is able to stimulate endogenous OPT-anchored p97 expression in BV-2 mouse microglia cells. Non-activated BV-2 cells were transiently transfected with c-jun expression vector. The RT-PCR result show significant up- regulation of c-jun (a) in transfectants, and subsequently led to significant increase in endogenous p97 (b). SlS housekeeping gene was used as control.