Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effect of heme oxygenase-1 on breast cancer metastasis in vitro and in vivo Kim, Ada Yongyeon 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2014_september_kim_ada.pdf [ 4.32MB ]
Metadata
JSON: 24-1.0167212.json
JSON-LD: 24-1.0167212-ld.json
RDF/XML (Pretty): 24-1.0167212-rdf.xml
RDF/JSON: 24-1.0167212-rdf.json
Turtle: 24-1.0167212-turtle.txt
N-Triples: 24-1.0167212-rdf-ntriples.txt
Original Record: 24-1.0167212-source.json
Full Text
24-1.0167212-fulltext.txt
Citation
24-1.0167212.ris

Full Text

    THE EFFECT OF HEME OXYGENASE-1 ON BREAST CANCER METASTASIS IN VITRO AND IN VIVO  by  Ada Yongyeon Kim B.Sc. (Hons.), The University of British Columbia, 2012   THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in   The Faculty of Graduate and Postdoctoral Studies (Pathology and Laboratory Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) May 2014    © Ada Yongyeon Kim, 2014 ii  Abstract  The metastatic spread of cancer is linked to over 90% of cancer-related deaths. Therapies designed to prevent the dissemination of metastatic cells from the primary tumour is a therapeutic strategy to improve outcome for patients at risk of developing metastatic disease. Heme oxygenase-1 (HMOX1) is the rate-limiting enzyme in heme catabolism and is induced by various stress stimuli. The role of HMOX1 in breast cancer metastasis is conflicting with some groups indicating that HMOX1 promotes metastasis while others indicating that HMOX1 reduces metastatic spread. Previously in our laboratory, HMOX1 expression was chemically induced with hemin, a potent and effective HMOX1 inducer, in murine mammary carcinoma cells to determine the effect of HMOX1 on tumour cell migration and invasion. Hemin reduced migration and invasion of all three cell lines when assessed by Boyden chamber transwell assays. Therefore, we hypothesize that HMOX1 reduces breast cancer metastasis by decreasing tumour cell migration and invasion.  We used three different murine mammary carcinoma cell lines as models for breast cancer: 67NR – noninvasive/nonmetastatic, 4TO7 – invasive/metastatic, 4T1 - highly aggressive metastatic behavior. We increased HMOX1 expression genetically or by chemical induction with another HMOX1 inducer, cobalt protoporphyrin (CoPP), to determine the effect of HMOX1 on migration and invasion. CoPP and HMOX1 overexpression reduced migration of all tumour cells in vitro. Interestingly, HMOX1 overexpression decreased invasion of 4T1 cells, but increased the invasion of 67NR and 4TO7 cells, modeling the dichotomy of HMOX1’s influence on breast cancer cell invasion. We also assessed the effect of HMOX1 overexpression and knockdown on lung metastases in vivo. In 4T1 tumours, HMOX1 overexpression had no effect on primary tumour growth or lung metastases. On the other hand, in 4T1 tumours with HMOX1 knockdown, there was a reduction in tumour growth and in lung metastases.  iii  Preface  I developed the hypothesis, the aims of this study and the experimental design with guidance from Dr. Kevin Bennewith. All in vitro experiments were performed by myself. For in vivo experiments, orthotopic implants were conducted by Dr. Kevin Bennewith and Nancy LePard while murine tissue harvests were assisted by Nancy LePard, Momir Bosiljcic, and Bryant Harbourne. Clonogenic assays were completed by Nancy LePard and the flow cytometry of lung samples were completed by Denise McDougal. I was responsible for all of the data collection, analysis and thesis composition.  All mouse work and methods were approved by the University of British Columbia’s Committee on Animal Care; project title: Promotion of Metastasis by Tumour Hypoxia and Myeloid Cells. Certificate # A13-0223 and A09-0251.   iv  Table of Contents  Abstract .................................................................................................................... ii Preface ..................................................................................................................... iii Table of Contents .................................................................................................... iv List of Tables .......................................................................................................... vii List of Figures ....................................................................................................... viii List of Symbols, Abbreviations, or Other .............................................................. x Acknowledgements ................................................................................................ xi Dedication............................................................................................................... xii Chapter 1 Introduction ............................................................................................ 1 1.1 Cancer .............................................................................................................. 1 1.1.1 Brief overview of the “hallmarks of cancer” ............................................... 1 1.2 Breast Cancer ................................................................................................... 2 1.2.1 Subtypes ................................................................................................... 3 1.2.2 Standard of care ....................................................................................... 5 1.2.3 Tumour model ........................................................................................... 6 1.3 Metastasis ........................................................................................................ 7 1.3.1 The metastatic process ............................................................................. 8 1.3.2 Migration and invasion ............................................................................ 12 1.3.3 Tumour microenvironments .................................................................... 17 1.4 Heme Oxygenase-1 (HMOX1) ........................................................................ 19 1.4.1 Structure ................................................................................................. 20 1.4.2 Regulation of HMOX1 ............................................................................. 21 1.4.3 Physiological role of HMOX1 .................................................................. 25 v  1.4.4 Role of HMOX1 in diseases .................................................................... 26 1.4.5 HMOX1 and cancer ................................................................................ 27 1.4.6 HMOX1 and breast cancer...................................................................... 29 1.4.7 Previous data on the role of HMOX1 in breast cancer metastasis .......... 30 1.5 Hypothesis and Aims ...................................................................................... 34 Chapter 2 Materials and Methods ......................................................................... 36 2.1 Cell Culture ..................................................................................................... 36 2.2 Genetically Modified Cell Lines ....................................................................... 36 2.2.1 Genetic overexpression of HMOX1 ......................................................... 36 2.2.2 RNA interference by small hairpin RNA of HMOX1 ................................ 37 2.3 Hemin and CoPP treatment ............................................................................ 37 2.4 Protein Analysis .............................................................................................. 37 2.5 In Vitro Migration and Invasion Assays ........................................................... 38 2.6 Confocal Microscopy ...................................................................................... 39 2.7 Gelatinase Zymography .................................................................................. 40 2.8 In Vivo Tumour Implants ................................................................................. 40 2.9 Tissue Processing .......................................................................................... 41 2.9 Flow Cytometry ............................................................................................... 41 2.10 Clonogenic assays ........................................................................................ 41 2.11 Immunofluorescence .................................................................................... 42 2.12 Statistical Analysis ........................................................................................ 42 Chapter 3 The Effect of HMOX1 on Migration and Invasion In Vitro ................. 43 3.1 Introduction ................................................................................................. 43 3.2  Results ........................................................................................................ 44 vi  3.2.1 Comparison of migration and invasion of chemically and genetically induced HMOX1 in 67NR and 4TO7 cells ........................................................ 44 3.2.2 Comparison of migration and invasion of chemically and genetically induced HMOX1 in 4T1 cells ........................................................................... 56 3.3 Discussion .................................................................................................. 66 Chapter 4 The Effect of HMOX1 on Metastasis In Vivo – 4TO7 and 4T1 Tumours ................................................................................................................................. 73 4.1  Introduction ................................................................................................ 73 4.2   Results ....................................................................................................... 73 4.2.1 Characterization of primary tumour growth and metastasis in the 4TO7 tumour model – HMOX1 overexpression and knockdown ............................... 73 4.2.2 Characterization of primary tumour growth and metastasis in the 4T1 tumour model – HMOX1 overexpression and knockdown ............................... 78 4.3 Discussion .................................................................................................. 85 Chapter 5 Summary and Future Directions ......................................................... 88 References.............................................................................................................. 94 Appendices........................................................................................................... 104 Appendix A: Migration and Invasion of murine mammary carcinoma cells. ........ 104  vii  List of Tables  Table 1 Description of mouse murine mammary carcinoma cell lines – 67NR, 4TO7, 4T1. ............................................................................................................................ 6   viii  List of Figures  Figure 1.1 The metastatic process. ........................................................................... 9 Figure 1.2 Tumour cell invasion and migration........................................................ 14 Figure 1.3 Structures of HMOX1 chemical inducers. .............................................. 23 Figure 1.4 Hmox1 mRNA in normoxia versus hypoxia versus hemin treatment...... 31 Figure 1.5 HMOX1 protein expression in normoxia versus hypoxia versus hemin treated. ..................................................................................................................... 32 Figure 1.6 Decrease in migration and invasion after hemin treatment in murine mammary carcinoma cells. ...................................................................................... 33 Figure 3.1 Treatment of 67NR and 4TO7 cells with CoPP decreases migration in vitro.. ........................................................................................................................ 45 Figure 3.2 Treatment of 67NR and 4TO7 cells with CoPP reduces invasion in vitro.  ................................................................................................................................. 47 Figure 3.3 HMOX1 knockdown does not rescue wild type migration phenotype in 67NR and 4TO7 cells. .............................................................................................. 48 Figure 3.4 HMOX1 knockdown does not rescue the wild type invasion phenotype in 67NR and 4TO7 cells.. ............................................................................................. 50 Figure 3.5 HMOX1 overexpression decreases migration in 67NR and 4TO7 cells similar to chemical induction.. .................................................................................. 51 Figure 3.6 HMOX1 overexpression enhances invasion in 67NR and 4TO7 cells in vitro.  ........................................................................................................................ 53 Figure 3.7 Immunofluorescent stainings of HMOX1 show non-differential cellular localization of HMOX1. . .......................................................................................... 54 Figure 3.8 MMP and MMP2 levels decrease in HMOX1 expressing 67NR and 4TO7 cells.  ........................................................................................................................ 55 Figure 3.9 Treatment of 4T1 cells with CoPP decreases migration in vitro.  ........... 57 Figure 3.10 CoPP decreases invasion in 4T1 cells in vitro. .................................... 58 Figure 3.11 HMOX1 knockdown does not rescue wild type migration phenotype in 4T1 cells. ................................................................................................................. 60 ix  Figure 3.12 HMOX1 knockdown does not rescue the wild type invasion phenotype in 4T1 cells. .............................................................................................................. 61 Figure 3.13 HMOX1 overexpression decreases migration in 4T1 cells similar to chemical induction. .................................................................................................. 62 Figure 3.14 HMOX1 overexpression reduces invasion in 4T1 similar to chemical induction.. ................................................................................................................ 63 Figure 3.15 Immunofluorescent staining of HMOX1 show non-differential cellular localization of HMOX1 in 4T1 cells. ......................................................................... 64 Figure 3.16 MMP and MMP2 levels decrease in HMOX1 expressing 4T1 cells. .... 66 Figure 3.17 Illustration of possible mechanism of action for the effect of HMOX1 on murine mammary carcinoma cell migration and invasion. ....................................... 72 Figure 4.1 HMOX1 overexpression has no effect on 4TO7 tumour growth or lung metastasis. ............................................................................................................... 74 Figure 4.2 Immunofluorescent staining of HMOX1 in 4TO7 tumour. ....................... 76 Figure 4.3 HMOX1 knockdown has no effect on 4TO7 tumour growth or lung metastasis. ............................................................................................................... 77 Figure 4.4 HMOX1 overexpression has no effect on 4T1 tumour growth or lung metastasis. ............................................................................................................... 78 Figure 4.5 Immunofluorescent staining of HMOX1 in a 4T1 tumour. ....................... 79 Figure 4.6 HMOX1 knockdown has an effect on 4T1 tumour growth and metastasis.  ................................................................................................................................. 80 Figure 4.7 HMOX1 levels in tumours correlate more with metastatic growth than primary tumour size. ................................................................................................ 82 Figure 4.8 Immunofluorescent staining of HMOX1 and normal cells in 4T1 tumours.  ................................................................................................................................. 84 Figure 5.1 Immunofluorescent staining of hypoxia, vasculature, apoptotic and proliferation markers in 4T1 tumours expressing shGFP or shHMOX1. .................. 90  x  List of Symbols, Abbreviations, or Other  AP-1  activator protein 1 ARE antioxidant-response element BACH1 BTB and CNC homolog 1 CAF carcinoma-associated fibroblasts  CAIX carbonic anhydrase IX  CO carbon monoxide CoPP cobalt(III) protoporphyrin IX chloride ECM extracellular matrix EGF epidermal growth factor EMT epithelial-mesenchymal transition ER estrogen receptor FAK focal adhesion kinase FBS fetal bovine serum FCS fetal calf serum HIF hypoxia-inducible factor HER2  human epidermal growth factor receptor 2 HMOX1 heme oxygenase-1 LOX lysyl oxidase  MMP matrix metalloproteinase MT1-MMP membrane type 1 metalloproteinase NOS NO synthase  Nox4 NADPH oxidase 4  NRF2 nuclear factor (erythroid-derived 2)-like  OCT optimal cutting temperature  PDT photodynamic therapy  PFA paraformaldehyde  PR progesterone receptor PXDN peroxidasin  ROS reactive oxygen species  SCC squamous cell carcinoma  SEM standard error of the mean sER smooth endoplasmic reticulum shRNA short hairpin RNA SNP single nucleotide polymorphism SSRBC sickled erythrocytes  TNBC triple-negative breast cancer Tregs regulatory T cells  TSS transcription start site  VEGF  vascular endothelial growth factor ZnPP  zinc protoporphryrin  xi  Acknowledgements  I would like to acknowledge the members of the Bennewith Lab for assisting me with the work included in this thesis as well as providing a wonderful place to work for the last 3 years. My greatest gratitude goes to Dr. Kevin Bennewith for being the most supportive and amazing mentor who has positively contributed to my growth as a young scientist. I would also like to thank my committee members, Dr. Sandra Dunn, Dr. Robert Nabi, Dr. Hayden Pritchard, and Dr. Marianne Sadar for their insightful guidance that helped shape this project. I would also like to acknowledge Dr. Andrew Weng and Sam Gusscott for providing the overexpression vector, Dr. Robert Nabi and Peter Wang for assisting in confocal microscopy, and Dr. Sam Aparicio, Elena Ostroumov and Charles Soong for the use of their fluorescent microscope.  Lastly, I would like to thank my family and friends for their unyielding support and encouragement throughout this process.   xii  Dedication    To my family.    Chapter 1 Introduction  1.1 Cancer  Cancer is a broad collection of diseases in which normal cells transform into neoplastic states. As the leading cause of death in Canada, cancer will have killed an estimated 75 500 Canadians, and 187 600 more will have developed cancer in 2013 (1).  Breast cancer is the most prevalent cancer in Canadian women, making up 26% of new cases, and 1 in 9 women will likely develop breast cancer in their lifetime (1).   A tumour or neoplasm forms when normal cells undergo uncontrolled growth. The abnormal tissue mass, which can be classified as benign, pre-malignant or malignant, can be derived from the same cell types, reach similar sizes, and be induced by equivalent causes that allow them to sustain growth signaling (2). The ability to invade tissues and to metastasize distinguishes malignant from benign tumours (3).   1.1.1 Brief overview of the “hallmarks of cancer”  As previously mentioned, cancer is defined as the transition of normal cells to a neoplastic state. Hanahan and Weinberg first proposed six “acquired capabilities” that are essential for cells to become malignant during tumour development and referred to these capabilities as the “hallmarks of cancer” (4). The first four hallmarks are associated with the capability of tumour cells to grow uncontrollably. These include sustained proliferative signaling permitting chronic proliferation in cancer cells, evading growth suppressors through alterations in tumour suppressor genes, resisting cell death by changing regulation of apoptotic machinery, and enabling replicative immortality via counteracting telomeric erosion (5). Induction of angiogenesis is another hallmark that complements this expansion of tumour cells 2  by generating neovasculature to provide oxygen and nutrients to the proliferating cells. This so-called “angiogenic switch” is activated early during tumour progression (6).  These hallmarks can also be characteristics of benign tumours since they are only concerned with the uncontrolled growth of tumour cells. In addition to the five hallmarks just described, the last of the original hallmarks of cancer is the activation of invasion and metastasis, which is reflected in higher pathological grades of malignancy, not seen in benign tumours (5). Metastasis will be comprehensively introduced in section 1.3.   More than a decade after they first proposed the six hallmarks of cancer, Hanahan and Weinberg added two more emerging hallmarks and two enabling characteristics of tumour cells (5). The two emerging hallmarks include reprogramming of energy metabolism, seeing as adjustment is necessary to accommodate excessive cell proliferation, as well as evasion of immune destruction.  The first enabling characteristic of cancer is genome instability and mutation. Given that cancer cells acquire mutation and epigenetic changes in genes related to tumorigenesis, they often increase their mutation rate (7). The second enabling characteristic is tumour-promoting inflammation. Such inflammation contributes to many of the previously described hallmarks by providing factors such as proangiogenic molecules and extracellular matrix (ECM) modifying enzymes to the tumour microenvironment (5). Collectively, these hallmarks and enabling characteristics are functionally distinct and yet complementary in their roles to ensure cancer cell survival and metastatic tumour development (4).   1.2 Breast Cancer  Breast cancer is a vastly heterogenous disease varying in histology, molecular profile, and clinical outcome. In 10-15% of breast cancer patients, distant metastases are present within 3 years of primary tumour diagnosis (8). It is not, however, atypical for metastatic disease to appear after 10 or more years. The most common sites of breast cancer metastases are the lung, liver and bone (8). A study 3  quantifying breast cancer data from the US National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) database has shown that the incidence of metastatic breast cancer (in contrast to locoregional breast cancer) is increasing in women younger than 40 years of age (9). This increase has been found to be independent of race, ethnicity, and location. The 5-year survival for women with metastatic breast cancer is 31.4% and for women with locoregional breast cancer is 86.8%. Therefore, therapies designed to prevent the dissemination of metastatic cells from the primary breast tumour hold great promise to improve the outcome for patients at risk of developing metastatic disease.  Breast cancer is most widely classified by a combination of histological grading and TNM classification (10). Staging breast cancer by TNM classification is defined by primary tumour size (T), dissemination to regional lymph nodes (N), and occurrence of distant metastases (M). With the accessibility of molecular analysis, hormone receptor and human epidermal growth factor receptor 2 (HER2) status determinations are also used in conjunction with the traditional classification systems.   1.2.1 Subtypes  Under histopathological classification, there are twenty different major breast tumour types and eighteen minor subtypes based on histomorphological characteristics (11).  Most breast cancers originate from epithelial ductal or lobular cells. However, there are also some that develop from the supporting stroma. Although histopathological classification defines numerous subtypes of breast cancer, 50-80% of breast cancers are diagnosed as “invasive ductal carcinoma not otherwise specified” (IDC NOS).  This diagnosis is often by default, as tumours that fail to be classified into one of the special histological types are classified as IDC NOS (11). The overall 10-year survival rate of patients diagnosed with IDC NOS is 35-50%.   4  Although the majority of patients are diagnosed as IDC NOS, the apparent heterogeneity that exists in breast cancer can be found through molecular profiling. There are five distinct breast cancer subtypes based on molecular profiling (12, 13). Tumours that highly express many of the genes normally expressed by breast luminal cells are grouped as “luminal” (12, 13). This group is estrogen receptor (ER)-positive and/or progesterone receptor (PR)-positive. Estrogen binds to the ER and stimulates cell proliferation leading to an increase in cells with possible oncogenic mutations. Luminal subtype has the best prognosis and lowest rate of relapse (14). The luminal subtype is further classified as “A” and “B”. Luminal A tumours express high levels of cytokeratin 8/18 and are HER2-, while luminal B tumours express low levels of genes expressed by luminal A tumours but have a higher level of HER2. The subtype in which there is high expression of genes characteristic of basal epithelial and adipose cells, and low expression of genes characteristic of luminal epithelial cells, reflecting a gene expression pattern of a normal breast, is labeled “normal-like” (12, 13). “Normal-like” tumours are often smaller and have good prognosis (15).   The next subtype is “HER2+” tumours. HER2/neu (or Erb-B2) is a receptor tyrosine kinase that activates many targets in the mTOR and MAPK pathway to promote cell proliferation and survival. HER2/neu is amplified in 20-30% of breast cancers (12, 13, 16). The overexpression of HER2/neu is associated with low levels of ER and genes associated with basal-like tumours. HER2 type is associated with poor prognosis and increased relapse and metastasis (14). Trastuzumab (Herceptin) is one of the first humanized antibody against HER2/neu and inhibits proliferation of Her2/neu overexpressing breast cancer cells by reducing signaling of the PI3K-Akt pathway (17).   Tumours with many gene characteristics of breast basal epithelial cells are grouped together and classified as “basal-like”. Basal-like tumours express cytokeratin 5/6 and fail to express estrogen receptor (ER), progesterone receptor (PR), and HER2/neu (12, 13). Triple negative breast cancer (TNBC) classification, which is 5  described as tumours with the absence of ER, PR, and HER2/neu, shares similarities to the basal-like subtype, and the two are often interchangeably classified. Basal-like and TN breast tumours account for 15% of breast tumours. These tumours have the worst prognosis out of all molecular subtypes and cannot be treated with hormone therapies or trastuzumab due to the absence of the receptors. Although within a certain histopathological classification, tumours often have similar molecular expression, there are instances in which heterogenous molecular profiles exist between tumours of the same histopathological subtype (18).  Comparing patterns of metastasis between each breast cancer subtypes show high rates of metastases in the brain, lung, bone, and distant nodes for basal-like breast tumours, while TNBC has similar patterns to basal-like breast cancer, but also has an increased propensity to the liver (19). In comparison, HER2-positive subtype metastasizes to the brain, lung and liver.   1.2.2 Standard of care  Screening for breast cancer is achieved by mammography. Since mammography screening was implemented in 1988, the incidence of breast cancer dramatically increased through the 1990s (1). Once the breast tumour is detected, the standard practice is to determine the ER, PR, and HER2/neu status of the newly diagnosed breast tumour. Identification of receptor status is standard practice for management of breast cancer as it is a predictive marker for response to cytotoxic chemo- and endocrine therapy. For early stage and low risk patients, tumours are locally controlled by removing the primary tumour by mastectomy, or lumpectomy followed by radiation therapy (20). As mentioned, metastasis is the main cause of death in cancer patients, therefore to target the disseminated breast tumour cells, chemo- and/or endocrine therapy is given after removal of the primary tumour (8). Over 80% of patients receive adjuvant therapy after primary tumour removal due to the inability to accurately predict metastatic development. In addition, patients should receive bilateral mammography at the time of their diagnosis since BRCA1/BRCA2 mutation carriers are at a risk of contralateral breast cancer (18). TNBC is insensitive to 6  therapies directed at breast cancer with receptor status such as trastuzumab, a HER2-directed therapy, and tamoxifen, an endocrine therapy.   1.2.3 Tumour model  The tumour models used in this thesis were three murine mammary carcinoma cell lines, 67NR, 4TO7, and 4T1. These cell lines were isolated from a spontaneous mammary tumour from a single female BALB/cfC3H mouse (21). Several distinct populations of cells were obtained from the original morphologically heterogeneous culture and 5 major cell lines (4T1, 66c14, 4TO7, 168FARN, 67NR) are now utilized in most studies. In 3D organotypic systems, 67NR and 4TO7 cells form highly branched 3D structures, while 4T1 cells form spheroids lacking branching (22).   Table 1 Description of mouse murine mammary carcinoma cell lines – 67NR, 4TO7, 4T1.  67NR 4TO7 4T1 Organotypic structures Highly branched Highly branched Spheroids lacking branching Primary Tumour Growth + + + Invasion - + + Sites of metastases - Lung, liver (micrometastases) Lung, liver, bone, brain (macrometastases)  These three tumour sublines are not only heterogeneous in morphology but they also display varying degrees of metastatic potential (23). When 4T1 cells are implanted in the mammary fat pad, tumour cells metastasize primarily by a hematogenous route rather than by disseminating to draining lymph nodes. 4T1 cells are capable of metastasizing to the lungs, liver, bone, and brain. These metastases are found earliest in the lungs, followed by the liver. 4TO7 cells spread hematogenously to the lungs and occasionally to the liver, but are unable to grow macrometastases in sygeneic BALB/c mice. 4TO7 cells are highly immunogenic in immunocompetent mice and therefore decrease in survival, restricting metastatic 7  growth. This is evident when 4TO7 cells are implanted in immunocompromised mice and develop metastatic nodules. Finally, although 67NR tumours can grow to sizes exceeding that of 4T1 tumours, 67NR cells are unable to intravasate as demonstrated by the absence of tumour cells in blood, lymph nodes, lungs or livers. Therefore, on the metastatic spectrum, 4T1 cells have the most aggressive metastatic behavior, 4TO7 cells are systemically invasive but unable to form macrometastases, and 67NR cells are noninvasive and therefore nonmetastatic. These cell lines are often referred to as isogenic cell lines. However,  transcriptome analysis at the single exon level by Dutertre et al. shows that between the primary tumours of these cell lines, there are genes expressed with differential exons (24). The number of genes with alternative exons between 67NR and 4T1 is the greatest at 257 genes, compared to 4TO7 and 4T1 which had 25 genes with alternative exons. Eckhardt et al. have also shown the occurrence of altered gene expression between 4T1 and 67NR tumours, and as expected the corresponding genes have functional roles in metastasis (25). 4T1 cell line models stage IV breast cancer in humans and has negative status for ER, PR, and HER-2. It is of ductal differentiation shown by positive E-cadherin staining and cytokeratin 5/6 expression similar to basal-like molecular subtype (26).  1.3 Metastasis  Metastatic cancer is caused by the spread of tumour cells from the primary tumour site to a distant tissue. It is associated with an increase in morbidity and is associated with 90% of cancer-related deaths (27). In order to manage metastatic disease, systemic chemotherapy is given before (neoadjuvant) or after (adjuvant) surgery. The goal of systemic therapy is to target disseminated tumour cells that are clinically undetectable or inaccessible at diagnosis (28, 29).  At what point in tumour progression do tumour cells disseminate from the primary tumour remains to be answered. There are two models of metastasis, linear progression and parallel progression, that hypothesize the metastatic progression 8  process (28). As the name suggests, the linear progression model proposes that the primary tumour proceeds to full malignancy and develops a tumour microenvironment before the tumour cells leave the primary tumour. In the linear model, the most malignant cancer cells are selected and metastasis arises from the advanced primary tumour. The parallel progression model depicts an earlier metastatic event, in which cancer cells leave the primary tumour well before the disease is detected clinically. Unlike the linear progression model, tumour cells that depart in the parallel progression model have independent genetic and epigenetic alterations compared to the cells in the primary tumour. Also, the tumour cells may continuously seed to different sites in parallel and adapt to the different niches provided by the metastatic sites. As suggested in the two models of metastasis, the timeline of genetic alteration occurrences that are responsible for metastasis is still poorly understood. An example of genetic alteration occurring in the primary human tumour is Ack1, a non-receptor tyrosine kinase. In ovarian, pancreatic, lung and esophageal cancers, copy number gains of the Ack1 gene and Ack1 mRNA overexpression are seen in both metastatic tumours and advanced-stage primary tumours, but absent in early-stage tumours (30). In contrast, genetic alterations can be found in metastatic cancer cells independent of modifications acquired in the primary tumour. For example, changes in expression of certain “metastamirs”, miRNAs that regulate formation of metastases, such as miR-10b can be found only in metastasized cancer cells (31, 32). This suggests that cancer cells acquire mutations late in tumour development that predispose them to metastasis. Regardless of the model, in order for metastasis to occur, it requires the following series of steps: invasion, intravasation, survival, arrest, extravasation, and metastatic colonization (Figure 1.1). 1.3.1 The metastatic process  The first step in the metastatic process is invasion, which consists of a series of steps that moves the tumour cell from its surrounding tissue to the lymphatics or directly into the blood circulation. The details of tumour cell invasion will be 9  discussed in the next section, but briefly, invasion consists of tumour cell adherence to the ECM, followed by proteolytic degradation of surrounding tissue and finally migration of the tumour cell through the tissue.     Figure 1.1 The metastatic process. A, the basement membrane is intact until tumour cells detach and invade the basement membrane and the stroma shown in B. C and D, tumour cells intravasate to lymph or blood vessels. E and F, tumour cells arrest and extravasate at a distant organ and establish micrometastases. G, proliferation and angiogenesis results in macrometastases. Figure from (33).   Once the tumour cell invades through the basement membrane and the stromal tissue, the cell migrates towards the blood or the lymphatic vessels, where it intravasates through the vessel walls. Tumour cells may also enter the blood circulation indirectly via the lymphatic system. Once the tumour cell enters the 10  vasculature, it must survive the harsh environment of the bloodstream before it extravasates at a distant site. Not only does the tumour cell need to survive the attack from various immune cells within the blood, it must also survive the mechanical stress from the turbulent flow of the bloodstream (34).   The next step in the metastatic process is the arrest and the extravasation of the tumour cell. Organs such as the bone, lung and liver, which have extensive capillary networks, are common metastatic targets. One explanation for organ specificity is the mechanical factors, including the spatial restriction that the small vessels pose on the tumour cell leading to efficient arresting of the flow of cancer cells (35). In addition to size restrictions, there are nonspecific arrests caused by tumour cells binding to coagulation factors.  Another explanation for organ specificity is the path that cancer cells travel within the blood circulation. For example, in breast cancer, after intravasation to the blood circulation, tumour cells travel to the heart followed by the lungs, then to the other organs such as the bone, liver, and brain. When the tumour cells intravasate into the lymphatics, the tumour cells drain into the lymph nodes and form metastatic colonies. There are also key tumour cell signaling complexes that contribute to specific homing of tumour cells to metastatic targets. An example is the stromal derived factor-1(SDF-1)/chemokine CXC receptor type 4 (CXCR4) complex (35). CXCR4 is expressed in human breast tumour cells, but not in healthy breast tissue, while SDF-1 is expressed in high levels in common breast cancer metastatic sites. Therefore, the interaction between SDF-1 ligand expressed at metastatic sites and CXCR4 receptor expressed in breast cancer cells promotes the homing of these cells to the metastatic sites. Interestingly, SDF-1 and CXCR4 interaction has also been implicated in recruitment and migration of endothelial progenitor cells and the induction of angiogenesis, a required step in metastatic colonization(36). SDF-1 that is expressed in this mechanism is mediated by heme oxygenase-1 (HMOX1).    Once the tumour cell reaches the secondary site, the cell extravasates and attaches to the subendothelial ECM. The microenvironment of the distant tissue is crucial for 11  successful metastatic colonization. Organ specificity of breast cancer metastatic colonization can be described by the “seed and soil” hypothesis, which states “when a plant goes to seed, its seeds are carried in all directions; but they can only live and grow if they fall on congenial soil” (37). According to the “seed and soil” hypothesis, the tissue environment (soil) containing various cells and ECM may facilitate organ specificity seen in metastatic colonization of tumour cells (seed). Furthermore, new findings show that tumour microenvironments also influence the metastasis of a tumour cell. An example of this is the concept of “pre-metastatic niche” formation, in which tumour secreted factors predeterminedly prime the metastatic site prior to tumour cell arrival and also indirectly mobilize bone marrow-derived progenitor cells (BMDC) (38). Therefore, both mechanical factors and seed-soil compatibility contribute to metastatic colonization at specific sites (39).  There are similarities and differences in molecular characteristics of tumour cells that metastasize to particular metastatic sites. For example, gene expression profiling of a MDA-MB-231 sub-line with increased lung metastatic activity show that multiple secreted and membrane-organized proteins are jointly expressed and induce lung metastasis, providing growth advantages exclusively in the lung microenvironment (40). There were two classes of genes within the lung metastasis signature. One subset, including matrix metalloproteinase 1 (MMP1), contributes to both primary tumourigenicity and lung metastatic predisposition. The other subset, includes genes such as CXCR4, and MMP2, contributes solely to facilitating lung metastasis. The liver is a suitable site for metastatic colonization due to stromal cells and the ECM (29). In bone metastatic breast cancer, there is an increase in MMP-1 and IL-11, and an increase in the interaction between the tumour cells and the bone microenvironment (41).   In mice, out of the large number of tumour cells intravasating into the circulation, only an estimated 0.02% of cells will actually form metastatic colonies (39). This evident elimination of most tumour cells is termed metastatic inefficiency (42). There are many steps of the metastatic process that makes it inefficient. Anoikis, which is 12  cell death upon detachment from the ECM, is one of the proposed mechanisms for metastatic inefficiency (43). Other groups propose that the inefficiency is at the early stage of the distant site attachment. When GFP-labeled metastatic HT1080 fibrosarcoma cells are tail-vein injected, most of the cells underwent apoptosis in the lungs within 24 hours of injection (42). Fidler and Nicolson observed the same inefficiencies in radioactive-labeled highly metastatic B16 melanoma tumour cells that were injected intravenously. By 24 hours, 85% of the cells that initially arrested in the lungs were lost (44). Chambers et al. found that early steps before the extravasation in hematogenous dissemination are efficient in experimental models (39). It is after the arrival of tumour cells at the organ that only a small subset of cells is able to proliferate and form micrometastases and from that small number of micrometastases, a small fraction will survive and vascularize to form macroscopic metastases.   As mentioned previously, it is not uncommon for breast cancer patients to develop metastatic disease 10 years after diagnosis of the primary tumour. This occurs due to tumour dormancy in which the patient is asymptomatic but still carries undetected disseminated tumour cells, supporting the metastatic model in which dissemination is an early event (45). Tumour dormancy can occur by cellular dormancy, in which cells arrest at G0-G1 in addition to developing mechanisms to escape immune surveillance, and by angiogenic dormancy which results from a balance between pro- and anti-angiogenic factors (45). During tumour dormancy, breast cancer cells can evolve into a “differentiated” state by inhibiting signals such as β1-integrins and EGF, leading to change in tumour cells to non-proliferating acinar structures (46). A switch in the dormancy can lead to micrometastatic growth and ultimately metastatic disease.   1.3.2 Migration and invasion  Invasion is an early step in the metastatic process. In order to gain access to the lymph and blood vessels, the tumour cell needs to invade the basement membrane, 13  consisting of a dense matrix of collagen, glycoproteins, proteoglycans, and the underlying interstitial stroma. Therefore, invasion involves changes in cell-cell and cell-ECM adherence followed by migration of the tumour cell towards the vessels.   First, a moving cell becomes polarized and elongated by modifying its shape and stiffness through actin filaments that grow to push the cell membrane in an outward direction and to interact with the ECM (47). This allows pseudopod extensions at the cell’s leading edge (Figure 1.2). Chemokines and growth-factor receptors such as EGF activate PI3K, generating phosphoinosites which are responsible for engaging small GTPases such as RAC and RHO (48).    Key mediators in cell-matrix interaction are integrins, which are heterodimeric transmembrane proteins that bind to specific ECM proteins (49). Specific heterodimers bind to a specific ECM substrate. For example, α5β1 integrin binds to fibronectin, while α6β1 and α6β4 integrins bind to laminin, and this enables signal transmission into or out of the cell (50, 51). Once the integrin binds to a ligand, they form clusters at the membrane, and interact with pertinent proteins involved in focal contact assembly such as focal adhesion kinase (FAK). This in turn recruits actin-binding and -regulatory molecules. Another example of an ECM protein receptor on the tumour cell is CD44 which binds to hyaluronan, a glycosaminoglycan found in the ECM (47).   Binding of ECM proteins to tumour cell receptors recruit proteases such as matrix metalloproteinases (MMP) and plasmin that degrade ECM at the attachment site (47).  Soluble proteases such as MMP1 (collagenase) and MMP2 (gelatinase) can bind to integrins such as α2β1 and αvβ3 respectively (52). This in turn sequentially induces surface proteases to become more concentrated near the binding sites. Collagenases such as MMP1 and MT1-MMP cleave collagen and other ECM macromolecules into smaller pieces, which are subsequently degraded by MMP2 and MMP9 (gelatinases). During the process of ECM degradation, proteases free chemokines and growth factors that are embedded within the ECM. This contributes 14  to the infiltration of macrophages and lymphocytes to this tumour region, and thereby augments the release of proteases and other inflammatory factors. The pro-angiogenic factor vascular endothelial growth factor (VEGF) is another example of a molecule released from the ECM. VEGF can accumulate in latent forms in the ECM and be released and activated by MMPs during ECM degradation (53).    Figure 1.2 Tumour cell invasion and migration. Figure from (47). 15  ECM degradation results in a creation of a migration track as the tumour cell moves dynamically through the ECM. As the tumour cell adheres to the matrix at the leading edge of the cell, it detaches at the trailing edge by disassembling the focal contact. This is facilitated by the binding of receptor tyrosine kinase to integrins which then activates the FAK-Src complex (54). Along with the shift in the focal adhesion to the leading edge of the tumour cell, active myosin II binds to actin filaments causing cell contraction (47). This leads to a gliding forward movement of the tumour cell. Adhesion and migration mechanisms, including the number of focal adhesions and integrin-ECM interactions, may differ from one cell type to another and even between cells of the same type depending on the environmental conditions.   Invasive tumour cells have protrusions called invadopodia formed by actin polymerizations that contain matrix-degrading organelles (55). Functional invadopodia that degrade the ECM are associated with motility proteins such as cortactin and cofilin, adhesion proteins, signaling proteins that regulate membrane remodeling, and membrane-associated proteases such as MMP2 and MT1-MMP (56). Invadopodia can mediate polarization of the cell by signaling directional protrusions during chemotaxis and both invadopodia and lamellipodia (the mobile edge of the cell) are found together at the cell front in 3D (55). Invadopodia can also be influenced by the rigidity of the ECM as seen in breast cancer in which cells vary their invadopodia formation depending on the ECM rigidity (57). In breast cancer cells, ECM degradation is greatest in ECM with rigidity similar to that of the stroma rather than of the basement membrane. The effect that ECM rigidity has on the formation of invadopodia is an example of how the tumour microenvironment can affect different motility behaviour. The normal cells within the tumour microenvironment may also regulate the formation of invadopodia as well as other protrusions (55).   Tumour cells can migrate individually or collectively in cell strands, sheets, and clusters (47). Tumour cells that migrate as individual cells change adherence from 16  neighbouring cells to the ECM. Epithelial-mesenchymal transition (EMT) is a characteristic alteration in which the expression of a key cell-to-cell adhesion molecule, E-cadherin, shifts to N-cadherin and vimentin, which are expressed in mesenchymal cells. The loss of E-cadherin or the inactivation of E-cadherin due to a mutation is frequently seen in human carcinomas (58). EMT may also mark the transition from collective to individual migration (47). TWIST is a transcription factor that regulates EMT and is highly expressed in metastatic breast cancer (59). Likewise, compared to 67NR and 4TO7 cells, 4T1 cells express higher levels of TWIST (59). Unexpectedly, highly invasive and metastatic 4T1 cells express E-cadherin in contrast to nonmetastatic 67NR cells that express vimentin and N-cadherin but not E-cadherin (60). In addition, 67NR cells exhibit a mesenchymal-like phenotype, while the 4T1 cells resemble an amoeboid in 3D culture. Interestingly, TGF-β, a mediator of EMT, is able to increase migration and invasion in 4T1 and 4TO7 cells, and treatment with TGF-β changes the morphology of 4TO7 cells to one that is similar to 4T1 cells (22, 61). Also, comparing miRNA expression between 4TO7 and 4T1 cells reveals an upregulation of miR200 in 4T1 cells (62). miR200 targets Zeb2 which inhibits E-cadherin expression. When miR200 expression is induced in 4TO7 cells, the tumours metastasize more frequently. Dykxhoorn et al. propose that this is attributed to the re-establishment of an epithelial-like phenotype at distant tissue sites enabling a more effective metastatic colonization (62).This suggests that the metastatic ability of these murine mammary carcinoma cell lines do not correlate with the genotypic or the phenotypic properties of EMT, but that other mechanisms may control their metastatic capabilities(60, 63).    The advantages of collective migration are the production of high concentrations of promigratory factors by the large cell mass, the protection of inner cells from immune cells, and the ability to maintain contact with the primary tumour while locally invading (47). Collective-cell movement has been detected histologically in invasive epithelial cancers as well as in primary murine mammary carcinoma cells grown in Matrigel, which is a reconstituted basement membrane (47, 64). In squamous cell carcinoma (SCC), collective invasion occurs when the cells maintain epithelial marks 17  such as E-cadherin and are unable to degrade the surrounding matrix. These SCC cells rely on stromal fibroblasts to remodel the ECM, and collectively follow behind the leading fibroblast (65). Along with force-mediated remodeling of the ECM by cell migration, fibroblasts degrade ECM with MMPs and generate tracks deposited with fibronectin and tenascin C in the ECM, which are then used by the cancer cells to invade.   Live cell examination of nonmetastatic and metastatic primary tumours shows that metastatic MTLn3 cells (highly invasive breast adenocarcinoma cells) have an average velocity of approximately 3.4 µm/min and move individually in a linear fashion while interacting with collagen fibres (66). However, in non-metastatic MTC tumours, cells move over each other and in a nonlinear pattern without interacting with collagen fibres. In another study, the relationship between TGF-β and motility in MTLn3 cells was examined by live imaging of the primary tumour (67). TGF-β was transient, and activated disseminating single cells. The inhibition of TGF-β switched the cells from moving individually to collectively. They also found that single cell motility occurred during intravasation into the blood vessels, while cells moving cohesively moved slower and spread to the lymphatic vessels.  There is also another type of transition called mesenchymal-amoeboid transition, in which tumour cells change from spindle shaped to roundish elliptoid (47). This transition is acquired in MDA-MB-231 during protease-independent migration, where the round shape of the cancer cell allows it to squeeze through narrow matrix regions following pre-existing fibre strands (68).  1.3.3 Tumour microenvironments  Historically, tumours had been thought to be a homogenous population of tumour cells. This view is far from accurate, as tumours are now known to contain complex heterogeneous microenvironments containing various cell types along with diversified subpopulations of tumour cells, and various environmental conditions. 18  Normal cells that are present in the tumour microenvironment contribute to the success of tumour survival. Endothelial cells form tumour-associated vasculature and are a prominent cell type within the tumour. Pericytes are a specialized mesenchymal cell type that “wrap” around the endothelium. The functional role of pericytes in the tumour is unknown. However, a study by Cooke et al. suggests that they are gatekeepers against tumour progression and metastasis, and that the lack of pericytes leads to an increase in hypoxia (69). Carcinoma-associated fibroblasts (CAF) within the tumour-associated stroma have been shown to enhance tumourigenesis by promoting growth, invasion, and angiogenesis (5). Tumour-associated stromal cells can arise from adjacent normal tissue, recruitment of bone marrow-derived stem and progenitor cells, and proliferation of existing stromal cells (5). Virtually all tumour lesions contain immune cells at varying densities and both tumour-promoting and tumour-suppressing leukocytes can be found (5, 70). Fully differentiated immune cells such as macrophage subtypes, neutrophils, and T and B lymphocytes are known to be tumour-promoting. In addition to fully differentiated immune cells, partially differentiated myeloid progenitors such as myeloid derived suppressor cells infiltrate the tumour and suppress cytotoxic T lymphocytes aiding tumour cells in escaping immune destruction (46). Immune cells are also major sources of various factors including matrix remodeling enzymes, pro-angiogenic factors, and growth factors, which enhance tumour progression.            Hypoxia (areas of low oxygen tension) is a characteristic of many solid cancers since the rate of tumour growth often exceeds the ability of the existing blood vessels to supply oxygen and nutrients. Hypoxia plays a large role in tumour progression by promoting aggressive tumour phenotypes leading to metastasis through the activation of hypoxia-inducible pathways (71, 72). The levels of hypoxia can range from severe (0.1% O2) to approaching normoxia (3-5% O2), and in tumours, anoxia also occurs at 0% O2. Adaptation to hypoxia is predominantly regulated by hypoxia-inducible factors (HIF). Approximately 1-1.5% of the genome is transcriptionally regulated by hypoxia and many of the hypoxia-inducible genes are involved in proliferation, metabolism, angiogenesis and metastasis (73). Carbonic 19  anhydrase IX (CAIX) is a well-known hypoxia-inducible gene and it regulates both intra- and extracellular pH in hypoxic conditions. Analysis of human breast cancers shows CAIX expression as an independent prognostic marker for distant metastases (74). Another gene regulated by hypoxia is a secreted protein called lysyl oxidase (LOX). LOX contributes to metastasis by mediating FAK activity and cell-to-matrix adhesion. It is also responsible for forming a pre-metastatic niche by cross-linking collagen in the basement membrane and recruiting bone marrow-derived cells (75).   Low oxygen levels typically impair aerobic metabolism and ATP production, decreasing the viability of the cell. However in tumour cells, adaptive changes, including a switch from aerobic to anaerobic metabolism, allows survival and proliferation of tumour cells in hypoxic conditions by upregulating glucose transporters and enzymes in the glycolytic pathway (76, 77). Since anaerobic metabolism is not enough to maintain metabolic activity, the release of angiogenic factors such as VEGF promotes vascularization. Compared to normal vasculature, which is organized and provides oxygen and nutrient supply to all cells, tumour vessels are disordered, far apart from one another, and are hyperpermeable due to incomplete or absent basement membranes (78, 79). Therefore tumour regions distant from the vessels often become necrotic. Unlike apoptotic cells, necrotic cells bloat and explode, releasing proinflammatory signals into the tumour microenvironment and therefore recruiting immune cells (5).   1.4 Heme Oxygenase-1 (HMOX1)  Among the many genes responsible for tumour progression, heme oxygenase-1 (HMOX1) has been studied in various types of cancer (80-105). Levels of HMOX1 mRNA and protein are increased by stress stimuli including lipopolysaccharide, heme, and peroxides. Therefore HMOX1 is thought to protect cells under oxidative stress. In the literature, the most prevalent role of HMOX1 in cancer is cytoprotection of tumour cells (82, 93, 94, 96, 97, 101, 105). The effect of HMOX1 on invasive phenotype of tumour cells have also been examined, however, there are 20  discrepancies in the conclusions and the mechanism behind HMOX1-mediated changes in invasion is still unknown (87, 90, 91, 100, 102, 104).   Heme oxygenase-1 is the first and rate-limiting enzyme in heme catabolism by cleaving the heme ring to form biliverdin along with free iron, and carbon monoxide (CO). The catalytic reaction, shown below, is unique because there are no cofactors required since heme serves as both the substrate and the cofactor.  heme + NAD(P)H + H+ + 3 O2  ↔ biliverdin + Fe2+ + CO + NAD(P)+ + 3 H2O (106)  The main functions of HMOX1 are to recycle iron molecules, primarily from senescent erythrocytes, for erythropoiesis, and to maintain cellular homeostasis during oxidative stress through production of antioxidants such as biliverdin. As an enzyme, it is traditionally known to be anchored to the membrane of the smooth endoplasmic reticulum (sER) and exists as a homodimer in most organisms. However, there are studies that suggest the existence of HMOX1 as a free protein in the cytoplasm and the localization of HMOX1 to the nucleus (84, 107, 108).  In the body, HMOX1 is most abundant in the spleen since it is the location of highest heme breakdown. There are three isoforms of heme oxygenase. HMOX1 is inducible in response to various stress stimuli including heme, reactive oxygen species (ROS), NO, heat shock, and hypoxia. In contrast, heme oxygenase-2 is constitutive and ubiquitously expressed in the spleen, brain, kidney and liver, and heme oxygenase-3 is catalytically inactive. All three isoforms are transcribed by different genes, which are located on separate chromosomes. HMOX1 and HMOX2 share 43% amino acid sequence identity, while HMOX1 and HMOX3 share less than 50% sequence identity, however, all isozymes differ in their amino acid structure (109, 110).  1.4.1 Structure  HMOX1 (EC 1.14.99.3) is a 288 amino acid long protein with a molecular mass of 32 kDa. The structure of HMOX1 is a single compact domain consisting mostly of α-21  helices with a conserved core (111). The fold is novel compared to the other known structures found in the Protein Data Bank. Within the core, two helices, labeled proximal and distal, sandwich heme creating a “heme pocket” that is the active site of HMOX1. The C-terminal hydrophobic tail of HMOX1 extends through the microsomal membrane, anchoring the enzyme on the sER membrane (112). The lack of a C-terminal hydrophobic tail does not interfere with the enzymatic mechanism of HMOX1.   1.4.2 Regulation of HMOX1  The mouse Hmox1 gene is approximately 7 kbp long with 5 exons and 4 introns (113). The proximal promoter of the mouse HMOX1 gene contains several conserved sequence elements found in the human Hmox1 promoter region, and multiple consensus binding sites of transcription factors such as AP-1 and cMyc.   1.4.2.1 Transcriptional regulation  The expression of Hmox1 is species- and cell-specific. Since HMOX1 is induced by various stress stimuli including heme, ROS, NO, heat shock, and hypoxia, it is regulated by numerous transcription factors (103, 106, 114). One of the main transcriptional regulation complexes of Hmox1 is Nrf2 (Nuclear factor (erythroid-derived 2)-like) and BACH1 (BTB and CNC homolog 1). Nrf2 is part of the cap-n-collar sub family of basic leucine zipper transcriptional factors and BACH1 is a distant member of the family (115, 116). Nrf2 and BACH1 bind to a cis-acting enhancer element known as an antioxidant-response element (ARE). In the Hmox1 promoter, Nrf2 and BACH1 both bind to two sites containing multiple ARE motifs; one labeled E1, at 3928 bp upstream of TSS, and the other more distally located 8979 bp upstream, labeled E2 (117). By forming a heterodimer with a small Maf protein, BACH1 binds to the ARE and represses the transcription of Hmox1. This repression is dominant over Nrf2 activation of Hmox1 transcription and this repression cannot be overcome by excess nuclear accumulation of Nrf2 (117). 22  Therefore, in order for Hmox1 to be induced, BACH1 needs to be inactivated and unbound from the promoter, allowing basal Nrf2 that is already present in the nucleus to bind to the promoter as a heterodimer with a Maf protein and enhance Hmox1 transcription.   Variation in ARE motifs exists and leads to binding of other regulatory factors to ARE regions in addition to Nrf2 binding (118). Therefore this variation increases the number of transcription factors aside from BACH1 that can also compete to bind to ARE regions bound by Nrf2. In addition, polymorphisms are found in the promoter region of the Hmox1 gene. Microsatellite (GT)n dinucleotide repeats and single nucleotide polymorphisms are present in the promoter producing alternative 5’UTR splicing, influencing mRNA and ultimately affecting the protein expression of HMOX1 (119). There are multiple transcription start sites (TSS) with polymorphism clusters located downstream of the annotated TSS but upstream of the translation initiation codon. Aside from the E1 and E2 ARE motifs found upstream of the typical first exon, there is a possible second promoter with 2 clusters of transcription factor binding sites 1 kb upstream of exon 1 (119). This suggests that there may be binding of other transcription factors other than Nrf2 and BACH1 at this second promoter that are regulating Hmox1 transcription.   1.4.2.2 Hemin and cobalt protoporphyrin  HMOX1 can be induced by metalloporphyrins such as hemin. Hemin (Iron (III) protoporphyrin IX chloride) is a substrate (heme) analogue of HMOX1. It contains an iron atom in a coordinate bond at the centre of a protoporphyrin IX ring with a chloride ligand (Figure 1.3A). Hemin is a well-known inducer of HMOX1 expression at both mRNA and protein levels (120). Similar to the mechanism of HMOX1 induction caused by high cellular concentration of heme, hemin binds directly to BACH1, which leads to a conformational change and the release of BACH1 from the enhancer region (117). Decrease in DNA binding by BACH1 permits Nrf2-Maf to occupy the ARE, activating HMOX1 transcription. BACH1 is predominantly localized 23  to the nucleus. However within 30 minutes of hemin treatment, BACH1 can be found in the cytoplasm (117). Shan et al. have compared the induction of HMOX1 expression between hemin-treated cells and cells with BACH1 knocked down, and found that hemin-treated cells have a greater induction of HMOX1 expression than BACH1 repression (121). Therefore hemin may be affecting HMOX1 transcription outside of its interaction with BACH1.   A          B                       Figure 1.3 Structures of HMOX1 chemical inducers. A, hemin, Iron (III) protoporphyrin IX chloride. B, CoPP, Cobalt (III) protoporphyrin IX chloride.  Apart from hemin, CoPP (Cobalt (III) protoporphyrin IX chloride) is another metalloporphyrin that potently and effectively induces HMOX1 expression at both mRNA and protein levels (122, 123). CoPP has a similar structure and chemistry to hemin but a cobalt ion replaces the iron found in the heme ring (Figure 1.3B). Induction of HMOX1 by both hemin and CoPP is time and dose dependent. CoPP differs from hemin in that it is not a substrate of HMOX1. CoPP induces HMOX1 expression by the destabilization of BACH1 protein post-transcriptionally and has no effect on the mRNA expression of BACH1 (122). Conversely, CoPP stabilizes the half-life of Nrf2 from 2.5 to 9 hours. Although it is unclear whether CoPP and hemin have the same effect on BACH1 and Nrf2, CoPP has a greater effect on HMOX1 24  induction than does hemin, suggesting a difference in their mechanism of action (122).   1.4.2.3 Induction by other stressors  HMOX1 is highly inducible by various stress stimuli, including heme, ROS, NO, heat shock, hypoxia, arsenicals and transition metals such as cadmium and cobalt. Arsenicals such as sodium arsenite induce chemical oxidative stress in cells. Sodium arsenite activates MAP kinase which leads to the increase in AP-1, followed by binding of AP-1 to multiple AP-1 consensus elements found on the HMOX1 promoter (124). Transcriptional activation by AP-1 is not seen in hemin or CoPP induction. HMOX1 is also referred to as heat shock 32 protein, and in mouse and rat, hyperthermia induces HMOX1 expression, however in human HMOX1, neither mRNA nor protein change is seen after heat shock (120, 125). HMOX1 is induced in oxidative stress such as peroxides and ROS accumulation since biliverdin and bilirubin, products of heme breakdown, are antioxidants that scavenge singlet oxygen and act as reducing agents to peroxides (126).  The iron in heme is necessary for transport of oxygen to cells and tissues, therefore, iron deprivation and hypoxia have similar consequences by inducing genes necessary to adapt to low oxygen conditions (116). In mouse HMOX1, two functional hypoxia response elements are found in a 163 bp region located 9.5 kb away from the TSS (127). HIFs can bind to and activate HMOX1 transcription to provide an increased amount of iron to enhance oxygen uptake and delivery. However, the general response of HMOX1 to hypoxia is cell- and species-specific. For example, in human glioblastoma cells and lung cancer cells, hypoxia represses HMOX1 expression, while in other cells and species such as rat smooth muscle cells, HMOX1 is induced by hypoxia (116, 128, 129). Hypoxia-inducible repression of HMOX1 expression may be mediated by BACH1, as BACH1 transcript and protein levels increase in hypoxia in certain cell types, which lead to complete reduction of HMOX1 expression (127).    25  1.4.2.4 Post-transcriptional regulation and protein turnover  Post-transcriptional regulation of HMOX1 includes mRNA targeting by miRNAs. miR217 and miR377 target the 3’-UTR of HMOX1 and the knockdown of these two miRNAs upregulates HMOX1 expression while  hemin treatment decreases miR217 and miR377 expression (130). Interestingly, miRNA signatures of ER+ breast cancer consist of miR217, while miR377 is a signature of PR status with 100% prediction accuracy (131). Regulation of deadenylation may change HMOX1 mRNA turnover in response to various cellular stresses (132). Exposure of NIH3T3 cells to NO increases HMOX1 mRNA stability up to 11 more hours compared to cells without NO exposure, while chemical and peroxide exposure increases stability by 5 hours.   TRC8, an endoplasmic reticulum associated E3 ligase, regulates HMOX1 turnover by ubiquitinating HMOX1 for proteasomal degradation (133, 134). The transmembrane domain of TRC8 binds to the transmembrane portion of HMOX1 during ubiquitination. In rat vascular smooth muscle cells, expression of HMOX1 induced by CoPP treatment or HMOX1 overexpression, the half life of HMOX1 is 8 hours (133).    1.4.3 Physiological role of HMOX1   The physiological roles of HMOX1 have been indicated by Hmox1 deficiency studies in vivo. Mating between heterozygous mice only yields 20% of the expected Mendelian ratio of Hmox1-/- mice, suggesting Hmox1 deficiency is partially lethal (135). As mentioned previously, normal functions of HMOX1 include iron homeostasis and protection against cellular stress. Approximately 70% of the body’s total iron is found in hemoglobin, therefore, HMOX1 mediates recycling of iron by metabolizing heme released by senescent red blood cells. The crucial role of HMOX1 in iron homeostasis is reflected in Hmox1-/- mice and Hmox1 deficient humans, in which iron deficient anemia is present (135, 136). Hmox1-/- mice and humans have reduced serum iron levels, however, iron depositions are found in the 26  liver and kidneys, which is not seen in Hmox1 wild type or heterozygous tissues. Tissue damage is also found in the liver of Hmox1-/- mice due to the inability of macrophages to degrade heme after phagocytosing senescent red blood cells (137).   In addition to iron homeostasis, induction of HMOX1 occurs in the presence of stress stimuli in order to adapt and respond to cellular stress and tissue injury (81, 94, 138-141). HMOX1 deficient mice also show this particular role of HMOX1, as Hmox1-/- mice have increased oxidized proteins and lipid peroxidation leading to chronic inflammatory disease, as well as increase in free radical generation after treatment with stress stimuli (135, 142).     1.4.4 Role of HMOX1 in diseases  Heme oxygenase-1 is an enzyme studied in a wide range of diseases due to its various functions. Cytoprotection is one of the most prevalent roles of HMOX1. For example, in tissues of Hmox1-/- mice, there are excessive levels of γ-histone H2AX, a histone modification found in double-strand break repair (143). Irradiation of HMOX1 deficient cells leads to DNA damage, while HMOX1 induction induces homologous recombination-mediated DNA repair (143). The cytoprotective role of HMOX1 is also found in the kidneys, where HMOX1 provides protection from oxidative injury in kidney epithelial cells, and is implicated in renal disease (94).  Another role that HMOX1 plays in many diseases is through its potent anti-inflammatory effect. An in vitro study showed that inhibition of HMOX1 by tin protoporphyrin induces the activation, proliferation, and maturation of naïve CD4+ and CD8+ T cells by interacting with CD14+ monocytes (144). This reveals a role of HMOX1 in T cell homeostasis. HMOX1 produced by mesenchymal stem cells is also implicated in induction of regulatory T cells (Tregs) by promoting the formation of Type 1 regulatory T cells and T helper 3 cells, as well as the production of anti-inflammatory cytokine IL-10 (145). In murine macrophages, HMOX1 expression is induced by IL-10 and the production of CO from heme degradation is involved in the anti-inflammatory effect of IL-10 (146). NADPH oxidase 4 (Nox4) is the main source 27  of ROS in chondrocytes, and HMOX1 downregulates Nox4 expression which in turn decreases MMP1 expression and ultimately reduces cartilage matrix breakdown (147).  The implication of HMOX1 in angiogenesis has been found by wound healing studies. For example, after myocardial infarction, HMOX1 increases cardiac regeneration by increasing vascular smooth muscle, capillary and vascular densities (140). In endothelial cells, VEGF induces HMOX1, however the effects of HMOX1 on angiogenesis differ between in vitro and in vivo (148). Blocking HMOX1 inhibits VEGF-induced angiogenesis in vitro but in contrast increases angiogenesis in vivo. Additionally, blocking HMOX1 in vivo promotes dense leukocytic infiltration while HMOX1 activation leads to inhibition of leukocyte recruitment and inflammation-related angiogenesis. Leukocyte infiltration has also been seen in wound healing studies, where influx of leukocytes in wounds is significantly elevated after inhibition of HMOX1 by tin protoporphyrin (149).   These examples reflect the varying roles of HMOX1 on a variety of models of disease and the complexity involved in understanding the effect of HMOX1 on disease. This complexity can be translated into the role of HMOX1 in cancer. The literature on HMOX1 and cancer demonstrates the unpredictable nature of this well-studied enzyme, as its role and expression in cancer differ from one cancer to another.  1.4.5 HMOX1 and cancer  As a protein known to have cytoprotective, anti-inflammatory and pro-angiogenic functions in a wide range of disease, it is not unexpected for HMOX1 to have a role in cancer as well. In certain cancers, such as pancreatic cancer, HMOX1 mRNA and protein are upregulated in cancer cells compared to normal cells and within the tumour, both cancer cells and tumour-associated immune cells express HMOX1 (82). When pancreatic cancer cells are treated with gemcitabine or radiation, 28  HMOX1 is strongly induced (97). The sensitivity to these treatments increases when HMOX1 expression is repressed in these cells. HMOX1 also stimulates angiogenesis of pancreatic carcinoma and promotes tumour growth (99). Increased HMOX1 is seen in renal cancer, where the induction promotes tumour cell survival and inhibits ERK-induced apoptosis (81).   In prostate cancer, androgen-sensitive cells express higher levels of HMOX1 compared to androgen-insensitive cells (87). Induced expression of HMOX1 in PC3 cells (androgen-insensitive prostate cancer cell line), generates less vascularized tumours with decreased cell proliferation, and lower MMP9 expression (86, 87). Forced expression of HMOX1 in liver cancer cells decreases migration in vitro and tumour growth in vivo in part due to reduction in IL-6 production (104). Genome-wide expression profiling of uterine cancer cells shows that genes including MMP2, peroxidasin (PXDN), and TGF-β are upregulated in HMOX1 expressing cells compared to knockdown cells (100). Comparing the expression profile of the uterine cancer cells to 14 other human cancer types through data mining of preexisting expression databases by Tauber et al. revealed a differential expression pattern of HMOX1 and possible target genes throughout the cancer types as well as within each cancer type.  Aside from tumour cells, HMOX1 expression in normal cells present in the tumour microenvironment may also contribute to changes in tumour progression. For instance, the inhibition of inducible isoform of NO synthase (NOS), expressed by macrophages that do not express HMOX1, decreases HMOX1 mRNA expression in hepatoma cells (83). Unlike hepatomas where macrophages do not express HMOX1, in gliomas, infiltrating macrophages express HMOX1 and this closely correlates to vascular density (150). In addition, HMOX1 specific CD8+ T cells can be found in situ, and isolation of these cells from peripheral blood from cancer patients inhibits cytokine release, proliferation, and cytotoxicity of other immune cells (80). These HMOX1-specific CD8+ T cells have a significantly greater inhibitory effect than that of CD4+CD25+CD127- Tregs.  29  In addition to the examination of HMOX1 in cancer, groups have established a link between therapy resistance in cancer cells and the protection that HMOX1 provides cells against oxidative stress. In chronic myeloid leukemia, BCR/ABL-transformed cells have enhanced proliferation and reduced apoptosis, as well as BCR-ABL dependent HMOX1 expression (151). Resistance towards imatinib and other BCR/ABL tyrosine kinase inhibitors in CML may be attributed to HMOX1 expression, and treatment of HMOX1 inhibitors along with imatinib inhibits growth of imatinib-resistant cells (92). Chemoresistance to gemcitabine in pancreatic cancer increases with HMOX1 expression due to biliverdin and iron produced by heme degradation by HMOX1 (97). Therefore, inhibition of HMOX1 in pancreatic tumours with high expression increases chemotherapeutic efficacy of gemcitabine. Photodynamic therapy (PDT) induces stress-related genes, and in colon adenocarcinoma cells, induction of HMOX1 is seen after PDT and augments resistance to therapy (96). Inhibition of HMOX1 by a HMOX1 inhibitor, zinc protoporphyrin (ZnPP), decreases this resistance, and potentiates cytotoxic effects of PDT. In contrast to the expected role of HMOX1 in chemoresistance, in adult T-cell leukemia, induction of HMOX1 with CoPP increases bortezomib-induced apoptosis and the opposite effect is seen after HMOX1 induction with ZnPP (88).   1.4.6 HMOX1 and breast cancer  As presented above, the role of HMOX1 in cancer is varied. The role of HMOX1 in breast cancer has not been as widely studied as other cancers within the literature. The effect of HMOX1 on breast cancer metastasis is still under investigation. All studies that have examined the implications of HMOX1 on migration and invasion in breast cancer thus far have used a nonmetastatic human cancer cell line, MCF-7. In addition, the effect of HMOX1 has been only studied in vitro. Two groups have found that HMOX1 inhibits MCF-7 invasion by repressing MMP9 (91, 102). Wang et al. found that bone morphogenetic protein 6, belonging to the TGF-β superfamily, induces HMOX1 expression which leads to decreases in migration and invasion (102). This reduction is suggested to be attributed to the decrease in MMP9 30  expression found in HMOX1 overexpressing cells. Lin et al. observed similar decreases in MMP9 and in invasion of MCF-7 cells after induction of HMOX1 by hemin (91). Addition of CO but not ferric iron, biliverdin or bilibrubin reduced invasion by suppressing MMP9 expression through the suppression of the ERK/AP-1 signal pathway. Moreover, not much is known about the role of HMOX1 on proliferation and apoptosis breast cancer cells. Hill et al. found that protoporphyrin-induced and genetically induced expression of HMOX1 in MCF-7 cells decreased cell proliferation and induced apoptosis in vitro (89).   HMOX1-mediated therapies have also been studied in breast cancer models. Cermak et al. proposed that an increase in HMOX1 protein levels induces the resistance of tumours to chemotherapeutic agents due to the accumulation of iron from heme breakdown by HMOX1 (152). Terman et al. suggested using sickled erythrocytes (SSRBC), which selectively target hypoxic tumour vascular microenvironments, along with ZnPP to treat hypoxic solid tumours (101). In the 4T1 tumour model, they have shown that trapped SSRBCs induce the production of H2O2 in endothelial cells. Therefore by inhibiting HMOX1 by ZnPP, the tumour cells underwent apoptosis due to the lack of protection from a pro-oxidant that otherwise would have been provided by HMOX1.    1.4.7 Previous data on the role of HMOX1 in breast cancer metastasis  Previous work in our lab examined the effect of hemin on migration and invasion of mouse murine carcinoma cell lines, 67NR, 4TO7, and 4T1 (153). The in vitro expression of HMOX1 in these cell lines at normoxia was undetectable at both mRNA and protein level (Figure 1.5). Since HMOX1 is potentially a hypoxia-inducible gene in these cell lines, we treated the cells in hypoxia (1% O2) for 18 hours and analyzed HMOX1 expression levels. In hypoxia, mRNA levels were induced in 67NR and 4TO7 cells, but not in 4T1 cells, and HMOX1 protein expression was undetected in all three cell lines. Therefore, by utilizing hemin to 31  chemically induce HMOX1, we found significantly increases in mRNA and protein levels of HMOX1 in all three cell lines.       Figure 1.4 Hmox1 mRNA in normoxia versus hypoxia versus hemin treatment. Total RNA isolated from 67NR, 4TO7, and 4T1 cells that were incubated in normoxia, hypoxia, normoxia and hemin treated, or hypoxia and hemin treated. The Ct values of Hmox1 were normalized using Gapdh and the expression of Hmox1 was quantified by the ∆∆Ct method relative to the calibrator sample, the respective cell line incubated only in normoxia. Statistically significant differences in panel A were detected using one-way ANOVA, * represents p≤0.05. Data from (153).      32    Figure 1.5 HMOX1 protein expression in normoxia versus hypoxia versus hemin treatment. Protein was extracted from 67NR, 4TO7, and 4T1 cells that were incubated in normoxia, hypoxia, normoxia and hemin treated, or hypoxia and hemin treated. The expected molecular weight of HMOX1 is 34.6 kDa, however the actual band was specifically found at 30 kDa. HIF-1α levels were assessed to verify hypoxic conditions. Tubulin was used as the endogenous control. Data from (153).                33  A  B  Figure 1.6 Decrease in migration and invasion after hemin treatment in murine mammary carcinoma cells. Control and hemin-treated 67NR, 4TO7, and 4T1 cells were grown in 1 % FBS containing medium for 24 h before being placed in the upper chamber of the Boyden chamber. The lower chamber was filled with RPMI with 10 % FBS to stimulate the cells to migrate and invade the membrane separating the upper and lower chamber. Membranes in invasion assays were coated with Matrigel. After 24 h, cells that migrated to the lower chamber were fixed and stained with DAPI. A, migration. B, invasion. Images were obtained with a fluorescent microscope. The number of migrated cells was counted by ImageJ. Each bar represents a relative number of cells migrated per field compared to the untreated wild type control. Data given as mean ± SEM of independent experiments, N=6, one-tailed t-test, * represents p≤0.05. Data from (153).   34  Treatment of 67NR, 4TO7, and 4T1 cells with hemin yields a significant reduction in the number of tumour cells that have migrated and invaded as assessed by Boyden chamber transwell assays (Figure 1.6). These results from the migration and invasion assays support conclusions made in MCF-7 cells, in which the overexpression of HMOX1 exhibits a decrease in the invasive ability of MCF-7 cells (91, 102). A comparison of migration and invasion between the three murine carcinoma cells are shown in Appendix A. A limitation in our conclusion is that we cannot state for certain whether the decrease in migration and invasion is due to the induction of HMOX1 expression or whether it is due to possible off-target effects of hemin. Therefore, we can only conclude that the hemin treatment decreased migration and invasion. To verify that the reduction in migration and invasion is due to the increase in HMOX1 expression in these murine mammary cancer cells, further studies need to be conducted to verify the migration and invasion assays using genetic overexpression of HMOX1. In addition, this previous study only examined the possible role of HMOX1 in migration and invasion in vitro. As shown in various diseases and cancers presented above, the regulation and effect of HMOX1 are multifaceted, differences existing between cells and species, as well as discrepancies between the effects of HMOX1 in vitro compared to in vivo. Therefore, examining the effect of HMOX1 in vivo will be crucial in understanding the role of HMOX1 in breast cancer metastasis.   1.5 Hypothesis and Aims  The role of HMOX1 in breast cancer metastasis is controversial with some groups indicating that HMOX1 promotes invasion while others stating that HMOX1 reduces invasion (90, 91, 102).  We are interested in determining how HMOX1 affects breast cancer metastasis, and whether HMOX1 represents a potential therapeutic target for patients with metastatic disease. We hypothesize that heme oxygenase-1 reduces breast cancer metastasis by decreasing tumour cell migration and invasion.   35  We will address this hypothesis and assess the importance of HMOX1 in breast cancer metastasis by studying two specific aims: 1. To determine the effect of HMOX1 on tumour cell migration and invasion in vitro.  Murine mammary carcinoma cells will be treated with CoPP to examine whether another chemical inducer can reduce migration and invasion similar to hemin. Genetic knockdown of HMOX1 in murine mammary carcinoma cells will be utilized in combination with hemin to verify whether off-target effects of hemin are affecting migration or invasion. In addition, we will use murine mammary carcinoma cells genetically overexpressing HMOX1 to validate the results from hemin-treated cells.   2. To examine the effect of HMOX1 on primary tumour growth and metastasis in vivo.  4TO7 and 4T1 cells with genetic HMOX1 overexpression and knockdown will be orthotopically implanted into mice to determine how genetic manipulation of HMOX1 levels in tumour cells affects primary tumour growth and metastasis in vivo.  36  Chapter 2   Materials and Methods  2.1 Cell Culture  We obtained murine mammary carcinoma cell lines, 67NR, 4TO7, and 4T1 from Dr. Fred Miller (Karmanos Cancer Institutes, Detroit, MI). These cells were routinely cultured in RPMI 1640 medium supplemented with 1% sodium pyruvate, 10mM HEPES, and 10% fetal bovine serum at 37°C with 5% CO2.    2.2 Genetically Modified Cell Lines  67NR, 4TO7, and 4T1 cell lines were stably transfected to overexpress or knockdown HMOX1 by lentiviral transfection. 293T cell lines were transfected with RRE, REV, VSVG lentiviral packaging plasmids along with the plasmid of interest using LipofectAMINETM 2000 (Invitrogen, Carlsbad, CA). After 2 days, 293T conditioned medium was collected and filtered with a 0.45 µm Millipore filter. The filtered supernant containing lentivirus was added to approximately 70% confluent 67NR, 4TO7, and 4T1 cells. Stable transfectants of 67NR, 4TO7, and 4T1 cells were selected with 6, 10, 20 µg/mL puromycin respectively (Sigma-Aldrich, St. Louis, MO).     2.2.1 Genetic overexpression of HMOX1  Mouse HMOX1 cDNA was cloned into a mammalian expression lentiviral vector with a puromycin resistance marker, pRRLSIN.cPPT.PGK-PURO.WPRE (G537 Puro) provided by Dr. Andrew Weng (BC Cancer Agency, Vancouver, BC). The cDNA was obtained by PCR using pCMV-SPORT6-HMOX1 (Thermo Scientific, Waltham, MA) as template and forward 5’- AAAGGATCCAGTCCGGTGATGGAGCGTCCA-3’, reverse 5’-TTTGAATTCTTTATTATTTCACACAGAAGTTAGAG-3’ as primers.   37   2.2.2 RNA interference by small hairpin RNA of HMOX1  Five mouse pLKO.1 lentiviral shRNA constructs for HMOX1 (Thermo Scientific) were tested for knockdown efficiency by protein analysis through a western blot. Clone ID# TRCN0000071578 and #TRCN0000071581 respectively called shHMOX1 #1 and #4 were chosen for best knockdown in 67NR, 4TO7, and 4T1 cells.     2.3 Hemin and CoPP treatment  30 mM stock solution of hemin (Sigma-Aldrich) was made in 20 mM sterile NaOH and utilized for all hemin treatments. 25 mM stock solution of CoPP (Frontier Scientific, Logan, UT) was made in 20 mM sterile NaOH and utilized for all CoPP treatments. The protoporphyrins were prepared in dim light and stock solutions were stored in 4°C and away from light due to their light sensitivity. Cells were treated with hemin and CoPP after cells had adhered to the culture plate. For migration and invasion assays, cells were treated for 18 h prior to plating in transwell chambers and treated for an additional 24 h for the assays.   2.4 Protein Analysis  Cells were washed with PBS then lysed in 300-400 μL urea lysis buffer (9 M urea, 75 mM Tris-HCl) and immediately stored in -80°C overnight. After thawing, the lysates were sonicated for 5 sec to shear the DNA, and then centrifuged for 10 min at 20800 rcf to remove cell debris. The protein lysates were stored in -80°C. The protein concentrations of the cell lysates were measured by a Bradford protein assay in a 96-well format (Bio-Rad, Mississauga, Canada) using the manufacturer’s protocol.  38  A total of 30-45 μg of the cell lysate was separated by 10% SDS-polyacrylamide gel electrophoresis, and then transferred to a nitrocellulose membrane. The membrane was blocked in blocking buffer (1X TBS, pH 7.6, 0.1% Tween-20 with 5% w/v nonfat dry milk) for 30 min at room temperature. Mouse monoclonal antibody directed against HMOX1 was from Abcam (Abcam, Cambridge, MA), and mouse monoclonal anti-Tubulin was from Sigma. The membranes were probed with, anti–HMOX1 and Tubulin at 1/250 and 1/40000 dilutions respectively in BSA buffer (1X TBS-T [pH 7.6, 0.1% Tween-20], 20% BSA, 0.4% sodium azide). After the primary antibody incubations, the blots were washed 3 times with TBS-T for 3 min each. Secondary HRP (horseradish peroxidase)-linked antibodies were incubated for 45 min at 1/40000 dilution in blocking buffer for anti-tubulin and 1/3000 for anti-HMOX1 primary antibodies. After secondary antibody incubations, the blots were washed 5 times with TBS-T for 3 min each time. The antibody reactions were detected by enhanced chemiluminescence (ECL) (PerkinElmer Inc., Waltham, MA).   2.5 In Vitro Migration and Invasion Assays  Migration of 67NR, 4TO7, and 4T1 cells were assessed by Boyden chamber transwell assays. Cell culture inserts with a transparent polyethylene terephthalate (PET) membrane (8 μm pore size, 6.4 mm diameter) (BD Biosciences, Mississauga, Canada) were used. 1x106 cells were plated on 10-cm culture dishes and incubated overnight in standard medium. After 24 h, the cells were rinsed in medium containing 1% FBS and grown in this low-serum containing medium for 24 h. 1% FBS was used to starve the cells because it was the lowest concentration of FBS in which the cells were viable. Then, the cells were treated with trypsin and resuspended in 1% FBS containing medium to a concentration of 4x105 (67NR), 1x105 (4TO7),  0.5x105 (4T1) cells. In the lower chamber of the Boyden Chamber (a well of 24-well plate), 600 μL of complete medium was placed, while 300 μL of the medium containing the starved cells was placed in the upper chamber. The cells were incubated in the chambers for 24 h.  39  To stain the cells that have migrated across the membrane to the lower chamber side of the membrane for imaging, the cells remaining on the upper chamber side of the membrane were removed with a cotton swab moistened with PBS, and then the cell culture insert was washed with PBS three times. Subsequently, the insert was incubated in 100% methanol for 10 min, followed by 10 min in DAPI nucleic acid stain (0.5 μg/mL). After staining, the insert was washed with PBS and the membrane was removed from the insert by a sharp scalpel blade. The removed membrane was placed on a microscope slide and immersed in mounting medium. Invasion of 67NR, 4TO7, and 4T1 cells were assessed similarly to the migration assays, however BD BioCoatTM MatrigelTM Invasion Chambers (BD Biosciences) were used according to manufacturer’s protocol. The cell seeding densities for invasion assay are as follows: 4x105 (67NR), 2x105 (4TO7), 0.5x105 (4T1) cells. The migrated and invaded cells were imaged with Retiga EX camera (QImaging, Surrey, BC) at 10X original magnification. The DAPI stained cells were counted using ImageJ software.  2.6 Confocal Microscopy  Cells were cultured on sterile microscope coverslips to approximately 80% confluency. Cells were fixed with 4% PFA at room temperature for 15 min. Cells were permeablized with 0.2% Triton X-100 in PBS. After blocking in 0.5% BSA in PBS (blocking buffer), primary goat polyclonal anti-calnexin (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and primary rabbit polyclonal anti-HMOX1 (Abcam) were applied for 1 h, diluted 1:50 and 1:100 respectively in blocking buffer. Secondary detection reagent was Alexa Fluor® 594 Donkey Anti-Goat IgG and Alexa Fluor® 488 Chicken Anti-Rabbit IgG (Invitrogen), applied for 45 min at room temperature. Images were captured at 100X original magnification with a spinning-disk confocal microscope (Zeiss M200; assembled by Intelligent Imaging Innovation) and with SlideBook 5.0 operating software.      40  2.7 Gelatinase Zymography  The enzyme level of MMP2 and MMP9 was analyzed by gelatin zymography. The cells were cultured in RPMI 1640 + 1% FBS for 48 h with a starting seeding density of 1x106 cells. After 48 h, the conditioned medium was collected and the protein concentration was measured by a Bradford protein assay. With an exception of MMP9 zymogram of 4T1 cells, conditioned medium containing 15 µg of total protein was mixed with non-reducing loading buffer for each sample and loaded onto 10% SDS-polyacrylamide gel containing 1mg/mL collagen. Electrophoresis was performed at 125 mV, 40 mA for 8 h at 4°C. Gels were washed twice with 2.5% Triton X-100 at room temperature then incubated at 37°C for 41 h in the activation buffer (10mM Tris-HCl, pH 7.5, 1.25% Triton X-100, 4.9 mM CaCl2, 9.8 µM ZnCl2). The gel was stained with Coomassie Blue R-250 for 2 h and destained with a destaining solution (50% MeOH, 16.6% acetic acid) for 30 min. For MMP9 zymogram of 4T1 cells, medium containing 0.75 µg of total protein was utilized and the zymogram was activated for 24 h.   2.8 In Vivo Tumour Implants  Female BALB/c mice, 8-10 weeks of age, and female NOD/scid mice, 8-10 weeks of age were used in the experiments and purchased from Simonsen Laboratories (Gilroy, CA) or in house from the BC Cancer Agency Animal Resource Centre breeding colony respectively. 105 4T1 cells or 106 4TO7 cells were injected orthotopically in the fourth mammary fat pad. After 1 week post-implant, tumours were monitored by measuring tumour volume, pi/6 x L x W x H, every 2 days. All mice were housed under specific-pathogen free conditions in the Animal Resource Centre at the BC Cancer Agency’s Research Centre. All of the animal studies were performed in accordance with the Canadian Council on Animal Care Guidelines and the UBC Committee on Animal Care; Ethics Certificates A09-0251 and A13-0223.    41  2.9 Tissue Processing  After 21 days post-implant, tumours and lungs were excised immediately after CO2 euthanization of the mice. Tumours were embedded in Optimal Cutting Temperature (OCT) medium (Sakura Finetek, Torrance, CA) and frozen in -80°C. Lungs were finely minced with scalpels followed by an enzymatic digest in a PBS solution containing 0.5% trypsin and 0.08% collagenase for 40 min at 37°C on an agitator. After the digest, 0.06% DNAse was added then the cell suspension was filtered through 30 µm nylon mesh to remove cell clumps. Then the cell suspensions were treated with ammonium chloride for erythrocyte lysis. Cells were then utilized in clonogenic assays or fixed in 70% EtOH and stored at -20°C for flow cytometry.  2.9 Flow Cytometry  Metastasized tumour cells in lung tissue were determined by analyzing 5x105 cells from ethanol-fixed single-cell suspension of lung tissue. The cells were washed in 4% FCS in PBS then stained with polyclonal rabbit anti-cytokeratin antibody (Dako, Markham, ON), 1:400 dilution, at 4°C overnight on an agitator. Cells were incubated in Alexa Fluor® 488 Goat Anti-Rabbit IgG (Invitrogen), 1:100 dilution, at 4°C for 30 min on an agitator, followed by staining with 1 µg/mL DAPI for 10 min.  4T1-HMOX1 lung tissue was analyzed by flow cytometry using a dual laser Epics Elite-ESP flow cytometer (Coulter Corp., Hialeah, FL) and WINLIST software package (Verity Software House Inc., Topsham, ME). 4TO7-HMOX1 lung tissue analysis was performed on the FACSCalibur flow cytometer (BD Biosciences) and data was collected using CellQuestPro software.  2.10 Clonogenic Assays  Single-cell suspensions of lung tissue after enzymatic digest were washed in PBS and resuspended in culture medium. Aliquots of 3x103, 3x104, and 3x105 4T1 cells or 42  104, 105, and 106 4TO7 cells were plated in triplicate in culture dishes containing 60 µM (4T1 selection)  or 15 µM 6-thioguanine (4TO7 selection) to select for tumour cells. Cells were incubated for 10 days in 5% CO2 at 37°C prior to staining surviving tumour colonies with malachite green. Colonies on each plate were counted. The total number of clonogenic tumour cells in the lungs was calculated by multiplying the percentage of colony-forming tumour cells by the total number of cells recovered from the lungs. Plating efficiency of cells obtained from lung tissue is approximately 30%.  2.11 Immunofluorescence  Tumours frozen in OCT medium were cut in 8-10 µm serial sections and stored in 4°C. Tumour sections were fixed in 2% PFA and dipped in methanol. After 20 min blocking in 4% calf serum and 0.2% TritonX-100 in PBS, sections were stained for 1 h 30 min at room temperature with primary antibodies against: rabbit polyclonal anti-HMOX1 (Abcam), 1:100 dilution, rat monoclonal anti-CD11b (eBioscience), 1:100 dilution, rabbit polyclonal anti-active Caspase 3 (Abcam), 1:200 dilution, rat monoclonal anti-CD31 (BD Pharmingen), 1:200 dilution, rabbit polyclonal anti-CA9 (Novus Biologicals, Littleton, CO), 1:500 dilution. Secondary detection reagent was Alexa Fluor® IgG, incubated for 45 min at room temperature. The tumour sections were stained with 1 µg/mL DAPI. Images were acquired on Zeiss AxioObserver.Z1 microscope with AxioVision Rel 4.0 software.   2.12 Statistical Analysis All results are given as mean ± SEM of “n” separate independent experiments unless stated otherwise. Student’s t-test was used to determine statistical significance with a threshold of p≤0.05.  43  Chapter 3 The Effect of HMOX1 on Migration and Invasion In Vitro  3.1 Introduction  Invasion of tumour cells is the first major step of metastasis. For tumour cells to invade the basement membrane and the stromal tissue, the cells need both ECM remodeling and migratory characteristics. Heme oxygenase-1 has been implicated in the metastatic process of breast cancer, however, the findings have been inconsistent (90, 91, 102). Therefore, we previously examined the possible role of HMOX1 in murine mammary carcinoma cell migration and invasion by using hemin as a HMOX1 inducer (Figure 1.6). Migration and invasion can be studied in vitro by using Boyden chamber assays. By starving the cells in medium containing 1% serum prior to the assay, chemotactic migration is utilized to observe both migration and invasion through a cell permeable membrane. The invasion assay differs from the migration assay by having a Matrigel-coated membrane, therefore cells need matrix degrading enzymes such as MMPs before migrating through the membrane.   As previously shown, the chemical induction of HMOX1 by hemin treatment reduces the migration and invasion of 67NR and 4TO7 cells (Figure 1.6). To further investigate the effect of HMOX1 on tumour cell migration and invasion in vitro, assessing possible off-target effects of hemin treatment was necessary. Three approaches were taken to validate the impact of hemin on migration and invasion. Firstly, we examined the effect of a second chemical inducer of HMOX1 on migration and invasion. Secondly, we stably transduced 67NR and 4TO7 cells with shRNA knocking down HMOX1 to determine whether these cells would phenocopy the migration and invasion of the untreated wild type cells. Finally, 67NR and 4TO7 cells genetically overexpressing HMOX1 were utilized to determine whether HMOX1 overexpression would mimic the decrease in migration and invasion that was seen in hemin-treated cells.     44  3.2  Results  3.2.1 Comparison of migration and invasion of chemically and genetically induced HMOX1 in 67NR and 4TO7 cells    We examined the changes in migration and invasion of 67NR and 4TO7 cells after treatment with CoPP. CoPP is a protoporphyrin similar to hemin that induces HMOX1 through affecting the Nrf2/BACH1 regulatory system of HMOX1 (117, 122). In the literature, between 25-50 μM CoPP has been used to induce HMOX1 in cells (89, 122, 148). We found that at both 25 and 50 μM, CoPP induced HMOX1 expression in 67NR cells drastically more than hemin (Figure 3.1A and B). However, in 4TO7 cells, HMOX1 was expressed only at 50 μM CoPP. 4TO7 cells treated with 60 μM hemin expressed similar levels of HMOX1 as 4TO7 cells treated with 50 μM CoPP (Figure 3.1A and B). Therefore, 50 μM CoPP was used to induce HMOX1 in subsequent assays in both 67NR and 4TO7 cells. We used Boyden chamber transwell assays to assess migration. Different numbers of 67NR and 4TO7 cells (67NR – 4x105 cells, 4TO7 – 1x105 cells) were plated in the assay due to the differences in migratory ability between the two cell lines (Figure 3.1B). In both cell lines, tumour cell migration decreased after CoPP treatment, similar to hemin-treated cells (Figure 3.1C). CoPP-treated 67NR cells migrated comparably to hemin-treated cells. On the other hand, in 4TO7 cells, CoPP treatment had a greater effect on migration than hemin treatment.   45   Figure 3.1 Treatment of 67NR and 4TO7 cells with CoPP decreases migration in vitro. The indicated cell lines were treated with 60 μm hemin or 50 μm cobalt protoporphyrin (CoPP), and transwell migration assays were performed using 10% FBS as a chemoattractant. Cells were fixed and stained with DAPI. Three fields of view were imaged per membrane. WT – wild type, WT+ hemin – wild type treated with hemin, WT + CoPP – wild type treated with CoPP. A and B, levels of HMOX1 expression were analyzed by Western Blot, with tubulin as a loading control. A, cells were treated with 30, or 60 μm hemin for 18 hours. B, cells were treated with 25, or 50 μm CoPP for 18 hours. C, representative images of cells that have migrated onto the lower surface of the membrane. D, quantification of cell migration. Number of cells per field was counted using ImageJ. Each bar represents a relative number of cells migrated per field compared to the untreated wild type control. N=6, one-tailed t-test, * represents p≤0.05 compared to respective wild type cell. Scale bar = 200 μm in B.  46  Next, we examined invasion in vitro by transwell assays. Invasion assays are similar to migration assays except the membrane is coated with Matrigel. Matrigel is composed of ~60% laminin, ~30% collagen IV, ~8% entactin, heparin sulfate proteoglycan, growth factors, and matrix metalloproteinases, and provides tumour cells with a basement membrane-like environment (154).  Therefore, the cells must degrade the Matrigel through the expression of proteolytic proteins such as MMP9 before being able to dynamically migrate through the membrane. As with migration, different numbers of cells (67NR – 4x105 cells, 4TO7 – 2x105 cells) were plated in the invasion assays (Fig 3.2A). In both cell lines, CoPP treatment decreased the number of cells that invaded, and at a greater extent than after hemin treatment (Fig 3.2B). It is important to note that CoPP or hemin treatment was non-toxic to the cells as assessed by cell counts. The cell growth remained the same between the CoPP- and hemin-treated cells compared to the untreated control cells before plating the cells in the migration and invasion assays. Therefore, CoPP induced HMOX1 and reduced both migration and invasion of 67NR and 4TO7 cells and this result supports the phenotype observed after hemin treatment.   To further investigate whether hemin is acting on migration and invasion solely due to the induction of HMOX1, we used 67NR and 4TO7 cells stably transduced with shRNA targeting HMOX1. Since wild type 67NR and 4TO7 cells do not express HMOX1 at detectable levels in vitro, in order to verify that the knockdown was present, we induced HMOX1 expression with hemin (Fig 3.3A). With the aim to use the lowest concentration of hemin in which HMOX1 expression was detectable in the control cells expressing shRNA against GFP, different levels of hemin were used to induce HMOX1 between 67NR (22.2 μM) and 4TO7 (60 μM) cells. In 4TO7 cells, both shHMOX1 #1 and #4 successfully knocked down HMOX1 (Figure 3.3A). However, the knockdown was minimal in the 67NR cell line and therefore migration and invasion of 67NR cells with HMOX1 knocked down were not assessed. Since hemin has a potent effect on the induction of HMOX1 particularly in 67NR cells, shRNA against HMOX1 may have been unable to suppress this effective induction.   47   Figure 3.2 Treatment of 67NR and 4TO7 cells with CoPP reduces invasion in vitro. Transwell invasion assays were performed on cell lines treated with 60 μm hemin or 50 μm CoPP using 10% FBS as a chemoattractant. Invasion is assessed by the ability of the cells to invade through Matrigel on top of the membrane. Cells were fixed and stained with DAPI. Three fields of view were imaged per membrane. A, representative images of cells that have invaded onto the lower surface of the membrane. B, quantification of cell invasion. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells invaded per field compared to the untreated wild type control. N=6, one-tailed t-test, * represents p≤0.05 compared to respective wild type cell. Scale bar = 200 μm in A. 48   Figure 3.3 HMOX1 knockdown does not rescue wild type migration phenotype in 4TO7 cells. The indicated cell lines were stably transduced to knockdown HMOX1 using shRNA. The cells were treated with 60 μm hemin to induce HMOX1 expression. shGFP – control, shH#1 – shHMOX1 clone 1, shH#4 – shHMOX1 clone 4. Same protocol was performed as in Figure 3.1. A, levels of HMOX1 expression were analyzed by Western Blot, tubulin acting as a loading control. B, representative images of cells that have migrated onto the lower surface of the membrane. C, quantification of cell migration. Number of cells per field was counted using ImageJ. Each bar represents a relative number of cells migrated per field compared to the control. N=6, one-tailed t-test. Scale bar = 200 μm in B. Knocking down HMOX1 in 4TO7 cells was expected to increase migration to untreated control levels. In 4TO7 cells treated with hemin, there were no differences in the number of knockdown cells that migrated compared to the shGFP control (Fig 3.3B and C). In 4TO7 cells, where HMOX1 levels were knocked down close to untreated control levels, the absence of the expected increase in migration compared to the control suggests in the presence of hemin, HMOX1 levels were not related to 4TO7 cell migration.  49   4TO7 cells with HMOX1 knockdown were assayed for invasion. In 4TO7-shHMOX1 #1 and #4 cells, there was unexpectedly a further decrease in invasion compared to the hemin-treated control. Treating 4TO7 wild type cells with hemin, increased HMOX1 levels and decreased invasion (Figure 1.6). On the other hand, hemin treatment of 4TO7 cells with HMOX1 knocked down, reduced HMOX1 levels but resulted in a further decrease in invasion. Cell counts prior to the migration and invasion assays showed that the growth of HMOX1 knockdown cells remained unchanged compared to the control cells even after hemin treatment. Although the cell growth was not affected in hemin treated cells, the absence of HMOX1 to metabolize excess hemin in the culture medium may have been a factor in the additional decrease in invasion in the knockdown cells by possibly affecting cellular functions such as a decrease in functional invadopodia. Free heme has some oxidative effects on membrane-bound proteins along with lipid components of the cell (155). The effect of hemin on proteins may be associated with membrane-bound proteases such as MT1-MMP found in invadopodia, therefore affecting invasion independent of HMOX1 expression.   As seen in Figure 3.3 and 3.4, hemin decreased 4TO7 tumour cell migration and invasion even in the presence of HMOX1 knockdown. Hemin was affecting migration and invasion independently from HMOX1 levels, therefore, we stably transduced 67NR and 4TO7 cells with a HMOX1 overexpression vector to examine if the decreases in migration and invasion observed in hemin-treated cells would be phenocopied. The genetic overexpression of HMOX1 in 67NR cells was comparable to the HMOX1 expression found in hemin-treated cells. However, 4TO7 cells overexpressing HMOX1 had higher HMOX1 expression than hemin-treated cells (Fig 3.5A). Consistent with hemin treatment, overexpressing HMOX1 in these cells decreased migration (Fig 3.5B and C).  50   Figure 3.4 HMOX1 knockdown does not rescue the wild type invasion phenotype in 4TO7 cells. The same protocol was performed as in Fig 3.2. A, representative images of cells that invaded onto the lower surface of the membrane. B, quantification of cell invasion. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells invaded per field compared to the control. N=6, one-tailed t-test, * represents p≤0.05 compared to 4TO7-shGFP treated with hemin. Scale bar = 200 μm in A. 51   Figure 3.5 HMOX1 overexpression decreases migration in 67NR and 4TO7 cells similar to chemical induction. The indicated cell lines were genetically modified to express HMOX1 and transwell migration assays were performed. The same protocol was performed as in Fig 3.1. WT – wild type, WT + 60 µM hemin – wild type treated with 60 µM hemin, Empty – empty vector, HMOX1 – HMOX1 overexpression vector. A, levels of HMOX1 expression were analyzed by Western Blot, with tubulin acting as a loading control. B, representative images of cells that migrated onto the lower surface of the membrane. C, quantification of cell migration. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells migrated per field compared to the respective controls (WT vs. WT + hemin, Empty vs. HMOX1). N=6, one-tailed t-test, * represents p≤0.05 compared to respective controls. Scale bar = 200 μm in B.  52   Next, we examined the changes in invasion after HMOX1 was overexpressed in 67NR and 4TO7 cells. Unexpectedly, HMOX1 overexpression resulted in an increase in invasion in both cell lines, which is opposite from the reduction in invasion of 67NR and 4TO7 cells after hemin treatment (Fig 3.6A and B). 67NR-HMOX1 cells had 4.7 fold more invaded cells compared to 67NR-Empty control cells, while 4TO7-HMOX1 cells invaded 3.1 fold more than the 4TO7-Empty control cells (Fig 3.6B). Thus, the induction of HMOX1 expression by both hemin treatment or genetic overexpression of HMOX1 produced different effects on invasion. This difference in invasion may be due to the difference in expression of ECM-degrading factors, such as MMP9 and MMP2, triggered by the hemin or the genetic overexpression.   The mechanism of how HMOX1 affects migration and invasion in cancer is still relatively unknown. There are two schools of thought on the mechanism of action of HMOX1 outside of its role as an enzyme involved in heme metabolism. One view is that the products of heme breakdown such as carbon monoxide, iron, biliverdin, and bilirubin are responsible for the various implicated effects of HMOX1. The other view is that HMOX1 directly affects the transcription of its downstream targets. HMOX1 typically resides on the sER but has been shown to localize in to the nucleus as well (107). Given that the invasive phenotype of HMOX1 overexpressing cells differed from the hemin-treated cells, we assessed the localization of HMOX1 to determine whether there is a difference in cellular localization between the two groups of cells. 67NR and 4TO7 cells were grown on coverslips and stained with HMOX1 and calnexin, an endoplasmic reticulum marker.   In both murine mammary carcinoma cell lines, there were no discrepancies in the localization of HMOX1 between the cells genetically overexpressing HMOX1 and those that had been chemically induced (Fig 3.7). In addition, these cells expressed HMOX1 in the endoplasmic reticulum rather than in the nucleus. HMOX1 mostly co-localized with calnexin, however in 4TO7- Empty cells, HMOX1 was also expressed 53  outside of the sER. Using immunofluorescence, we were able to visualize the heterogeneous expression of HMOX1 at low levels in the wild type and empty vector control cells. Resembling intracellular localization of HMOX1 between hemin-treated and HMOX1 overexpressing 67NR and 4TO7 cells suggests HMOX1 expressed chemically or genetically may have similar mechanism of action. Examination of cellular localization of HMOX1 did not resolve the discrepancies found in invasion of hemin-treated and HMOX1 overexpressing 67NR and 4TO7 cells.  Figure 3.6 HMOX1 overexpression enhances invasion in 67NR and 4TO7 cells in vitro. A, the same protocol was performed as in Fig 3.2. A, representative images of cells that have invaded onto the lower surface of the membrane. B, quantification of cell invasion. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells invaded per field compared to the control (WT vs. WT + hemin, Empty vs. HMOX1). N=6, one-tailed t-test, * represents p≤0.05 compared to respective controls. Scale bar = 200 μm in A. 54   Figure 3.7 Immunofluorescent stainings of HMOX1 show non-differential cellular localization of HMOX1. 67NR and 4TO7 cells were grown on coverslips and stained with HMOX1 and calnexin (endoplasmic reticulum marker). Original magnification, 100X.  Previous groups have seen a decrease in MMP9 levels in the non-metastatic human breast cancer cell line, MCF-7, after hemin treatment as well as after HMOX1 overexpression  (90, 91, 102). Therefore, we assessed the MMP9 and MMP2 levels in 67NR and 4TO7 cells to determine if there were differences in the levels of proteinases between tumour cells that had chemically or genetically induced HMOX1 expression. For both cell lines, gelatinase zymography was conducted in which 15 μg of total protein secreted into culture medium was subjected to electrophoresis.  The pro-MMP2 found at 72 kDa was not observed in the zymogram. Nevertheless, active MMP2 expression decreased in both 67NR-HMOX1 cells and hemin-treated 67NR cells compared to the control cells. However, hemin-treated cells secreted significantly less active MMP2 compared to the untreated control cells (Figure 3.8). MMP9 levels in 67NR cells slightly decreased only in hemin-treated cells, but remained relatively the same between all of the groups. In contrast, pro- and active MMP9 were notably reduced in 4TO7-HMOX1 cells and 55  hemin-treated 4TO7 cells. In all of the HMOX1 induced 4TO7 cells, MMP2 expression was slightly decreased.. Overall, both MMP9 and MMP2 levels reduced in both cell lines with genetic overexpression and after hemin treatment.. The reduction in MMP2 and MMP9 may lead to a decrease in invasion by hemin-treated 67NR and 4TO7 cells since MMP9 and MMP2 are responsible for ECM degradation. However, the number of invaded 67NR and 4TO7 cells increased after genetic overexpression. This suggests that not only are HMOX1 overexpressing cells not solely dependent on MMP9 and MMP2 for invasion, but upregulation of factors involved in invasion such as MMP1 may be compensating for the loss of MMP9 and MMP2 and also further enhance invasion of HMOX1 overexpressing 67NR and 4TO7 cells. Future work addressing the discrepancy between invasion by hemin-treated and HMOX1 overexpressing 67NR and 4TO7 cells will be discussed in Chapter 5.     Figure 3.8 MMP and MMP2 levels decrease in HMOX1 expressing 67NR and 4TO7 cells. Respective cells were grown in serum-free medium for 48 hours before conditioned media were collected. Media containing 15 μg of total protein were examined by gelatin zymography for MMP9 and MMP2 and the zymogram was activated for 41 h. Similar results were obtained in 3 different experiments.     56  3.2.2 Comparison of migration and invasion of chemically and genetically induced HMOX1 in 4T1 cells  Out of the three murine mammary carcinoma cell lines utilized in this project, 4T1 cells are the most metastatic and have a highly invasive phenotype (23). Similar to 67NR and 4TO7 cells, wild type 4T1 cells do not express detectable levels of HMOX1 protein in vitro and as observed in 67NR and 4TO7 cells, hemin treatment decreased the number of migrated and invaded 4T1 cells (Figure 1.6). Therefore, we took the same three approaches as we did with 67NR and 4TO7 cells to validate the results from hemin-treated cells.   Although the induction of HMOX1 was less than after hemin treatment, at both 25 and 50 μM CoPP, HMOX1 was induced in 4T1 cells (Fig 3.9A). 4T1 cells were treated with 50 μM CoPP for the subsequent assays. In agreement with the decrease in migration after hemin treatment, CoPP treatment reduced migration of 4T1 cells (Fig 3.9B). Despite the expression of HMOX1 being lower in CoPP-treated compared to hemin-treated cells, the number of migrated cells was 93% less than the control after CoPP treatment (Fig 3.9C). Additionally, invasion also decreased in CoPP-treated 4T1 cells (Fig 3.10A and B). The number of invaded cells was comparable between hemin- and CoPP-treated cells. Therefore another chemical inducer of HMOX1 reduced migration and invasion of 4T1 cells analogous to the reduction seen with hemin treatment. CoPP-treated 4T1 cells resulted in a consistent reduction in migration and invasion also observed in 67NR and 4TO7 cells treated with CoPP.    57   Figure 3.9 Treatment of 4T1 cells with CoPP decreases migration in vitro. The same protocol was performed as in Figure 3.1. A, B, levels of HMOX1 expression after hemin (A) or CoPP (B) were analyzed by Western Blot, with tubulin acting as a loading control. C, representative images of cells that have migrated onto the lower surface of the membrane. D, quantification of cell migration. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells migrated per field compared to the untreated wild type control. N=6, one-tailed t-test, * represents p≤0.05 compared to wild type 4T1 cells. Scale bar = 200 μm in B.  58   Figure 3.10 CoPP decreases invasion in 4T1 cells in vitro. The same protocol was performed as in Figure 3.2. A, representative images of cells that have invaded onto the lower surface of the membrane. B, quantification of cell invasion. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells invaded per field compared to the untreated wild type control. N=6, one-tailed t-test, * represents p≤0.05 compared to wild type 4T1 cells. Scale bar = 200 μm in A. Next, we knocked down HMOX1 in 4T1 cells and treated the cells with hemin to examine whether the knockdown would rescue migration and invasion phenotype of untreated 4T1 cells. At 60 μM hemin, HMOX1 expression was detectable in 4T1-shGFP cells but knocked down in 4T1-shHMOX1 #1 and #4 cells (Fig 3.11A). Unexpectedly, the knockdown did not increase the number of migrated cells compared to the control. In comparison to 4TO7 cells with HMOX1 knockdown in which no differences were found between the control and HMOX1 knockdown cells, an additional reduction in the number of migrated 4T1-shHMOX1 cells was observed compared to the shGFP control cells (Fig 3.11B). 4T1-shHMOX1 #1 cells, which did 59  not establish complete HMOX1 knockdown, migrated 46% less than the control. 4T1-shHMOX1 #4 cells, which had complete HMOX1 knockdown, migrated 79% less than the control. This showed that irrespective of HMOX1 expression, hemin reduced migration in 4T1 cells. Furthermore, the lack of HMOX1 expression in the presence of hemin may additionally decrease the migration of 4T1 cells.   Subsequently, invasion assays showed no differences in the number of invaded cells between the 4T1-shGFP cells and 4T1-shHMOX1 cells treated with hemin (Fig 3.12A and B). Cell growths of 4T1-shGFP, -shHMOX1 #1, and –shHMOX1 #4 cells after hemin treatment were comparable as assessed by cell counts prior to plating for the assay. Similar to the migration of 4T1 cells, hemin reduced the number of invaded cells independent from the level of HMOX1. Unlike the migration of 4T1 cells, there was not a further decrease in invading 4T1-shHMOX1 cells compared to the control. This suggests hemin reduces 4T1 migration and invasion, similar to reduction of migration and invasion by hemin-treated 4TO7 cells. 4T1 and 4TO7 cells with HMOX1 knocked down demonstrated that hemin does not decrease migration and invasion by exclusively increasing HMOX1 expression.   As shown in Figure 3.11 and 3.12, it is likely that treating 4T1 cells with hemin is having an effect on migration and invasion apart from the effect caused by the induction of HMOX1. Therefore, we genetically overexpressed HMOX1 in 4T1 cells to see if the genetic overexpression would mimic the effect of hemin on the migration and invasion of 4T1 cells. 4T1-HMOX1 cells expressed a similar level of HMOX1 to 4T1 wild type cells treated with hemin (Fig 3.9A and 3.13A). As expected, HMOX1 overexpression decreased migration in 4T1 cells (Fig 3.13B). However, the number of migrated cells reduced to 55% of the control compared to hemin which reduced the number of migrated 4T1 cells to 20% of the untreated 4T1 cells (Fig 3.13C). Similar to the effect of hemin on 4T1 cell migration that was observed in 4T1-shHMOX1 cells, the greater reduction of migration by hemin-treated 4T1 cells indicates that hemin may influence 4T1 cell migration apart from the induction in HMOX1 expression.   60   Figure 3.11 HMOX1 knockdown does not rescue wild type migration phenotype in 4T1 cells. The same protocol was performed as in Fig 3.3. A, levels of HMOX1 expression were analyzed by Western Blot, with tubulin acting as a loading control. B, representative images of cells that have migrated onto the lower surface of the membrane. C, quantification of cell migration. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells migrated per field compared to the control. N=6, one-tailed t-test, * represents p≤0.05 compared to 4T1-shGFP treated with hemin. Scale bar = 200 μm in B. 61    Figure 3.12 HMOX1 knockdown does not rescue the wild type invasion phenotype in 4T1 cells. The same protocol was performed as in Fig 3.4. A, representative images of cells that have invaded onto the lower surface of the membrane. B, quantification of cell invasion. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells invaded per field compared to the control. N=6, one-tailed t-test, * represents p≤0.05 compared to 4T1-shGFP treated with hemin. Scale bar = 200 μm in A.     62   Figure 3.13 HMOX1 overexpression decreases migration in 4T1 cells similar to chemical induction. The same protocol was performed as in Fig 3.5. A, levels of HMOX1 expression were analyzed by Western Blot, with tubulin acting as a loading control. B, representative images of cells that have migrated onto the lower surface of the membrane. C, quantification of cell migration. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells migrated per field compared to the respective controls (WT vs. WT + hemin, Empty vs. HMOX1). N=6, one-tailed t-test, * represents p≤0.05 compared to respective controls. Scale bar = 200 μm in B. 63   Figure 3.14 HMOX1 overexpression reduces invasion in 4T1 similar to chemical induction. A, the same protocol was performed as in Fig 3.6. Representative images of cells that have invaded onto the lower surface of the membrane are shown in A. B, quantification of cell invasion. Number of cells per field was counted using ImageJ. Each bar represents the relative number of cells invaded per field compared to the control (WT vs. WT + hemin, Empty vs. HMOX1). N=6, one-tailed t-test, * represents p≤0.05 compared to respective controls. Scale bar = 200 μm in A.   64   Figure 3.15 Immunofluorescent staining of HMOX1 show non-differential cellular localization of HMOX1 in 4T1 cells. Cells were grown on coverslips and stained with HMOX1 and calnexin (endoplasmic reticulum marker). Original magnification, 100X. Consistent with the results from hemin-treated 4T1 cells, HMOX1 overexpression reduced the number of invaded cells (Fig 3.14A and B). 4T1-HMOX1 cells had a 79% reduction in the number of invaded cells compared to the control. Expression of HMOX1 reduced both migration and invasion of 4T1 cells. Furthermore, in 4T1 cells, compared to genetic HMOX1 overexpression, hemin had a greater effect reducing migration. In contrast, HMOX1 overexpression and hemin treatment had similar effects on invasion. This trend was also observed in the 4T1 cells with HMOX1 knocked down in Figures 3.11C and  3.12B. After hemin treatment, migration by 4T1-shHMOX1 cells was further reduced compared to the control cells, however, invasion by 4T1-shHMOX1 cells decreased only to control cell levels. Overall, the number of migrated and invaded 4T1 cells overexpressing HMOX1 decreased compared to the number of invaded 4T1 cells expressing the empty vector. This was consistent with 4T1 cells treated with hemin, in which reductions in migration and invasion were also observed.   To assess whether hemin-treated and HMOX1 overexpressing 4T1 cells have the same mechanism of reducing migration and invasion, we initially examined the cellular localization of HMOX1. Similar to 67NR and 4TO7 cells, chemically induced HMOX1 was expressed in the endoplasmic reticulum in 4T1 cells (Figure 3.15). Likewise, genetically induced HMOX1 also co-localized with calnexin, indicating that 65  HMOX1 is primarily localized in the sER. In both control wild type and empty vector expressing 4T1 cells, there is a small number of cells expressing HMOX1. Western blots in Figure 3.9A and B showed that in 4T1 cells, CoPP induced HMOX1 less than hemin. This was also evident in the immunofluorescent staining of HMOX1 in 4T1 cells treated with CoPP compared to hemin. Hemin induced HMOX1 expression in majority of 4T1 cells. However, in CoPP-treated cells, there was a heterogenous population of HMOX1 expressing cells. Nonetheless, CoPP decreased the number of migrated and invaded cells similar to hemin (Figure 3.9C).  To compare MMP9 and MMP2 levels in 4T1-HMOX1 and hemin-treated 4T1 cells, gelatinase zymography was performed. The zymogram showed that in hemin-treated cells, active MMP2 levels were significantly diminished compared to the untreated cells (Figure 3.16A). The active MMP2 levels were also slightly reduced in 4T1-HMOX1 cells compared to 4T1-Empty cells. In the zymogram for MMP9, 5% of the total amount of protein that was loaded for 67NR and 4TO7 cells was utilized. In addition, MMP9 was activated for 24 hours rather than 41 hours to be able to visualize the difference in MMP9 levels between HMOX1 expressing and control 4T1 cells (Figure 3.16B). For comparison, Figure 3.16B shows the same loading and activation time that were used for 67NR and 4TO7 cells.   Vast amounts of MMP9 were expressed by 4T1 cells compared to 67NR and 4TO7 cells. This is expected due to the highly invasive nature of these cells. MMP9 levels also decreased in both 4T1-HMOX1 and hemin-treated 4T1 cells. Although there was a reduction in MMP9 and MMP2 in both 4T1-HMOX1 and hemin-treated cells, a larger decrease was observed in the hemin-treated cells. Likewise, there was a decrease in the number of invaded cells in both 4T1-HMOX1 and hemin-treated wild type cells compared to the control cells. However, MMP9 levels remained relatively high in 4T1 cells expressing HMOX1. This suggests that the reduction of MMP9 found in hemin-treated and HMOX1 overexpressing 4T1 cells was likely not affecting cell invasion. Therefore, a further investigation on the regulation of proteins such as MT1-MMP that are involved in MMP zymogen activation may inform the mechanism in which HMOX1 decreases 4T1 cell invasion. 66   Figure 3.16 MMP and MMP2 levels decrease in HMOX1 expressing 4T1 cells. Respective cells were grown in serum-free medium for 48 hours before conditioned media were collected. A, media containing 0.75 μg and 15 μg of total protein for MMP9 and MMP2 zymography respectively. MMP9 was activated for 24 h compared to MMP2 which was activated for 41 h. B, media containing 15 μg of total protein were examined by gelatinase zymography for MMP9, and the zymogram was activated for 41 h. Similar results were obtained in 3 different experiments.   3.3 Discussion  In our study, we have utilized three murine mammary carcinoma cell lines that have different metastatic characteristics. Induction of HMOX1 expression varied between the three cell lines. 67NR cells, which are noninvasive, readily expressed HMOX1 after hemin or CoPP treatment. In contrast, 4TO7 and 4T1 cells, which are invasive, required a higher concentration of hemin and CoPP in order to achieve a detectable 67  level of HMOX1. HMOX1 expression was also significantly lower in 4TO7 and 4T1 cells compared to 67NR cells after treatment.  Both hemin and CoPP induce HMOX1 expression by affecting the Nrf2/BACH1 regulatory system of HMOX1 (117, 122). If 4T1 and 4TO7 cells have a higher expression of BACH1 that is available to repress HMOX1 transcription, a higher dose of protoporphryins may be necessary to elicit an induction. Therefore, establishing BACH1 protein levels in the three cell lines may help characterize the upstream regulation of HMOX1 affecting HMOX1 expression in the cells.  There also may be post-transcriptional regulation of HMOX1 in 4T1 and 4TO7 cells that is not present in 67NR cells, thereby reducing the amount of HMOX1 being translated. shRNA used to knockdown HMOX1 in these cell lines also demonstrates how potently effective hemin is in inducing HMOX1 in 67NR cells compared to 4T1 and 4TO7 cells. Both shHMOX1 #1 and shHMOX1 #4 were able to knockdown HMOX1 in 4TO7 and 4T1 cells, however, in 67NR cells, HMOX1 levels were not affected by the shRNA constructs.   CoPP is suggested to be the better choice as a chemical inducer of HMOX1 because it is still able to induce HMOX1 to similar levels as hemin without being a substrate analogue. Hemin is analogous to heme, the substrate of HMOX1, and differs only by the chloride ligand bonded to the iron ion at the centre of the protoporphyrin. In contrast, CoPP replaces the iron ion found in hemin with a cobalt ion, which prevents the binding of the protoporphyrin to the active site of HMOX1. Since CoPP reduced migration and invasion of all three cells lines to similar levels observed after hemin treatment, verifying actions of CoPP with knockdown studies may ascertain whether CoPP is a more rational choice as an HMOX1 inducer than hemin for subsequent studies.   As mentioned, knockdown of HMOX1 was not achieved in 67NR cells, and therefore as expected, we did not observe a change in migration and invasion of these cells compared to the control. In contrast, knockdown was attained in 4TO7 cells. However the results were unexpected. We anticipated a reversal in the reduction of migration and invasion after hemin treatment in the 4TO7 knockdown cells. Instead, 68  the number of migrated cells remained unchanged and a further decrease of invasion by 4TO7-shHMOX1 cells compared to the shGFP control cells was found. The additional reduction in invasion that was seen in 4TO7-shHMOX1 cells suggests that hemin is affecting factors involved in invasion regardless of HMOX1 expression levels. In contrast, invasion of 4T1-shHMOX1 cells was unaffected but a decrease in migration was observed. Since 4T1 cells have the most invasive phenotype, it is possible that hemin does not affect the critical mechanisms involved in invasion in the 4T1 cells as it does in the 4TO7 cells. There is also an ample amount of MMP9 being expressed by 4T1 cells and therefore hemin may not affect invasion as much as it affects migration.   Free heme has been implicated in oxidation of certain lipids and proteins and therefore the induction of HMOX1 is necessary to metabolize heme to pro-oxidant products (155-158). Oxidation of membrane-bound proteins such as MT1-MMP or integrins may alter their function and therefore affecting cell invasion and migration.   The cell growths remained unchanged during cell culture between hemin-treated shGFP and shHMOX1 cells, which suggest free hemin in the culture medium due to the lack of HMOX1, may have had an effect on cellular function of 4TO7 and 4T1 cells with HMOX1 knocked down rather than cell viability. Therefore recovery of the wild type phenotype may have not been observed.   When we compared genetically overexpressed cells to the hemin-treated cells, we were able to observe effects on migration and invasion solely based on HMOX1 expression independent of any off-target effects of the chemical inducer. In 67NR and 4TO7 cells, HMOX1 overexpression resulted in a decrease in migration similar to hemin-treated cells. On the contrary, overexpressing cells invaded through the Matrigel significantly better than hemin-treated cells. This supports the observation in 4TO7-shHMOX1 cells, in which hemin seemingly had an effect on invasion independent of increased HMOX1 expression. Likewise, the number of invaded HMOX1 overexpressing 67NR cells was higher than the number of invaded hemin-treated 67NR cells.   MMP9 has been previously shown to be decreased in HMOX1 expressing MCF-7 cells (91, 102). Similar to MCF-7 cells, MMP9 levels were significantly decreased in both the hemin-treated and HMOX1 overexpressing 4TO7 69  cells, which do not explain the increase in invasion displayed by HMOX1 overexpressing cells. It also demonstrates that in 4TO7 cells, MMP9 levels alone do not dictate cell invasion, as demonstrated by an increase in invasion but a decrease in MMP9. In 67NR cells, hemin treatment significantly reduced MMP2 levels and HMOX1 overexpression had a slight decrease in MMP2 levels. The ablated levels of MMP2 in hemin-treated 67NR cells correspond to the decreased invasion seen in these cells. Therefore, addition of exogenous MMP2 to hemin-treated 67NR cells may elucidate possible mechanism of hemin-mediated decrease in invasion. In contrast, HMOX1 overexpression significantly increased the invasion of 67NR cells, which do not have an invasive phenotype. This enhanced invasion was not deduced by MMP9 and MMP2 zymography, since minimal changes were seen in 67NR-HMOX1 compared to 67NR-Empty cells. This indicates that HMOX1 overexpression upregulates factors other than MMP9 and MMP2 that are able to trigger invasion by 67NR cells. These targets may include MMP1, which have been shown to be upregulated after HMOX1 induction in MCF-7 cells (90).      It is intriguing that overexpression of HMOX1 was able to decrease the migration of 67NR and 4TO7 cells, yet increase invasion when cell migration is a necessary step in invasion. Depending on the components found in the basement membrane, cancer cells can change their invasive phenotypes and utilize different mechanisms of migration and invasion by changing protrusions or transforming to different types of migration (64). Therefore, it is possible that genetic overexpression of HMOX1 in 67NR and 4TO7 cells enhances the invasive phenotype by affecting structures involved in invasion including integrins that bind better to the substrate within the Matrigel. Along with HMOX1 expression, examination of integrins such α6β1 and α6β4 that bind to laminin in 67NR and 4TO7 overexpressing HMOX1, may reveal potential downstream targets of HMOX1 that affect invasion in 67NR and 4TO7 cells.   In 4T1 cells, genetic overexpression of HMOX1 had a similar affect on both migration and invasion as hemin-induced HMOX1 expression. Invasion by 4T1-70  HMOX1 cells was comparable to hemin-treated 4T1 cells. However, hemin treatment reduced migration of 4T1 cells significantly more than the genetic overexpression of HMOX1. Hemin also had a greater effect on migration than on invasion of 4T1-shHMOX1 cells. Invasion by 4T1-shHMOX1 cells remained similar to the shGFP control cells after hemin treatment. However the number of migrated 4T1 cells with HMOX1 knockdown further decreased after hemin treatment compared to the control. This suggests that hemin independently has a potent effect of decreasing migration by 4T1 cells in addition to the reduction in migration affected by HMOX1 expression.    HMOX1 is typically an enzyme that resides on the sER membrane anchored at the carboxyl-end of the protein. Therefore, in determining the molecular mechanism of HMOX1, groups have utilized CO, iron, biliverdin, and bilirubin the products of heme breakdown, as treatment to assess whether the enzymatic activity of HMOX1 is responsible for phenotypic change after HMOX1 induction (91, 97). For example, treatment of MCF-7 cells with [Ru(CO3)Cl2]2 decreases migration similar to hemin treatment (91). However, treatment of cells with Fe(SO4), biliverdin or bilirubin did not affect MCF-7 cell migration, suggesting CO from heme breakdown may be responsible for the effect of HMOX1 on breast cancer cell migration.   In contrast, instead of its enzymatic role, HMOX1 has been suggested to act on downstream targets as a transcription factor (107, 158). Interestingly, mutant HMOX1 lacking catalytic activity is still able to provide cytoprotection against oxidative stress through mechanisms involving upregulation of stress response genes such as catalase (158). Lin et al. have also detected HMOX1 in the nucleus after exposure to heme and hypoxia, and have determined that carboxyl-terminal deletion is necessary for nuclear transport affecting the transcription of genes involved in oxidative stress response through interaction with AP-1 (107). In prostate cancer, ChIP-qPCR revealed that HMOX1 is enriched at the MMP9 promoter when HMOX1 is induced in LNCaP cells, and is also bound to the PSA proximal promoter region by direct interaction with a transcription factor, STAT3 (84). In the murine 71  mammary carcinoma cell lines that we utilized, HMOX1 was predominantly expressed in the sER and there were no differences in HMOX1 localization between the hemin-induced HMOX1 compared to the genetically induced HMOX1. Although, a detailed analysis on the downstream targets of HMOX1 is warranted, the cellular localization of HMOX1 to the sER suggests that HMOX1 may be affecting migration and invasion through its enzymatic activity in these cell lines.   Conducting three experiments assessing the effect of HMOX1 on tumour cell migration and invasion in vitro revealed that in all three cells, hemin is having an effect on migration and invasion independent of HMOX1 expression (Figure 3.17). However in 4T1 cells overexpressing HMOX1 migrated and invaded less than the control cells, consistent with hemin-treated cells. In contrast, in 67NR and 4TO7 cells overexpressing HMOX1, enhanced invasion was observed compared to the control cells, illustrated in Figure 3.17A. Hemin continues to be utilized as a HMOX1 inducer in various studies examining the role of HMOX1 in different pathologies. Our in vitro study shows the importance of validating results derived from hemin-treated cells to establish that the expression of HMOX1 is entirely responsible for changes observed in the phenotype. In the next chapter, we examined the role of HMOX1 in breast cancer metastasis in vivo to determine whether similar in vitro effects of HMOX1 would be observed in vivo.    72   Figure 3.17 Illustration of in vitro results and potential mechanism of action for HMOX1 on murine mammary carcinoma cell migration and invasion. A, summary of experiments in 67NR and 4TO7 cells. B, summary of experiments in 4T1 cells. C, possible mechanism of action for the effect of HMOX1 on murine mammary carcinoma cell migration and invasion 73  Chapter 4  The Effect of HMOX1 on Metastasis In Vivo – 4TO7 and 4T1 Tumours   4.1  Introduction  The phenotypes observed in vitro are sometimes incongruent with in vivo phenotypes. This is mainly due to the tumour microenvironment that is absent in a culture dish. Within a tumour, cancer cells interact with normal cells present in the microenvironment, and adapt to changes in oxygen tension and energy metabolism. Therefore it was crucial to verify whether the phenotypes observed in the migration and invasion assays, discussed in Chapter 3, agree with the metastatic phenotype in vivo.   Using the 4TO7 and 4T1 tumour models, we examined the effect of HMOX1 on primary tumour growth and metastasis in vivo. 4TO7 tumours form micrometastases and disseminated tumour cells can be found in the blood, and lungs. 4T1 tumours have a more aggressive metastatic phenotype and disseminated cells can be found in the blood, lung, bone, and brain. Although, the 4T1 cells do not model the lymphatic spread found in patients, dissemination of 4T1 cells resembles the metastatic spread found in human breast cancer outside the lymph nodes. 4TO7 and 4T1 cells express undetectable levels of HMOX1 in vitro, In contrast to HMOX1 expression in vitro, HMOX1 was expressed in tumours in vivo, therefore when we implanted HMOX1 overexpressing 4TO7 and 4T1 cells, we observed no changes in the primary tumour or metastasis. Therefore, we implanted cells with HMOX1 knocked down to determine if HMOX1 has an effect on breast cancer metastasis.   4.2   Results  4.2.1 Characterization of primary tumour growth and metastasis in the 4TO7 tumour model – HMOX1 overexpression and knockdown 74   Figure 4.1 HMOX1 overexpression has no effect on 4TO7 tumour growth or lung metastasis. 106 4TO7 cells (WT, Empty, or HMOX1) were orthotopically implanted into the mammary fat pad of female BALB/c mice. A, levels of HMOX1 expression were analyzed by Western Blot on the day of the implant, with tubulin acting as a loading control. B, the mice were euthanized as they approached ethical endpoints, when the primary tumours reached a size of 5% of an animal’s body weight, or 57 days post transplant. C, fixed single-cell lung samples were stained with cytokeratin and DAPI. Representative flow plot for tumour cells in lung samples, the tumour cells being cytokeratin positive and hyperploid cells. D, relative number of 4TO7 tumour cells in lungs determined by flow cytometry analysis compared to wild type control. Four 4TO7-Empty lung samples were lost during processing. WT N=4, Empty N=4, HMOX1 N=8, one-tailed t-test.    To examine the effect of HMOX1 on 4TO7 primary tumour growth and metastasis, we orthotopically implanted 4TO7-HMOX1 cells in BALB/c mice. Before implant, 4TO7-HMOX1 cells, grown parallel to the cells for implant, were verified for HMOX1 expression (Figure 4.1A). The tumours in all groups had variable growth rates due to 75  4TO7 tumours being highly immunogenic in syngeneic BALB/c mice, which often produces variable take-rates and growth rates of wild type 4TO7 cells. As a result, the endpoints depended on ethical tumour sizes rather than the intended 21 days post-implant (Figure 4.1B). The endpoints ranged from 18 to 57 days. The effect of HMOX1 overexpression on primary tumour growth could not be determined because variable tumour growth rates were observed in all groups due to the immunogenicity of 4TO7 cells.  To compare the number of metastases in the lungs from the 4TO7-HMOX1 and control tumours, single lung cell suspensions of tumour-bearing mice were analyzed by flow cytometry. The cells were fixed and stained for cytokeratin and DAPI. Cytokeratin-positive hyperdiploid cells identified the tumour cell populations in the lungs, as previously established in our laboratory (Figure 4.1C). The number of tumour cells in the lungs is displayed as the relative number of tumour cells in the lungs compared to mice with wild type 4TO7 tumours (Figure 4.1D). There were no differences in the number of tumour cells in the lungs between the groups. The variable endpoints due to the differences in tumour growth led to variable length of time in which tumours were present in the mice. This may have affected the numbers of tumour cells metastasizing to the lungs.     Immunofluorescence of 4TO7 wild type tumour sections revealed that in vivo there is HMOX1 expression in the tumours, despite the lack of HMOX1 expression under normal tissue culture conditions in vitro (Figure 4.2). Next, to determine whether the loss of HMOX1 affects tumour growth and metastasis, we implanted 4TO7-shHMOX1 cells. To examine tumour growth without the concern of immunogenicity, we orthotopically implanted 4TO7 tumour cells in immunocompromised NOD/SCID mice. This allowed a proper examination of the effect of HMOX1 on metastasis because tumour cells were given the same length of time to metastasize to the lungs. HMOX1 protein expression was analyzed in an equivalent culture to the cells used for implant, and demonstrated a decrease in HMOX1 expression in both 4TO7-shHMOX1 #1 and #4 cells after hemin treatment compared to 4TO7-shGFP cells 76  (Figure 4.3A). Tumours grown in NOD/SCID mice grew consistently within each group as opposed to BALB/c mice. Tumour growth remained unchanged between 4TO7-shGFP, 4TO7-shHMOX1 #1, 4TO7-shHMOX1 #4 groups (Figure 4.3 B). At 21 days post-implant, the tumours and lungs were excised and assessed. In agreement with the growth rate data, the tumour weights at endpoint also showed no difference between the groups (Figure 4.3 C).    Figure 4.2 Immunofluorescent staining of HMOX1 in 4TO7 tumour. A, representative image of frozen tumour sections stained with HMOX1, and DAPI. Scale bar = 200 μm.   Instead of flow cytometry to measure the number of tumour cells that have metastasized to the lungs, we performed clonogenic assays, which is a more sensitive method of detecting 4TO7 tumour cells present in the lungs. Across all groups, the numbers of tumour cells in the lungs were highly variable and no differences were seen between the number of metastases between the control and the knockdown tumours (Figure 4.3D). The number of metastases from 4TO7 tumours is normally relatively small since 4TO7 cells are unable form macrometastases. Despite the possibility that the effect of HMOX1 knockdown may not be significant due to absence of complete HMOX1 knockdown in 4TO7-shHMOX1 cells, the clonogenic assay is able to detect minor changes in the small 77  number of tumour cells present in the lungs. The limitation of detecting the effect of HMOX1 knockdown on lung metastases may be the heterogeneous population of tumour cells in the lungs of the mice bearing 4TO7-shGFP tumours seen in Figure 4.3D.    Figure 4.3 HMOX1 knockdown has no effect on 4TO7 tumour growth or lung metastasis. 106 4TO7 cells (shGFP, shHMOX1 #1, or shHMOX1 #4) were orthotopically implanted into the mammary fat pad of female NOD/SCID mice. A, levels of HMOX1 expression were analyzed by Western Blot on the day of implant, with tubulin acting as a loading control. B, tumour growth was examined by tumour measurements prior to the endpoint. C, tumours were excised at 21 days post-implant. D, the number of 4TO7 cells that metastasized to the lungs was assessed by clonogenic assay. N=8, one-tailed t-test.  78   Figure 4.4 HMOX1 overexpression has no effect on 4T1 tumour growth or lung metastasis. 106 4T1 cells (WT, Empty, or HMOX1) were orthotopically implanted into the mammary fat pad of female BALB/c mice. A, levels of HMOX1 expression were analyzed by Western Blot on the day of implant, with tubulin acting as a loading control. B, tumour growth was examined by tumour measurements prior to the endpoint. C, tumours were excised at 21 days post-implant. D, the number of 4T1 cells in the lungs was determined by flow cytometry analysis. N=8, one-tailed t-test.   4.2.2 Characterization of primary tumour growth and metastasis in the 4T1 tumour model – HMOX1 overexpression and knockdown  Similar to 4TO7 cells, 4T1 cells do not express HMOX1 in detectable levels in vitro, therefore we implanted HMOX1 overexpressing 4T1 cells in BALB/c mice. Unlike 4TO7 tumours, 4T1 cells are not immunogenic in BALB/c mice and generally form uniformly sized tumours. Before implant, 4T1-HMOX1 cells were verified for HMOX1 overexpression (Figure 4.4A). 4T1-HMOX1 tumours grew at the same rate as the controls and there were no differences in tumour weight at 21 days post-implant 79  (Figure 4.4B and C). The number of tumour cells in the lungs was analyzed by flow cytometry for hyperdiploid cyotokeratin-positive tumour cells. Differences were not observed in the amount of metastatic cells in the lungs between the groups (Figure D). This was an unexpected result, since in vitro, overexpression of HMOX1 decreased both cell migration and invasion. The tumour microenvironment is much more complex than cells grown in culture, therefore the decrease in migration and invasion phenotype in 4T1 cells by HMOX1 may have a lesser impact on the whole metastatic process in vivo. Alternately, the mechanisms of HMOX1 may be completely different in solid tumours. In addition, immunofluorescence of wild type 4T1 tumour sections revealed that 4T1 tumours express HMOX1 in vivo and therefore an overexpression of HMOX1 may have had no additional effect on tumour metastasis (Figure 4.5).    Figure 4.5 Immunofluorescent staining of HMOX1 in a 4T1 tumour. Representative image of frozen tumour sections stained with HMOX1, and DAPI. Scale bar = 200 μm.  Consequently, to continue examining the effect of HMOX1 on primary tumour growth and metastasis, we orthotopically implanted 4T1-shHMOX1 cells into BALB/c mice. Both shHMOX1 #1 and #4 achieved successful knockdown at the point of implant (Figure 4.6A). Tumour volume measurements showed that shHMOX1 #1 had a slower growth rate than shGFP and shHMOX1 #4 (Figure 4.6B). At the endpoint, 21 80  days post-implant, 4T1-shHMOX1 #1 tumours had significantly smaller tumours than shGFP as well as shHMOX1 #4 tumours (Figure 4.6C). Although the knockdown of HMOX1 was equally successful in both knockdown populations of cells prior to tumour cell implant, 4T1-shHMOX1 #4 did not have the same effect on tumour growth or weight as shHMOX1 #1. In addition to the lower tumour weight, 4T1-shHMOX1 #1 tumours had lower numbers of lung metastases compared to the control (Figure 4.6D). In contrast, 4T1-shHMOX1 #4 had no difference in the number of metastases compared to the control. Additionally, 4T1-shHMOX1 #4 tumours had a wide range of values of tumour weight (Figure 4.6C).  Figure 4.6 HMOX1 knockdown has an effect on 4T1 tumour growth and metastasis. 106 4T1 cells (shGFP, shHMOX1 #1, or shHMOX1 #4) were orthotopically implanted into the mammary fat pad of female BALB/c mice. A, levels of HMOX1 expression were analyzed by Western Blot on the day of implant, with tubulin acting as a loading control. B, tumour growth was examined by tumour measurements prior to the endpoint. C, tumours were excised at 21 days post-implant. shGFP N=10, shHMOX1 #1 and #4 N=11. D, number of 4T1 cells that have metastasized to the lungs was assessed by clonogenic assay. shGFP N=6, shHMOX1 #1 and #4 N=7. One-tailed t-test, * represents p≤0.05 compared to shGFP. 81  We investigated 4T1-shHMOX1 #4 tumours in detail due to the presence of variable tumour weights and the dissimilar results to shHMOX1 #1. We examined the frozen tumours for HMOX1 expression by immunofluorescence. Prior to HMOX1 expression analysis by immunofluorescence, we attempted to determine in vivo HMOX1 expression by utilizing western blots, but failed to obtain discernible differences in HMOX1 protein expression between any particular tumours. Analysis of tumour sections stained with antibody against HMOX1 showed the knockdown was lost in some 4T1-shHMOX1 #4 tumours but retained in others (Figure 4.7A). When we examined shHMOX1 #1 tumour sections for HMOX1 expression, there was consistent knockdown throughout the tumour samples, which is reflected in the consistent phenotype seen in Figure 4.6C and D. Tumour sections were graded for HMOX1 expression by comparing HMOX1 to DAPI signals. Figure 4.7A shows examples of high and low HMOX1 expression represented by the ratio of HMOX1 and DAPI expression. The red and blue circles in Figure 4.7 indicate the two representative tumours shown in Figure 4.7A. Figure 4.7 B compares HMOX1 expression to the number of tumour cells present in the lungs.   Our data suggest that higher HMOX1 expression may be correlated with higher number of tumour cells in the lungs. This agrees with shHMOX1 #1 tumours, since HMOX1 knockdown present in shHMOX1 #1 tumours decreased the number of tumour cells in the lungs. Therefore, this may indicate a possible role for HMOX1 in promoting metastasis in 4T1 tumours.  Since large tumours are often associated with larger numbers of metastases, we also compared tumour weights to lung metastases to determine whether HMOX1 knockdown had an effect on metastatic growth independent from the differences in primary tumour size (Figure 4.7C). Tumours of comparable sizes did not have similar number of metastases, and larger tumours did not necessarily have higher lung metastases, suggesting tumour size does not always dictate the number of lung metastases if tumour cells do not have the potential to metastasize. This is exemplified in large 67NR tumours that do not metastasize (23). HMOX1 expression (R2 = 0.5622) correlated more strongly with the number of metastases than it did with tumour weight in 4T1-shHMOX1 #4 82  tumours (R2 = 0.0004098) (Figure 4.7C). This suggests that the propensity to metastasize in 4T1 cells may be altered by HMOX1 expression and HMOX1 expression may a larger effect on lung metastasis than 4T1 tumour size. A further examination of HMOX1 knockdown tumours through additional experiments will provide a larger sample size and required to make a firm conclusion on the correlation between HMOX1 expression and the number of metastases.         Figure 4.7 HMOX1 levels in tumours correlate more with metastatic growth than primary tumour size. Frozen 4T1 shHMOX1 #4 tumour sections were stained with HMOX1 and DAPI.  A, examples of high (left) and low (right) HMOX1 expression. B, the number of tumour cells that have metastasized to the lungs was compared to the HMOX1 expression of the primary tumour. Red and blue circles represent the tumour sections in A. The ratio of HMOX1 and DAPI staining was analyzed using ImageJ. R2 = 0.5622. C, primary tumour weight was compared to the number of lung metastasis. R2=0.0004098. Original magnification 10X in A. 83   Certain normal cells such as macrophages and endothelial cells are known to express HMOX1 (16, 144, 148). Tumour cells make up the majority of the cell population in 4T1 tumours, with the next highest cell population in the tumour being endothelial cells at <4.5% of the tumour (Unpublished data, Bosiljcic et al.). Therefore, verifying HMOX1 levels by immunofluorescence in tumour sections does not necessarily show which cells are expressing HMOX1. To address this problem, we identified the sources of HMOX1 expression by co-staining 4T1-shGFP and 4T1-shHMOX1 #1 tumour sections with HMOX1 and a pan-myeloid cell marker, CD11b or an endothelial marker, CD31 (Figure 4.8).   In 4T1-shGFP tumours, an area of co-localization between HMOX1 and CD11b is shown in the inset in Figure 4.8A. The inset shows an area of immune cell infiltration and is likely a necrotic region of the tumour, based on tissue morphology. Outside of this area, HMOX1 is expressed from cells other than CD11b+ cells (Figure 4.8A, arrow). In contrast to the shGFP control tumour section, cells in the 4T1-shHMOX1 tumour generally lack HMOX1 expression. These tumour sections also demonstrated that in general, CD11b+ cells outside of the necrotic region, express undetectable levels of HMOX1 in 4T1 tumours. Tumour sections co-stained with HMOX1 and CD31 also revealed that CD31+ cells do not express HMOX1 (Figure 4.8B). This suggests that HMOX1 in a 4T1 tumour is largely expressed by tumour cells rather than CD11b+ or CD31+ cells. Therefore, validating HMOX1 knockdown through immunofluorescence is a suitable method of measuring the level of HMOX1 knockdown in the tumour. In addition, these preliminary observations revealed shHMOX1 tumours with fewer CD11b+ cells and less areas of necrosis compared to shGFP tumours. These data will be further discussed in Chapter 5.   84   Figure 4.8 Immunofluorescent staining of HMOX1 and normal cells in 4T1 tumours. A, frozen tumour sections were stained with HMOX1, CD11b, and DAPI. The inset is the magnified area from the boxed area. Arrow indicates areas with HMOX1 expression that lack CD11b expression. B, frozen tumour sections were stained with HMOX1, CD31, and DAPI. Left-hand side – original magnification 10X, right-hand side – original magnification 20X in A and original magnification 10X in B. Scale bar = 200 μm.       85  4.3 Discussion  As a major player in cellular stress response, HMOX1 is expected to have a role in tumour progression and many groups have examined its function in various types of cancers. Despite the lack of expression under normal tissue culture conditions, the expression of HMOX1 in 4TO7 and 4T1 tumours exemplifies the intricacy of the tumour microenvironment and how it can affect HMOX1. Firstly, hemorrhages and neovascularization robustly occur in tumours, which lead to a large influx of heme derived from extravasated red blood cell deposits that in turn induce HMOX1 (152). Secondly, HMOX1 is part of multiple signaling pathways that are augmented in tumourigenesis including AP-1 and NF-κB (114). Although there is some functional overlap between these pathways, each regulator plays a different role in controlling cellular stress response and thereby adjusting HMOX1 expression. HMOX1 is also induced by many of the factors involved in the metastatic cascade including VEGF, and IL-1 (149, 159). In 4T1 tumours, HMOX1 was primarily expressed by tumour cells instead of normal cells such as macrophages and endothelial cells. In addition to the varying HMOX1 expression in tumour cells of different cancers, the expression of HMOX1 varies in the normal cells present in tumour microenvironments of different tumours (83, 150). Whether the presence or the lack of HMOX1 expression in normal cells influences the metastatic phenotype of the tumour cell or metastasis in general is unknown. Since in 4T1 tumours, CD11b+ cells express low levels of HMOX1 outside of necrotic areas, induction of HMOX1 expression in macrophages with IL-10 in wild type tumours with or without the use of ZnPP, a HMOX1 inhibitor, may be a potential study to examine the influence of HMOX1 expressing macrophages on primary tumour growth and metastasis.   Investigating the effect that HMOX1 has on migration and invasion in vitro showed that invasion was increased in 67NR and 4TO7 cells overexpressing HMOX1, while invasion was reduced in 4T1-HMOX1 cells. This revealed a complex dichotomy in the role of HMOX1 in mammary tumour cell invasion similar to what other groups have seen in human breast cancer in vitro (90, 91, 102). Observations in vitro 86  however did not translate in vivo. Since invasion was decreased in 4T1-HMOX1 cells in vitro, we expected 4T1-shHMOX1 cells to exhibit increased lung metastases in vivo. Instead, the loss of HMOX1 expression led to a decrease in tumour growth and in lung metastases. It is possible that in vivo, the cytoprotective role of HMOX1 is amplified due to the number of stress stimuli and tumour cells may manipulate this role of HMOX1 to survive. Therefore the lack of HMOX1 may lead to an absence of cytoprotection in knockdown cells and in turn cell death. HMOX1 may have a dual function of mediating cytoprotection and affecting metastasis. Establishing if HMOX1 knockdown is potentially selected against during tumour growth, as found in some 4T1-shHMOX1 #4 tumours, will help in understanding which tumourigenic property of the tumour cells is affected by HMOX1. In addition, assessing which proportion of the cells that reach the lungs maintain HMOX1 knockdown may be valuable in determining whether tumour cells exhibiting a metastatic phenotype express HMOX1 or not in a potentially heterogeneous population of tumour cells in 4T1-shHMOX1 tumours.   Studies that examine the effect of upstream regulators of HMOX1 in breast cancer metastasis also show a dichotomy. For example, orthotopic implants of 4T1 cells overexpressing miR217, which targets the 3’-UTR of HMOX1, show a decrease in tumour mass and lung metastases similar to 4T1-shHMOX1 tumours (160). However, in MDA-MB-231 cells, increased Raf kinase inhibitory protein leads to a decrease in the repressor of HMOX1, BACH1 and subsequently decreased bone metastasis (161). Interestingly, this decrease in BACH1 induced MMP1 expression in these cells, which mirrors the increase in MMP1 seen in MCF-7 cells treated with hemin or CoPP in vitro (90). The dichotomy of HMOX1’s influence on metastasis exists within in vivo findings as well.     Although 4TO7 tumours do form micrometastases, the heterogeneity in metastatic growth between mice makes it very difficult to observe effective therapeutic interventions such as HMOX1 knockdown in this model. The number of lung metastases in 4TO7-shHMOX1 #1 tumour bearing mice was the most consistent. 87  However, as seen in Figure 4.3, the number of tumour cells detected in the lungs was variable in the control group. Therefore we were unable to detect possible decrease in the number of lung metastases in 4TO7-shHMOX1 #1 group.   4T1-shHMOX1 #1 and #4 tumours offer an interesting view of how HMOX1 may have an effect on primary tumour growth and metastasis. Although both shHMOX1 #1 and #4 tumours had successful HMOX1 knockdown before implant, in shHMOX1 #1 tumours, the HMOX1 expression was sustained in vivo. These tumours were smaller and had lower numbers of tumour cells in the lungs. The decrease in lung metastases may be due to slower tumour growth and therefore metastatic tumours appearing later. Tumours within the 4T1-shHMOX1 #4 group with varying HMOX1 expression may provide a possible indication of whether the decrease in HMOX1 in shHMOX1 #1 tumours led to the decrease in lung metastases by changing the tumour cells metastatic phenotype or by decreasing primary tumour growth. HMOX1 levels in shHMOX1 #4 tumours were more strongly correlated with metastatic growth than primary tumour size. Depending on at what point knock down was lost in some of the tumours, the effect of HMOX1 knockdown on primary tumour growth and metastasis may be variable. Given that the decrease in lung metastases was observed in 4T1-shHMOX1 #1 tumour bearing mice, supporting experiments involving therapeutic inhibition of HMOX1 are necessary to conclude whether breast cancer metastasis correlates better with HMOX1 expression than it does with tumour weight. Future studies involving therapeutic HMOX1 inhibitors will be described further in Chapter 5.   88  Chapter 5 Summary and Future Directions  Our study examined the effect of HMOX1 in breast cancer metastasis in vitro and in vivo by using murine mammary carcinoma cell lines with varying degrees of metastatic potential. Results from both the HMOX1 knockdown and overexpression experiments showed that hemin, which decreased migration and invasion in 67NR, 4TO7, 4T1 cells, may have effects that do not depend on the induction of HMOX1 expression. Interestingly, overexpression of HMOX1 did lead to a decrease in migration in all three cell lines. However, HMOX1 increased invasion in 67NR and 4TO7 cells, but decreased the number of invaded 4T1 cells. This demonstrates a cell type specific role of HMOX1. Understanding the molecular mechanism of the effect of HMOX1 on migration and invasion will help to interpret why this discrepancy exists between the cell lines.   We began to answer this question by examining the cellular localization of HMOX1 to ascertain whether HMOX1 played a role on migration and invasion through its enzymatic mechanism or whether HMOX1 is possibly acting as a transcription factor in the nucleus. In all of the cell lines, overexpression of HMOX1 was consolidated in the endoplasmic reticulum, suggesting that HMOX1 is not localized to the nucleus for direct transcriptional function. To confirm that HMOX1 is not acting as a transcription factor in the murine mammary cell lines treated with hemin or overexpressing HMOX1, potential expression arrays of oxidative stress response genes, such as Stat1 and Stat3 can be conducted. Although groups have observed direct binding of HMOX1 to promoter regions of oxidative stress response genes, it is important to note that CO, a product of HMOX1 enzymatic activity, is also capable of modulating gene expression of transcription factors such as STAT1 and STAT3 (162). Additionally, unbound HMOX1 lacking the C-terminal hydrophobic tail can be found in the cytoplasm and has been observed to retain of catalytic activity (84, 107, 108, 112). Therefore an analysis of HMOX1 enzymatic activity through measuring CO production through a gas analyzer or spectrophotometric analysis of bilirubin production can be conducted in vitro (163, 164). This will provide confirmation that 89  may support that the localization of HMOX1 in the sER and not the nucleus suggest that HMOX1 expressing cells are affecting migration and invasion through its enzymatic mechanism.       We also evaluated MMP9 and MMP2 levels. In 4T1 cells, HMOX1 overexpressing cells had a decrease in MMP9, however, the level of MMP9 was still high in all of the 4T1 cells. This suggests that there may be other factors involved in invasion that are altered or HMOX1 may be affecting 4T1 cell migration, and thereby affecting the migration of 4T1 cells after the degradation of the Matrigel during the invasion assay. In 67NR and 4TO7 cells, the decrease in MMP9 and MMP2 did not result in a decrease in invasion, but instead invasion was increased after HMOX1 overexpression. Using specific inhibitors against MMP9 or MMP2 in 67NR and 4TO7 cells overexpressing HMOX1 may demonstrate whether HMOX1 overexpression can increase invasion independently from MMP9 or MMP2.   In addition, a proteomics-based approach  can be taken to determine changes in the expression of proteins involved in invasion such as integrins or MMPs other than MMP9 and MMP2 between cells expression high and low HMOX1 levels. Results from the proteomics may give us a clue on which downstream target of HMOX1 is affecting migration and invasion in each of these cell lines.   The effect of HMOX1 was cell-specific and moreover, within a cell type, HMOX1 expression was different in vivo and in vitro. HMOX1 was expressed in 4TO7 and 4T1 tumours, and HMOX1 overexpressing tumours had no additive effect on metastasis of these tumours. Knockdown of HMOX1 in 4T1 tumours suggest an effect on both the primary lesion and the number of lung metastases.    90   Figure 5.1 Immunofluorescent staining of hypoxia, vasculature, apoptotic and proliferation markers in 4T1 tumours expressing shGFP or shHMOX1. Frozen tumour sections from Fig 4.6 were stained with CA9 (hypoxia), CD31 (endothelial cells), and DAPI (nuclear) in A and PCNA (proliferation), Caspase3 (apoptosis), and DAPI in B. Scale bar = 200 μm.     Additionally, a preliminary analysis of frozen tumour sections of 4T1 tumours with HMOX1 knocked down revealed areas of hypoxia and necrosis are reduced in 4T1-shHMOX1 tumours despite no apparent differences in blood vessel density based on CD31-positive staining (Figure 4.8 and Figure 5.1A). Since hypoxia is known to promote metastatic and cancer progression, further investigation on the effect of HMOX1 on tumour hypoxia is a logical next step in studying the role of HMOX1 in breast cancer metastasis.  Consequently, studying vasculature and perfusion in 4T1 tumours in comparison to expression of HMOX1 will inform us on whether a decrease in hypoxic regions in 4T1-shHMOX1 tumours exists and the mechanism behind an increase in tumour oxygenation or potentially the effect of HMOX1 on HIF-91  1α activity. Furthermore, apoptosis and proliferation within the tumour are shown to be reduced in the 4T1-shHMOX1 tumour section, indicating differences in tumour cell turnover between tumours with HMOX1 knocked down and the control tumours (Figure 5.1B). Therefore, we will examine 4T1 tumour sections to study possible correlations between HMOX1 expression and apoptosis. Additionally, to supplement the observations made in the tumour sections, further study of tumour cell turnover will be conducted by investigating the role of HMOX1 in 4T1 cell proliferation and apoptosis. This preliminary data may help to uncover a novel role for HMOX1 in affecting the tumour microenvironment and ultimately tumour progression and metastasis of breast cancer.  Since HMOX1 knockdown was lost in vivo for some of the tumours, an HMOX1 inhibitor such as ZnPP is an alternative method to establishing the efficacy of targeting HMOX1 in breast cancer. ZnPP is a well-known selective inhibitor of HMOX1 and has been used as an anti-tumour reagent in various cancers (83, 85, 93, 101, 105). HMOX1 catalysis involves three oxygenation steps that bind the iron molecule in the middle of the heme ring to molecular oxygen. ZnPP is a non-metabolizable protoporphyrin that competitively inhibits HMOX1 by occupying the active site but inefficiently binding to molecular oxygen, therefore preventing HMOX1 from metabolizing metalloporphyrins (165). Intraperitoneal or intra-arterial injections of ZnPP have been shown to increase apoptosis of tumour cells and diminish tumour growth by decreasing HMOX1 activity in colon, urothelial, and renal cancer (83, 93, 105). ZnPP has low solubility in water. Pharmacokinetic studies between pegylated ZnPP and native ZnPP show that the majority of both molecules sequester in the spleen and the liver after an intravenous injection (85). However, pegylated ZnPP accumulates better in the tumour than native ZnPP and the maximum accumulation is seen at 48 hours compared to native ZnPP which reaches maximum accumulation in the tumour at 8 hours post injection.   The effect of ZnPP on metastasis has not been investigated. Using ZnPP to treat 4T1 wild type tumours and assessing primary tumour growth as well as lung 92  metastasis may enhance our understanding of the role of HMOX1 on breast cancer metastasis. Furthermore, investigating the level of HMOX1 expression of the disseminated tumour cells compared to cells in the primary tumour may provide an indication on whether HMOX1 affects metastasis by changing the invasive properties of the tumour cells or by altering the primary tumour environment as a whole. ZnPP has also been suggested as a potentially new therapeutic compound that may be used clinically in addition to already existing therapeutics such as gemcitabine (82, 93, 97). However for ZnPP to be an effective compound, tumour specific inhibition of HMOX1 must be considered, as HMOX1 is a critical enzyme in many organs.     The incongruity between our in vitro and in vivo results calls for further investigation on the role of HMOX1 in breast cancer metastasis in vivo.  4T1 cells overexpressing HMOX1 have decreased motility and invasive properties in vitro. However, this did not translate in vivo. Additional 4T1-shHMOX1 #1 tumours may need to be analyzed to see if the reduced metastasis after HMOX1 knockdown can be reproduced. In addition, how HMOX1 affects metastasis still needs to be investigated since in vivo, HMOX1 may not be necessarily affecting metastasis through decreasing tumour cell migration and invasion. Tail vein assay involving intravenous injections of tumour cells assesses the propensity of the tumour cells to complete the subsequent steps of metastasis after invasion and intravasation. Therefore, intravenously injecting tumour cells expressing luciferase and quantifying the accumulation of tumour cells expressing shHMOX1 compared to control cells in the lungs could be used to study the role of HMOX1 in extravasation and metastatic colonization. Furthermore, examining the expression of HMOX1 in the population of tumour cells at the primary tumour and the distant metastases in both wild type and shHMOX1 tumours may determine whether expression of HMOX1 predisposes these tumours to metastasis.   Using murine mammary carcinoma cell lines may give us an indication of the role of HMOX1 in human breast cancer metastasis. However, the cell- and species-specificity of HMOX1’s mechanism of action prompt a need to investigate the role of 93  HMOX1 on breast cancer in human breast cancer cell lines in vivo. A human breast cancer cell line such as MDA-MB-231, which has similar molecular signatures to 4T1 cells including a triple negative status and the ability to form macrometastases in the lungs may be a suitable model (166).   The data in this thesis suggest that HMOX1 may have a role in breast cancer metastasis. Specifically, in 4T1 tumours, HMOX1 knockdown led to a decrease in primary tumour growth and a reduction in lung metastases. Our data may support the therapeutic strategies to inhibit HMOX1 in breast cancer to enhance the outcome of patients at the risk of developing metastatic disease. Improving our understanding of the development and growth of metastatic tumours is crucial to effectively treat or prevent tumour metastases. This study represents an important first step in determining the importance and viability of HMOX1 as a therapeutic target for metastatic breast cancer.  94  References  1. Statistics CCSsACoC. Canadian Cancer Statistics 2013. Toronto, ON: Canadian Cancer Society; 2013. 2. Lazebnik Y. What are the hallmarks of cancer? Nature reviews Cancer. 2010;10:232-3. 3. Giordano A. DFG, Rubin E., and Rubin R. . "Rubin’s Pathology: Clinicopathologic Foundations of Medicine". Baltimore, Philadelphia: Wolters Kluwer, Lippincott Williams & Williams; 2008. 4. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100:57-70. 5. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144:646-74. 6. Hanahan D, Folkman J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell. 1996;86:353-64. 7. Negrini S, Gorgoulis VG, Halazonetis TD. Genomic instability--an evolving hallmark of cancer. Nature reviews Molecular cell biology. 2010;11:220-8. 8. Weigelt B, Peterse JL, van 't Veer LJ. Breast cancer metastasis: markers and models. Nature reviews Cancer. 2005;5:591-602. 9. Johnson RH, Chien FL, Bleyer A. Incidence of breast cancer with distant involvement among women in the United States, 1976 to 2009. JAMA : the journal of the American Medical Association. 2013;309:800-5. 10. de Ruijter TC, Veeck J, de Hoon JP, van Engeland M, Tjan-Heijnen VC. Characteristics of triple-negative breast cancer. Journal of cancer research and clinical oncology. 2011;137:183-92. 11. Tavassoli F.A. DP. World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of the Breast and Female Genital Organs. Lyon: IARC Press; 2003. 12. Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA, et al. Molecular portraits of human breast tumours. Nature. 2000;406:747-52. 13. Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proceedings of the National Academy of Sciences of the United States of America. 2001;98:10869-74. 14. Voduc KD, Cheang MC, Tyldesley S, Gelmon K, Nielsen TO, Kennecke H. Breast cancer subtypes and the risk of local and regional relapse. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2010;28:1684-91. 15. Carey LA, Perou CM, Livasy CA, Dressler LG, Cowan D, Conway K, et al. Race, breast cancer subtypes, and survival in the Carolina Breast Cancer Study. JAMA : the journal of the American Medical Association. 2006;295:2492-502. 16. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science. 1987;235:177-82. 17. Delord JP, Allal C, Canal M, Mery E, Rochaix P, Hennebelle I, et al. Selective inhibition of HER2 inhibits AKT signal transduction and prolongs disease-free survival in a micrometastasis model of ovarian carcinoma. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO. 2005;16:1889-97. 95  18. Weigelt B, Horlings HM, Kreike B, Hayes MM, Hauptmann M, Wessels LF, et al. Refinement of breast cancer classification by molecular characterization of histological special types. The Journal of pathology. 2008;216:141-50. 19. Kennecke H, Yerushalmi R, Woods R, Cheang MC, Voduc D, Speers CH, et al. Metastatic behavior of breast cancer subtypes. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2010;28:3271-7. 20. Fisher B, Anderson S, Redmond CK, Wolmark N, Wickerham DL, Cronin WM. Reanalysis and results after 12 years of follow-up in a randomized clinical trial comparing total mastectomy with lumpectomy with or without irradiation in the treatment of breast cancer. The New England journal of medicine. 1995;333:1456-61. 21. Dexter DL, Kowalski HM, Blazar BA, Fligiel Z, Vogel R, Heppner GH. Heterogeneity of tumor cells from a single mouse mammary tumor. Cancer research. 1978;38:3174-81. 22. Wendt MK, Smith JA, Schiemann WP. Transforming growth factor-beta-induced epithelial-mesenchymal transition facilitates epidermal growth factor-dependent breast cancer progression. Oncogene. 2010;29:6485-98. 23. Aslakson CJ, Miller FR. Selective events in the metastatic process defined by analysis of the sequential dissemination of subpopulations of a mouse mammary tumor. Cancer research. 1992;52:1399-405. 24. Dutertre M, Lacroix-Triki M, Driouch K, de la Grange P, Gratadou L, Beck S, et al. Exon-based clustering of murine breast tumor transcriptomes reveals alternative exons whose expression is associated with metastasis. Cancer research. 2010;70:896-905. 25. Eckhardt BL, Parker BS, van Laar RK, Restall CM, Natoli AL, Tavaria MD, et al. Genomic analysis of a spontaneous model of breast cancer metastasis to bone reveals a role for the extracellular matrix. Molecular cancer research : MCR. 2005;3:1-13. 26. Bao L, Haque A, Jackson K, Hazari S, Moroz K, Jetly R, et al. Increased expression of P-glycoprotein is associated with doxorubicin chemoresistance in the metastatic 4T1 breast cancer model. The American journal of pathology. 2011;178:838-52. 27. Chaffer CL, Weinberg RA. A perspective on cancer cell metastasis. Science. 2011;331:1559-64. 28. Klein CA. Parallel progression of primary tumours and metastases. Nature reviews Cancer. 2009;9:302-12. 29. Steeg PS. Tumor metastasis: mechanistic insights and clinical challenges. Nature medicine. 2006;12:895-904. 30. van der Horst EH, Degenhardt YY, Strelow A, Slavin A, Chinn L, Orf J, et al. Metastatic properties and genomic amplification of the tyrosine kinase gene ACK1. Proceedings of the National Academy of Sciences of the United States of America. 2005;102:15901-6. 31. Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature. 2007;449:682-8. 32. Tavazoie SF, Alarcon C, Oskarsson T, Padua D, Wang Q, Bos PD, et al. Endogenous human microRNAs that suppress breast cancer metastasis. Nature. 2008;451:147-52. 33. Steeg PS. Metastasis suppressors alter the signal transduction of cancer cells. Nature reviews Cancer. 2003;3:55-63. 34. Weiss L, Elkin G, Barbera-Guillem E. The differential resistance of B16 wild-type and F10 cells to mechanical trauma in vitro. Invasion & metastasis. 1993;13:92-101. 35. Muller A, Homey B, Soto H, Ge N, Catron D, Buchanan ME, et al. Involvement of chemokine receptors in breast cancer metastasis. Nature. 2001;410:50-6. 96  36. Deshane J, Chen S, Caballero S, Grochot-Przeczek A, Was H, Li Calzi S, et al. Stromal cell-derived factor 1 promotes angiogenesis via a heme oxygenase 1-dependent mechanism. The Journal of experimental medicine. 2007;204:605-18. 37. Paget S. The distribution of secondary growths in cancer of the breast. 1889. Cancer metastasis reviews. 1989;8:98-101. 38. Peinado H, Lavotshkin S, Lyden D. The secreted factors responsible for pre-metastatic niche formation: old sayings and new thoughts. Seminars in cancer biology. 2011;21:139-46. 39. Chambers AF, Groom AC, MacDonald IC. Dissemination and growth of cancer cells in metastatic sites. Nature reviews Cancer. 2002;2:563-72. 40. Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, Giri DD, et al. Genes that mediate breast cancer metastasis to lung. Nature. 2005;436:518-24. 41. Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, Cordon-Cardo C, et al. A multigenic program mediating breast cancer metastasis to bone. Cancer cell. 2003;3:537-49. 42. Wong CW, Lee A, Shientag L, Yu J, Dong Y, Kao G, et al. Apoptosis: an early event in metastatic inefficiency. Cancer research. 2001;61:333-8. 43. Zhan M, Zhao H, Han ZC. Signalling mechanisms of anoikis. Histology and histopathology. 2004;19:973-83. 44. Fidler IJ, Nicolson GL. Fate of recirculating B16 melanoma metastatic variant cells in parabiotic syngeneic recipients. Journal of the National Cancer Institute. 1977;58:1867-72. 45. Aguirre-Ghiso JA. Models, mechanisms and clinical evidence for cancer dormancy. Nature reviews Cancer. 2007;7:834-46. 46. Weaver VM, Petersen OW, Wang F, Larabell CA, Briand P, Damsky C, et al. Reversion of the malignant phenotype of human breast cells in three-dimensional culture and in vivo by integrin blocking antibodies. The Journal of cell biology. 1997;137:231-45. 47. Friedl P, Wolf K. Tumour-cell invasion and migration: diversity and escape mechanisms. Nature reviews Cancer. 2003;3:362-74. 48. Price JT, Tiganis T, Agarwal A, Djakiew D, Thompson EW. Epidermal growth factor promotes MDA-MB-231 breast cancer cell migration through a phosphatidylinositol 3'-kinase and phospholipase C-dependent mechanism. Cancer research. 1999;59:5475-8. 49. Guo W, Giancotti FG. Integrin signalling during tumour progression. Nature reviews Molecular cell biology. 2004;5:816-26. 50. Cukierman E, Pankov R, Stevens DR, Yamada KM. Taking cell-matrix adhesions to the third dimension. Science. 2001;294:1708-12. 51. Rabinovitz I, Mercurio AM. The integrin alpha6beta4 functions in carcinoma cell migration on laminin-1 by mediating the formation and stabilization of actin-containing motility structures. The Journal of cell biology. 1997;139:1873-84. 52. Giannelli G, Falk-Marzillier J, Schiraldi O, Stetler-Stevenson WG, Quaranta V. Induction of cell migration by matrix metalloprotease-2 cleavage of laminin-5. Science. 1997;277:225-8. 53. Kessenbrock K, Plaks V, Werb Z. Matrix metalloproteinases: regulators of the tumor microenvironment. Cell. 2010;141:52-67. 54. Mitra SK, Hanson DA, Schlaepfer DD. Focal adhesion kinase: in command and control of cell motility. Nature reviews Molecular cell biology. 2005;6:56-68. 55. Bravo-Cordero JJ, Hodgson L, Condeelis J. Directed cell invasion and migration during metastasis. Current opinion in cell biology. 2012;24:277-83. 56. Artym VV, Zhang Y, Seillier-Moiseiwitsch F, Yamada KM, Mueller SC. Dynamic interactions of cortactin and membrane type 1 matrix metalloproteinase at invadopodia: 97  defining the stages of invadopodia formation and function. Cancer research. 2006;66:3034-43. 57. Parekh A, Ruppender NS, Branch KM, Sewell-Loftin MK, Lin J, Boyer PD, et al. Sensing and modulation of invadopodia across a wide range of rigidities. Biophysical journal. 2011;100:573-82. 58. Cavallaro U, Christofori G. Cell adhesion and signalling by cadherins and Ig-CAMs in cancer. Nature reviews Cancer. 2004;4:118-32. 59. Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, Come C, et al. Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell. 2004;117:927-39. 60. Lou Y, Preobrazhenska O, auf dem Keller U, Sutcliffe M, Barclay L, McDonald PC, et al. Epithelial-mesenchymal transition (EMT) is not sufficient for spontaneous murine breast cancer metastasis. Developmental dynamics : an official publication of the American Association of Anatomists. 2008;237:2755-68. 61. Taylor MA, Sossey-Alaoui K, Thompson CL, Danielpour D, Schiemann WP. TGF-beta upregulates miR-181a expression to promote breast cancer metastasis. The Journal of clinical investigation. 2013;123:150-63. 62. Dykxhoorn DM, Wu Y, Xie H, Yu F, Lal A, Petrocca F, et al. miR-200 enhances mouse breast cancer cell colonization to form distant metastases. PloS one. 2009;4:e7181. 63. Maeda M, Johnson KR, Wheelock MJ. Cadherin switching: essential for behavioral but not morphological changes during an epithelium-to-mesenchyme transition. Journal of cell science. 2005;118:873-87. 64. Nguyen-Ngoc KV, Cheung KJ, Brenot A, Shamir ER, Gray RS, Hines WC, et al. ECM microenvironment regulates collective migration and local dissemination in normal and malignant mammary epithelium. Proceedings of the National Academy of Sciences of the United States of America. 2012;109:E2595-604. 65. Gaggioli C, Hooper S, Hidalgo-Carcedo C, Grosse R, Marshall JF, Harrington K, et al. Fibroblast-led collective invasion of carcinoma cells with differing roles for RhoGTPases in leading and following cells. Nature cell biology. 2007;9:1392-400. 66. Wang W, Wyckoff JB, Frohlich VC, Oleynikov Y, Huttelmaier S, Zavadil J, et al. Single cell behavior in metastatic primary mammary tumors correlated with gene expression patterns revealed by molecular profiling. Cancer research. 2002;62:6278-88. 67. Giampieri S, Manning C, Hooper S, Jones L, Hill CS, Sahai E. Localized and reversible TGFbeta signalling switches breast cancer cells from cohesive to single cell motility. Nature cell biology. 2009;11:1287-96. 68. Wolf K, Mazo I, Leung H, Engelke K, von Andrian UH, Deryugina EI, et al. Compensation mechanism in tumor cell migration: mesenchymal-amoeboid transition after blocking of pericellular proteolysis. The Journal of cell biology. 2003;160:267-77. 69. Cooke VG, LeBleu VS, Keskin D, Khan Z, O'Connell JT, Teng Y, et al. Pericyte depletion results in hypoxia-associated epithelial-to-mesenchymal transition and metastasis mediated by met signaling pathway. Cancer cell. 2012;21:66-81. 70. Pages F, Galon J, Dieu-Nosjean MC, Tartour E, Sautes-Fridman C, Fridman WH. Immune infiltration in human tumors: a prognostic factor that should not be ignored. Oncogene. 2010;29:1093-102. 71. Cairns RA, Kalliomaki T, Hill RP. Acute (cyclic) hypoxia enhances spontaneous metastasis of KHT murine tumors. Cancer research. 2001;61:8903-8. 72. Graeber TG, Osmanian C, Jacks T, Housman DE, Koch CJ, Lowe SW, et al. Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours. Nature. 1996;379:88-91. 98  73. Lunt SJ, Chaudary N, Hill RP. The tumor microenvironment and metastatic disease. Clinical & experimental metastasis. 2009;26:19-34. 74. Lou Y, McDonald PC, Oloumi A, Chia S, Ostlund C, Ahmadi A, et al. Targeting tumor hypoxia: suppression of breast tumor growth and metastasis by novel carbonic anhydrase IX inhibitors. Cancer research. 2011;71:3364-76. 75. Erler JT, Bennewith KL, Nicolau M, Dornhofer N, Kong C, Le QT, et al. Lysyl oxidase is essential for hypoxia-induced metastasis. Nature. 2006;440:1222-6. 76. Warburg O. On respiratory impairment in cancer cells. Science. 1956;124:269-70. 77. Semenza GL. HIF-1: upstream and downstream of cancer metabolism. Current opinion in genetics & development. 2010;20:51-6. 78. Brown JM, Giaccia AJ. The unique physiology of solid tumors: opportunities (and problems) for cancer therapy. Cancer research. 1998;58:1408-16. 79. Baluk P, Morikawa S, Haskell A, Mancuso M, McDonald DM. Abnormalities of basement membrane on blood vessels and endothelial sprouts in tumors. The American journal of pathology. 2003;163:1801-15. 80. Andersen MH, Sorensen RB, Brimnes MK, Svane IM, Becker JC, thor Straten P. Identification of heme oxygenase-1-specific regulatory CD8+ T cells in cancer patients. The Journal of clinical investigation. 2009;119:2245-56. 81. Banerjee P, Basu A, Datta D, Gasser M, Waaga-Gasser AM, Pal S. The heme oxygenase-1 protein is overexpressed in human renal cancer cells following activation of the Ras-Raf-ERK pathway and mediates anti-apoptotic signal. The Journal of biological chemistry. 2011;286:33580-90. 82. Berberat PO, Dambrauskas Z, Gulbinas A, Giese T, Giese N, Kunzli B, et al. Inhibition of heme oxygenase-1 increases responsiveness of pancreatic cancer cells to anticancer treatment. Clinical cancer research : an official journal of the American Association for Cancer Research. 2005;11:3790-8. 83. Doi K, Akaike T, Fujii S, Tanaka S, Ikebe N, Beppu T, et al. Induction of haem oxygenase-1 nitric oxide and ischaemia in experimental solid tumours and implications for tumour growth. British journal of cancer. 1999;80:1945-54. 84. Elguero B, Gueron G, Giudice J, Toscani MA, De Luca P, Zalazar F, et al. Unveiling the association of STAT3 and HO-1 in prostate cancer: role beyond heme degradation. Neoplasia. 2012;14:1043-56. 85. Fang J, Sawa T, Akaike T, Akuta T, Sahoo SK, Khaled G, et al. In vivo antitumor activity of pegylated zinc protoporphyrin: targeted inhibition of heme oxygenase in solid tumor. Cancer research. 2003;63:3567-74. 86. Ferrando M, Gueron G, Elguero B, Giudice J, Salles A, Leskow FC, et al. Heme oxygenase 1 (HO-1) challenges the angiogenic switch in prostate cancer. Angiogenesis. 2011;14:467-79. 87. Gueron G, De Siervi A, Ferrando M, Salierno M, De Luca P, Elguero B, et al. Critical role of endogenous heme oxygenase 1 as a tuner of the invasive potential of prostate cancer cells. Molecular cancer research : MCR. 2009;7:1745-55. 88. Hamamura RS, Ohyashiki JH, Kurashina R, Kobayashi C, Zhang Y, Takaku T, et al. Induction of heme oxygenase-1 by cobalt protoporphyrin enhances the antitumour effect of bortezomib in adult T-cell leukaemia cells. British journal of cancer. 2007;97:1099-105. 89. Hill M, Pereira V, Chauveau C, Zagani R, Remy S, Tesson L, et al. Heme oxygenase-1 inhibits rat and human breast cancer cell proliferation: mutual cross inhibition with indoleamine 2,3-dioxygenase. FASEB journal : official publication of the Federation of American Societies for Experimental Biology. 2005;19:1957-68. 90. Kim DH, Kim JH, Kim EH, Na HK, Cha YN, Chung JH, et al. 15-Deoxy-Delta12,14-prostaglandin J2 upregulates the expression of heme oxygenase-1 and subsequently matrix 99  metalloproteinase-1 in human breast cancer cells: possible roles of iron and ROS. Carcinogenesis. 2009;30:645-54. 91. Lin CW, Shen SC, Hou WC, Yang LY, Chen YC. Heme oxygenase-1 inhibits breast cancer invasion via suppressing the expression of matrix metalloproteinase-9. Molecular cancer therapeutics. 2008;7:1195-206. 92. Mayerhofer M, Gleixner KV, Mayerhofer J, Hoermann G, Jaeger E, Aichberger KJ, et al. Targeting of heat shock protein 32 (Hsp32)/heme oxygenase-1 (HO-1) in leukemic cells in chronic myeloid leukemia: a novel approach to overcome resistance against imatinib. Blood. 2008;111:2200-10. 93. Miyake M, Fujimoto K, Anai S, Ohnishi S, Nakai Y, Inoue T, et al. Inhibition of heme oxygenase-1 enhances the cytotoxic effect of gemcitabine in urothelial cancer cells. Anticancer research. 2010;30:2145-52. 94. Morimoto K, Ohta K, Yachie A, Yang Y, Shimizu M, Goto C, et al. Cytoprotective role of heme oxygenase (HO)-1 in human kidney with various renal diseases. Kidney international. 2001;60:1858-66. 95. Noh SJ, Bae JS, Jamiyandorj U, Park HS, Kwon KS, Jung SH, et al. Expression of nerve growth factor and heme oxygenase-1 predict poor survival of breast carcinoma patients. BMC cancer. 2013;13:516. 96. Nowis D, Legat M, Grzela T, Niderla J, Wilczek E, Wilczynski GM, et al. Heme oxygenase-1 protects tumor cells against photodynamic therapy-mediated cytotoxicity. Oncogene. 2006;25:3365-74. 97. Nuhn P, Kunzli BM, Hennig R, Mitkus T, Ramanauskas T, Nobiling R, et al. Heme oxygenase-1 and its metabolites affect pancreatic tumor growth in vivo. Molecular cancer. 2009;8:37. 98. Ruiz-Ramos R, Lopez-Carrillo L, Rios-Perez AD, De Vizcaya-Ruiz A, Cebrian ME. Sodium arsenite induces ROS generation, DNA oxidative damage, HO-1 and c-Myc proteins, NF-kappaB activation and cell proliferation in human breast cancer MCF-7 cells. Mutation research. 2009;674:109-15. 99. Sunamura M, Duda DG, Ghattas MH, Lozonschi L, Motoi F, Yamauchi J, et al. Heme oxygenase-1 accelerates tumor angiogenesis of human pancreatic cancer. Angiogenesis. 2003;6:15-24. 100. Tauber S, Jais A, Jeitler M, Haider S, Husa J, Lindroos J, et al. Transcriptome analysis of human cancer reveals a functional role of heme oxygenase-1 in tumor cell adhesion. Molecular cancer. 2010;9:200. 101. Terman DS, Viglianti BL, Zennadi R, Fels D, Boruta RJ, Yuan H, et al. Sickle erythrocytes target cytotoxics to hypoxic tumor microvessels and potentiate a tumoricidal response. PloS one. 2013;8:e52543. 102. Wang C, Hu F, Guo S, Mi D, Shen W, Zhang J, et al. BMP-6 inhibits MMP-9 expression by regulating heme oxygenase-1 in MCF-7 breast cancer cells. Journal of cancer research and clinical oncology. 2011;137:985-95. 103. Was H, Dulak J, Jozkowicz A. Heme oxygenase-1 in tumor biology and therapy. Current drug targets. 2010;11:1551-70. 104. Zou C, Zhang H, Li Q, Xiao H, Yu L, Ke S, et al. Heme oxygenase-1: a molecular brake on hepatocellular carcinoma cell migration. Carcinogenesis. 2011;32:1840-8. 105. Nowis D, Bugajski M, Winiarska M, Bil J, Szokalska A, Salwa P, et al. Zinc protoporphyrin IX, a heme oxygenase-1 inhibitor, demonstrates potent antitumor effects but is unable to potentiate antitumor effects of chemotherapeutics in mice. BMC cancer. 2008;8:197. 106. Tenhunen R, Marver HS, Schmid R. Microsomal heme oxygenase. Characterization of the enzyme. The Journal of biological chemistry. 1969;244:6388-94. 100  107. Lin Q, Weis S, Yang G, Weng YH, Helston R, Rish K, et al. Heme oxygenase-1 protein localizes to the nucleus and activates transcription factors important in oxidative stress. The Journal of biological chemistry. 2007;282:20621-33. 108. Namba F, Go H, Murphy JA, La P, Yang G, Sengupta S, et al. Expression level and subcellular localization of heme oxygenase-1 modulates its cytoprotective properties in response to lung injury: a mouse model. PloS one. 2014;9:e90936. 109. McCoubrey WK, Jr., Huang TJ, Maines MD. Isolation and characterization of a cDNA from the rat brain that encodes hemoprotein heme oxygenase-3. European journal of biochemistry / FEBS. 1997;247:725-32. 110. Muller RM, Taguchi H, Shibahara S. Nucleotide sequence and organization of the rat heme oxygenase gene. The Journal of biological chemistry. 1987;262:6795-802. 111. Schuller DJ, Wilks A, Ortiz de Montellano PR, Poulos TL. Crystal structure of human heme oxygenase-1. Nature structural biology. 1999;6:860-7. 112. Yoshida T, Sato M. Posttranslational and direct integration of heme oxygenase into microsomes. Biochemical and biophysical research communications. 1989;163:1086-92. 113. Alam J, Cai J, Smith A. Isolation and characterization of the mouse heme oxygenase-1 gene. Distal 5' sequences are required for induction by heme or heavy metals. The Journal of biological chemistry. 1994;269:1001-9. 114. Alam J, Cook JL. How many transcription factors does it take to turn on the heme oxygenase-1 gene? American journal of respiratory cell and molecular biology. 2007;36:166-74. 115. Alam J, Stewart D, Touchard C, Boinapally S, Choi AM, Cook JL. Nrf2, a Cap'n'Collar transcription factor, regulates induction of the heme oxygenase-1 gene. The Journal of biological chemistry. 1999;274:26071-8. 116. Chepelev NL, Willmore WG. Regulation of iron pathways in response to hypoxia. Free radical biology & medicine. 2011;50:645-66. 117. Reichard JF, Motz GT, Puga A. Heme oxygenase-1 induction by NRF2 requires inactivation of the transcriptional repressor BACH1. Nucleic acids research. 2007;35:7074-86. 118. Amoutzias GD, Veron AS, Weiner J, 3rd, Robinson-Rechavi M, Bornberg-Bauer E, Oliver SG, et al. One billion years of bZIP transcription factor evolution: conservation and change in dimerization and DNA-binding site specificity. Molecular biology and evolution. 2007;24:827-35. 119. Kramer M, Sponholz C, Slaba M, Wissuwa B, Claus RA, Menzel U, et al. Alternative 5' untranslated regions are involved in expression regulation of human heme oxygenase-1. PloS one. 2013;8:e77224. 120. Yoshida T, Biro P, Cohen T, Muller RM, Shibahara S. Human heme oxygenase cDNA and induction of its mRNA by hemin. European journal of biochemistry / FEBS. 1988;171:457-61. 121. Shan Y, Lambrecht RW, Ghaziani T, Donohue SE, Bonkovsky HL. Role of Bach-1 in regulation of heme oxygenase-1 in human liver cells: insights from studies with small interfering RNAS. The Journal of biological chemistry. 2004;279:51769-74. 122. Shan Y, Lambrecht RW, Donohue SE, Bonkovsky HL. Role of Bach1 and Nrf2 in up-regulation of the heme oxygenase-1 gene by cobalt protoporphyrin. FASEB journal : official publication of the Federation of American Societies for Experimental Biology. 2006;20:2651-3. 123. Shan Y, Pepe J, Lu TH, Elbirt KK, Lambrecht RW, Bonkovsky HL. Induction of the heme oxygenase-1 gene by metalloporphyrins. Archives of biochemistry and biophysics. 2000;380:219-27. 101  124. Lu TH, Shan Y, Pepe J, Lambrecht RW, Bonkovsky HL. Upstream regulatory elements in chick heme oxygenase-1 promoter: a study in primary cultures of chick embryo liver cells. Molecular and cellular biochemistry. 2000;209:17-27. 125. Li L, Li CM, Wu J, Huang S, Wang GL. Heat shock protein 32/heme oxygenase-1 protects mouse Sertoli cells from hyperthermia-induced apoptosis by CO activation of sGC signalling pathways. Cell biology international. 2014;38:64-71. 126. Snyder SH, Baranano DE. Heme oxygenase: a font of multiple messengers. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology. 2001;25:294-8. 127. Lee PJ, Jiang BH, Chin BY, Iyer NV, Alam J, Semenza GL, et al. Hypoxia-inducible factor-1 mediates transcriptional activation of the heme oxygenase-1 gene in response to hypoxia. The Journal of biological chemistry. 1997;272:5375-81. 128. Kitamuro T, Takahashi K, Ogawa K, Udono-Fujimori R, Takeda K, Furuyama K, et al. Bach1 functions as a hypoxia-inducible repressor for the heme oxygenase-1 gene in human cells. The Journal of biological chemistry. 2003;278:9125-33. 129. Morita T, Perrella MA, Lee ME, Kourembanas S. Smooth muscle cell-derived carbon monoxide is a regulator of vascular cGMP. Proceedings of the National Academy of Sciences of the United States of America. 1995;92:1475-9. 130. Beckman JD, Chen C, Nguyen J, Thayanithy V, Subramanian S, Steer CJ, et al. Regulation of heme oxygenase-1 protein expression by miR-377 in combination with miR-217. The Journal of biological chemistry. 2011;286:3194-202. 131. Lowery AJ, Miller N, Devaney A, McNeill RE, Davoren PA, Lemetre C, et al. MicroRNA signatures predict oestrogen receptor, progesterone receptor and HER2/neu receptor status in breast cancer. Breast cancer research : BCR. 2009;11:R27. 132. Leautaud V, Demple B. Regulation of heme oxygenase-1 mRNA deadenylation and turnover in NIH3T3 cells by nitrosative or alkylation stress. BMC molecular biology. 2007;8:116. 133. Lin PH, Lan WM, Chau LY. TRC8 suppresses tumorigenesis through targeting heme oxygenase-1 for ubiquitination and degradation. Oncogene. 2013;32:2325-34. 134. Lin PH, Chiang MT, Chau LY. Ubiquitin-proteasome system mediates heme oxygenase-1 degradation through endoplasmic reticulum-associated degradation pathway. Biochimica et biophysica acta. 2008;1783:1826-34. 135. Poss KD, Tonegawa S. Heme oxygenase 1 is required for mammalian iron reutilization. Proceedings of the National Academy of Sciences of the United States of America. 1997;94:10919-24. 136. Yachie A, Niida Y, Wada T, Igarashi N, Kaneda H, Toma T, et al. Oxidative stress causes enhanced endothelial cell injury in human heme oxygenase-1 deficiency. The Journal of clinical investigation. 1999;103:129-35. 137. Kovtunovych G, Eckhaus MA, Ghosh MC, Ollivierre-Wilson H, Rouault TA. Dysfunction of the heme recycling system in heme oxygenase 1-deficient mice: effects on macrophage viability and tissue iron distribution. Blood. 2010;116:6054-62. 138. Nath KA. Heme oxygenase-1: a provenance for cytoprotective pathways in the kidney and other tissues. Kidney international. 2006;70:432-43. 139. Ferris CD, Jaffrey SR, Sawa A, Takahashi M, Brady SD, Barrow RK, et al. Haem oxygenase-1 prevents cell death by regulating cellular iron. Nature cell biology. 1999;1:152-7. 140. Lakkisto P, Kyto V, Forsten H, Siren JM, Segersvard H, Voipio-Pulkki LM, et al. Heme oxygenase-1 and carbon monoxide promote neovascularization after myocardial infarction by modulating the expression of HIF-1alpha, SDF-1alpha and VEGF-B. European journal of pharmacology. 2010;635:156-64. 102  141. Mougiakakos D, Jitschin R, Johansson CC, Okita R, Kiessling R, Le Blanc K. The impact of inflammatory licensing on heme oxygenase-1-mediated induction of regulatory T cells by human mesenchymal stem cells. Blood. 2011;117:4826-35. 142. Poss KD, Tonegawa S. Reduced stress defense in heme oxygenase 1-deficient cells. Proceedings of the National Academy of Sciences of the United States of America. 1997;94:10925-30. 143. Otterbein LE, Hedblom A, Harris C, Csizmadia E, Gallo D, Wegiel B. Heme oxygenase-1 and carbon monoxide modulate DNA repair through ataxia-telangiectasia mutated (ATM) protein. Proceedings of the National Academy of Sciences of the United States of America. 2011;108:14491-6. 144. Burt TD, Seu L, Mold JE, Kappas A, McCune JM. Naive human T cells are activated and proliferate in response to the heme oxygenase-1 inhibitor tin mesoporphyrin. J Immunol. 2010;185:5279-88. 145. Okita Y, Kamoshida A, Suzuki H, Itoh K, Motohashi H, Igarashi K, et al. Transforming growth factor-beta induces transcription factors MafK and Bach1 to suppress expression of the heme oxygenase-1 gene. The Journal of biological chemistry. 2013;288:20658-67. 146. Lee TS, Chau LY. Heme oxygenase-1 mediates the anti-inflammatory effect of interleukin-10 in mice. Nature medicine. 2002;8:240-6. 147. Rousset F, Nguyen MV, Grange L, Morel F, Lardy B. Heme oxygenase-1 regulates matrix metalloproteinase MMP-1 secretion and chondrocyte cell death via Nox4 NADPH oxidase activity in chondrocytes. PloS one. 2013;8:e66478. 148. Bussolati B, Ahmed A, Pemberton H, Landis RC, Di Carlo F, Haskard DO, et al. Bifunctional role for VEGF-induced heme oxygenase-1 in vivo: induction of angiogenesis and inhibition of leukocytic infiltration. Blood. 2004;103:761-6. 149. Wagener FA, van Beurden HE, von den Hoff JW, Adema GJ, Figdor CG. The heme-heme oxygenase system: a molecular switch in wound healing. Blood. 2003;102:521-8. 150. Nishie A, Ono M, Shono T, Fukushi J, Otsubo M, Onoue H, et al. Macrophage infiltration and heme oxygenase-1 expression correlate with angiogenesis in human gliomas. Clinical cancer research : an official journal of the American Association for Cancer Research. 1999;5:1107-13. 151. Mayerhofer M, Florian S, Krauth MT, Aichberger KJ, Bilban M, Marculescu R, et al. Identification of heme oxygenase-1 as a novel BCR/ABL-dependent survival factor in chronic myeloid leukemia. Cancer research. 2004;64:3148-54. 152. Cermak J, Balla J, Jacob HS, Balla G, Enright H, Nath K, et al. Tumor cell heme uptake induces ferritin synthesis resulting in altered oxidant sensitivity: possible role in chemotherapy efficacy. Cancer research. 1993;53:5308-13. 153. Kim AY, and Bennewith, K.L. The Role of Heme Oxygenase-1 in Breast Cancer Metastasis. Unpublished Undergraduate Thesis. 2012:1-46. 154. Kosovsky MJ. A Synthetic Hydrogel for 3D Cell Culture: BD™ PuraMatrix™ Peptide Hydrogel. 2009. 155. Vincent SH. Oxidative effects of heme and porphyrins on proteins and lipids. Seminars in hematology. 1989;26:105-13. 156. Nath KA, Vercellotti GM, Grande JP, Miyoshi H, Paya CV, Manivel JC, et al. Heme protein-induced chronic renal inflammation: suppressive effect of induced heme oxygenase-1. Kidney international. 2001;59:106-17. 157. Wagener FA, Feldman E, de Witte T, Abraham NG. Heme induces the expression of adhesion molecules ICAM-1, VCAM-1, and E selectin in vascular endothelial cells. Proc Soc Exp Biol Med. 1997;216:456-63. 103  158. Hori R, Kashiba M, Toma T, Yachie A, Goda N, Makino N, et al. Gene transfection of H25A mutant heme oxygenase-1 protects cells against hydroperoxide-induced cytotoxicity. The Journal of biological chemistry. 2002;277:10712-8. 159. Terry CM, Clikeman JA, Hoidal JR, Callahan KS. Effect of tumor necrosis factor-alpha and interleukin-1 alpha on heme oxygenase-1 expression in human endothelial cells. The American journal of physiology. 1998;274:H883-91. 160. Faraji F, Hu Y, Wu G, Goldberger NE, Walker RC, Zhang J, et al. An integrated systems genetics screen reveals the transcriptional structure of inherited predisposition to metastatic disease. Genome research. 2014;24:227-40. 161. Yun J, Frankenberger CA, Kuo WL, Boelens MC, Eves EM, Cheng N, et al. Signalling pathway for RKIP and Let-7 regulates and predicts metastatic breast cancer. The EMBO journal. 2011;30:4500-14. 162. Zhang X, Shan P, Alam J, Fu XY, Lee PJ. Carbon monoxide differentially modulates STAT1 and STAT3 and inhibits apoptosis via a phosphatidylinositol 3-kinase/Akt and p38 kinase-dependent STAT3 pathway during anoxia-reoxygenation injury. The Journal of biological chemistry. 2005;280:8714-21. 163. Vreman HJ, Stevenson DK. Heme oxygenase activity as measured by carbon monoxide production. Analytical biochemistry. 1988;168:31-8. 164. Turcanu V, Dhouib M, Poindron P. Determination of heme oxygenase activity in murine macrophages for studying oxidative stress inhibitors. Analytical biochemistry. 1998;263:251-3. 165. Drummond GS. Control of heme metabolism by synthetic metalloporphyrins. Annals of the New York Academy of Sciences. 1987;514:87-95. 166. Iorns E, Drews-Elger K, Ward TM, Dean S, Clarke J, Berry D, et al. A new mouse model for the study of human breast cancer metastasis. PloS one. 2012;7:e47995.     104  Appendices  Appendix A: Migration and Invasion of murine mammary carcinoma cells.  Control and hemin treated 67NR, 4TO7, and 4T1 cells were grown in 1% FBS containing medium for 24 h before being placed in the upper chamber of the Boyden chamber, 1.4x105 cells per chamber. The lower chamber was filled with RPMI with 10 % FBS to stimulate the cells to migrate and invade the membrane separating the upper and lower chamber. Membranes in invasion assays were coated with Matrigel. After 24 h, cells that migrated to the lower chamber were fixed and stained with DAPI. A, migration. B, invasion. Images were obtained with a fluorescent microscope. The number of migrated cells was counted by ImageJ. Data given as mean ± SEM of independent experiments, N=6, one-tailed t-test, * represents p≤0.05. Data from (153).   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0167212/manifest

Comment

Related Items