Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effect of hypoxia-inducible secreted protein, tenascin c, on 4T1 tumour cells in vitro and in vivo Harbourne, Bryant Thomas 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_spring_harbourne_bryant.pdf [ 1.51MB ]
Metadata
JSON: 24-1.0165859.json
JSON-LD: 24-1.0165859-ld.json
RDF/XML (Pretty): 24-1.0165859-rdf.xml
RDF/JSON: 24-1.0165859-rdf.json
Turtle: 24-1.0165859-turtle.txt
N-Triples: 24-1.0165859-rdf-ntriples.txt
Original Record: 24-1.0165859-source.json
Full Text
24-1.0165859-fulltext.txt
Citation
24-1.0165859.ris

Full Text

 EFFECT OF HYPOXIA-INDUCIBLE SECRETED PROTEIN,  TENASCIN C, ON 4T1 TUMOUR CELLS IN VITRO AND IN VIVO  by Bryant Thomas Harbourne B.Sc. (Spec.), University of Alberta, 2007  A 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) February 2014   ? Bryant Thomas Harbourne, 2014  ii  Abstract  Introduction: Metastatic cancer is responsible for 90% of cancer related deaths. Past research focused mainly on the primary tumour, leaving the process of metastasis poorly understood.  Poorly oxygenated (hypoxic) tumour cells express hypoxia inducible proteins and have a more aggressive and invasive phenotype correlated with poorer prognosis. Hypoxic tumour cells are responsible for increased angiogenesis, invasion, matrix deposition and remodelling along with many other functions. We hypothesize hypoxic tumour secreted proteins are responsible for promoting metastasis. Our aim is to identify these hypoxia inducible secreted proteins and determine mechanisms promoting metastasis.  Methods: Mammary carcinoma cells (4T1 ? metastatic and 67NR ? non-metastatic) were placed in 1% O2 (hypoxic) or 21% O2 (normoxic) for 24 hours. Stable Isotope Labelling of Amino acids in Cell culture (SILAC) and mass spectrometry were used to perform a quantitative proteomic screen of conditioned medium. Proteins in 4T1 conditioned media, up-regulated in hypoxia and absent from the 67NR results represented candidate secreted proteins in metastasis.  Tenascin C (TNC), a candidate protein identified from the proteomic screen was stably knocked down and over-expressed. In vitro, the Boyden chamber and wound healing assay were used to study invasion and migration. In vivo, metastasis was assessed using flow cytometry-based quantification of metastasized tumour cells in the murine lungs. Results: (TNC) was identified as a secreted protein with a role promoting metastasis in vivo through enhanced migratory ability. In vitro, knockdown of TNC in 4T1 enhanced migratory ability whereas over-expression decreased migratory ability. These results were contradictory to the expected results based on the hypothesized in vivo role. However, in vivo knockdown of TNC in 4T1 tumour cells resulted in a significant decrease in lung metastases. These results are consistent with the expected role of TNC in vivo.   Conclusions: Despite the contradictory results in vitro, TNC had a positive metastatic role potentially through a migratory mechanism. TNC represents a potential new therapeutic drug target.  iii  Given the 4T1 cell line results, these data support further examination of the migratory role of TNC and how it promotes metastasis. In addition, TNC expression in other tumour cell lines including human breast cancer should be examined.   iv  Preface  All experiments were performed by Bryant T. Harbourne except where noted. Experimental design, data collection and analysis, and thesis composition and edits were performed by me. Murine tissue harvests were assisted by Momir Bosiljcic, Nancy LePard, Ada Kim and Judit Banath. Dr. Kevin Bennewith assisted with experimental design and data analysis.  Mass spectrometry of SILAC labelled secreted proteins was performed by the Dr. Leonard Foster lab at UBC. Dr. Leonard Foster assisted in the design of the SILAC labelled secreted protein proteomic screen.  Mice were housed in specific pathogen free conditions within the Animal Resource Centre (ARC) at the BC Cancer Research Centre. 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.   v  Table of Contents  Abstract ............................................................................................................................ii Preface ............................................................................................................................ iv Table of Contents ............................................................................................................. v List of Tables.................................................................................................................. viii List of Figures .................................................................................................................. ix List of Symbols and Abbreviations ..................................................................................... xi Glossary ......................................................................................................................... xii Acknowledgments .......................................................................................................... xiii Introduction ...................................................................................................................... 1 Cancer ......................................................................................................................... 1 Hypoxia ........................................................................................................................ 1 HIF Transcription Factors ........................................................................................... 3 Breast Cancer ............................................................................................................... 4 Subtypes ................................................................................................................... 4 Tumour Model ........................................................................................................... 5 Metastasis .................................................................................................................... 6 Invasion and Migration ............................................................................................... 6 Seed and Soil Hypothesis .......................................................................................... 8 Tissue Specificity ....................................................................................................... 8 Pre-metastatic Niche Hypothesis ................................................................................ 9 CD11b and Gr1 Cell Markers ................................................................................... 10 Secreted Proteins ....................................................................................................... 11 Secretome Analysis ................................................................................................. 11 Stable Isotope Labelling of Amino Acids in Cell Culture ............................................. 11 Indirect Effects of Tumour Secreted Proteins ............................................................ 13 vi  Direct Effects of Tumour Secreted Proteins .............................................................. 13 Tenascin C ................................................................................................................. 15 History .................................................................................................................... 16 Structure ................................................................................................................. 17 Expression and Role ................................................................................................ 18 TNC in Cancer and Metastasis ................................................................................. 19 Hypothesis and Aims .................................................................................................. 22 Materials and Methods ................................................................................................... 24 Tumour Cell Lines ....................................................................................................... 24 In Vivo Mouse Models ................................................................................................. 24 Tissue Processing....................................................................................................... 24 Flow Cytometry ........................................................................................................... 25 Clonogenic Assay ....................................................................................................... 25 Immuno-Fluorescence................................................................................................. 25 Statistical Analysis ...................................................................................................... 26 Protein Analysis .......................................................................................................... 26 Protein Isolation and Quantification .......................................................................... 26 Western Blot............................................................................................................ 27 Migration and Invasion ................................................................................................ 28 Results .......................................................................................................................... 32 Conditioned Media ...................................................................................................... 32 Conditioned Media in BALB/c Mice ........................................................................... 34 Conditioned Media in NOD/scid Mice........................................................................ 39 S.I.L.A.C. Proteomic Screen ........................................................................................ 42 Invasion and Migration ................................................................................................ 50 TNC Knockdown...................................................................................................... 50 TNC Over-Expression .............................................................................................. 54 vii  Recombinant TNC ................................................................................................... 56 TNC In vivo Experiments ............................................................................................. 57 TNC Knockdown...................................................................................................... 58 TNC Over-expression .............................................................................................. 64 Discussion ..................................................................................................................... 66 Conditioned Media ...................................................................................................... 66 S.I.L.A.C. Proteomics .................................................................................................. 68 Tenascin C Expression ............................................................................................... 69 Tenascin C Knockdown and Over-expression .............................................................. 70 Migration and Invasion Assays .................................................................................... 71 Tenascin C In Vivo ...................................................................................................... 75 Conclusions ................................................................................................................... 79 Future Directions ........................................................................................................ 80 Bibliography ................................................................................................................... 82     viii  List of Tables  Table 1: Top 10 selected secreted proteins from the 4T1 S.I.L.A.C proteomic data based on level of hypoxic induction and known cancer or metastasis role. ...................... 44    ix  List of Figures  Figure 1 Oxygen diffusion from blood vessel and resulting intra-tumour cellular oxygen status. .............................................................................................................. 2 Figure 2 HIF1 regulation and action under normoxic and hypoxic oxygen conditions ........... 3 Figure 3 Heavy isotope forms of lysine and arginine used in SILAC. ................................. 12 Figure 4 Representative structure of a Tenascin C monomer. .......................................... 17 Figure 5 Effects of 4T1 tumours on female BALB/c mice. ................................................. 33 Figure 6 Representative flow plot for CD11b+/Gr1+ cells. .................................................. 34 Figure 7 Conditioned Media Trial 3 with female na?ve BALB/c mice. ................................. 36 Figure 8 Conditioned Media Trial 4 with female na?ve BALB/c mice. ................................. 38 Figure 9 Conditioned Media Trial 2 with female na?ve NOD/scid mice. .............................. 40 Figure 10 Conditioned Media Trial 3 with female na?ve NOD/scid mice. ............................ 41 Figure 11 Western blot validation of SILAC proteomic results. .......................................... 45 Figure 12 4T1 wild type TNC western blot under normoxic and hypoxic conditions. ........... 46 Figure 13 TNC western blot with Brefeldin A. ................................................................... 47 Figure 14 TNC western blot of knockdown in 4T1 cell lines with shGFP control. ................ 48 Figure 15 TNC western blot of 4T1 TNC over-expression cell line. ................................... 49 Figure 16 Western blot of recombinant TNC. ................................................................... 49 Figure 17 Boyden chamber migration assay under normoxic conditions with hypoxic conditioned media pre-treatment of membrane. ................................................ 50 Figure 18 Boyden chamber invasion assay performed under hypoxic conditions. .............. 51 Figure 19 Boyden chamber invasion assay under normoxic conditions with hypoxic conditioned media pre-treatment of Matrigel membrane. ................................... 52 Figure 20 Wound healing assay of 4T1 shGFP, KD1 and KD5 cells in hypoxia. ................ 53 Figure 21 Representative images of 4T1 TNC KD wound healing assay (Fig. 20) at 8, 16 and 24 hours. .................................................................................................. 54 x  Figure 22 Boyden chamber migration assay with 4T1 TNC over-expression cell line under normoxic conditions......................................................................................... 55 Figure 23 Wound healing of 4T1 TNC over expressing or empty vector cells in normoxia.  . 55 Figure 24 Representative images of 4T1 TNC over-expression and empty vector wound healing assay (Fig.23) at 8, 12 and 16 hours. ................................................... 56 Figure 25 Wound healing assay of 4T1 wild type cells with recombinant Tenascin C in normoxia......................................................................................................... 57 Figure 26 Harvested tissue characteristic of In Vivo 4T1 TNC knockdown experiment. ...... 59 Figure 27 Post harvest TNC knockdown tumour phenotype verification. ........................... 60 Figure 28 Flow cytometry analysis of CD11b+/Gr1+ cells from the in vivo TNC knockdown experiment ...................................................................................................... 61 Figure 29 Flow cytometry analysis of cytokeratin positive cells in lungs from 4T1 TNC knockdown experiment. ................................................................................. 62 Figure 30 TNC expression in lungs from 4T1 TNC knockdown experiment. ...................... 63 Figure 31 Harvested tissue characteristic of In Vivo 4T1 TNC over expression experiment. 64 Figure 32 Clonogenic assay of 4T1 TNC over-expression experiment. ............................. 65   xi  List of Symbols and Abbreviations  B16 B16 melanoma BMDC bone marrow derived cells BM basement membrane CM   conditioned media ECM   extracellular matrix EGF  epidermal growth factor EV  empty vector Fn   fibronectin IP  intra-peritoneal IF  immuno-fluorescent kDa  kilo Dalton KD  knockdown LLC   Lewis Lung Carcinoma Lox lysyl oxidase MDSC myeloid derived suppressor cell OE over-expression Pimo pimonidazole PyMT Polyoma Middle T Antigen SILAC stable isotope labelling of amino acids in cell culture TNC   tenascin c WT  wild type   xii  Glossary  4T1 EV 4T1 cell line transduced with the over-expression vector without the tenascin c gene 4T1 OE 4T1 cell line transduced with the over-expression vector with the Tenascin C gene Conditioned media media collected from tissue culture containing secreted factors produced by cells in vitro  Hypoxic conditioned media conditioned media from tissue culture which had been exposed to 1% O2 for 24 hours  Secretome a catalogue of all secreted proteins from a specific cell line shGFP shRNA specific to green fluorescent protein used as a transduction control TNC KD1 shRNA construct #1 specific to Tenascin C TNC KD5 shRNA construct #5 specific to Tenascin C    xiii  Acknowledgments  I would like to acknowledge the Bennewith lab members who assisted me with this work. In particular I would like to thank Momir Bosiljcic and Nancy LePard. I would like to thank Dr. Kevin Bennewith for his assistance and guidance during my research as well as Dr. Shoukat Dedhar for acting as my co-supervisor. I would like to acknowledge Dr. Leonard Foster and his lab, as well as Jenny Moon for the collaboration on the secreted protein proteomic screen, Dr. Donald Yapp and Dr. Andrew Minchinton for the use of their hypoxia chambers; Dr. Aly Karsan for reagents and microscope imaging, Dr. Andrew Weng and Sam Gusscott for their over-expression vector and Dr. Gerry Krystal for reagents provided. Lastly, I would like to thank my committee members for their thoughts and guidance.   1  Introduction Cancer In 2013, there were 187,600 Canadians diagnosed with cancer and cancer was the leading cause of death in Canada in 2009.[1] Tumours of the breast represent 25% of all new diagnoses, making it the most common and prevalent tumour afflicting women.[1]  However, primary breast tumours are not responsible for the majority of the deaths associated with breast cancer. Metastasis accounts for 90% of cancer associated deaths as a result of macro-metastases in secondary locations.[2]  There are two kinds of tumours: benign and malignant (cancer). Benign tumours form masses at primary growth locations but rarely spread to other locations in the body. As a result of this lack of spread, benign tumours tend to have a better prognosis. In contrast, malignant tumours invade the tissue surrounding the primary growth site and spread throughout the body. For many years research focused on primary malignancy and the development of new and better treatments including chemotherapy, radiation and surgery. Consequently, less focus was directed at how metastasis occurs and specifically what are the differences between benign and malignant tumours and how do these differences enable malignant tumours to spread.  Hypoxia As a result of the early characterization of primary tumours it was noted that the intra-tumoural micro-environment often had areas with very low oxygen tension (hypoxia).[3] Head and neck and cervical cancers showed that tumours with this low oxygen characteristic had a strong correlation with increased metastatic spread.[4, 5] Moreover, people with hypoxic primary tumours have been shown to have worse outcomes.[6]  Hypoxia results from an insufficient supply of dissolved oxygen to meet the needs of all cells. This creates regions of the primary tumour where cells are in an oxygen deficient micro-environment. Hypoxia occurs as a gradient of oxygen tensions or concentrations ranging from 4% O2 (30 mmHg) near well perfused blood vessels to hypoxic regions of 1% O2 (8-10 mmHg) to regions of necrosis where the oxygen levels are 0%.[7, 8] In contrast, physiological oxygen tension in tissues can range from 11% (83.6 mmHg) to 4% (30 2  mmHg).[9, 10] This gradient occurs between 40 and 170 ?m away from blood vessels depending on the metabolic rate of the cells in the oxygen diffusion zone. The fraction of the   Figure 1 Oxygen diffusion from blood vessel and resulting intra-tumour cellular oxygen status. tumour exposed to hypoxic conditions can range from 0 - 25% of the total tumour. [11][Unpublished data, Bennewith et al]  Increased transcription of hypoxia inducible proteins occurs at the 1% O2 (8-10 mmHg) hypoxia range.[7]  Hypoxia is caused by the concerted action of numerous factors. The two primary factors are poor tumoural vasculature and high metabolic rate of cancerous cells. Simplistically, cancer is a deregulation of cell proliferation resulting in uncontrolled growth. This rapid growth creates an enormous requirement for nutrients to sustain the growing tumour. In the rush to create and form the network of blood vessels needed to supply nutrients to the tumour, vessel quality is compromised for speed of creation. This creates blood vessels with abnormalities: excessive vessel diameter leading to collapse, insufficient vessel diameter leading to blockages, leaky or porous walls, or structurally corrupt vessels with dead ends.[12] These defects result in regions of tumour tissue which do not receive adequate nutrient, oxygen and blood supply. In response to the hypoxic stress state, genes which can counteract some of the stresses are activated. Hypoxia can increase expression of genes involved in oxygen transport, angiogenesis, glycolysis and glucose uptake, metabolism and 3  pH control, growth factors, cytokines as well as cell adhesion, extra-cellular matrix and secreted proteins.[13]  HIF Transcription Factors All cells have a built in response system for stressful conditions present in a microenvironment. Under normal physiological conditions these stress responses aren?t required and therefore cells have systems in place to repress the expression of them. In the case of hypoxia, the presence of oxygen prevents the increased transcription of hypoxia inducible genes.  The hypoxia inducible factor (HIF) transcription factor is responsible for the expression of many genes associated with the hypoxic stress response. HIF-1 is a heterodimer consisting of an alpha unit (HIF-1?) and a beta unit (HIF-1?).[13] HIF-1? is constitutively expressed whereas HIF-1? is tightly regulated and would be degraded when tissues are better oxygenated, above 15 mmHg. Prolyl hydroxylases in well oxygenated cells hydroxylate proline residues of HIF-1? and enable interaction with the von Hippel Lindau (vHL) protein.[13] The HIF-1?/vHL complex along with other proteins causes HIF-1? to become ubiquitinated and targeted for proteosomal degradation.[13]     Figure 2 HIF1 regulation and action under normoxic and hypoxic oxygen conditions  Under normoxic conditions HIF1? is hydroxylated by prolyl hydroxylases allowing vHL interaction and subsequent ubiquitination and degradation by the proteosome. Under hypoxic conditions, prolyl hydroxylases are unable to interact with HIF1? allowing the HIF1?? heterodimer to form and interact with HRE?s within the nucleus. This transcription factor binding enables transcription of a variety of hypoxia regulated proteins  such as LDH ? lactate dehydrogenase, GLUT1 ? glucose transporter 1, VEGF ? vascular endothelial growth factor, LOX ? lysyl oxidase, CA IX ? carbonic anhydrase 9 and TNC ? tenascin c. 4  However, during times of low oxygen tension, prolyl hydroxylases are unable to modify HIF-1? allowing HIF-1? to bind. Together HIF-1?/? creates a functional HIF1 heterodimer transcription factor which can translocate to the nucleus.[13] Genes with roles in the stress responses to low oxygen have hypoxia response elements (HRE?s) allowing the HIF transcription factor to bind to and increase gene expression of these genes. HIF-1 can transcriptionally induce the expression of glycolytic enzymes, glucose transporters, angiogenesis proteins, metastasis associated proteins, pH regulation proteins and matricellular proteins (Fig.2).[13]  In cancer, these newly expressed hypoxia induced genes result in a phenotypic change in cells exposed to low oxygen conditions. The expression of these new genes promotes a more aggressive and invasive primary tumour that can produce secreted proteins which assist in metastatic spread. Intra-tumoural hypoxia is an important factor involved in making a tumour more metastatic.  Breast Cancer  Subtypes Breast cancer is a remarkably heterogeneous disease. Tumour types can be broken down by classification of molecular patterning. There are three main subtypes of breast cancer: luminal, HER2+, and basal.[14, 15] These subtypes are separated primarily by receptor expression on the tumour cells: estrogen receptor, progesterone receptor, and HER2/neu status.[16] HER2/neu or ErbB2 is a member of the epidermal growth factor receptor family.[17]  The luminal subtypes, A and B, represent tumours with positive expression of both the estrogen and progesterone receptors but differ in HER2 expression.[15] The luminal B type has an increased level of HER2 expression and a greater proliferative capacity.[15] Overall, the luminal subtypes have the best clinical prognosis with the Luminal A types having a lower relapse rate than the B sub-group.[15] The HER2 subtype represents tumours in which the HER2/neu gene is over expressed but are negative for the estrogen and progesterone receptors.[15] The basal subtype is characterized by the absence of the estrogen, progesterone and HER2/neu receptors along with the presence of other criteria such as high molecular weight cytokeratins.[18] In addition to the basal subgroup, another group of breast cancers, triple 5  negative, also lack expression of the three main receptors which leads to the imperfect grouping of the basal subtype and the triple negative subtype.[15, 18] The basal or triple negative subgroups together represent approximately 15% of breast cancers as well as the subgroup with the worst prognosis. [5, 7][19]  Tumour Model The primary model of cancer utilized in this research was the 4T1 murine mammary carcinoma. It is negative for the estrogen, progesterone and HER2/neu receptors making it triple negative and in addition positive for cytokeratin 5/6 further classifying it as a basal subtype.[20, 21] The 4T1 cell line is hyper diploid and expresses cytokeratins at levels 2 or 3 fold higher than most tissues. This difference can be utilized to identify tumour cells in disaggregated tissues such as lungs. The 4T1 cell line is a result of a spontaneous breast carcinoma in a BALB/cfC3H mouse strain that was removed and processed to produce a number of tumour sub-clones.[22] There are 5 major tumour cell lines that arose from that original tumour, including the 4T1, 66c14, 4T07, 168FARN, and 67NR.[23] The 4T1 tumour cell line is the most metastatic of the group and spontaneously forms macro-metastases in the lung, liver, bone and brain. The 66c14 spontaneously forms macro-metastases in the lung while the other three cell lines do not commonly form metastases but are highly tumourigenic.[23] The 4T1 cell line is a suitable model for late stage human breast cancer given the locations of the metastases. The 4T1 tumour cell line forms primary tumours with significant regional hypoxia.[24] This hypoxia is not due to an absence of blood vessels, but rather the lack of vessels with sufficient quality and functional ability to provide for the demands of the tumour. As a consequence, 4T1 primary tumours express hypoxia inducible genes and are able to mobilize significant numbers of bone marrow derived cells (BMDC?s) and are metastatic. In contrast, the 67NR tumour cell line forms primary tumours with very minimal hypoxia. Presumptively, the lack of hypoxia is due to a superior quality and abundance of blood vessel as compared with the 4T1 tumour. The 67NR primary tumours have no hypoxia, do not show the same myeloid derived suppressor cell (MDSC) mobilization or accumulation and are non-metastatic.[24] This contrast in syngeneic tumour cell lines allowed an intriguing comparison of what is responsible for the metastatic progression seen in the 4T1.Tumoral hypoxia status has been 6  positively correlated with metastatic ability.[24] In addition, accumulation of MDSC?s correlate with increased hypoxia and higher metastatic burden. Metastasis Metastasis is the most harmful aspect of cancer. Metastasis refers to the usually distant spread of tumour cells after escape from the confined primary location. The process of metastasis is a step-wise process in which all steps must be satisfied in order for a tumour cell to successfully spread and form secondary tumours. After primary tumour formation, a cell must be able to: locally invade surrounding tissue, intravasate into the circulatory system (or lymphatic system), survive transit in the circulatory system, arrest at a suitable secondary location, extravasate into the parenchyma, form micro-metastases able to establish a neo-vasculature system and successfully drive outgrowth of a secondary tumour.[25] The process is extremely inefficient and less than 0.02% of circulating tumour cells are capable of forming secondary tumours.[26] Despite this inefficiency, a patient can concurrently present with significant metastatic disease at the time of diagnosis of the primary tumour or conversely, metastatic disease can have a period of dormancy for years after remission of the primary tumour and at an unknown stimulus present with secondary tumour outgrowth.  Of the entire process, the two most important and limiting steps are local invasion into the surrounding stroma of the primary tumour and the successful establishment of a tumour at a secondary location.[26, 27] These two steps are important determinants of whether the tumour cells are capable of spreading but do not influence where metastases would occur. Secondary locations are dependent upon secreted factors from the tumour and host cells, such as MDSC?s interacting with the stroma. Often, secondary tumours occur in multiple areas such as breast cancer commonly spreading to the lungs but also spreading to the brain, bones, kidneys or liver.[23] Invasion and Migration The process of invasion and migration is not unique to cancer cells. Many cells such as neuronal cells during embryogenesis, immune cells during inflammation or fibroblasts during wound healing exhibit these physiological abilities.[28-30] However, immune cells also utilize this ability during pathologic conditions such as atherosclerosis.[29] While immune cells maintain the ability to invade and migrate, the majority of cells after terminal differentiation such as epithelial cells, do not retain that ability. However, carcinomas which derive from epithelial cells regain this invasive and migratory phenotype indicating a switch 7  from a non-migratory epithelial phenotype has taken place. This switch represents a change from an epithelial to a mesenchymal phenotype and is a key regulatory step which cells must complete during the transformative process prior to becoming an invasive cancer.[28]  Epithelial cells normally grow in an organized sheet interconnected with other epithelial cells through cell-cell junctions and adhesions.[31] These cells have a polarized morphology and express specific proteins common to epithelial cells; E-cadherin being the prototypical epithelial protein seen.[31] However, as a cell begins the change from an epithelial to a mesenchymal cell, changes in morphology, adhesion characteristics and migratory ability will occur.[31] Among those changes is a shift from the uniform epithelial shape to a spindle mesenchymal shape, E-cadherin is down regulated while N-cadherin is increased and vimentin is expressed along with many other proteins and transcription factors.[31] Cytoskeletal reorganization also occurs and acts as an important factor in increasing migration speed and the production of matrix digestion proteins used for ECM degradation and remodelling.[32] The net result of these changes is reduced adhesion between cells, increased migratory and invasive ability and resistance to anoikis.[31, 32] While these changes are required for the cancer cells to escape the primary tumour, once they reach an appropriate metastatic site the establishment of metastases requires an epithelial- like phenotype. There is a process called mesenchymal to epithelial transition (MET) where the cells that went through EMT reverse the process.[31] While complete shifts from epithelial to mesenchymal cell types may be unlikely, the existence of a metastable cell type which shares characteristics of both is likely.[31]  Invasion and migration are key processes in metastasis and are intrinsically coupled together. Invasion has three major components: cell adhesion to the ECM, commonly the basement membrane; proteolysis of the ECM/BM; and migration through the degraded matrix.[30] Migration enables cell movement allowing contact between the cell and the ECM/BM to occur as well as movement of the cell through the degraded BM. Migration also refers to movement of cells through the stroma. However, when a cell moves through a three dimensional substrate some degree of proteolysis is involved as the stroma must be degraded to make room for the cell to progress.[30] As a cell begins to migrate, actin rearrangements form pseudopods as the leading edge of the cell begins to protrude.[30] The extending pseudopods make contact with the ECM which binds adhesion molecules such as integrins on the cell membrane.[30] This ECM/integrin binding causes clustering of other cell membrane/matrix interactions and forms small focal complexes that can form 8  stable focal adhesions.[30] Integrins are composed of an alpha and a beta chain which provide specificity to the kinds of ECM proteins that they can bind; the ?5?3 integrin can bind fibronectin (Fn) among others.[33] Integrin binding and stable focal adhesion formation result in membrane matrix metalloproteinase (MMP) and soluble MMP recruitment to these sites.[30] As the ECM degrades around the focal adhesion sites, the cell contracts and the strong binding to the ECM through the integrins results in the cell body being pulled towards the sites of focal adhesion attachment.[30] At the same time, the focal adhesion interactions with the ECM at the trailing edge of the cell begin to be broken down and releases the cell to be pulled forward by the contracting cell.[30]  Invasion and migration as a result of the EMT process provide key features to cancers, the ability to invade surrounding local tissues and spread distantly throughout the body. Seed and Soil Hypothesis In 1889, an English surgeon, Stephen Paget wrote an article in The Lancet titled Distribution of Secondary Growths in Cancer of the Breast in which he argues that if metastasis was a matter of chance then the spread of the tumour should be equal in all tissues and organs proportional to the amount of blood flow through that tissue.[34] Instead based on his observations he drew the analogy of plants going to seed where outside influences such as wind and weather will direct the seeds randomly but will only germinate if they land on congenial soil.[34] Likewise he suggested that metastatic spread is due to a ?seminal influence? of the primary tumour on the body predisposing sites for secondary tumours.[34] It is now well known that there are metastatic profiles specifically associated with metastatic spread of most malignant cancers.  Tissue Specificity The study of cancer and metastasis has progressed significantly and it is now appreciated that for primary tumours, there are common organs or tissues in which metastasis occurs. Inevitably there is some overlap between different tumour types but interestingly some organs can be common sites in one type of cancer and be very rare in another. This disparity would indicate the fact that the primary tumour is playing some role in ?choosing? the metastatic sites or producing factors causing that organ or tissue to become permissive to colonization and growth by tumour cells.  Certainly, tissue specificity plays a role in dictating the locations of spreading tumour cells but blood flow is also an important factor. Tumour cells in general tend to be larger than 9  most normally circulating blood cells. Red blood cells and platelets can easily pass through a capillary bed when encountered. White blood cells are larger than the diameter of capillaries but have the ability to stretch and contort their cell shape to pass through the small diameter of the vessel. White blood cells can also enter and exit the blood vasculature using the process of intra- and extravasation. Cancerous cells can also intra- and extravasate in a process similar to white blood cells but first must digest the extra-cellular matrix and basement membranes prior to successful invasion. However, due to this larger size of cancerous cells, capillaries can trap those cells too large to pass through. Organs and tissues which are highly vascular or supplied by major vessels, such as the lungs and liver, should be exposed to significant numbers of metastasizing tumour cells, resulting in the majority of secondary tumours. While these organs are common sites of metastases, an organ such as the spleen, highly vascularised and designed to filter the blood is a rare site of metastasis. Observations like this make the argument that both tumour specificity and blood flow play a role in dictating where secondary tumours will form. Pre-metastatic Niche Hypothesis The pre-metastatic niche hypothesis grew out of the seed and soil concept put forth by Paget. The pre-metastatic niche represents an area in a distant tissue or organ that has been prepared by the primary tumour to allow colonization by circulating primary tumour cells. The tumour creates this niche through direct action of tumour secreted proteins at metastatic sites and indirect actions such as mobilization of host BMDC?s. Together these actions create sites where tumour secreted proteins have remodelled the ECM allowing the accumulation of host BMDC?s, as well the deposition of addition niche-associated secreted proteins.[35] The architectural change in ECM and the presence of immunosuppressive cells creates an area permissive to the invasion, survival and growth of circulating tumour cells. The formation of pre-metastatic niches can be studied by observing a number of factors: an accumulation of tumour specific secreted proteins, an accumulation of host derived cells or BMDC?s, or the co-localization of tumour secreted proteins and a pre-metastatic niche associated cell type, such as the previously mentioned BMDC?s, in a metastatic target organ. The first example of an accumulating cell population was vascular endothelial growth factor receptor 1 positive (VEGFR1+) bone marrow derived hematopoietic progenitor cells.[36] These cells arrived in the lungs of Lewis Lung Carcinoma (LLC) tumour bearing mice prior to the arrival of tumour cells.[36] Since then, CD11b+ Gr1+ granulocytes have 10  been shown to accumulate in the pre-metastatic niche.[37] These niche sites include the lungs, liver and kidneys of 4T1 tumour bearing mice which are all metastatic target organs.[37]  Published results have also shown the use of conditioned media in the study of tumour secreted proteins and their effect on pre-metastatic niche development. Medium when harvested from in vitro tissue culture which has had tumour cells growing for a period of time is said to have been conditioned. This conditioned media contain the secreted factors the tumour cells have produced since they were plated in the dish and presumably ones similar to those produced in vivo. It has been shown that injection of conditioned media from MDA-MD-231 cells treated for 24 hours in hypoxia (1% O2) can create accumulation of niche-like areas of CD11b+ bone marrow derived cells to accumulate in the lungs of naive immunocompromised mice.[35] This has also been shown in experiments with hypoxic conditioned medium from the spontaneously generating polyoma middle T antigen (PyMT) mammary carcinoma cell line. Daily intra-peritoneal injections of hypoxic conditioned medium caused a significant increase in CD11b+/Ly6Cmed/Ly6G+ (granulocytic MDSC) in metastatic organs.[38] These reports show that conditioned medium, and specifically hypoxia treated cells, produce tumour secreted proteins which can establish pre-metastatic-like niches in naive mice. This indicates that in vitro tumour cells produce secreted proteins similar to those produced in vivo. CD11b and Gr1 Cell Markers The cells shown to accumulate in the pre-metastatic niches were identified using CD11b and Gr1 proteins. CD11b is a cell surface marker which generically identifies all myeloid cell types. It is one half of an integrin heterodimer with a role in cellular adherence.[39] CD11b is also known as alpha M integrin or Mac-1 ? and complexes with the ? 2 integrin also known as CD18.[39]  Gr1 is a marker for early myeloid lineage commitment in mice.[40] Cells which are Gr1+ yield a heterogeneous population as the marker actually refers to two antigens on the surface of cells: Ly6C and Ly6G. Using these markers, it is possible to identify more specific cell types as cells Ly6C+ mid and G+ bright are granulocytes whereas Ly6C+ bright but G- cells are monocytes.[37] Immature myeloid cells and myeloid derived suppressor cells (MDSC) are among the cell types identified by CD11b and Gr1. However, while MDSC?s are CD11b+ and 11  Gr1+, they can only be called suppressive after they have shown a suppressive phenotype in an immunosuppression assay.[41] Secreted Proteins Secreted proteins play an important role in the pathogenesis of solid cancers as well as the creation of the pre-metastatic niche. Regarding the niche areas, secreted proteins play two main functions. Firstly, they mobilize BMDC?s, such as CD11b+ Gr1+ cells and secondly they are responsible for the stromal extracellular matrix breakdown and remodelling. Through analyzing conditioned media and the blood/serum from tumour bearing mice, it is possible to discover which proteins are secreted by tumour cells and at what levels.  Secretome Analysis Given the importance of secreted proteins in the creation and establishment of the metastatic process, a method of secreted protein analysis and quantification was required. Mass spectrometry enabled the identification of the secretome or complete list of all secreted proteins from a given cell line. Coupling this ability with Stable Isotope Labelling of Amino Acids in Cell Culture (S.I.L.A.C.) enabled quantitative comparison between the secreted proteins in multiple in vitro conditions. SILAC was performed on the metastatic 4T1 cell line and the non-metastatic 67NR cell line in both hypoxic and normoxic oxygen conditions.    Stable Isotope Labelling of Amino Acids in Cell Culture During cell culture, cells obtain all required amino acids and nutrients from the media. Using media which had all amino acids removed allowed the selective addition of amino acids with specifically created molecular characteristics. In this research, the amino acids arginine and lysine were removed from the regular RPMI1640 media and in their places heavy isotope forms of arginine and lysine with were added back. For each cell line in vitro, the proteins secreted during hypoxia and normoxia were compared for differential expression indicating hypoxic induction. It was suspected that the hypoxic tumour of the 4T1 cell line is producing secreted proteins that are in part responsible for the creation of the pre-metastatic niche. In order to investigate this, parallel passage lines were created using the same cell culture source in which one would receive normal arginine and lysine and the other would receive heavy isotope labelled forms of the amino acids. After 5 generations the modified amino acids are incorporated into every protein made by the cell.[42] The heavy form of arginine is formed by replacing the carbon-12  12 with carbon-13. This creates an amino acid with an extra 6 neutrons and consequently will appear on mass spectrometry as having a shift in molecular weight allowing quantitative comparison of the same protein if also present in the normal isotope labelled normoxic cell culture conditions.    A heavy isotope form of lysine was also added where the carbons in position 4 and 5 had deuterium atoms instead of normal hydrogen atoms resulting in 4 additional neutrons. Trypsin, which cleaves after arginine and lysine residues, is used to digest the proteins ensuring that every fragment will contain at least one heavy Figure 3 Heavy isotope forms of lysine and arginine used in SILAC.  Lysine-4 created by substitution of 4 hydrogen atoms for deuterium atoms. Arginine-6 created by substitution of carbon-12 atoms for carbon-13 atoms. labelled amino acid.[43, 44] This allows every fragment to accurately be attributed to the hypoxic cell culture conditions when run in tandem mass spectrometry with the normoxic samples The SILAC media was used for both the 4T1 and 67NR cell lines. Parallel passage plates of each cell line allowed for medium with different amino acid isotopes (heavy and light) to be used and enabled different oxygen conditions to be compared. Overall, SILAC and mass spectrometry enabled quantitative comparison of 4T1 secreted proteins induced during hypoxia (1% O2) compared to 4T1 proteins secreted during normoxia (21% O2). In addition a non-quantitative comparison of secreted proteins secreted by the metastatic 4T1 cells verses non-metastatic 67NR cells was possible. The secretome of a primary tumour cell lines showed that many different kinds of proteins were produced. Given the diverse results, it is not likely that one secreted protein is exclusively responsible for the creation of the pre-metastatic niche. Tumour secreted 13  proteins function as an interconnected series often building upon one another to create the necessary changes in the ECM that allow BMDC?s and later circulating tumour cells to invade. These proteins also have multiple different mechanisms acting upon similar ECM components at the same time. Each protein is required to do its part to achieve an ECM which will allow invasion and colonization. It is possible to interrupt this process through knockdown of secreted proteins and decrease metastasis but removal of one protein cannot completely inhibit metastasis. In the future, interventions aimed at secreted proteins could hopefully reduce or prevent metastatic disease in a clinical setting.   Indirect Effects of Tumour Secreted Proteins A major secreted protein responsible for release of BMDC?s from the bone marrow is granulocyte colony stimulating factor (G-CSF).[37] It is responsible for myeloid cell mobilization and is secreted by multiple metastatic tumour types including the 4T1, MDA-MB-231, and PyMT but absent in non-metastatic tumour types.[37] [unpublished data, Bennewith et al.] In this way, G-CSF acts indirectly to help establish the pre-metastatic niche by mobilizing the BMDC?s that will accumulate in the future metastatic sites. G-CSF also causes expansion of circulating BMDC?s, including immature myeloid cells, which accumulate within the spleen. The spleen is used as a reservoir and the cellular accumulation results in significant splenomegaly. [unpublished data, Bennewith et al.]   Direct Effects of Tumour Secreted Proteins Many secreted proteins play a much more direct role in establishing the niche. These proteins directly change the ECM of the metastatic target organ and often become sites of future metastases. As collagen represents the most abundant ECM glycoprotein in the body it is no surprise that multiple tumour secreted proteins act on this ECM protein.[45] One of the earliest proteins found to have a role in establishing the pre-metastatic niche is Fn.[36] In both the LLC and the B16 melanoma models in mice, Fn levels were increased in the lungs prior to arrival of VEGFR1+ BMDC?s.[36] Moreover, the locations of increased Fn were near the areas of future pre-metastatic niche formation.[36] In addition to tumour bearing models, conditioned media from B16 melanoma cells up regulated the expression of Fn in the lung, oviduct and intestine of na?ve mice.[36] The deposition of Fn has proven to be an important step in the creation of niches as it can act as a ligand for cell receptors. The VEGFR1+ BMDC?s express VLA4 (integrin ?4 ?1) which is a Fn ligand and aids in clustering of VEGFR1+ VLA4+ BMDC?s at future pre-metastatic niches.[36] Many other cell types 14  including immature myeloid cells and CD11b+/Gr1+ cells also express integrin receptors specific to Fn. In addition to acting as a cell receptor ligand, Fn also co-localizes with many secreted proteins potentially acting as an anchor to direct niche formation. Lysyl oxidase (Lox) is a hypoxia induced secreted protein and has been shown to have an important role in metastasis and in the formation of the pre-metastatic niche.[35, 46] It co-localizes with Fn in the lungs and is essential to create a mature ECM through its ability to cross link collagen IV in the basement membrane of tissues.[35, 46] The cross-linking of collagen enables CD11b+ cells to adhere and accumulate in the tissue.[35] Lox, like Fn, was one of the early secreted proteins shown to have a role in BMDC accumulation. Activation of Lox from the secreted pro-peptide form of the protein is done by another tumour secreted protein, bone morphogenetic protein-1 (BMP-1).[47] In addition to Lox, there are other proteins from the Lox family of proteins which have roles in the establishment of metastatic niches. Lox-like 2 and 4 (Loxl 2/4) are hypoxia inducible secreted proteins which like Lox catalyze intra and inter-molecular collagen cross-linking.[48] Lox, Loxl 2 and 4 share the ability to stabilize collagen fibres and fibrils enhancing collagen strength which aids in focal adhesion formation and signal pathway activation promoting invasion of BMDC?s.[48]   Pro-collagen lysine, 2-oxoglutarate, 5-dioxygenase or pro-collagen lysyl hydroxylase 2 (PLOD2) is a protein with intra and extra-cellular effects on collagen.[45] Intracellularly it hydroxylates lysine residues on pro-collagen molecules providing glycosylation sites.[45] When secreted, these modified collagen molecules are cross linked by the Lox family of proteins and having hydroxylated lysines create more stable crosslinks which are more resistant to enzymatic degradation.[45] PLOD2 when secreted creates an ECM with increased collagen fibre formation in the primary tumour and through secretion of modified collagen molecules enhances the architecture remodelling leading to increased metastasis.[45]  Concurrent with the changes in the ECM architecture, the spleen acts as a major repository for CD11b+ cells [unpublished data, Bennewith et al]. Fn, PLOD2 and Lox have modified and cross linked collagen fibres in the ECM enabling BMDC?s from the spleen to accumulate in the remodelled sites establishing pre-metastatic niches. The BMDC?s in the niche area will continue the architectural changes to further enable BMDC accumulation and circulating tumour cell invasion and survival. 15  Matrix metalloproteinases (MMP?s) are often induced through co-opting stromal cells or BMDC?s by the primary tumour.[35, 37, 49, 50]  MMP-2 has been shown to be expressed by CD11b+ cells which have accumulated after cross linking of collagen IV by Lox.[35] MMP-2 would then cleave the collagen fibres allowing enhanced invasion of BMDC?s and metastasizing tumour cells.[35] MMP-9 has multiple sources of expression and multiple roles in metastatic tissues. MMP-9 expressed from lung endothelial cells and macrophages increases invasion ability into the lung.[50] In addition to endothelial cells, Ly6C+/Ly6G+ cells are responsible for expression of MMP-9 in the pre-metastatic niche [37] and similarly CD11b+/Gr1+ cells express MMP-9 and promote vascular remodelling[49]. Many secreted proteins required in the creation of the pre-metastatic niche are enzymes but there are also non-enzymatic proteins which act as matricellular proteins. They represent non-structural matrix proteins as a subset of the ECM.[51] Context dependent micro-environment cues can alter the function of these proteins from: ECM remodelling, modulation of cell-matrix interaction, to growth factor and ligand binding interactions with various receptor mediated signalling pathways.[51]  Periostin is a secreted matricellular glycoprotein capable of binding with several other known ECM proteins such as Fn and collagen I as well as matricellular proteins like Tenascin C (TNC).[51, 52] Periostin plays multiple roles with regard to its mechanism of promoting metastasis. It has been shown to play a role in ECM remodelling and specifically assists the incorporation of TNC into the ECM.[52] In colon cancer model of metastasis, periostin increases metastatic growth by reducing apoptosis of tumour cells and by increasing angiogenesis of endothelial cells in metastatic target organs.[53] Periostin also has interactions directly with BMP-1 and through it with Lox. BMP-1 was shown to activate Lox from its pro-peptide form into the enzymatically active mature form.[47] Interaction with BMP-1 enhances periostin deposition on the ECM enabling it to proteolytically activate more LOX pro-peptide which increases collagen cross-linking in the niche area.[54]   TNC is another matricellular protein secreted by primary tumours that is deposited in areas of ECM remodelling. Its role as a secreted protein in niche formation will be introduced later.  Tenascin C TNC was a top result in terms of fold over-expression during hypoxia in the 4T1 cell line. In addition, TNC was absent from the 67NR results indicating a potential role in metastasis.  16  History  In 1984, Erickson and Inglesias were looking at fibronectins in tissue culture supernatant. They discovered that among the fibronectins produced there were also oligomers with an elaborate and well defined structure that could be isolated from the fibronectins. They called the oligomers, ? hexabrachion?.[55]  Also in 1984, Chiquet et al. had discovered ?myotendinous antigen? which they characterized to be a novel extracellular matrix component present during embryogenesis which they suggested was involved in developing tendon connections with bone and muscle fibres.[56] In 1985, Grumet et al. characterized an extracellular matrix glycoprotein and gave the name ?cytotactin?.[57]  The initial characterization of cytotactin was from a 14 day chicken embryo brain, however after further study it was found to be expressed in a variety of tissues during embryogenesis.[57] In this early study, the role of cytotactin was one of cell-substrate adhesion molecules particularly in neuron-glia adhesion with the possibility of supporting migration in the brain and potentially in the other tissues with selective expression of cytotactin.[57]  In 1985, Kruse et al. isolated a novel nervous system cell adhesion molecule that they termed J1glycoprotein after the antibodies that reacted with the protein.[58] In 1986, Chiquet-Ehrismann et al. working with myotendinous antigen protein described its role in fetal development but also in oncogenesis.[59] They also noted that hexabrachion was the same as this protein.[59] It was in this publication that the modern name, Tenascin from Latin verbs tenere, to hold - root of the English word tendon, and nasci, to be born- root of the English word nascent, describing its location and its development expression, was given to these proteins.[59] After being characterized numerous times, Vaughan et al. in 1987 showed evidence that in fact hexabrachion and myotendinous antigen were the same glycoprotein and also posited that J1-glycoprotein and cytotactin based on the reports in the literature were also likely to be one in the same.[60] After this consolidation of knowledge, they did not object to the new name given to the protein by Chiquet-Ehrismann et al. a year earlier.  There are now a number of separate proteins under the Tenascin family name including: TN-C,TN-R[61], TN-Y[62] TN-W[63], and TN-X[64]. 17  Structure TNC is a modular protein. Secreted functional TNC is a hexamer comprised of six monomer subunits joined at an amino-terminal tenascin assembly domain.[65] In humans, the entire gene is 80 kilobases of DNA representing 27 exons separated by 26 introns.[66] Each TNC monomer is made up of three major parts: a C-terminus fibrinogen like domain, an area of fibronectin type III repeats, and an N-terminus consisting of primarily epidermal growth factor (EGF) like repeats.[66, 67] The fibrinogen like domain is comprised of 210 amino acids encoded by 5 exons.[66] The EGF like repeats are a much conserved portion of the gene. There are 14.5 repeats encoded in one exon.[66] On the other hand the type III fibronectin like repeats while still conserved are split up into 17 repeats over 20 exons and are alternatively spliced to create isoforms of the protein.[68] Eight of the repeats are universally expressed and present in all splice variants as represented by exons 3-9 and 17-22, or fibronectin repeats 1-5 and 6-8. The additional 9 repeats can be spliced out of the transcript shown as nine variants: A1, A2, A3, A4, B, AD1, AD2, C and D in Fig.4.[67]   Figure 4 Representative structure of a Tenascin C monomer.  Exon organization is below and corresponding protein schematic shown above. EGF-like repeats form the majority of the N-terminal region; fibronectin type-III repeats, constitutive repeats in white and alternatively spliced repeats in grey; C-terminal fibrinogen. Image used from [68].  TNC has a heparin binding domain which has been determined to be mainly a part of the fifth type III repeat.[69] This region has been to shown to very strongly bind VEGF, PDGF, fibroblast growth factor (FGF), TGF beta, and insulin-like growth factor-binding proteins (IGF-BP?s).[70] 18  Expression and Role TNC has been extensively studied since its discovery in the 1980?s both in terms of the physiological and pathophysiological functions. Primary expression of TNC is during embryogenesis where it is expressed in the neural crest.[71] In addition, embryonic expression is seen in certain organ and tissue vasculature like lung and corneal development.[71] TNC, although tightly regulated, can be found persistently within adult tissues in primary and secondary lymphoid organs [72] However, the majority of adult TNC observed is in response to specific conditions where expression is induced.[70, 73] These conditions are time and location specific such as during tissue remodelling or repair, wound healing, nerve regeneration and inflammation.[70, 73] TNC is an early and transient event in the aforementioned situations and when the initiating stimulus dissipates, expression of TNC decreases, as seen with mature collagen scar formation after wound healing.[29] During these types of events, TNC acts in a physiological role but when the conditions inducing expression become prolonged or chronic TNC becomes involved in pathophysiological mechanisms. Abnormal expression can be associated with conditions such as: chronic inflammation, fibrosis, cardiac pathologies, heart failure, ischemia, atherosclerosis, and cancer.[29, 73] TNC?s physiological role is one of an anti-adhesive or adhesion modifying ECM glycoprotein.[71, 74] More recently TNC has been identified as a member of a family of non-structural ECM proteins with regulatory roles called matricellular proteins.[75] A main TNC mechanism is inhibition of cellular binding with fibronectin.[76] This inhibition can result in reduced adhesion to Fn.[76] The normally strong adhesion generated by cellular binding with Fn is decreased when TNC is present, generating a state of intermediate adhesion characterized by focal adhesion disruption and actin stress fibre reorganization.[77] However, these changes only have minimal effect to the shape of the cell.[78] Strong adhesion is characterized by focal adhesions and actin stress fibres. An intermediate adhesion state maintains the cell morphology, integrin matrix connections and talin, but vinculin and ?-actinin is lost.[78] As mentioned, the actin-matrix connections are maintained but the actin stress fibres and the integrin proteins are no longer linked.[78]  The alternatively spliced region of the Fn type III domain of TNC has been shown to be the region responsible for the creation of the intermediate adhesion state through binding annexin II on the cell membrane.[78-80] Receptor binding modifies the adhesion state of cells through basal cGMP-dependent protein kinase (PKG).[78] Importantly, the 19  intermediate adhesion state is the most favourable to cell migration; strong adhesions make it harder for the cell to release the focal adhesion - matrix interactions.[78] This process of cellular de-adhesion is seen in many biological situations where TNC is known to be expressed, including tissue remodelling during morphogenesis and wound healing, cellular transformation and metastasis.[78] In wound healing, TNC has been shown to influence fibroblast cell phenotype through interaction with syndecan-4 and Fn.[81] When TNC is present, Fn signalling through ?V?1integrin and syndecan-4 was inhibited.[81] This prevented activation of RhoA and FAK which altered stress fibre and focal adhesion formation thereby reducing matrix contraction at the wound edges.[81] The reduced contraction prevented excess wound healing at the edges where tissue is already appropriately healed.[81]  While TNC has a helpful role during vessel damage and wound healing, if it is expressed for a prolonged period it becomes harmful. During atherosclerosis, TNC is expressed in response to chronic vessel damage. The changes in the endothelium, at sites of damage, recruit immune cells which produce TNC.[29] Immune cells, such as macrophages, migrate into the walls of the vessels and activate smooth muscle cells (SMC) and myofibroblasts, which in turn produce additional TNC.[29] TNC changes the phenotype of the SMC?s and results in migration and replication of non-proliferative SMC?s.[82] In vitro, TNC results in rat SMC adherence to TNC coated plates through an interaction of the fibrinogen binding domain of TNC and the ?V?3 integrin on SMC.[83, 84]  Binding of the ?V?3 integrin resulted in focal adhesion formation.[85] In addition, TNC inhibits the adhesion of rat SMC to Fn further modifying the ability of SMC to migrate in areas of atherosclerosis.[86]  These changes enable additional cells to migrate to the sites of damage and ECM deposition continues and advances plaque growth leading to narrowed vessels and eventually vessel rupture.[29] TNC is found especially in lipid rich plaques, areas of macrophage accumulation and areas of rupture. [29] TNC in Cancer and Metastasis  TNC?s role as a matricellular and adhesion modifying protein lends wide and diverse effects to cancer and metastasis through many different regulatory and cell-matrix modifying interactions.[75] In addition to the adhesion and migratory role, evidence for TNC in EMT, proliferation, invasion, angiogenesis, immune cell migration and activation, as well as signalling and stem cell niches have been shown.[75] However, precise determination of a 20  singular role remains unclear and likely multiple functions exist determined by site of action, stage of cancer progression or specific spice variants present. Cancer cell adhesion to Fn was decreased when TNC was present but this also resulted in increased proliferation.[87] Similar to the fibroblast matrix interactions shown in wound healing, the TNC mechanism of anti-adhesion was through binding Fn.[87] It was shown that TNC binds the thirteenth Fn type III repeat in Fn and blocked the HepII/syndecan-4 cellular binding  site within Fn.[87] Syndecan-4 binding in fibroblasts is required for focal adhesion and stress fibre formation.[88] Previously, Fn type III repeats 1 ? 5 of TNC have been shown to bind integrin ?5?1 and inhibit adhesion.[89] This action of TNC was suggested to be responsible for the increased proliferation as reduced Fn matrix binding is known to increase tumourigenesis.[87] Colon cancer cell invasion could be regulated through hepatocyte growth factor and TNC.[90] In vitro invasion resulted in a change from a non-migratory phenotype characterized by high RhoA and low Rac activity into a migratory phenotype with low RhoA and high Rac activity.[90] The EGF like repeats of TNC caused the RhoA inactivation allowing HGF to activate the invasive characteristics of Rac via c-Met.[90]  Similarly, TNC was shown to be important in melanoma metastasis by creating a niche that elicited stem cell-like characteristics.[91] Melanoma cells with TNC knockdown showed reduced metastases to the lungs, reduced chemotherapy resistance and a decrease in cells that showed these stem cell-like characteristics.[91] With regards to angiogenesis, TNC knockout mice have shown reduced tumour capillary nets compared to wild type mice after transplantation with melanoma cells.[92] This effect was attributed to the role of TNC in cell-cell communication between mesenchymal cells and cancer cells as well as stromal TNC inducing diffusible factors that enabled cancerous cells to secrete VEGF and tumour derived TNC.[92]  With regard to the varied roles that exist for TNC, it has been shown that specific splice variants or isoforms appear to play an important role in regulating the effect of TNC in disease progression.[68] Full length isoforms are commonly found within breast cancer and fully truncated isoforms are found in normal breast tissue.[68] TNC isoforms with alternatively spliced repeats are thought to be required for the focal adhesion disassembly.[78] Indeed, alternatively spliced TNC containing exons 14/16 and exon 16 21  alone were shown to have enhanced proliferation and invasion as well as suggested to be important for migration.[93] Repeats 3-6 of the FN type III domain have been found to play a role in cellular adhesion, while repeats 7-8 have been found to be counter-adhesive.[68] The alternatively spliced region, grey shaded boxes in Fig.4 are susceptible to cleavage by MMP?s.[68] Degradation by MMP2 has been associated with tumour recurrence. MMP2 cleavage reveals cryptic binding sites on TNC leading to ?1 integrin activation via syndecan 4.[94] MMP7 also cleaves TNC potentially revealing cryptic sites which could attribute additional mechanism of actions to full length TNC.[95] However, MMP9 does not cleave TNC but MDA-MB-231 cells co-stimulated with TNC and TGF? had higher MMP9 production and greater invasion that with either of the two alone.[95, 96] In addition to cleavage, TNC has also been shown to induce the expression of MMP-13 and tissue inhibitor of matrix metaloproteinase-3 (TIMP-s) in an isoform independent manner.[93] The diverse role of TNC is also seen when looking specifically at breast cancer and breast cancer metastasis.  TNC can induce EMT like changes in breast cancer cells.[97] Soluble TNC added to MCF-7 media resulted in E-cadherin and ?-catenin delocalization from membrane cell-cell contact to the cytoplasm.[97] It was later shown that these EMT like changes were a result of binding ?v?1 and ?v?6 integrins.[98] In these reports, TNC mediated EMT correlated with FAK phosphorylation by SRC which resulted in a loss of inter-cellular adhesion and increased migration.[97] TNC has been shown to have a role in regulating expression of cancer stem cell signalling components in MDA231-LM2 cells.[99] Specifically, the ability to induce positive regulators of the signalling pathways, NOTCH and WNT.[100] This effect was suggested to work by TNC suppression of JAK2-STAT5 signalling in breast cancer cells to enhance the expression of the NOTCH pathway component MSI1 and the pro-metastatic function for this pathway.[100] This effect is not solely due to tumour derived TNC. Tumour cells initially produced TNC after IV injection. Drug induced TNC shRNA expression at an early time point caused apoptosis and inhibited metastatic outgrowth of the tumour cells.[100] However, if the tumour cells were allowed to remain for a longer period prior to drug induced TNC shRNA induction, the effect of TNC knockdown was abrogated as the stroma was induced to produce TNC. No deleterious effect on metastatic growth occurred.[100] 22  While specific mechanisms for TNC have been suggested, the effects of TNC are not always clear. S100A4+ cells, suggested to be fibroblasts could express TNC in the metastatic foci of lungs of 4T1 metastases.[101] It was noted that in the lungs, a reduction in S100A4+ cells resulted in a reduction of TNC and VEGF and a reduction in metastasis.[101] It was suggested in this research that TNC was potentially acting as an anti-apoptotic protein potentially through binding TGF-? receptors or by inducing cancer stem cell markers to promote tumour cell survival.[101]  Clinical breast cancer studies have shown that high levels of TNC are associated with a shortened time to lung metastasis relapse.[100] Gene arrays indicate that TNC was associated with a set of genes that mark and mediate breast cancer metastasis to the lungs.[102] Another study created a transgenic mouse strain which readily formed pulmonary metastases and compared those mice to a similar transgenic strain which only formed micro-metastases.[103] When the tumour gene profiles were compared, a set of deregulated genes including TNC was obtained.[103] This study followed up and knocked down TNC and saw anchorage independent cell proliferation and impaired cell migration.[103] When TNC was knocked down in MDA-MB-435 cells and implanted in mice, the result was a decrease in primary tumour growth, decreased time to tumour relapse post primary tumour surgical removal, and decreased lung metastasis.[103] These results were consistent with findings that TNC enhanced metastasis in both humans and mice through alteration of tumour cell migration and proliferation resulting in poorer prognosis. Hypothesis and Aims Given the information on the primary tumour characteristics of the 4T1 versus 67NR cell lines and along with the corollary data of hypoxia relating to metastatic disease, it was hypothesized that hypoxia induced expression and secretion of specific proteins were responsible for promoting metastasis. These proteins have a number of roles, including enabling tumour cell migration and invasion as well as creating regions within metastatic target organs that were permissive to circulating tumour cells enabling the creation of metastases. The aim was to identify the secreted proteins created by the 4T1 cell line under normoxic and hypoxic conditions. This would allow specific hypoxia induced secreted proteins to be identified and studied for their role in the establishment of metastases. TNC was identified 23  and selected as a hypoxia-inducible secreted protein which may act to promote metastasis through the enhancement of tumour cell migration in vivo.   24  Materials and Methods Tumour Cell Lines 4T1 and 67NR murine mammary carcinoma cells were provided by Dr. Fred Miller (Karmanos Cancer Institutes, Detroit, MI).  Orthotopic tumour injections were performed with 105 4T1 cells or 2x105 67NR cells in the fourth mammary fat pad. These cell numbers produce consistent tumor growth rates, with tumor volumes that approach ethical restrictions 4 weeks after implantation.  For all tumour cell lines, cell culture was performed using RPMI 1640 supplemented with 10% fetal bovine serum as well as additional glucose, HEPES, sodium pyruvate and L-glutamine (standard media used). Routine passage was a 1:10 dilution performed every other day using 0.1% trypsin.  In Vivo Mouse Models All mice were housed under specific-pathogen free conditions in the Animal Resource Centre at the BC Cancer Agency Research Centre. All animal experiments 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. Intra-peritoneal injections were performed according to UBC written procedures. Euthanization was performed using CO2 according to UBC written procedures. Female BALB/cfC3H mice (8 ? 10 weeks of age) were purchased from Simonsen Laboratories (Gilroy, CA). Female NOD/scid mice (8 - 10 weeks of age) were purchased in house from the Animal Resource Center breeding colony. Tissue Processing Spleens were pushed through 100?m and 40?m mesh filters with PBS to create single cell suspensions. Lungs were finely minced with scalpels prior to agitation for 40 minutes at 37oC with an enzyme suspension containing 0.5% trypsin and 0.08% collagenase in PBS. After incubation, 0.06% DNAse was added and the cell suspension was gently vortexed and filtered through 30?m nylon mesh to remove clumps. Cell suspensions were treated with 25  NH4Cl for erythrocyte lysis. Cells were fixed in 70% EtOH and stored at -20oC for subsequent flow cytometry analysis.  Flow Cytometry 5x105 ? 1x106 alcohol-fixed cells were rehydrated and washed in PBS + 4% FCS prior to contact with the appropriate antibodies, including: CD11b-PE, Gr1-Alexa 488, unconjugated F4/80 (eBioscience), pan-cytokeratin (Dako, Markham, ON, Canada), and Alexa 488 or 594 secondary antibodies (Invitrogen, Burlington, ON, Canada). DAPI was used to analyze DNA content in all samples; when analyzing lung samples that contain 4T1 tumor cells, the hyperdiploid DNA content of the tumor cells enabled their exclusion from analysis of the diploid normal cells.  List mode files were collected using a dual laser Epics Elite-ESP flow cytometer (Coulter Corp., Hialeah, FL) and were subsequently reprocessed for analysis. Doublet correction and bitmap gating were used to select the cell populations of interest with the WINLIST software package (Verity Software House Inc., Topsham, ME). Flow cytometry was also performed on FACSCalibur machines with 488 nm and 633 nm lasers. Data collection was done using CellQuestPro software and data analysis was performed using FloJo software.  Clonogenic Assay Monodispersed lung cells (derived by enzymatic disaggregation of lung tissue- refer to tissue processing) were washed by centrifugation before NH4Cl erythrocyte lysis. Cells were washed in PBS, resuspended in RPMI 1640 with 10% FBS, and aliquots of 3 ? 103 , 3 ? 104 and 3 x 105 cells were plated into tissue culture plates containing 60 ?M 6-thioguanine (to specifically allow growth of 4T1 cells). Cells were incubated for 9?12 d (37?C, 5% CO2) before staining colonies with malachite green for enumeration. The total number of clonogenic tumor cells in the lungs was calculated by multiplying the proportion of colony-forming tumor cells by the total number of cells recovered from the lungs. Immuno-Fluorescence Tumors were frozen in Optimal Cutting Temperature (OCT) medium (Sakura Finetek, Torrance, CA) and 8-10?m serial sections were cut and stained in PBS + 4% FCS with antibodies against: Tenascin C (ab6346, Abcam) or CD31 (PharMingen, Mississauga, ON) with Alexa 488 or 594 secondary antibodies. Slides stained for pimonidazole were 26  processed in PBS + 4% FCS + 0.1% Triton-X with pimonidazole-FITC antibody (HPI Inc.). Images were captured with a Retiga EXi camera (QImaging, Surrey, BC, Canada) using an Axiovert S100 microscope (Carl Zeiss Canada Ltd., Toronto, ON, Canada).  Statistical Analysis Student?s T-tests were used for all comparisons, with p<0.05 being considered significant. One-tailed and two-tailed T-tests were used as indicated. Hypoxia Hypoxic conditions were 1% O2 with 5% CO2 and 94% N2 at 37?C.Hypoxia chamber belonged to the Experimental Therapeutics department at the B.C. Cancer Research Centre. An additional hypoxia chamber was used belonging to the Integrative Oncology Department at the B.C. Cancer Research Centre. Protein Analysis Protein Isolation and Quantification Cells were washed with PBS then lysed in 300 ? 400 ?L urea lysis buffer (9 M urea, 75mM Tris-HCL) and placed in -80?C overnight. After thawing, the lysate was sonicated for 5 seconds to shear DNA, and then centrifuged for 15 min at 14500 rpm to remove cell debris. The protein isolate was stored at -80?C. The protein concentrations of cell lysates were measured by a Bradford protein assay in a 96 well format (Bio-Rad, Mississauga, Canada) using the manufacturer?s protocol. Standards were made with BSA at 2, 1, 0.5, 0.25, and 0.125 mg/ml. The absorbance was measured at 650 nm with a GENios Microplate Reader (Tecan, Switzerland).  Conditioned Media/ Secreted Protein Isolation 4T1 and 67NR cells were grown to ~ 80 % confluence in 10 cm plates in standard RPMI 1640 media. Plates were washed once with PBS and 6 ml of phenol red free RPMI 1640 supplemented with additional glucose, HEPES, sodium pyruvate and L-glutamine. Plates were then incubated for 24 hours in standard incubator conditions or in hypoxia chamber as described.Conditioned media was collected using a 5 ml syringe and filtered using a 0.20 ?m filter.  27  To precipitate the proteins in the conditioned media it was diluted with 99% ethanol in a 1:4 ratio volume by volume. 1.25 M sodium acetate (pH 5.0) was added to achieve a final concentration of 50mM sodium acetate. 20 ?g of glycogen was added from a 10 ?g/?l stock solution. Conditioned media solution was mixed well and let stand overnight at room temperature (21?C). Next day, the solution was spun down sequentially using the same microcentrifuge tube at 10000 rpm for 10 minutes. Pellet was resuspended using 1% SDS in dH20. Alternatively, conditioned media was collected using a 5 ml syringe and filtered using a 0.20 ?m filter and transferred to a Vivaspin 6 column with a 5 kDa molecular weight cut off. Concentration columns were spun at 4000 g at 20 ?C for 45 minutes or until ~500 ?l remained. Concentrated conditioned media was then removed and stored at -80?C.  Western Blot A total of 30 ? 40 ?g of the cell lysate was separated by 10% or 12% SDS-polyacrylamide gel electrophoresis, and then transferred to a nitrocellulose membrane. The membrane was blocked with TBS-T (1X TBS, pH 7.6, 0.1% Tween 20) and 5% w/v non-fat milk powder (blocking buffer) for 30 min at room temperature with rocking.  The membranes were probed with primary antibodies in TBS-T with 5% w/v non-fat milk powder. Mouse monoclonal anti-tubulin was from Sigma and used in a 1:40000 dilution. Rabbit monoclonal anti-tenascin c was from Abcam (ab108930) and used in a 1:2500 dilution. Rabbit polyclonal anti-periostin was from Abcam (ab14041) and used at 0.3 ?g/ml. Membranes with primary antibodies were probed at room temperature for 90 minutes or overnight at 4?C with rocking. Membranes were washed twice with TBS-T for 15 seconds followed by two washes with TBS-T for 5 minutes with rocking. Goat anti-rabbit and goat anti-mouse secondary HRP (horseradish peroxidise) linked antibodies were incubated for 60 minutes in TBS-T with 5% w/v non-fat milk powder with rocking. Secondary antibody for tubulin was used at 1:40000. Secondary antibody for tenascin c and periostin were used at 1:2500 and 1:3500 respectively. Membranes were washed as before. The antibodies were detected using enhanced chemiluminescence (ECL) (PerkinElmer Inc., Waltham, MA).    28  Protein Secretion Inhibitor BD GolgiPlug? containing Brefeldin A was used to inhibit protein secretion by blocking protein transit in the Golgi complex. 12 hours prior to the end of the 24 hour incubation period for secreted proteins 6 ?l of BD GolgiPlug was added to the media. This was briefly mixed by swirling the plate. BD GolgiPlug? was gifted by Dr. Gerry Krystal.  Migration and Invasion Wound Healing/ Scratch Assay 4T1 (1x106 cells) were plated on 6 cm culture dishes and grown overnight in standard medium to create a confluent monolayer. Next day the cells were washed twice with PBS. Using a p200 pipette tip, a straight scratch was made in the cell monolayer. The cells were washed again in PBS and growth medium to remove debris and smooth the scratch edge. Images were acquired in the same field at 10 x under an inverted microscope with a Retiga EXi camera (QImaging, Surrey, Canada). The images were taken 0, 8, 12, 16 and 24 hours after the scratch. Boyden Chamber Migration Assay Migration of 4T1 cells were assessed on cell culture inserts with a transparent polyethylene terephthalate (PET) membrane (8?m pore size, 6.4 mm diameter) (BD Biosciences, Mississauga, Canada). When the inserts are placed in the wells of a 24 well plate, a Boyden chamber (or transwell chamber) is made with the well of the plate being the lower chamber and the insert being the upper chamber. 4T1 (1.4x106 ) were plated on 10 cm culture dishes and incubated overnight in standard media. The next day, cells were rinsed in medium containing 1%FBS and grown in this media to serum starve them for 24 hours. The cells were treated with trypsin and resuspended in 1% FBS containing media to a concentration of 3.3x105 cells/ml. In the lower chamber of the Boyden chamber using a 24 well plate, 600 ?L of standard media supplemented with 10% FBS was added, while 300 ?L of medium containing the serum starved cells were placed in the upper chamber. The cells were incubated in the chambers for 24 hours. To stain the cells that migrated across the PET membrane, the upper chamber side of the membrane was swabbed with a Q-tip cotton swab moistened with PBS. The cell insert was then washed 3 times with PBS, followed by 10 minutes in 100% methanol. Cell nuclei were 29  stained using 0.5 ?g/ml DAPI nucleic stain. After staining, the membranes were washed twice with PBS and inserts were cut out. The cells were imaged using Retiga EX camera (QImaging) and the DAPI stained cells were counted using ImageJ. PET pre-coating with conditioned media. A 10 cm tissue culture plate with 4T1 KD1, KD5 or shGFP cells were allowed to grow to be 80% confluent. Media was changed to serum free RPMI1640 and placed into 1% O2 for 24 hours. Conditioned media was removed using a 5 ml syringe and filtered with a 0.2 ?m filter. Conditioned media was then placed into the upper and lower chamber of the Boyden chamber for at least 2 hours at 37 ?C prior to cell plating in the upper chamber. Boyden Chamber Invasion Assay Invasion of 4T1 cells were assessed the same as for the migration assays with the exception that the membranes of the cell culture inserts were pre-coated with Matrigel? (BD Biosciences). Matrigel inserts were used according to manufacturer?s directions. Conditioned media pre-coating was performed the same as for migration assay. Proteomic Screen  Stable Isotope Labelling of Amino Acids in Culture (S.I.L.A.C.) 4T1 and 67NR cells were grown in RPMI 1640 lacking arginine and lysine and supplemented with dialyzed FBS to remove any potential arginine and lysine from the serum. Normal arginine and lysine was added back to create a ?light? media; L-[13C]6-Arg and L-[2H]4-Lys were added back to create a ?heavy? media. The heavy arginine and lysine isotopes were added back at a ratio of 1:4000. For both cell lines the normal media was used to label the cells that will receive the normoxic treatment and the heavy media was used to label the cells which would receive the hypoxic treatment. Cells were passaged for 5 generations to effectively ensure that all proteins would be labelled with the heavy isotope amino acids. At the end of 5 generations, the 4T1 and 67NR cell lines containing normal media were replaced with normal amino acid, serum free RPMI 1640 and the heavy media were replaced with heavy amino acid, serum free RPMI 1640. The heavy media cells were then placed into a hypoxia chamber at 1% O2 for 24 hours, while the normal media cells were left in standard incubator conditions. After 24 hours the conditioned media was collected as previously described and secreted proteins were precipitated using the ethanol 30  and sodium acetate method, also previously described. Media supplies and labelled amino acids were provided by Dr. Leonard Foster, UBC. Precipitated secreted proteins were analyzed by the Dr. Leonard Foster lab, UBC.  TNC Knockdown Knockdown of tenascin c was performed using lentiviral transduction of shRNA. Five shRNA constructs were purchased from Sigma. The two constructs (KD1 - TRCN0000312138 and KD5 - TRCN0000110735) with the best knockdown as determined by expression in 4T07 cell line were chosen for use.  Creation of 4T1 TNC KD cells was performed in a level III containment facility. DNA for lentiviral particles and shRNA constructs were transduced into the 293T cell line using Lipofectamine 2000 from Invitrogen. Protocol was provided by Invitrogen with purchase of Lipofectamine 2000. Media from the infected 293T cells was filtered using a 0.45 ?m filter and then transferred to 4T1 cells. The media contained TNC shRNA DNA inside lentivirus particles. This was done three times at which point the 4T1 cells were removed from the level III facility back to standard incubators. Cells were passaged 1:10 and 48 hours later 100 ?l of 2mg/ml puromycin was added to cell media. Cells were passaged as normal from this point on. Puromycin was added for all subsequent passages. Lipofectamine 2000 and level III protocols called for DMEM with 10% FBS unless noted otherwise. Knockdown was confirmed by western blot of conditioned media after 24 hour incubation in hypoxia. TNC Over-Expression The over-expression vector used was a mammalian expression lentiviral vector with the gene of interest under the MNDU3 promoter (pRRLSIN.cPPT.PGK-PURO.WPRE (G537 Puro). Puromycin was used as a selectable marker. Vector was provided by Dr Andrew Weng. TNC cDNA was purchased from Thermoscientific, catalog number MMM1013-211693798, clone ID BC117979.Cloning services were provided by CD BioSciences Inc. 45-16 Ramsey Road, Shirley, NY, 11967, USA. Over expression of TNC was performed using lentiviral transduction of the over expression vector containing the TNC gene or the vector lacking the gene (empty vector).  31  Creation of 4T1 TNC over-expression cells was performed using lentiviral expression as previously described in creation of 4T1 TNC KD cells. Over-expression was confirmed by western blot of conditioned media after 24 hour incubation in normoxic conditions.  32  Results 4T1 hypoxic conditioned media (CM) was tested to recapitulate the results previously published with MDA-MB-231 and polyoma middle T antigen (PyMT) mammary carcinoma cell lines. 4T1 CM was analyzed to identify secreted proteins expressed during hypoxia. A candidate metastasis secreted protein was selected to genetically manipulate to assess the role in metastasis.   Conditioned Media It has been previously shown that CM when injected can induce elements of the pre-metastatic niche to form in na?ve mice.[35, 36, 38] Kaplan et al. first showed that B16 melanoma CM could mobilize BMDC?s to the lungs and cause Fn to increase in metastatic organs.[36] They also showed that the reaction of na?ve mice to IP injections of CM was tumour cell line specific. In addition, it was shown that CM from B16 tumour cells can re-direct Lewis lung carcinoma metastatic niche development.[36] Erler et al. have shown that hypoxic MDA-MB-231 CM injected into female Nude (nu/nu) mice would cause BMDC?s and LOX to accumulate in the lungs of na?ve mice.[35] Most recently, Sceneay et al. showed that hypoxic CM from PyMT-WT cells can cause an accumulation of BMDC?s in the lungs of the non-syngeneic C57/BL6 na?ve mice.[38]  These reports suggest that in vitro tumour cell CM from metastatic tumour cell lines can elicit the foundation of the pre-metastatic niche - BMDC accumulation in metastatic target organs. Given the characteristics of the 4T1 tumour cell line, establishment of a pre-metastatic niche response using tumour cell CM in na?ve mice was tested. The hypothesis was that secreted proteins from a hypoxic primary tumour, represented by in vitro hypoxic CM, would be sufficient and able to mobilize and accumulate BMDC?s, specifically CD11b+/Gr1+, to the lungs and spleen in mice. In 4T1 tumour bearing mice, significant levels of CD11b+/Gr1+ cells accumulate and proliferate in the spleen of BALB/c mice (Fig.5b) which results in severe splenomegaly, up to 300% enlargement.         33    a)      b)  N a iv e   B a lb /c% CD11b+/Gr1+spleenlung012345 4 T 1  T u m o u r B e a r in g  B A L B /c  M o u s e% CD11b+/Gr1+spleenlung01 02 03 04 0   c)   Figure 5 Effects of 4T1 tumours on female BALB/c mice.  a) CD11b+/Gr1+ cell populations in the spleen and lungs.  as a percent (%) of total cells from a na?ve female BALB/c mouse. b) CD11b+/Gr1+ cell populations in the spleen and lungs  as a percent (%) of total cells from a from a 4T1 tumour bearing female BALB/c mice c) Spleen weights as a function of increasing 4T1 tumour weights. Additional mouse data points in c) provided by Bennewith lab.   The spleen is thought to act as repository and site of proliferation for CD11b+/Gr1+ cells prior to accumulation in metastatic target organs. The lungs of 4T1 tumour bearing mice also have a significant accumulation of CD11b+/Gr1+ cells (Fig.5b). These act as immune-suppressive cells that aid in the establishment of metastases. CD11b+/Gr1+ cells can be detected from these organs using flow cytometry and can act as an indicator of metastatic niche formation.  34  Example CD11b+/Gr1+ flow plots Figure 6 Representative flow plot for CD11b+/Gr1+ cells.  Left frame shows gating for viable cells. Right frame shows selection of double positive cells. Conditioned Media in BALB/c Mice The initial experiment consisted of daily IP injections with 300 ?l of CM from hypoxic (1% O2) and normoxic (21% O2) 4T1 and 67NR cells into na?ve BALB/c mice. The concentration of secreted protein in this CM was typically under 0.5 mg/ml. After 17 days the spleens and lungs were harvested from the na?ve mice and examined for the presence of CD11b+/Gr1+ cells. No significant differences were observed between the hypoxic and normoxic samples or between the 4T1 and 67NR cell lines (Data not shown). Moreover, there was no increase in CD11b+/Gr1+ cells verses the media only control (Data not shown). The results of the initial protocol did not yield an accumulation of CD11b+/Gr1+ cells. This perhaps suggested that the secreted protein concentration in the CM was too low to initiate the process.  Conditioned media trial #2 addressed this concern by increasing the number of daily CM injections to two and the experiment was repeated. There were now twice daily IP injections of 300 ?l for 11 days. The experiment was performed only with 4T1 hypoxic CM and a media only IP injection control. Flow cytometry of the spleen and lungs showed no accumulation of CD11b+/Gr1+ cells in the spleen or lungs of the na?ve BALB/c receiving the hypoxic 4T1 CM verses the media only control (Data not shown).  Once again, the concern was that the concentration of secreted proteins was insufficient to initiate the process. Conditioned media trial #3 addressed this concern by utilizing an ex vivo 35  source of CM. 4T1 tumours were grown orthotopically and harvested to be used as ex vivo tissue culture. The tumours were minced and plated in tissue culture plates. The following day they were incubated for 24 hours in serum free, phenol red free RPMI under hypoxic conditions. This method had the additional benefit of including tumour stromal tissue with the 4T1 tumour cells. The hypoxic CM was collected and injected IP into na?ve mice. Utilizing this process, both the 4T1 tumour cells and the tumour associated stromal cells would be producing the constituents of the hypoxic CM. Once daily IP injections of 500 ?l of hypoxic ex vivo 4T1 CM was performed for 13 days. Protein concentration quantification was not performed on the ex vivo CM.                 36   a)  S p le e nWeight (g)Media ControlEx vivo CM0 .0 00 .0 50 .1 00 .1 5N S b)      c)  S p le e n% CD11b+/Gr1+Media ControlEx vivo CM0 .00 .51 .01 .5 **  L u n g% CD11b+/Gr1+Media ControlEx vivo CM012345N S Figure 7 Conditioned Media Trial 3 with female na?ve BALB/c mice. Mice were given daily 500 ?l intra-peritoneal injections of hypoxic (1% O2) conditioned media from an ex vivo 4T1 tumour for 13 days. The tumour was removed from female BALB/c mice and processed into finely chopped pieces and maintained in regular serum containing RPMI 1640. This ex vivo cell culture was exposed to hypoxic conditions (1% O2 for 24 hours) prior to conditioned media collection. a) Spleen weights b) Percent (%) CD11b+/Gr1+ cells from the spleen c) Percent (%) CD11b+/Gr1+ cells from the lungs There was no significant difference in the size of spleens between the media only control group and the hypoxic ex vivo 4T1 CM group (Fig.7a). When the spleen was analyzed for the proportion of CD11b+/Gr1+ cells as a percent of total cell number (% CD11b+/Gr1+ ), the percent of CD11b+/Gr1+ cells present within the spleen was unexpectedly lower in the ex vivo treatment mice (Fig.7b). However, when the lungs were examined this reduced percentage of CD11b+/Gr1+ cells was absent and no significant difference was seen (Fig.7c). 37  Despite the unexpected decrease in percent CD11b+/Gr1+ cells within the spleen during the third CM trial, no other effects of the hypoxic ex vivo CM were seen.  A final attempt to elicit an accumulation of CD11b+/Gr1+ cells was performed using protein concentration columns with a 5 kDa cut off, allowing all proteins larger than 5 kDa to be concentrated into 300 ?l. Concentrated 4T1 CM from hypoxic and normoxic treated tissue culture was injected IP for 10 days. Previously an 80% confluent, 10 cm tissue culture plate would produce 5 ml of CM that would be shared between all mice receiving the CM treatment. Using concentration columns, each mouse would receive the equivalent of 5 ml from an 80% confluent, 10 cm tissue culture plate concentrated into 300 ?l of CM. Secreted protein concentrations of the concentrated CM typically were between 1 and 2 mg/ml. The mice were sacrificed and examined for spleen weights and CD11b+ Gr1+ cell accumulation in the spleen and lungs as before. Both the normoxic and hypoxic 4T1 CM groups had a significant increase in spleen weights versus the na?ve control group (Fig.8a). The percent CD11b+ Gr1+ cells in the spleen did not change (Fig.8b) but there was a significant increase in the total number of CD11b+ Gr1+ cells in the spleens of the hypoxic CM treatment mice (Fig.8c). Importantly, these effects in spleen size and number of CD11b+ Gr1+ cells were only observed when compared to the na?ve controls and not between the normoxic and hypoxic treatment groups. Despite the increase seen in spleen size and total number of CD11b+ Gr1+ cells in the spleen, there was no significant accumulation of CD11b+ Gr1+ cells in the lungs of na?ve female BALB/c mice from either the hypoxic or normoxic concentrated CM (Fig.8 d and e).   38  a)  S p le e nWeight (g)NaiveControlNormoxicHypoxic0 .0 00 .0 50 .1 00 .1 50 .2 0*  b)      c)  S p le e n% CD11b+/Gr1+NaiveControlNormoxicHypoxic0 .00 .51 .01 .52 .02 .5N S S p le e n# CD11b+/Gr1+NaiveControlNormoxicHypoxic02 .0?1 054 .0?1 056 .0?1 058 .0?1 05*N S d)      e)  L u n g% CD11b+/Gr1+NaiveControlNormoxicHypoxic012345N S  L u n g# CD11b+/Gr1+NaiveControlNormoxicHypoxic01 .0?1 052 .0?1 053 .0?1 05N S Figure 8 Conditioned Media Trial 4 with female na?ve BALB/c mice.  Once daily 300 ?l intra-peritoneal injections of conditioned media of 4T1 cells from normoxic and hypoxic conditions for 10 days. The conditioned media was concentrated from a starting volume of 6 ml to the final injection volume. Control mice were na?ve mice a) Spleen weights b) Percent (%) CD11b+/Gr1+ cells from the spleen c) Number (#) CD11b+/Gr1+ cells from the spleen d) Percent (%) CD11b+/Gr1+ cells from the lungs e) Number (#) CD11b+/Gr1+ cells from the lungs.  39  Throughout the 4 trials, neither the normoxic nor the hypoxic 4T1 CM were capable of eliciting the accumulation of CD11b+ Gr1+ cells within the lungs of female BALB/c mice indicative of metastatic niche formation. However, in previously published papers where accumulation of BMDC?s was achieved, none of the experiments utilized the BALB/c mouse strain. Select trial conditions were chosen to be repeated using the NOD/scid mouse background.   Conditioned Media in NOD/scid Mice The NOD/scid strain of mouse was added to examine if the lack of CD11b+ Gr1+ cell accumulation in the lungs of BALB/c mice was related specifically to the BALB/c strain of mice. CM experiments have been performed previously with immune-compromised nude mice and resulted in successful accumulation of BMDC?s in metastatic target organs.[35]     40  a)        S p le e nWeight (g)Media ControlHypoxic0 .0 00 .0 20 .0 40 .0 6N S   b)      c)  S p le e n% CD11b+/Gr1+Media ControlHypoxic051 01 5   L u n g% CD11b+/Gr1+Media ControlHypoxic051 01 5  Figure 9 Conditioned Media Trial 2 with female na?ve NOD/scid mice.  Mice were given twice daily 300 ?l intra-peritoneal injections of conditioned media from 4T1 cells under hypoxic conditions for 11 days. Control was serum free phenol red free RPMI 1640. a) Spleen weights b) Percent (%) CD11b+/Gr1+ cells from the spleen c) Percent (%) CD11b+/Gr1+ cells from the lungs.  Trial #2 from the BALB/c mice was replicated using the NOD/scid mice and received twice daily IP injections of 300 ?l of hypoxic 4T1 CM for 11 days. There was no significant difference in spleen weights between the treatment group and the media only injection control (Fig.9a). However, when the percentage of CD11b+/Gr1+ cells in the spleen was assessed there was a trend indicating a twofold greater percentage of CD11b+/Gr1+ cells after treatment with hypoxic 4T1 CM (Fig.9b). Likewise, when the lungs of these mice were assessed, the percentage of CD11b+/Gr1+ cells present within exhibited a fivefold greater percentage of CD11b+/Gr1+ cells (Fig.9c).The NOD/scid mouse strain under the same protocol as the BALB/c strain yield a positive trend of accumulating CD11b+/Gr1+ cells in the spleen and lungs while the BALB/c mice did not. 41  Trial #3 was also replicated within the NOD/scid mouse background to further test the effect seen within the NOD/scid mice. Trial #3 involved the harvest of 4T1 tumours from BALB/c mice and the creation of an ex vivo 4T1 tissue culture. The 4T1 tumours were grown within the syngeneic BALB/c background. Trial #3 entailed daily 500 ?l IP injections of hypoxia treated conditioned media from a 4T1 ex vivo tumour for 13 days. There was no significant difference in spleen weights between the treatment group and the media only injection control (Fig.10a). When the percentage of CD11b+/Gr1+ cells in the spleen was assessed there again was a marked increase in percentage of CD11b+/Gr1+ cells within the spleen and lungs (Fig.10 b and c). The trends seen in the spleen and lung were similar to the ones seen in the NOD/scid mice after CM trial #2. a)  S p le e nWeight (g)Media ControlEx vivo CM0 .0 00 .0 10 .0 20 .0 30 .0 40 .0 5N S b)      c)  S p le e n% CD11b+/Gr1+Media ControlEx vivo CM012345   L u n g% CD11b+/Gr1+Media ControlEx vivo CM0 .00 .20 .40 .60 .8 Figure 10 Conditioned Media Trial 3 with female na?ve NOD/scid mice.  Mice were given daily 500 ?l intra-peritoneal injections of conditioned media from a 4T1 ex vivo tumour for 13 days. Tumours were removed from female BALB/c mice and minced. Ex vivo tumours were maintained in RPMI 1640 containing serum. This ex vivo cell culture was then exposed to hypoxic conditions prior to CM collection. a) Spleen weights b) Percent (%) CD11b+/Gr1+ cells from spleen c) Percent (%) CD11b+/Gr1+ cells from lungs 42  In summary, after considerable trial protocol modification, accumulation of CD11b+/Gr1+ cells was achieved in the spleens of BALB/c mice which resulted in increased spleen size and increased number of CD11b+/Gr1+ cells within the spleen (Fig.8 a and c). However, CD11b+/Gr1+ cell accumulation was not achieved within the lungs of BALB/c mice (Fig.8 d and e). Interestingly, when the BALB/c mouse background was exchanged for the NOD/scid background, accumulation of CD11b+/Gr1+ cells occurred readily in both the spleen and the lungs (Fig. 9 and 10). These results suggest 4T1 CM is capable of BMDC mobilization except in BALB/c mouse model, indicating some additional component of the CM is required. The BMDC mobilization and accumulation could only be attributable to CM and not specifically hypoxic CM.  S.I.L.A.C. Proteomic Screen While the 4T1 and 67NR cell lines were derived from the same spontaneous tumour they have marked differences regarding metastasis. 4T1 tumours in vivo are hypoxic and aggressively spread throughout the body; 67NR tumours are not hypoxic but tumourigenic while being non-metastatic. That dichotomy allows the study of potential differences which could be responsible for metastatic progression between genetically related tumours. Secreted factors from the primary tumour and host BMDC?s assist in metastatic progression.[35] My CM experiments in the NOD/scid mice and previously published data suggest that tumour secreted proteins play a role in the establishment of metastatic niches.[35, 38] Moreover, comparison of the 4T1 and 67NR tumour cell line suggests that intra-tumoural hypoxia is correlated with metastasis.[24] Similar findings have been seen by the Bennewith lab, unpublished data. Given this information, comparison of the secreted protein constituents within hypoxic 4T1 CM and normoxic 4T1 CM as well as between 4T1 CM and 67NR CM could identify secreted proteins unique to hypoxic 4T1 CM which could play a role in metastatic progression.  The 4T1 SILAC proteomic screen raw data identified 326 proteins. Of those, 20 proteins had at least a twofold greater expression in normoxic conditions; 129 proteins had at least a twofold greater expression in hypoxic conditions (hypoxia inducible). The 67NR SILAC proteomic screen raw data identified 628 proteins. Of those, 20 proteins which had at least two fold greater expression normoxia; 478 proteins which had at least two fold greater expression in hypoxia (hypoxia inducible). 43  Table 1 includes the ten secreted proteins from the 4T1 SILAC proteomic raw data shown to have the highest increased expression under hypoxic conditions as well as having known roles in cancer and metastasis. In addition, Table 1 indicates whether the identified protein was also found within the 67NR proteomic results.   44  Protein H/N Ratio Present in 67NR Results MW [kDa] Proteomic Result Description Known Function Tenascin C 16.625 NO 221.7 Isoform 2 of Tenascin OS=Mus musculus GN=Tnc - [TENA_MOUSE] Matricellular protein  Lysyl hydroxylase 2 9.312 YES 84.5 Procollagen-lysine,2-oxoglutarate 5-dioxygenase 2 OS=Mus musculus GN=Plod2 PE=2 SV=1 - [PLOD2_MOUSE] Stabilizes inter-collagen cross-links needed for stability and carbohydrate attachment Peroxidasin homolog 6.163 YES 160.5 MKIAA0230 protein (Fragment) OS=Mus musculus GN=Pxdn PE=2 SV=1 - [Q80U60_MOUSE] Extracellular matrix associated peroxidase  Matrix metallo- proteinase-9 5.867 NO 7.6 Matrix metalloproteinase 9 (Fragment) OS=Mus musculus GN=Mmp9 PE=2 SV=1 - [Q6GXA2_MOUSE] Extracellular matrix degradation/digestion Bone morphogenetic protein 1  4.086 NO 87.8 Bmp1 protein (Fragment) OS=Mus musculus GN=Bmp1 PE=2 SV=1 - [Q6P550_MOUSE] Metalloprotease used in processing pro-collagen Lysyl oxidase  like 4 3.812 YES 84.8 Lysyl oxidase-like 4 OS=Mus musculus GN=Loxl4 PE=2 SV=1 - [Q6NV59_MOUSE] Collagen cross-linking Periostin 3.154 NO 87.0 Isoform 5 of Periostin OS=Mus musculus GN=Postn - [POSTN_MOUSE] Matricellular protein Syndecan 4 3.134 NO 21.5 Syndecan-4 OS=Mus musculus GN=Sdc4 PE=1 SV=1 - [SDC4_MOUSE] Extracellular matrix protein Fibronectin 2.193 YES 260.0 Putative uncharacterized protein OS=Mus musculus GN=Fn1 PE=2 SV=1 - [Q3UHL6_MOUSE] Extracellular matrix protein Fibronectin 2.190 YES 239.6 Fn1 protein OS=Mus musculus GN=Fn1 PE=2 SV=1 - [B7ZNJ1_MOUSE] Extracellular matrix protein Table 1: Top 10 selected secreted proteins from the 4T1 S.I.L.A.C proteomic data based on level of hypoxic induction and known cancer or metastasis role. H/N Ratio: Hypoxic to Normoxic Ratio  45  High throughput proteomics data must be validated in terms of accurate protein identification as well as confirmation of hypoxic fold induction. TNC and periostin represented candidate proteins involved in metastasis which were induced during 4T1 hypoxia and absent from 67NR secreted protein results. As such, they were chosen to be used for proteomic results validation.   Figure 11 Western blot validation of SILAC proteomic results.  Tenascin c and periostin western blots indicating correct identification and hypoxic induction similar to results expected based on heavy to normal ratio in Table 1 Western blots were performed to confirm the presence of these proteins in the CM from 4T1 tissue culture as well as the level of hypoxic induction of the secreted proteins (Fig.11). There is no known protein that is secreted at a steady level through hypoxia and normoxia which could act as a loading control for the secreted samples. Controls for the secreted protein western blots included using an equal number of plated cells for the hypoxic and normoxic CM generation and BSA quantification of the secreted protein samples to enable equal protein loading based on protein concentration. Based on the SILAC proteomic results (Table 1) TNC had a 16.625 fold induction in hypoxia while periostin had a 3.154 fold induction in hypoxia.  The TNC western blot showed a pronounced induction under hypoxic conditions, (Fig.11 ?H? 4T1 top panel), as well as low levels of expression in 67NR cell line. The same western blot was used to create the top panel of Fig.11. The periostin western blot shows a slight induction in the hypoxic (Fig.11 ?H? bottom panel) condition. These western blots validated 46  the results generated from the proteomic screen. They confirm the presence of both TNC and periostin in the 4T1 CM. While the induction shown on the western blots was not 16.625 fold and 3.154 fold for TNC and periostin respectively, the relative induction is consistent. TNC was selected as the best candidate secreted protein to follow up and characterize. TNC exhibited a large induction under hypoxic conditions, expected from a secreted protein with a role in 4T1 metastasis. In addition, TNC was absent from the 67NR results. After a literature search, TNC had previously been shown to have a known role in cancer and therefore represented a good candidate protein to study for a potential role in metastasis.   Figure 12 4T1 wild type TNC western blot under normoxic and hypoxic conditions.  Western blot included cell lysates used as an alternate loading control. Other secreted protein loading controls included equal cell numbers plated for all secreted protein samples and BSA quantification to enable equal amount of protein loading. Expression of intracellular TNC was examined in addition to the secreted protein. Cell lysates with tubulin acted as an alternate loading control in addition to equal cell numbers and protein quantification prior to loading. In Fig.12 there are two TNC bands which were seen in the western blots produced by the 4T1 cell line. Multiple isoforms are commonly seen in western blots of TNC. A strong induction of TNC under hypoxic conditions was also seen in Fig.12, consistent with the TNC blot in Fig.11. The addition of the lysate showed that intracellular TNC was not present within the cell and tubulin showed equal loading was achieved. The protein quantification used for the lysates was the same as used for the secreted protein samples. To investigate whether the absence of the TNC band in the lysate from Fig.12 was due to rapid secretion of TNC or an inability of the antibody to detect TNC prior to secretion, a protein transport inhibitor, BD GolgiPlug TM, was used. This inhibitor 47  contains Brefeldin A which blocks intracellular protein transport and causes proteins to build up in the Golgi apparatus, preventing secretion.  Figure 13 TNC western blot with Brefeldin A. 4T1 wild type cells in 1% O2 with and without Brefeldin A added. Lysates as well as secreted protein in conditioned media were probed for Tenascin C.  CM without Brefeldin A exhibits TNC expression induced by hypoxia and absent in the lysate as seen in the other western blots (Fig. 13). However, when Brefeldin A is added secretion of TNC is markedly reduced and a TNC specific band appears in the lysate. The results from Fig. 13 suggest that the antibody is capable of detecting both intracellular and secreted TNC and the lack of intracellular TNC in the lysate could be due to rapid transit and secretion of TNC.  The 4T1 cell line expressed significant levels of TNC under hypoxic conditions (Fig.11) and subsequently was transduced with TNC specific shRNA to knockdown expression. A panel of 5 shRNA constructs was purchased and tested for TNC protein knockdown with the best two chosen for transduction into 4T1 cells (data not shown). Construct #1 and #5 had the best knockdown of TNC in hypoxia compared to wild type 4T1 (Fig.14a). Control shRNA specific to green fluorescent protein (GFP) was used as a transduction control. Expression levels of TNC in the shGFP cell line verses wild type 4T1 are comparable (Fig.14b).  KD1 and KD5 shRNA both effectively reduced the total TNC expression (Fig.14a). Interestingly this knockdown was specific to the larger isoform and leaves the smaller TNC isoform present (Fig.14a). The cell lines transduced with KD1 and KD5 also had higher expression of this smaller TNC band in normoxia at levels similar to hypoxia. The effect of this is unclear but 4T1 cells commonly have low levels of TNC present in normoxic CM. There was no difference in proliferation between KD 1, KD5 or the shGFP cell lines (data not shown).  48  a)  b)  Figure 14 TNC western blot of knockdown in 4T1 cell lines with shGFP control.  a) TNC levels in 4T1 wild type and 4T1 Knockdown cell line 1 and 5 (KD1 and KD5) in which TNC has been knocked down using short hairpin RNA (shRNA) KD1 and 5 represented for the most effective repression of Tenascin C expression b) TNC expression levels in 4T1 wild type compared to the 4T1 transfection control shGFP (short hairpin RNA specific to green fluorescent protein). In addition to the shRNA knockdown cells, TNC over-expressing (OE) 4T1 cells were constructed. These cells were transduced with a lentiviral over-expression vector having TNC under the control of a MNDU3 promoter. An accompanying 4T1 cell line with the same vector lacking the TNC gene was also transduced. The TNC OE cell line will constitutively produce TNC in the absence of hypoxia (Fig.15). The empty-vector (EV) 4T1 cells produced a moderate intensity band in the conditioned media sample under hypoxic conditions consistent with wild type 4T1 blots. The 4T1 OE cells produced a weak band in the lysate sample while the EV cell line produced no band. The 4T1 OE cells produced an interesting banding pattern not seen in any other blots. Unlike the normal pattern of two isoforms, with the top being the most intense, the OE cells produce three isoforms in which the bottom band is the most intense (Fig. 15). At this point, the effect of the altered isoform is not known. There is no difference in proliferation between these two cell lines (data not shown).  49    Figure 15 TNC western blot of 4T1 TNC over-expression cell line.  4T1 empty vector cells will act as the experimental control. Cells were exposed to standard incubator conditions at 21% O2.for 24 hours   A recombinant TNC (rTNC) protein was also used in a wound healing assay. The rTNC had two major bands present slightly above the 250 kDa marker (Fig.16). There is also a faint third band which was smaller than the other two major bands. The smaller third band was at 250 kDa. The rTNC banding pattern is most consistent with the wild type 4T1 expression which also showed a two band pattern but the sizes are most similar to the two large isoforms produced by the 4T1 OE cells. However, the third smaller band is more consistent with the minor bands seen in the 4T1 WT cells.   Figure 16 Western blot of recombinant TNC.  Full image of western blot with 0.1 ?g of protein loaded.  The results from the 4T1 SILAC proteomic secreted protein screen produced a list of 10 secreted proteins induced by hypoxia with known roles in cancer and metastasis (Table 1). From this list, secreted proteins also found in the 67NR proteomic screen results were no longer considered. TNC was selected as a candidate secreted protein with a role in 4T1 50  metastasis. Construction of TNC knockdown and over-expression cell lines was completed. The KD1 and KD5 4T1 cell lines prevented expression of the largest isoform (Fig.14); the 4T1 OE cell line over expressed a new isoform of TNC which was smaller than the two isoforms seen in the wild type 4T1 (Fig.15). The rTNC protein purchased exhibited isoforms consistent with 4T1 WT and OE expression (Fig.16). Invasion and Migration The hypothesis regarding TNC was that through enhanced migration, TNC was having a positive effect on metastasis. TNC was also hypothesized to have a role in the formation of the pre-metastatic niche. That hypothesis was based on known roles in vivo.[75, 78] As such the expected results of in vitro invasion and migration assays are based on TNC having the same effects in vitro as in vivo. At the outset, TNC knockdown was expected to inhibit migration while over-expression was expected to enhance migration.  TNC Knockdown The first assay was a Boyden chamber migration assay. This would examine the ability of the cells to migrate through a porous membrane using 10% FBS as a chemoattractant.   Number of CellsshGFPKD1KD501 0 0 02 0 0 03 0 0 04 0 0 0* Figure 17 Boyden chamber migration assay under normoxic conditions with hypoxic conditioned media pre-treatment of membrane.  Conditioned media from 4T1 shGFP, KD1 and KD5 cells generated through 24 hours of hypoxia from each of the three cell lines was incubated with the inserts for 6.5 hours. Equal numbers of 4T1 wild type cells were then plated into each well. Assay was performed in standard incubator conditions for 24 hours. Assay was performed in triplicate Two-tailed T test * ? 0.05 4T1 KD1 and KD5 both showed an increased number of cells that migrated through the porous membrane compared to the shGFP control. 4T1KD5 showed a statistically significant increase in migration compared to the control cell line (Fig. 17).  51  Given the effect seen in the migration assay, the next assay performed was a Boyden chamber invasion assay similar to the previous migration assay except a layer of Matrigel is present on top of the porous membrane to represent basement membrane. This invasion assay would simulate a tumour cell invading out of the primary tumour site or into a metastatic site. The first variation was to plate the shGFP, KD1 and KD5 cell lines directly onto the inserts and incubate in 1% O2 for 24 hours (Fig.18). The KD1 and KD5 cell lines invaded more than the shGFP control; however, there was considerable variation among the replicates within each group. The trends were not significant compared to the control shGFP. Number of CellsshGFPKD1KD501 0 02 0 03 0 0N S Figure 18 Boyden chamber invasion assay performed under hypoxic conditions.  Cells were serum starved for 24 hours and 1.5x105 cells were plated in serum free media in the upper well. RPMI 1640 with 10 % FBS was used as a chemoattractant in the lower well. They were incubated for 24 hours in 1% O2 5% CO2 94% N2 at 37?C. Assay was performed in triplicate. The second variation of the invasion assay was to pre-treat the Matrigel coated membrane with concentrated hypoxic conditioned media to allow secreted proteins, including TNC, to adhere to the Matrigel prior to the start of the assay (Fig.19). The inserts which were to receive the shGFP cells for example would be pre-incubated for 2.5 hours with hypoxic CM from shGFP cells at which time shGFP cells would be plated for the assay. Each cell line would be pre-treated with its own respective hypoxic conditioned media. The 24 hour incubation was then performed in normoxia. KD1 and shGFP were very similar in number of invaded cells while KD5 was lower but with considerable variation within the group. TNC pre-coating prior to invasion did not result in increased invasion. 52  Number of  CellsshGFPKD1KD505 0 01 0 0 01 5 0 02 0 0 02 5 0 0N S Figure 19 Boyden chamber invasion assay under normoxic conditions with hypoxic conditioned media pre-treatment of Matrigel membrane.  Conditioned media from 4T1 shGFP, KD1 and KD5 cells generated through 24 hours of hypoxia from each of the three cell lines was incubated with the inserts for 2.5 hours prior to addition of cells. Equal numbers of 4T1 wild type cells were then plated into each well. Assay was performed in standard incubator conditions for 24 hours . Assay was performed in triplicate  Knocking down TNC expression resulted in more cells migrating across the membrane, while it had no effect on the ability of cells to invade through a reconstituted basement membrane. Following the Boyden Chamber invasion and migration experiments, a wound healing assay was performed to assess migration using an alternate model. Using the tip of a pipette, a scratch was created in the monolayer of 4T1 cells and the time needed for the cells to close the gap was analyzed. Pictures were taken at the same location on the plate at 8, 16 and 24 hours to calculate the fraction of the gap that has been closed as compared to the 0 hour time point.  53  W o u n d  H e a lin g  A s s a y  in  1 %  O2H o u rsFraction ofGap Closed0 8 1 6 2 40 .00 .20 .40 .60 .81 .0s h G F PK D 1K D 5************NS Figure 20 Wound healing assay of 4T1 shGFP, KD1 and KD5 cells in hypoxia.  Each cell line was plated in triplicate with 1x106 cells per well of a 6 well plate and allowed to adhere overnight. A separate plate was used for each time point because when a plate was removed to image at a particular time point, it would no longer be exposed to a continuous hypoxic environment. A scratch was created using a p200 pipette tip. The plates were then placed into 1% O2 at 37?C for 8, 16, or 24 hours. Assay was performed in triplicate One-tailed T test * ? 0.05 ** ? 0.01 *** ? 0.001 The wound healing assay for migration showed that the gap for the TNC knockdown cells closed faster than shGFP control (Fig.20). At 8 hours, the cell lines had migrated approximately the same but by 16 hours the gap for KD1 and KD5 was almost closed. At 24 hours, the shGFP control cell line still had ~ 30% of the gap remaining whereas the KD cell lines were indeed closed. The potential reasons for this will be discussed later. 54   Figure 21 Representative images of 4T1 TNC KD wound healing assay (Fig. 20) at 8, 16 and 24 hours.  KD1 and KD5 had no significant differences between them. Area remaining at each time point was calculated and compared to the area calculated at time 0, directly after the scratch was made to create a fraction of gap closed.  Images presented in Fig.21 were representative of all images used in Fig.20. The KD and shGFP cells at each time point exhibited similar morphology (Fig.21). TNC Over-Expression As with the KD cell lines, the Boyden chamber migration assay was used first to assess the influence of TNC over-expression on cellular migration over a short distance. The assay was performed in 21% O2 for a period of 24 hours. The over-expression cell line had significantly fewer migrated cells (Fig.22) compared to the EV cell line, opposite as expected to the 4T1 TNC KD data in Fig.17. 55  Number of CellsEmpty VectorTNC Over Expression05 0 01 0 0 01 5 0 02 0 0 02 5 0 0* Figure 22 Boyden chamber migration assay with 4T1 TNC over-expression cell line under normoxic conditions.  Equal numbers of 4T1 empty vector control and 4T1 TNC over-expression cells were plated into each well. Assay was performed in standard incubator conditions for 24 hours. Assay was performed in triplicate. One-tailed T test * ? 0.05 Following the Boyden chamber assay, a wound healing assay was performed similar to 4T1 TNC KD would healing assay, except the hypoxic conditions were not required due to constitutive TNC expression and the oxygen levels used were 21% for 24 hours (Fig.23). The same plate was used for all images. W o u n d  H e a lin g  A s s a y  in  2 1 %  O2H o u rsFraction ofGap Closed0 8 1 2 1 6 2 40 .00 .20 .40 .60 .81 .0E m p ty  V e c to rT N C  O ve rE x p re s s io n* ** * Figure 23 Wound healing of 4T1 TNC over expressing or empty vector cells in normoxia.  Each cell line was plated in triplicate with 1x106 cells per well of a 6 well plate and allowed to adhere overnight. A scratch was created using a p200 pipette tip. Plates were placed in 21% O2 at 37?C; the same plate was used to image at each time point. Assay was performed in triplicate with two independent repeats. One-tailed T test ** ? 0.001 56  The EV cells closed the gap significantly faster than the OE cells (Fig.23). At 8 hours the EV gap was significantly more closed compared to the OE gap and this effect was maintained at the 12 hour time point. However by 16 hours, both gaps were approximately the same and at 24 hours, both gaps had completely closed (Fig.23), opposite as expected to the TNC KD wound healing assay in Fig.20. Figure 24 Representative images of 4T1 TNC over-expression and empty vector wound healing assay (Fig.23) at 8, 12 and 16 hours.  The 24 hour time point was now shown as both gaps were closed. Images presented in Fig.24 were representative all images used in Fig.23. The EV and OE cells at each time point exhibited similar morphology (Fig.24).  Recombinant TNC Soluble rTNC was used in an attempt to recreate the results seen with the 4T1 TNC OE cell line. This would help to determine if the effects seen were based upon changes in the migratory rate using TNC as a matrix or if TNC was affecting the adhesion between cells. Using 4T1 wild type cells in normoxia would prevent hypoxia induced expression of TNC during the assay. Pre-coating with rTNC would provide a matrix which 4T1 cells could migrate upon. Using 4T1 wild type cells in normoxia would reduce the production of soluble TNC that could influence inter-cellular adhesion. The concentration of TNC used was a range from 2 - 0.125 ?g.  57  W o u n d  H e a lin g  rT N CH o u rsFraction ofGap Closed0 8 1 60 .00 .20 .40 .60 .81 .021.5.2 5.1 2 50?g Figure 25 Wound healing assay of 4T1 wild type cells with recombinant Tenascin C in normoxia.  Two wells of a 6 well plate were pre-coated with 2, 1, 0.5, 0.25, 0.125, or 0 ?g total recombinant purified Tenascin C onto which 1x106 cells were plated. As this assay was performed in 21% O2 the same plate was used to image at each time point. A scratch was created using a p200 pipette tip. Assay was performed in triplicate. Pre-coating with rTNC under normoxic conditions did not change the migratory characteristics of wild type 4T1 cells (Fig. 25). Further experiments should be performed using rTNC discussed further in the future directions section. In summary, the in vitro experiments demonstrated that TNC affected the migration of 4T1 cells. TNC knockdown resulted in an increased migratory ability seen in the Boyden chamber and wound healing assay results TNC over-expression resulted in a decreased migratory ability seen in the Boyden chamber and wound healing assay results. These results were opposite to the expected hypothesis based on the known in vivo effect of TNC.   TNC In vivo Experiments  TNC was hypothesized to positively affect metastasis through enhanced migration of tumour cells. The in vitro results from invasion and migration experiments contradicted that hypothesis. However, the hypothesis was formed regarding the role of TNC in vivo so the curious results from the in vitro data does not preclude the hypothesis from still being accurate in vivo. To determine the effect of TNC knockdown and over-expression in vivo, the 4T1 KD1 and KD5 as well as the 4T1 TNC EV and OE cell lines were orthotopically 58  implanted into female BALB/c mice. Metastasis and primary tumour growth were studied for the affect that TNC had on these processes. TNC Knockdown   The 4T1 KD1 and KD5 cell lines were orthotopically implanted into the mammary fat pad of female BALB/c mice. The shGFP 4T1 cells were again used as the control. The tumours were grown for 3 weeks at which point the mice were sacrificed. At the end point, there were no significant differences in the mouse or primary tumour weights (Fig.26 a and d). Furthermore, the growth kinetics of the primary tumours was consistent between all cell lines and there were no differences in the tumour volumes at any of the measurement time points (Fig.26e). The spleen and lung weights were also comparable (Fig.26 b and c).  After sacrifice, the primary tumours were removed and whole cell lysates were created to examine the TNC knockdown of the tumours at the experimental end point. All tumours exhibited some loss of TNC knockdown, although many still had reduced TNC expression (Fig.27a). TNC can also be expressed by stromal cells within the tumour which would not be distinguishable from tumour derived TNC in whole tumour lysates. Consequently, sections of tumours were also taken and used as a secondary confirmation of TNC expression in the tumour. The 4T1 KD5 cell line had the best knockdown maintenance and TNC for 6 of 7 tumours was mostly absent. Based on the western blot, 4T1 TNC KD5 tumour #5 and 4T1 shGFP tumour #4 were chosen to confirm the western blot findings (Fig.27b).The immuno-fluorescent sections of shGFP tumour #4 (Fig.27b top panel) showed more extensive TNC deposition as well as brighter antibody staining compared to 4T1 TNC KD5 tumour #5 (Fig.27b bottom panel).    59  a)      b) M o u s e  W e ig h ts  (g )shGFPKD1KD5051 01 52 02 5N S  S p le e n  W e ig h t (g )shGFPKD1KD50 .00 .20 .40 .60 .8N S c)      d) L u n g  W e ig h ts  (g )shGFPKD1KD50 .00 .10 .20 .30 .40 .5N S T u m o u r W e ig h t (g )shGFPKD1KD50 .00 .51 .01 .5 N S e)  Figure 26 Harvested tissue characteristic of In Vivo 4T1 TNC knockdown experiment.  Physical characteristics of female BALB/c mice three weeks post orthotopic implant of 4T1 shGFP, KD1 and KD5. a) Total mouse body weight b) Spleen weight c) Lung weight d) Tumour weight e) Primary tumour growth kinetics. shGFP and KD1 n=6; KD5 n=7 One tailed T test ? Tumour; Two tailed T test ? all others.  0 200 400 600 800 1000 1200 3 7 10 14 17 21 Tumour Volume (mm3) Days shGFP TNC1 TNC5 60  a)  b)   Figure 27 Post harvest TNC knockdown tumour phenotype verification.  a) Western blot for Tenascin C of 4T1 wild type, shGFP, KD1 and KD5 tumours harvested. Tumour number indicates separate primary tumour used for each lysate. b) Immuno-fluorescent sections of primary tumours. 4T1 shGFP tumour #4 (top) and 4T1 TNC KD tumour 5-5 (bottom). Sections were stained for TNC (green) and DNA stain DAPI (blue). Gain and exposure settings were consistent for all images.  61  To study the effect that TNC knockdown in the primary tumour had upon metastasis, the percentage of CD11b+/Gr1+ cells was analyzed in the spleen and lungs. In the spleen, there was no difference in the percentage of CD11b+/Gr1+ cells between the shGFP and KD cell lines (Fig.28a). However, in the lungs, both KD1 and KD5 had a higher percentage of CD11b+/Gr1+ cells than did the shGFP cell line (Fig.28b). The percentage increase for KD5 was statistically significant, while KD1 was not (Fig.28b).  a)      b) S p le e n% CD11b+/Gr1+shGFPKD1KD502 04 06 0N S L u n g s% CD11b+/Gr1+shGFPKD1KD501 02 03 04 05 0*N S Figure 28 Flow cytometry analysis of CD11b+/Gr1+ cells from the in vivo TNC knockdown experiment  a) Percent (%) CD11b+/Gr1+ cells from the spleen b) Percent (%) CD11b+/Gr1+ cells from the lungs. KD1, shGFP n=6; KD5 n=7 Two-tailed T test * ? 0.05  The lungs were analyzed by flow cytometry for the presence of 4T1 tumour cells using positive cytokeratin staining as an identifiable marker (Fig.29). The pan-cytokeratin antibody was created using higher molecular weight antigens consistent with cytokeratins expressed by breast cancer, basal subtypes such as the 4T1 cell line. Mice which had the TNC KD1 and KD5 tumours had a significantly lower percentage of metastatic tumour cells in the lungs as a function of total lung cell number.         62     a)     b) % Cytokeratin +Cells in LungshGFPKD1KD50 .00 .51 .01 .5** Figure 29 Flow cytometry analysis of cytokeratin positive cells in lungs from 4T1 TNC knockdown experiment.  a) Representative flow cytometry gate for analysis of cytokeratin positive cells 4T1 cells in the lungs of female BALB/c mice. DNA content was determined by DAPI staining of cells. 4T1 cells are hyper-diploid. Image was generated by Nancy LePard. b) Percent (%) cytokeratin positive (tumour) cells in 4T1 KD1, KD5 and shGFP mice. KD1, shGFP n=6; KD5 n=7 One tailed T test. * ? 0.05 Based upon the level of TNC knockdown in the primary tumours at the experimental endpoint, lung sections were selected corresponding to tumours with the lowest TNC expression. Levels of TNC expression were determined by the TNC western blot in Fig.27. Lungs from the control shGFP tumour bearing mice were also stained for TNC expression. In the shGFP#5 lungs, localized areas throughout the section were positive for TNC (Fig.30 top). Conversely, when KD5-4 lungs were stained, there was no positive staining for TNC (Fig.30 bottom).   63      Figure 30 TNC expression in lungs from 4T1 TNC knockdown experiment.  Immuno-fluorescent image from mouse lung with 4T1 shGFP #5 tumour (top). Immuno-fluorescent image from mouse lung with 4T1 TNC KD5-4 tumour (bottom). Sections were stained for Tenascin C (green) and DNA stain DAPI (blue). White hatched box shows an area positive for TNC staining likely representing a metastatic tumour nodule. Gain and exposure settings were consistent for all images  In summary, TNC knockdown did not affect primary tumour size or affect spleen or lung weights (Fig.27 b-d). However, there was an increase in the percentage of CD11b+/Gr1+ 64  cells within the lungs of 4T1 KD5 tumour bearing mice (Fig.28b). In addition, TNC knockdown resulted in a significantly lower percentage of metastatic 4T1 tumour cells in the lungs (Fig.29b). Immuno-fluorescent imaging suggested that tumours which maintained the TNC knockdown at the endpoint (Fig.27) also had decreased TNC present within the lungs of these mice (Fig.30).  TNC Over-expression The 4T1 TNC EV and OE cell lines were orthotopically implanted into the mammary fat pad of female BALB/c mice. The 4T1 TNC EV cells were used as the control. Tumours were grown for 3 weeks at which point the mice were sacrificed. At the end point, there was no significant difference between the mouse weights of the EV and OE groups (Fig.31a). The lung weights were also comparable between the EV and OE groups (Fig.31c). However; the spleen and primary tumour weights of the TNC OE group were significantly increased compared to the TNC EV group (Fig.31b and d).  a)      b)  M o u s e  W e ig h t (g )EVOE051 01 52 02 5N S    S p le e n  W e ig h t (g )EVOE0 .00 .20 .40 .60 .8* c)      d)  L u n g  W e ig h t  (g )EVOE0 .0 00 .0 50 .1 00 .1 50 .2 00 .2 5N S    T u m o u r W e ig h t (g )EVOE0 .00 .20 .40 .60 .81 0* * * Figure 31 Harvested tissue characteristic of In Vivo 4T1 TNC over expression experiment.  Physical characteristics of female BALB/c mice three weeks post orthotopic implant of 4T1 TNC EV and OE cells. a) Total mouse body weight b) Spleen weight c) Lung weight d) Tumour weight. 4T1 EV and OE n=5. Tumour - one tailed T test. *** ? 0.001 Mouse, spleen and lung - two tailed T test * ? 0.05. 65  Metastatic burden was assessed using a clonogenic assay as opposed to flow cytometry based detection of high cytokeratin expressing cells as was used for the TNC KD in vivo experiment. Clonogenic assays are performed by disaggregating the lungs and plating a proportion of the lung cells. After plating, a selectable marker is added which allows only resistant cells, in this case 4T1 tumour cells, to grow. The plated cells are left for 9 -12 days where surviving cells will grow and form colonies which can be stained and counted. Each colony represents a tumour cell which was present within the lungs at the time of sacrifice. Using the number of colonies and the known proportion of the lung plated the total tumour cell number, indicative of metastatic burden in the entire lung, can be determined.  4T1 tumour cell #'s in lungEVOE1 001 011 021 031 04* Figure 32 Clonogenic assay of 4T1 TNC over-expression experiment.  Female BALB/c mice three weeks post orthotopic implant of 4T1 TNC EV and OE cells. 4T1 EV and OE n=5. One tailed T test. * ? 0.05. Mice with the TNC OE tumours had significantly more 4T1 tumour cells present within the lungs at the time of sacrifice as determined by clonogenic assay (Fig.32).  Orthotopically implanted 4T1 TNC OE cells resulted in spleen weights significantly increased than the EV controls (Fig.31b). In addition TNC OE primary tumour weights were significantly increased (Fig.31d) as well as the number of 4T1 tumour cells present within the lungs at the time of sacrifice (Fig.32). These results are consistent with the hypothesis that TNC positively affects metastasis. Moreover, the results from the TNC over-expression experiment were opposite of the TNC knock-down experiment as expected. Interestingly, the KD and OE cells yield consistently contrasting phenotypes but only the in vivo results conformed to the expected hypothesis.    66  Discussion Conditioned Media It has been previously reported that hypoxic CM is able to induce elements of the pre-metastatic niche, namely the accumulation of tumour derived secreted proteins and BMDC?s. [35, 36, 38] From these previous reports it was expected that hypoxic CM derived from 4T1 tumour cells in vitro would be sufficient to mobilize and accumulate BMDC?s, specifically CD11b+/Gr1+ cells, in the spleen and lungs of BALB/c mice. Contrary to published data, daily IP injections of hypoxic CM failed to yield accumulation of CD11b+/Gr1+ cells in the spleen or lungs of BALB/c mice when performed as previously reported. The initial protocol used was modified three times in an attempt to correct for what were perceived shortcomings. The initial protocol was daily IP injections of 300 ?l of hypoxic CM for 17 days and was consistent with Erler et al, 2009. When this failed to result in CD11b+/Gr1+ cell accumulation, the amount of secreted protein injected was believed to be too low to initiate the process. To address this, the frequency was doubled to twice daily IP injections of 300 ?l of hypoxic CM for 11 days. However, this protocol again resulted in no accumulation of CD11b+/Gr1+ cells in the spleen or lungs. The protein concentration of the hypoxic CM was again thought to be the problem as a result of insufficient cell number used to generate the CM. To better replicate the in vivo conditions and the cell number of an actual tumour, ex vivo 4T1 tumours were used as the source for CM. The protocol was changed back to daily IP injections of 500 ?l for 13 days. There was no change in spleen weight when compared to the 500 ?l media only IP injections (Fig.7a). However, the percentage of CD11b+/Gr1+ cells within the spleen decreased (Fig.7b). This change did not yield any difference in the percentage of CD11b+/Gr1+ cells within the lungs (Fig.7c), which was the result that would indicate the potential formation of a pre-metastatic niche. The last attempt to modify the protocol was the use of protein concentration columns to concentrate the CM prior to injection. This method enabled an entire 10 cm tissue culture plate to be concentrated into 300 ?l of injected CM whereas previously, one 10 cm tissue culture plate was shared for all mice receiving IP injections with that specific CM. Normoxic and hypoxic CM resulted in significantly increased spleen weights versus na?ve controls receiving no IP injections (Fig.8a) indicating that the CM caused a low level response. The number of CD11b+/Gr1+ cells within the spleen increased significantly in the mice which received hypoxic CM (Fig.8c) further suggesting that the hypoxic CM was eliciting a response.  67  The failure of hypoxic CM to cause an accumulation of CD11b+/Gr1+ cells in the lungs could be due to a number of factors. Necessary secreted proteins were absent or not produced by the 4T1 cells in vitro; protein concentrations were too low; factors were absent from the CM or a trait of BALB/c mice preventing accumulation. CM generated in vitro is assumed to produce the same secreted proteins as would be seen in vivo. However, a tumour cell alone in culture is greatly simplified compared to growth in vivo. The signalling or stimulation required by tumour cells to produce the full physiological secretome may not be present. This could result in key proteins for the accumulation of CD11b+/Gr1+ cells in BALB/c mice to be absent. Protein concentrations could have remained too low to elicit CD11b+/Gr1+ cell accumulation. This concern was addressed, although adequate concentration may not have been achieved in the CM. If the level of granulocyte colony stimulating factor (G-CSF), for example, was too low, cell mobilization could have been impeded. The higher protein concentration as a result of the concentration columns could explain the increased spleen size and CD11b+/Gr1+ cell number within the spleen treated with hypoxic CM if G-CSF was increased (Fig.8c).[37] There could also be factors absent which could be responsible for CD11b+/Gr1+ cell mobilization. Hypoxic 4T1 tumours develop necrotic regions [24] which could play a role in immune system activation required for CD11b+/Gr1+ cell mobilization.  Lastly, there could be an intrinsic difference within the BALB/c mouse line that prevents the mobilization or accumulation CD11b+/Gr1+ cells in metastatic organs by CM. Reviewing previously published data, all CM experiments were performed in mouse backgrounds which were different from the BALB/c background of 4T1 tumours. The lack of response could be due to some unknown characteristic which prevented accumulation of CD11b+/Gr1+ cells in the lungs. Select protocols were chosen to be repeated in the NOD/scid mouse cell line as Erler et al successfully used the immuno-compromised nude (nu/nu) mouse for their CM experiments. The NOD/scid mice lack a complete and functional immune system characterized by absent mature lymphocytes and reduced NK cell activity. No published reports of conditioned medium resulting in the accumulation of BMDC?s or CD11b+/Gr1+ cells have used the BALB/c mouse strain.[35, 36, 38] 68  NOD/scid mice received hypoxic CM injections as described previously. Results from both experiments show consistent data including an increase in percentage of CD11b+/Gr1+ cells in the spleen and lungs when compared to media only injections (Fig.9 and 10). Normoxic CM was not used during these experiments; consequently, the effect seen was not able to be directly attributed to hypoxic CM. Instead, the observed response was only attributed to CM injections in general. However, the successful accumulation of CD11b+/Gr1+ cells in the spleen and lungs of NOD/scid mice supports the theory the BALB/c background could be less responsive to CM IP injections. While switching to the NOD/scid mouse strain resulted in CD11b+/Gr1+ cell accumulation, further experiments will be performed using the BALB/c mouse strain. 4T1 cells are syngenic with BALB/c mice and when analyzing the effects of TNC on metastasis it is beneficial to use a mouse model with a functional immune system to yield results most similar and applicable to human breast cancer metastasis.   S.I.L.A.C. Proteomics The results from the CM experiments did not conclusively show that hypoxic 4T1 CM was required to yield an accumulation of CD11b+/Gr1+ cells in the spleen and lungs. However, in the NOD/scid mice, there was an accumulation of CD11b+/Gr1+ cells within the spleen and lungs indicating that the CM created similar conditions to those observed during the metastatic process of the 4T1 tumour cell line. The SILAC proteomic experiment aimed to identify the secreted proteins present within the 4T1 CM, both normoxic and hypoxic. In doing so, identification of the protein(s) responsible for the mobilization and accumulation of CD11b+/Gr1+ cells in NOD/scid mice could be identified as well as other secreted proteins involved in migration, invasion or metastasis.  Utilizing SILAC, quantitative comparisons of the normoxic versus hypoxic 4T1 CM was available as well as qualitative comparisons between 4T1 and 67NR CM. As stated earlier, identifying secreted proteins up-regulated in hypoxic 4T1 CM and absent from 67NR CM could represent potential secreted proteins with metastatic roles and provide insight into the correlation between tumour hypoxia and metastasis. The results from the SILAC data needed to be validated with a secondary method prior to selection of candidate secreted proteins. Protein identification as well as the hypoxic induction ratio of 4T1 tumour cells in vitro was independently confirmed using western blots. Candidate metastasis associated proteins were selected based on level of hypoxic induction, absence from the 67NR CM, and any known role in cancer or metastasis. TNC 69  and periostin were selected from the results based on a greater than 2 fold increase of expression during hypoxia, absent from the 67NR results and previously shown roles in cancer (Table.1). Western blots were performed to confirm the 3.154 and 16.625 fold inductions for periostin and TNC, respectively. The western blots in Fig. 10 show that both periostin and TNC were induced in hypoxia. TNC has also been previously shown to be induced by 4T1 cells in hypoxia at the mRNA level while not induced in the 67NR cell line.[24] The western blots do not show the 3.154 and 16.625 fold induction as seen in the SILAC proteomic results, however the relative level of induction is confirmed. TNC was considerably more induced in hypoxia as compared to periostin, which was consistent with the SILAC data. Based on these criteria, TNC was selected as the best candidate secreted protein to examine further regarding a metastatic role. Tenascin C Expression Upon selection of TNC, protein expression characterization under hypoxic and normoxic conditions in the 4T1 cell line was performed. Previous secreted protein western blots did not contain traditional western blot loading controls. Review of published methods regarding secreted protein western blots indicated no suitable secreted protein with stable and consistent expression for use as a loading control. Instead, the cell number plated to generate the CM and a Bradford total protein quantification to enable equal protein loading per well were used as controls. In addition, the cell lysate with tubulin as a loading control was included (Fig.12).  Secreted TNC produced two bands on the film, which are consistently generated in all samples produced by the 4T1 cell line; there is a lower band at 250 kDa and an upper band at 270 kDa (Fig.11 and 12). The upper band had a greater expression level than the lower band. This doublet expression is consistent with published western blots of TNC expression.[91, 104] The multiple bands are presumed to be a product of alternatively spliced isoforms, although post-translational glycosylation also occurs.[68, 104, 105]  TNC could not be detected in 4T1 cell lysates (Fig.12). This lack of TNC could be due to protein epitope degradation preventing proper antibody detection or rapid transit and secretion of TNC through the ER and golgi complex resulting in levels below threshold detection. To elucidate this, a protein transport inhibitor, BD GolgiPlug TM, was used. This inhibitor contains Brefeldin A which blocks intracellular protein transport and causes proteins to build up in the Golgi apparatus, preventing secretion. In Fig.13, samples which had 70  Brefeldin A added produced a band in the lysates and a reduced band in the CM sample, due to build up of intra-cellular TNC. For the samples which did not have Brefeldin A added, a strong band remained in the CM sample and none present in the lysate. It is clear that the antibody is capable of detecting the protein inside the cell and the lack of detection in untreated hypoxic cells is due to low intracellular protein concentration. Tenascin C Knockdown and Over-expression To further examine the role of TNC, knockdown and over-expression cell lines were created. TNC knockdown was performed using shRNA constructs in the 4T1 cell line given the high TNC expression induced by hypoxia. The 67NR cell line lacked expression in normoxic and hypoxic conditions (Fig.11 top). Five constructs were tested and the best two were chosen, designated TNC shRNA knockdown (KD) 1 and 5 (Fig.14a). ShRNA specific for GFP was used as a transduction control. The TNC expression of 4T1 shGFP control cells was similar to the level in WT 4T1 cells and was used as the control for the TNC KD 1 and 5 (Fig.14b).  The shRNA TNC KD1 and KD5 cell lines effectively inhibited the expression of the heavier TNC isoform at 270 kDa (Fig.14a). However, the lighter and less abundant 250 kDa TNC isoform was still present in KD1 and KD5 samples for both the normoxic and hypoxic samples. Given the presence of the lighter band in the normoxic conditions from both KD1 and KD5 cell lines, the lentiviral transduction may induce expression of the smaller isoform from these cell lines. If this was the case, it would also explain the presence of the smaller isoform which is present in the hypoxic samples. Despite the presence of the smaller band, the larger isoform is the predominate isoform produced by the 4T1 cell line and knockdown represents a reduction of the majority of the TNC protein produced. The 4T1 shGFP cell line had isoform expression similar to 4T1 WT (Fig.13b). The constitutive TNC over-expression construct was transduced into the 4T1 cell line which effectively produced a TNC protein in normoxia which was detectable by the same antibody used previously (Fig.15). The 4T1 OE cell line had a new isoform expression for secreted TNC. There are three TNC isoforms produced at approximately 250, 270 and 290 kDa. The isoform predominately produced was at 250 kDa, in contrast to the major isoform at 270 kDa seen in WT 4T1 cells. The 290 kDa isoform had not been seen in the western blots of TNC prior to this point.  TNC OE cells had TNC present within the lysates similar to the Brefeldin A western blot (Fig.15). The lysate expression is likely due to TNC being produced at a rate which exceeds 71  the ability to secret it and therefore a build up occurred within the cell. The TNC in the lysate was not present in the EV cell lysate further suggesting the TNC in the OE cell lysate was a result of an inability to adequately secret TNC. Expression of TNC by the EV cell line was similar to 4T1 WT expression during normoxic and hypoxic conditions (Fig.15 and 12 respectively). Recombinant TNC had three isoforms present (Fig.16). Two isoforms at 270 and 290 kDa were present equally while the third isoform at 250 kDa was present in very low levels. These three bands have all been seen previously although this was the first time the 290 kDa isoform was seen in an abundant level. The exon composition and impact of the resulting isoform size was not determined in this research, although it has been shown that the larger TNC isoforms correlate with poor patient prognosis.[68] In breast cancer, full length TNC and large isoforms containing the alternatively spliced Fn type III repeats were not present in normal breast tissue but are highly expressed in mammary carcinomas and correlated with increased migration and proliferation in vitro.[106, 107] TNC isoforms with alternatively spliced repeats are thought to be required for the focal adhesion disassembly.[78] Specifically, alternatively spliced TNC containing exons 14/16 and exon 16 alone were shown to have enhanced proliferation and invasion as well as suggested to be important for migration.[93] TNC isolated from tumour biopsy samples resulted in three TNC isoforms at molecular weights of 350, 250 and 210 kDa.[106] The TNC isoforms shown in all western blot figures are consistent with the molecular weights of previously reported full length TNC protein monomers. MDA-MB-231 cells stimulated with TGF? induced an isoform of TNC 350 kDa in weight suggesting TNC produced by 4T1 cells to be less than a full length transcript.[96] Migration and Invasion Assays We assessed the influence of TNC on cellular migration in vitro using a Boyden chamber migration assay given the known function of TNC in modifying cellular adhesion in vivo (Fig.16). The 4T1 KD1, KD5 and shGFP cell lines were used. The wells were pre-incubated with hypoxic CM produced by the cell line which would be plated for that particular well, i.e. wells plated with 4T1 KD1 cells were pre-incubated with hypoxic CM produced from 4T1 KD1 cells. The pre-incubation would allow secreted proteins in the CM to adhere to the porous membrane prior to the start of the migration assay.  72  At the end of 24 hours in 21% O2 the 4T1 KD5 cell line had significantly more cell migration across the membrane compared to the shGFP control cells. The 4T1 KD1 resulted in a similar trend although not statistically significant with a p-value of 0.056. A two-tailed T-test was used to analyze the results because the hypothesis regarding TNC function was based on published in vivo results and the in vitro effect was uncertain.  A Boyden chamber invasion assay was performed next (Fig.18). This assay is similar to the Boyden chamber migration assay with the addition of a layer of reconstituted basement membrane called Matrigel on top of the porous membrane. Again, the wells were pre-incubated with hypoxic CM from the respective cell lines to which the cells were plated. After 24 hours in 21% O2 there was no effect seen by either the KD1 or KD5 cell line compared to the shGFP cell line as determined by two-tailed T-test.  The Boyden chamber invasion assay was repeated without hypoxic CM pre-incubation and instead the assay was performed entirely in 1% O2 for 24 hours (Fig.19). Again, no effect of TNC knockdown was observed compared to the 4T1 shGFP cell line. There was no effect of TNC knockdown when invasion was assessed but TNC knockdown did have an effect on Boyden chamber migration. To further examine that effect, a wound healing assay was performed (Fig.20). The 4T1 KD1 and 5 cell lines were again compared to the 4T1 shGFP cells. Since TNC has minimal expression under normoxic conditions, the assay was performed in 1% O2 to examine the effect of reduced TNC expression on migration during hypoxic conditions. At the 8 hour time point, the migration of all three cell lines was comparable. However, at the 16 and 24 hour time points, the KD1 and KD5 cell lines had closed the gap significantly faster than the shGFP cell line. TNC significantly decreased the amount of time required for the 4T1 KD1 and KD5 cells to close the gap in a wound healing assay The lag in effect could be attributed to the delay in TNC production in hypoxic conditions. Since the media was changed immediately after the scratch was made in the monolayer of cells, there would be no TNC present. As a result, when the cells were placed in hypoxia there would be a delay in TNC expression as residual oxygen in the media must dissipate prior to HIF1 transcription factor stabilization. Once stabilized, TNC must still be transcribed, translated, secreted and accumulate to a functional level before any effect would be seen. This could be the reason that no effect was observed at the 8 hour time point. 73  The same migration assays were performed with the 4T1 TNC OE cell line. Given the lack of results seen with the knockdown cell lines, the Boyden chamber invasion assay that was performed with the 4T1 TNC KD cell lines was not performed with the TNC OE cell line. Given that TNC knockdown resulted in an increased migratory phenotype in the 4T1 cells, the 4T1 TNC over-expression cells were expected to have a decrease in migratory ability.  The Boyden chamber migration assay was performed in 21% O2 since the OE cells do not require hypoxia to express TNC. After 24 hours, the 4T1 TNC OE cells had significantly fewer cells migrate across the porous membrane (Fig.22). A one-tailed T-test was used in the analysis as the KD cell lines enabled a prediction as to how the OE cells would migrate. Indeed, the 4T1 OE cells had the converse effect to the 4T1 KD cells.  The wound healing assay was next performed with the 4T1 TNC OE and the 4T1 EV cell lines (Fig.23). A 12 hour time point was added for the TNC OE wound healing assay to gather a more complete assessment of the migratory ability. At the 8 hour time point, the 4T1 TNC OE cells had closed the gap slower than the 4T1 EV cells. This effect was also seen at the 12 hour time point. The effect at both time points had p-values less than 0.01 which were more significant than the p-value obtained from the Boyden chamber migration assay. The effect was not present at the 16 or 24 hour time point.  Contrary to the TNC KD wound healing assay, the effect was present at the 8 hour time point for the TNC OE cells. This could be due to constitutive expression of TNC in the 4T1 TNC OE cell line which allowed the effect of TNC to be evident sooner.  In an attempt to better understand the role of TNC in altering the migratory phenotype of 4T1 cells, recombinant TNC (rTNC) was used in a wound healing assay (Fig.25). 4T1 wild type cells were used in normoxia where TNC expression is present in low levels only. Pre-coating the surface used in the assay with rTNC could mean that an effect was due to altered migration as opposed to soluble TNC actively secreted by cells causing a change in interaction between the cells. No effect on the migratory ability was observed as a result of the assay. The rTNC isoforms present differed from the major isoform expressed by the OE cells which could be responsible for the lack of effect. However, the lack of effect could also be due to an insufficient or incorrect coating of the plate by rTNC. Testing of rTNC plate coating should be performed using antibody detection similar to an ELISA to determine whether rTNC had effectively coated the surface (refer to future directions). Alternatively, the 74  lack of effect from the rTNC could be an accurate observation due to some unknown difference between the rTNC and the other TNC forms produced the 4T1 cells. The concentration of rTNC could have also been too low. Concentrations of 20 ? 60 ?g/ml have been used previously to alter focal adhesion integrity of endothelial cells.[79] In vitro TNC appears to have a negative effect on the migratory phenotype of 4T1 cells. When knocked down, TNC enhances the migratory ability, while over-expression decreases that ability. The use of rTNC did not yield any effect on the 4T1 cell migration. These results contradict the hypothesis that TNC enhances migration of tumour cells. However, that hypothesis was generated with regard to the in vivo effect and while these results suggest an opposite role, they do not preclude the hypothesis from being accurate in a mouse model. This is not the only example of 4T1 cells behaving unpredictably in vitro. It has been reported that 4T1 cells have EMT markers in vitro opposite to expected mesenchymal markers.[108] Perhaps, the 4T1 cell line responds in unexpected ways during in vitro assays and could be a possible explanation for the contrary results observed. Tumour cell interactions in vivo are much more complex than what is presented in vitro. Interaction of tumour cells with other stromal cells provides significant signalling which is absent during in vitro assays. Moreover, cytokine and chemokine molecules absent in vitro could have significant effects on the phenotype of tumour cells observed. For these reasons, establishment of in vitro conditions as close to in vivo as possible would be beneficial. Future experiments should include determination of the mechanism by which TNC affects migration in vitro and creating conditions more physiologically relevant (refer to future directions).  Conflicting results are not uncommon for TNC. The in vitro role of TNC is not universal and can be pro-migratory for one cell type and anti-migratory for another. [109] How a cell reacts to TNC can depend upon the environmental context or the other ECM proteins present.[109] In breast cancer, TNC has been primarily shown to be pro-migratory. Knockdown of TNC using shRNA in MDA-MB-435 resulted in impaired migration during wound healing assays.[103] Similarly, anti-TNC antibodies inhibited the migration of MV630, a spontaneous breast cancer cells derived from a transgenic mouse.[103] There has been little published regarding the role of TNC with specific regard to 4T1 tumour in vitro. However, one report showed no enhanced migration by 4T1 cells upon a TNC coated membrane in a Boyden chamber migration assay.[104] Although, this report also indicated a minimal migratory 75  ability by 4T1 cells which contradicts the migratory phenotype shown by 4T1cells in this research.[104]  The pro-migratory in vitro effect is consistent among varying cancer types. Pancreatic tumour cell lines also exhibit an overall enhanced migration during wound healing assays when TNC had been pre-coated.[110] Interestingly, one of the three cell lines used, displayed inhibited migration during the wound healing assays highlighting the unpredictable nature of TNC in vitro.[110]  Tenascin C In Vivo The proposed role of TNC was based on the in vivo function of the protein and hypothesized to enhance metastasis through enhanced tumour cell migration. 4T1 TNC knockdown cell lines were orthotopically implanted into BALB/c mice with the expectation that the lack of TNC expression would reduce the metastatic burden in the lungs through decreased tumour cell migration.  Whole tumour lysates were created to examine the expression level of TNC in the primary tumour at the experimental endpoint. Lentiviral transduced tumour cells should stably express the shRNA construct and therefore result in a primary tumour with reduced TNC expression. Fig. 27 shows TNC expression in the 4T1 KD5 cell line to be reduced compared to the shGFP controls, especially in tumours 4 through 7. However, the 4T1 KD1 cell line had TNC levels comparable to the 4T1 shGFP controls. Potential reasons for loss of TNC knockdown include tumour cells having lost expression of the shRNA, a sub population of the implanted tumour cells lacking the shRNA being positively selected for growth within the tumour or finally that the tumour stroma produce TNC. Published results have indicated that TNC is commonly found in the stroma of primary breast tumours observed by immuno-fluorescence.[111] However, if the 4T1 tumours are inducing stromal production of TNC, expression would be expected in all tumours which was not seen. Although TNC expression was found in 9 of 13 knockdown tumours it is not possible to determine when the TNC suppression was lost. If the effect of TNC on metastasizing tumour cells occurred early and TNC knockdown was lost late in the experiment then it is possible that the effect of TNC knockdown would still be detectable despite the whole tumour lysate indicating a loss of TNC knockdown. To verify the TNC knockdown of these tumours observed with western blot, immuno-fluorescent sections of primary tumour were probed for TNC and DAPI, a nuclear stain. The 76  results of shGFP tumour #4 and KD5 tumour #5 indicate that the TNC levels shown in the western blot are consistent with the TNC expression indicated by the IF images. KD5 tumour #5 showed that TNC deposition was less extensive and the intensity of the antibody staining was reduced in the knockdown tumour section. To measure the effect of TNC on metastasis, the lungs were harvested and processed to look for the presence of tumour cells. 4T1 cells are hyper diploid and express elevated levels of cytokeratin. Flow cytometry can be used to identify 4T1 tumour cells with a high DNA ploidy and high cytokeratin expression. There are stromal cells which express similar levels of cytokeratin to 4T1 cells. During replication these cells will have DNA content similar to 4T1 tumour cells as well as high cytokeratin expression. Tumour cell gates will include these stromal cells but inclusion will be proportional among all mice (Fig.29a). 4T1 KD1 and KD5 cell lines both had a statistically significant reduction in the percentage of cytokeratin positive cells within the lungs (Fig.29b). This represents fewer 4T1 tumour cells present in the lungs of 4T1 tumour bearing mice at the end of three weeks. This is consistent with the hypothesis that TNC enhances metastasis. Quantification of metastases could be improved by utilizing alternative techniques of tumour cell identification. Clonogenic assays would provide enhanced detection of tumour cells when lung tumour burden is low. Alternatively, 4T1 tumour cells expressing luciferase would allow quantification of tumour burden using IVIS imaging. 4T1 cells with luciferase activity produce light as a by product of the enzymatic reaction. This light can be used to quantify the 4T1 tumour burden within the lungs.  Using the TNC whole tumour western blot, lungs corresponding to primary tumours with low TNC levels were chosen to be stained for TNC expression. The lungs of mice KD5-4 through KD5-7 were selected as acceptable based on low levels of TNC within the tumours; any of the shGFP lungs were also acceptable for use as controls. Lungs from the shGFP5 mouse resulted in localized regions positive for TNC throughout the entire lung section (Fig.30 top). Lungs from KD5-4 did not have any positive regions for TNC (Fig.30 bottom). The lack of TNC in the KD5-4 lungs could explain the reduced tumour cell percentage observed in the lungs of mice implanted with the TNC KD5 tumour cell line. However, to confirm this, step sections from all shGFP lungs would need to be stained for the presence of TNC and the absence of TNC would need to be confirmed in all TNC KD5 lungs. At present, only sections from select shGFP and TNC KD5 lungs have had TNC deposition determined. In addition, it would be interesting to identify locations of metastases in context of regions with positive TNC expression to see if co-localization occurs.  77  It was not expected that TNC knockdown would have an effect on CD11b+/Gr1+cell populations. There were no trends in the CD11b+/Gr1+cell populations within the spleen (Fig.28a). However, the lungs of KD5 tumour bearing mice had a significantly higher percentage of CD11b+/Gr1+cells (Fig.28b). A two-tailed T-test was used since there was no hypothesized effect of TNC on CD11b+/Gr1+cell populations. Further examination for the increase of CD11b+/Gr1+cells could be performed by isolating these cells and performing in vitro migration assays to determine the result of TNC knockdown.   Overall, TNC knockdown did not affect the primary tumour but resulted in a decreased number of metastatic tumour cells within the lungs. The decreased level of tumour cells is potentially due to reduced TNC within the lungs affecting the migratory ability of 4T1 tumour cells in vivo. Knockdown of TNC in vivo yielded results consistent with hypothesized role. Further examination of the role of TNC included orthotopic implantation of 4T1 cells which over-expressed TNC into BALB/c mice. 4T1 TNC OE tumour bearing mice had spleen weights significantly increased compared to the EV control (Fig.31b). In addition, the primary tumour weights were also significantly increased (Fig.31d). Total mouse weights and lung weights were comparable (Fig.31 a and c). The increased spleen and tumour weights were not seen during the TNC knockdown experiment (Fig.26). However, TNC has been reported to affect primary tumour size previously.[99] The increased spleen weight could be attributable to the increased tumour weight (Fig.5c).  To examine the metastatic burden in the lungs, a clonogenic assay was used. As discussed previously, clonogenic assays are a better method of assessing low tumour cell numbers within a metastatic organ. The TNC OE tumour bearing mice had significantly higher numbers of tumour cells present within the lungs at the endpoint (Fig.32). However, the cause of the increase could be due to either TNC having an effect within the lungs or as a result of the increased primary tumour weight resulting in a more metastatic tumour. Larger tumours would be more hypoxic, produce more hypoxia induced proteins, mobilize increased numbers of BMDC?s and shed more tumour cells from the primary tumours, all resulting in a more metastatic phenotype.  In vivo, the TNC KD and OE cell lines have opposite phenotypes, where TNC KD repressed metastasis and TNC OE enhanced metastasis. In addition, TNC OE also resulted in enhanced primary tumour growth. Together the 4T1 TNC KD and OE cells in vivo were 78  consistent with the hypothesis that TNC promotes metastasis. However, it is still unclear if this promotion is through enhanced tumour migration in vivo as hypothesized.  Published literature regarding the effect of TNC in vivo consistently reports enhanced primary tumour growth and metastasis. However, the mechanism or role attributed to TNC?s affect is diverse. In MDA-MD-435 tumour cells, TNC knockdown resulted in decreased primary tumour size, reduced relapse tumour volume, and fewer metastatic nodules.[103] Differences in the effect of TNC among breast cancer cell lines do still exist. In MDA-231-LM2 and CN34-LM1, TNC knockdown did not affect primary tumour size while metastatic burden was still reduced.[99] That data is similar to the results presented in this thesis with the 4T1 KD cells ; no affect in primary tumour growth and a reduction in metastasis. As with the in vitro data, there have been very few reports published regarding TNC and the 4T1 cell line in vivo. Given the ability of TNC to play a role in adhesion and anti-adhesion given the context of expression, this could be a possible explanation for the contradictory results seen in vitro and in vivo. TNC with regards to migration exerts a positive role through an alteration in cellular adhesion with other ECM proteins such as FN. In vivo, TNC could be pro-migratory and pro-metastatic through reduced-adhesion of 4T1 tumour cells with FN and other ECM proteins. Conversely, in vitro there are no ECM proteins present. As a result, TNC could be acting in an adhesive fashion. This would explain the increased migratory ability when TNC was knocked down and a decreased migratory ability when TNC was over-expressed. If indeed TNC was acting in both fashions it could explain how TNC was inhibiting migration in vitro while promoting metastasis presumably through promoting migration in vivo. 79  Conclusions Contrary to published data regarding the effects of tumour cell CM, the 4T1 tumour cell line was not capable of inducing CD11b+/Gr1+cell accumulation in the metastatic organ, the lungs, of BALB/c mice. However, 4T1 CM injected into NOD/scid mice resulted in an accumulation of CD11b+/Gr1+cells in the spleen and lungs. These results indicate that 4T1 CM can induce BMDC accumulation similar to metastatic niches but only in a mouse strain specific manner.  Proteomic examination of the 4T1 CM yielded candidate metastasis associated proteins which were hypoxia inducible and produced only by the metastatic 4T1 tumour cell line when compared to a genetically related non-metastatic 67NR tumour cell line. Tenascin c (TNC) was selected as a potential candidate protein for further study based on the high level of hypoxic induction, known to be a metastasis associated factor. TNC was hypothesized to increase the metastatic ability of 4T1 cells through enhancing migration in vivo.  In vitro examination revealed TNC decreased the migratory ability of 4T1 cells in multiple migration assays. This contradicts the hypothesized in vivo effect of TNC on tumour cells. However, TNC in vivo promoted metastasis yielding an increased metastatic burden in the lungs consistent with the hypothesized role.   Although the in vitro and in vivo findings prove contradictory here, published data are in agreement with the in vivo results of TNC having a pro-metastatic role. Moreover, published data regarding the in vivo role of TNC supports the potential mechanism of action being related to migratory enhancement.  Given the role in enhancing metastatic progression, TNC represents a therapeutic target to prevent metastasis. In addition, given the restricted expression in healthy adult tissues, it provides unique specificity for tumour cell associated niches, potentially resulting in reduced treatment related side effects.     80  Future Directions Given the inconclusive results obtained from the rTNC wound healing assay determination of adequate coating should be performed. Antibody based detection similar to an ELISA assay would be suitable to confirm rTNC coating. A wound healing assay should be repeated once coating has been established. As the rTNC wound healing assay was meant to determine the potential affect of TNC on migration an alternative approach would be to allow 4T1 wild type cells to adhere and add soluble rTNC after the scratch is made. This would test the affect of TNC on inter-cellular adherence where pre-coating with rTNC would test the affect TNC had on migration. These migration assays should also have the maximum concentration of rTNC increased.  In vitro results of TNC on migration were contradictory to the in vivo hypothesis. To address this, physiologically relevant matrices should be employed during in vitro assays. Migration assays should be repeated including pre-coating with ECM proteins such as Fn or collagen IV. This would allow TNC to function in a manner more similar to the expected role in vivo. Incorporation of physiological serum obtained from 4T1 tumour bearing mice for use during in vitro assays would further increase the relevance of in vitro migration. Mechanistic studies of TNC upon tumour cells in vitro should be performed. Given the adhesion modulation TNC is expected to have, examination of focal adhesions during migration could provide insight into how TNC alters migration. Characterization of the TNC protein produced by 4T1 tumour cells to determine which Fn type III exons are present given the differing isoform effects. This would be of particular interest to determine the differences between the isoforms produced by the wild type, the small isoform from the KD cell lines and the three isoforms produced by the OE cell line. Creation of 4T1 cells which express TNC isoforms different from the cell lines generated in this research would also be of benefit. Specifically, TNC containing the alternatively spliced B and D mRNA exons if not currently expressed could be used to assess the influence of isoforms on migration. These in vitro experiments could help to tease out the cause of the contradictory results previously obtained.  Further examination of the TNC OE tumour bearing mice should include data analysis similar to that of the TNC KD tumour bearing mice. This would include: CD11b+/Gr1+cell flow cytometry of the spleen and lungs, endpoint TNC western blots of the primary tumours and immuno-fluorescent examination of the primary tumour and lungs.  81  The unexpected observation of increased CD11b+/Gr1+cells in the lungs during the TNC KD experiment should be explored further. In vitro migration assays of cells isolated from tumour bearing mice in the presence and absence of TNC should be performed to see if TNC also affects the migration of cell types other than tumour cells.  Given that TNC KD reduced metastatic burden and TNC OE increased metastatic burden, a mechanism for the observed phenotype should be pursued. Examination of how and when TNC transits from the primary tumour to metastatic sites would be beneficial. Immuno-fluorescent staining of TNC in the lungs as it relates to tumour cell and CD11b+/Gr1+cell location in both TNC KD and OE tumour bearing mice should be investigated. In addition, examination of TNC expression in the mammary fat pad surrounding the tumour and especially at invasive fronts should be undertaken. Furthermore, given the known expression of TNC by stromal cells, identification of TNC expressing cells both in the primary tumour and metastases should be done. These studies should include endpoints prior to three weeks to see if tumour and metastasis progression alters TNC expression. Expansion of the tumour cell lines used to include human breast cancer models such as the MDA-MB-231 and a cell line generated from a metastatic tumour such as MDA-MD2313-LM2 would improve the clinical significance of TNC as a potential therapeutic target. The potential of TNC to be a therapeutic target as well as identification of targeting modalities should be investigated. Antibody based therapies would be a potential method to neutralize or block TNC function thereby reducing metastatic burden although pursuing other options such as small molecule inhibitors should also be done.   82  Bibliography 1. Toronto, O., Canadian Cancer Statistics 2013. Canadian Cancer Society's Advisory Committee on Cancer Statistics, Canadian Cancer Society; 2013. 2. Chaffer, C.L. and R.A. Weinberg, A perspective on cancer cell metastasis. Science, 2011. 331(6024): p. 1559-64. 3. Thomlinson, R.H. and L.H. Gray, The histological structure of some human lung cancers and the possible implications for radiotherapy. Br J Cancer, 1955. 9(4): p. 539-49. 4. Brizel, D.M., et al., Tumor hypoxia adversely affects the prognosis of carcinoma of the head and neck. Int J Radiat Oncol Biol Phys, 1997. 38(2): p. 285-9. 5. Hockel, M., et al., Intratumoral pO2 predicts survival in advanced cancer of the uterine cervix. Radiother Oncol, 1993. 26(1): p. 45-50. 6. Vergis, R., et al., Intrinsic markers of tumour hypoxia and angiogenesis in localised prostate cancer and outcome of radical treatment: a retrospective analysis of two randomised radiotherapy trials and one surgical cohort study. Lancet Oncol, 2008. 9(4): p. 342-51. 7. Hockel, M. and P. Vaupel, Tumor hypoxia: definitions and current clinical, biologic, and molecular aspects. J Natl Cancer Inst, 2001. 93(4): p. 266-76. 8. Vaupel, P., et al., Oxygenation of human tumors: evaluation of tissue oxygen distribution in breast cancers by computerized O2 tension measurements. Cancer Res, 1991. 51(12): p. 3316-22. 9. Hockel, M., et al., Oxygenation of carcinomas of the uterine cervix: evaluation by computerized O2 tension measurements. Cancer Res, 1991. 51(22): p. 6098-102. 10. Carreau, A., et al., Why is the partial oxygen pressure of human tissues a crucial parameter? Small molecules and hypoxia. J Cell Mol Med, 2011. 15(6): p. 1239-53. 11. Sun, J.D., et al., Selective tumor hypoxia targeting by hypoxia-activated prodrug TH-302 inhibits tumor growth in preclinical models of cancer. Clin Cancer Res, 2012. 18(3): p. 758-70. 12. Vaupel, P., et al., Hypoxia in breast cancer: role of blood flow, oxygen diffusion distances, and anemia in the development of oxygen depletion. Adv Exp Med Biol, 2005. 566: p. 333-42. 13. Harris, A.L., Hypoxia--a key regulatory factor in tumour growth. Nat Rev Cancer, 2002. 2(1): p. 38-47. 14. Perou, C.M., et al., Molecular portraits of human breast tumours. Nature, 2000. 406(6797): p. 747-52. 15. Strehl, J.D., et al., Invasive Breast Cancer: Recognition of Molecular Subtypes. Breast Care (Basel), 2011. 6(4): p. 258-264. 16. Langlands, F.E., et al., Breast cancer subtypes: response to radiotherapy and potential radiosensitisation. Br J Radiol, 2013. 86(1023): p. 20120601. 17. Eccles, S.A., The epidermal growth factor receptor/Erb-B/HER family in normal and malignant breast biology. Int J Dev Biol, 2011. 55(7-9): p. 685-96. 18. Badve, S., et al., Basal-like and triple-negative breast cancers: a critical review with an emphasis on the implications for pathologists and oncologists. Mod Pathol, 2011. 24(2): p. 157-67. 19. de Ruijter, T.C., et al., Characteristics of triple-negative breast cancer. J Cancer Res Clin Oncol, 2011. 137(2): p. 183-92. 20. Kaur, P., et al., A mouse model for triple-negative breast cancer tumor-initiating cells (TNBC-TICs) exhibits similar aggressive phenotype to the human disease. BMC Cancer, 2012. 12: p. 120. 83  21. Bao, L., et al., Increased expression of P-glycoprotein is associated with doxorubicin chemoresistance in the metastatic 4T1 breast cancer model. Am J Pathol, 2011. 178(2): p. 838-52. 22. Dexter, D.L., et al., Heterogeneity of tumor cells from a single mouse mammary tumor. Cancer Res, 1978. 38(10): p. 3174-81. 23. Aslakson, C.J. and F.R. Miller, Selective events in the metastatic process defined by analysis of the sequential dissemination of subpopulations of a mouse mammary tumor. Cancer Res, 1992. 52(6): p. 1399-405. 24. Lou, Y., et al., Targeting tumor hypoxia: suppression of breast tumor growth and metastasis by novel carbonic anhydrase IX inhibitors. Cancer Res, 2011. 71(9): p. 3364-76. 25. Valastyan, S. and R.A. Weinberg, Tumor metastasis: molecular insights and evolving paradigms. Cell, 2011. 147(2): p. 275-92. 26. Chambers, A.F., A.C. Groom, and I.C. MacDonald, Dissemination and growth of cancer cells in metastatic sites. Nat Rev Cancer, 2002. 2(8): p. 563-72. 27. Wu, Y. and B.P. Zhou, New insights of epithelial-mesenchymal transition in cancer metastasis. Acta Biochim Biophys Sin (Shanghai), 2008. 40(7): p. 643-50. 28. Hanahan, D. and R.A. Weinberg, Hallmarks of cancer: the next generation. Cell, 2011. 144(5): p. 646-74. 29. Midwood, K.S., et al., Advances in tenascin-C biology. Cell Mol Life Sci, 2011. 68(19): p. 3175-99. 30. Friedl, P. and K. Wolf, Tumour-cell invasion and migration: diversity and escape mechanisms. Nat Rev Cancer, 2003. 3(5): p. 362-74. 31. Lee, J.M., et al., The epithelial-mesenchymal transition: new insights in signaling, development, and disease. J Cell Biol, 2006. 172(7): p. 973-81. 32. Mathias, R.A., S.K. Gopal, and R.J. Simpson, Contribution of cells undergoing epithelial-mesenchymal transition to the tumour microenvironment. J Proteomics, 2013. 78: p. 545-57. 33. Leavesley, D.I., et al., Requirement of the integrin beta 3 subunit for carcinoma cell spreading or migration on vitronectin and fibrinogen. J Cell Biol, 1992. 117(5): p. 1101-7. 34. Paget, S., The Distribution of Secondary Growths in Cancer of the Breast. The Lancet, 1889. 1(3421): p. 571-573. 35. Erler, J.T., et al., Hypoxia-induced lysyl oxidase is a critical mediator of bone marrow cell recruitment to form the premetastatic niche. Cancer Cell, 2009. 15(1): p. 35-44. 36. Kaplan, R.N., et al., VEGFR1-positive haematopoietic bone marrow progenitors initiate the pre-metastatic niche. Nature, 2005. 438(7069): p. 820-7. 37. Kowanetz, M., et al., Granulocyte-colony stimulating factor promotes lung metastasis through mobilization of Ly6G+Ly6C+ granulocytes. Proc Natl Acad Sci U S A, 2010. 107(50): p. 21248-55. 38. Sceneay, J., et al., Primary tumor hypoxia recruits CD11b+/Ly6Cmed/Ly6G+ immune suppressor cells and compromises NK cell cytotoxicity in the premetastatic niche. Cancer Res, 2012. 72(16): p. 3906-11. 39. Hickstein, D.D., et al., Isolation and characterization of the receptor on human neutrophils that mediates cellular adherence. J Biol Chem, 1987. 262(12): p. 5576-80. 40. Fleming, T.J., M.L. Fleming, and T.R. Malek, Selective expression of Ly-6G on myeloid lineage cells in mouse bone marrow. RB6-8C5 mAb to granulocyte-differentiation antigen (Gr-1) detects members of the Ly-6 family. J Immunol, 1993. 151(5): p. 2399-408. 84  41. Hamilton, M.J., et al., Macrophages Are More Potent Immune Suppressors Ex Vivo Than Immature Myeloid-Derived Suppressor Cells Induced by Metastatic Murine Mammary Carcinomas. J Immunol, 2013. 42. Ong, S.E., et al., Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics, 2002. 1(5): p. 376-86. 43. Harsha, H.C., H. Molina, and A. Pandey, Quantitative proteomics using stable isotope labeling with amino acids in cell culture. Nat Protoc, 2008. 3(3): p. 505-16. 44. Vandermarliere, E., M. Mueller, and L. Martens, Getting intimate with trypsin, the leading protease in proteomics. Mass Spectrom Rev, 2013: p. 0. 45. Gilkes, D.M., et al., Procollagen lysyl hydroxylase 2 is essential for hypoxia-induced breast cancer metastasis. Mol Cancer Res, 2013. 11(5): p. 456-66. 46. Erler, J.T., et al., Lysyl oxidase is essential for hypoxia-induced metastasis. Nature, 2006. 440(7088): p. 1222-6. 47. Uzel, M.I., et al., Multiple bone morphogenetic protein 1-related mammalian metalloproteinases process pro-lysyl oxidase at the correct physiological site and control lysyl oxidase activation in mouse embryo fibroblast cultures. J Biol Chem, 2001. 276(25): p. 22537-43. 48. Wong, C.C., et al., Hypoxia-inducible factor 1 is a master regulator of breast cancer metastatic niche formation. Proc Natl Acad Sci U S A, 2011. 108(39): p. 16369-74. 49. Yan, H.H., et al., Gr-1+CD11b+ myeloid cells tip the balance of immune protection to tumor promotion in the premetastatic lung. Cancer Res, 2010. 70(15): p. 6139-49. 50. Hiratsuka, S., et al., MMP9 induction by vascular endothelial growth factor receptor-1 is involved in lung-specific metastasis. Cancer Cell, 2002. 2(4): p. 289-300. 51. Wong, G.S. and A.K. Rustgi, Matricellular proteins: priming the tumour microenvironment for cancer development and metastasis. Br J Cancer, 2013. 108(4): p. 755-61. 52. Kii, I., et al., Incorporation of tenascin-C into the extracellular matrix by periostin underlies an extracellular meshwork architecture. J Biol Chem, 2010. 285(3): p. 2028-39. 53. Bao, S., et al., Periostin potently promotes metastatic growth of colon cancer by augmenting cell survival via the Akt/PKB pathway. Cancer Cell, 2004. 5(4): p. 329-39. 54. Maruhashi, T., et al., Interaction between periostin and BMP-1 promotes proteolytic activation of lysyl oxidase. J Biol Chem, 2010. 285(17): p. 13294-303. 55. Erickson, H.P. and J.L. Inglesias, A six-armed oligomer isolated from cell surface fibronectin preparations. Nature, 1984. 311(5983): p. 267-9. 56. Chiquet, M. and D.M. Fambrough, Chick myotendinous antigen. II. A novel extracellular glycoprotein complex consisting of large disulfide-linked subunits. J Cell Biol, 1984. 98(6): p. 1937-46. 57. Grumet, M., et al., Cytotactin, an extracellular matrix protein of neural and non-neural tissues that mediates glia-neuron interaction. Proc Natl Acad Sci U S A, 1985. 82(23): p. 8075-9. 58. Kruse, J., et al., The J1 glycoprotein--a novel nervous system cell adhesion molecule of the L2/HNK-1 family. Nature, 1985. 316(6024): p. 146-8. 59. Chiquet-Ehrismann, R., et al., Tenascin: an extracellular matrix protein involved in tissue interactions during fetal development and oncogenesis. Cell, 1986. 47(1): p. 131-9. 60. Vaughan, L., et al., A major, six-armed glycoprotein from embryonic cartilage. EMBO J, 1987. 6(2): p. 349-53. 85  61. Rathjen, F.G., U. Norenberg, and H. Volkmer, Glycoproteins implicated in neural cell adhesion and axonal growth. Biochem Soc Trans, 1992. 20(2): p. 405-9. 62. Hagios, C., et al., Tenascin-Y: a protein of novel domain structure is secreted by differentiated fibroblasts of muscle connective tissue. J Cell Biol, 1996. 134(6): p. 1499-512. 63. Weber, P., et al., Zebrafish tenascin-W, a new member of the tenascin family. J Neurobiol, 1998. 35(1): p. 1-16. 64. Bristow, J., et al., Tenascin-X: a novel extracellular matrix protein encoded by the human XB gene overlapping P450c21B. J Cell Biol, 1993. 122(1): p. 265-78. 65. Jones, P.L. and F.S. Jones, Tenascin-C in development and disease: gene regulation and cell function. Matrix Biol, 2000. 19(7): p. 581-96. 66. Gulcher, J.R., et al., Structure of the human hexabrachion (tenascin) gene. Proc Natl Acad Sci U S A, 1991. 88(21): p. 9438-42. 67. Pas, J., et al., Analysis of structure and function of tenascin-C. Int J Biochem Cell Biol, 2006. 38(9): p. 1594-602. 68. Guttery, D.S., et al., Expression of tenascin-C and its isoforms in the breast. Cancer Metastasis Rev, 2010. 29(4): p. 595-606. 69. Weber, P., et al., Tenascin-C binds heparin by its fibronectin type III domain five. J Biol Chem, 1995. 270(9): p. 4619-23. 70. De Laporte, L., et al., Tenascin C Promiscuously Binds Growth Factors via Its Fifth Fibronectin Type III-Like Domain. PLoS One, 2013. 8(4): p. e62076. 71. Orend, G. and R. Chiquet-Ehrismann, Tenascin-C induced signaling in cancer. Cancer Lett, 2006. 244(2): p. 143-63. 72. Castanos-Velez, E., P. Biberfeld, and M. Patarroyo, Extracellular matrix proteins and integrin receptors in reactive and non-reactive lymph nodes. Immunology, 1995. 86(2): p. 270-8. 73. Midwood, K.S. and G. Orend, The role of tenascin-C in tissue injury and tumorigenesis. J Cell Commun Signal, 2009. 3(3-4): p. 287-310. 74. Orend, G. and R. Chiquet-Ehrismann, Adhesion modulation by antiadhesive molecules of the extracellular matrix. Exp Cell Res, 2000. 261(1): p. 104-10. 75. Chong, H.C., et al., Matricellular proteins: a sticky affair with cancers. J Oncol, 2012. 2012: p. 351089. 76. Chiquet-Ehrismann, R., et al., Tenascin interferes with fibronectin action. Cell, 1988. 53(3): p. 383-90. 77. Bornstein, P. and E.H. Sage, Matricellular proteins: extracellular modulators of cell function. Curr Opin Cell Biol, 2002. 14(5): p. 608-16. 78. Murphy-Ullrich, J.E., The de-adhesive activity of matricellular proteins: is intermediate cell adhesion an adaptive state? J Clin Invest, 2001. 107(7): p. 785-90. 79. Murphy-Ullrich, J.E., et al., Focal adhesion integrity is downregulated by the alternatively spliced domain of human tenascin. J Cell Biol, 1991. 115(4): p. 1127-36. 80. Chung, C.Y., J.E. Murphy-Ullrich, and H.P. Erickson, Mitogenesis, cell migration, and loss of focal adhesions induced by tenascin-C interacting with its cell surface receptor, annexin II. Mol Biol Cell, 1996. 7(6): p. 883-92. 81. Midwood, K.S., et al., Coregulation of fibronectin signaling and matrix contraction by tenascin-C and syndecan-4. Mol Biol Cell, 2004. 15(12): p. 5670-7. 82. Hedin, U., J. Holm, and G.K. Hansson, Induction of tenascin in rat arterial injury. Relationship to altered smooth muscle cell phenotype. Am J Pathol, 1991. 139(3): p. 649-56. 83. LaFleur, D.W., et al., Aortic smooth muscle cells interact with tenascin-C through its fibrinogen-like domain. J Biol Chem, 1997. 272(52): p. 32798-803. 86  84. Yokoyama, K., et al., Identification of amino acid sequences in fibrinogen gamma -chain and tenascin C C-terminal domains critical for binding to integrin alpha vbeta 3. J Biol Chem, 2000. 275(22): p. 16891-8. 85. Jones, P.L., J. Crack, and M. Rabinovitch, Regulation of tenascin-C, a vascular smooth muscle cell survival factor that interacts with the alpha v beta 3 integrin to promote epidermal growth factor receptor phosphorylation and growth. J Cell Biol, 1997. 139(1): p. 279-93. 86. LaFleur, D.W., et al., Cloning and characterization of alternatively spliced isoforms of rat tenascin. Platelet-derived growth factor-BB markedly stimulates expression of spliced variants of tenascin mRNA in arterial smooth muscle cells. J Biol Chem, 1994. 269(32): p. 20757-63. 87. Huang, W., et al., Interference of tenascin-C with syndecan-4 binding to fibronectin blocks cell adhesion and stimulates tumor cell proliferation. Cancer Res, 2001. 61(23): p. 8586-94. 88. Woods, A. and J.R. Couchman, Syndecan 4 heparan sulfate proteoglycan is a selectively enriched and widespread focal adhesion component. Mol Biol Cell, 1994. 5(2): p. 183-92. 89. Hauzenberger, D., et al., Tenascin-C inhibits beta1 integrin-dependent T lymphocyte adhesion to fibronectin through the binding of its fnIII 1-5 repeats to fibronectin. Eur J Immunol, 1999. 29(5): p. 1435-47. 90. De Wever, O., et al., Tenascin-C and SF/HGF produced by myofibroblasts in vitro provide convergent pro-invasive signals to human colon cancer cells through RhoA and Rac. FASEB J, 2004. 18(9): p. 1016-8. 91. Fukunaga-Kalabis, M., et al., Tenascin-C promotes melanoma progression by maintaining the ABCB5-positive side population. Oncogene, 2010. 29(46): p. 6115-24. 92. Tanaka, K., et al., Tenascin-C regulates angiogenesis in tumor through the regulation of vascular endothelial growth factor expression. Int J Cancer, 2004. 108(1): p. 31-40. 93. Hancox, R.A., et al., Tumour-associated tenascin-C isoforms promote breast cancer cell invasion and growth by matrix metalloproteinase-dependent and independent mechanisms. Breast Cancer Res, 2009. 11(2): p. R24. 94. Saito, Y., et al., A peptide derived from tenascin-C induces beta1 integrin activation through syndecan-4. J Biol Chem, 2007. 282(48): p. 34929-37. 95. Siri, A., et al., Different susceptibility of small and large human tenascin-C isoforms to degradation by matrix metalloproteinases. J Biol Chem, 1995. 270(15): p. 8650-4. 96. Ilunga, K., et al., Co-stimulation of human breast cancer cells with transforming growth factor-beta and tenascin-C enhances matrix metalloproteinase-9 expression and cancer cell invasion. Int J Exp Pathol, 2004. 85(6): p. 373-9. 97. Nagaharu, K., et al., Tenascin C induces epithelial-mesenchymal transition-like change accompanied by SRC activation and focal adhesion kinase phosphorylation in human breast cancer cells. Am J Pathol, 2011. 178(2): p. 754-63. 98. Katoh, D., et al., Binding of alphavbeta1 and alphavbeta6 integrins to tenascin-C induces epithelial-mesenchymal transition-like change of breast cancer cells. Oncogenesis, 2013. 2: p. e65. 99. Oskarsson, T., et al., Breast cancer cells produce tenascin C as a metastatic niche component to colonize the lungs. Nat Med, 2011. 17(7): p. 867-74. 100. Massague, J., Breast cancer cells produce tenascin C as a metastatic niche component to colonize the lungs. Nature Medicine, 2011. 17(7): p. 867-74. 87  101. O'Connell, J.T., et al., VEGF-A and Tenascin-C produced by S100A4+ stromal cells are important for metastatic colonization. Proc Natl Acad Sci U S A, 2011. 108(38): p. 16002-7. 102. Minn, A.J., et al., Genes that mediate breast cancer metastasis to lung. Nature, 2005. 436(7050): p. 518-24. 103. Calvo, A., et al., Identification of VEGF-regulated genes associated with increased lung metastatic potential: functional involvement of tenascin-C in tumor growth and lung metastasis. Oncogene, 2008. 27(40): p. 5373-84. 104. Scherberich, A., et al., Tenascin-W is found in malignant mammary tumors, promotes alpha8 integrin-dependent motility and requires p38MAPK activity for BMP-2 and TNF-alpha induced expression in vitro. Oncogene, 2005. 24(9): p. 1525-32. 105. Boersema, P.J., et al., Quantification of the N-glycosylated secretome by super-SILAC during breast cancer progression and in human blood samples. Mol Cell Proteomics, 2013. 12(1): p. 158-71. 106. Tsunoda, T., et al., Involvement of large tenascin-C splice variants in breast cancer progression. Am J Pathol, 2003. 162(6): p. 1857-67. 107. Borsi, L., et al., Cell-cycle dependent alternative splicing of the tenascin primary transcript. Cell Adhes Commun, 1994. 1(4): p. 307-17. 108. Lou, Y., et al., Epithelial-mesenchymal transition (EMT) is not sufficient for spontaneous murine breast cancer metastasis. Dev Dyn, 2008. 237(10): p. 2755-68. 109. Chiquet-Ehrismann, R. and M. Chiquet, Tenascins: regulation and putative functions during pathological stress. J Pathol, 2003. 200(4): p. 488-99. 110. Paron, I., et al., Tenascin-C enhances pancreatic cancer cell growth and motility and affects cell adhesion through activation of the integrin pathway. PLoS One, 2011. 6(6): p. e21684. 111. Tokes, A.M., et al., Tenascin expression in primary and recurrent breast carcinomas and the effect of tenascin on breast tumor cell cultures. Pathol Oncol Res, 2000. 6(3): p. 202-9.   

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.24.1-0165859/manifest

Comment

Related Items