Anchor or Accelerate – A Study on Cancer Cell Adhesion and Motility

Cell migration and adhesion to the extracellular matrix (ECM) are crucial in many biological and pathological processes such as morphogenesis, tissue repair, inflammatory responses, survival, and cancer. Cell-matrix adhesion is mediated by the integrin family of transmembrane receptors, which not only anchor cells to their surroundings, but also transmit bidirectional signalling at the cell surface and couple the ECM to the cytoskeleton. Another group of adhesion receptors are the syndecan proteoglycans, which engage the ECM and possess signalling activity in response to a variety of ligands. Cell migration is a complex process that requires spatial and temporal coordination of adhesion, cell contractility, intracellular traffic of integrins, and matrix turnover by matrix metalloproteinases (MMPs). Thus, integrins and syndecans, as well as MMPs, play essential roles in cancer cell migration and invasion. The understanding of the cooperation of syndecans and integrins was broadened in this thesis study. The results reveal that syndecan-1 functions in concert with 21 integrin in cell adhesion to collagen, whereas syndecan-4 is essential in 21 integrin-mediated matrix contraction. Finally, oncogenic K-Ras was shown to regulate 21 integrin, membrane-type 1 MMP, and syndecan-1 and -4 expression and their cooperation in cell invasion. Epithelial-mesenchymal transition (EMT) is fundamental during embryogenesis and organ development. Activation of EMT processes, including the upregulation of mesenchymal intermediate filament protein vimentin, has also been implicated in the acquisition of a malignant phenotype by epithelial cancer cells. Members of the protein kinase C (PKC) superfamily are involved in cell migration and various integrindependent cellular functions. One aim of this work was to shed light on the role of vimentin in the regulation of integrin traffic and cell motility. In addition, the mechanism by which vimentin participates in EMT was investigated. The results show that integrin recycling and motility are dependent on the PKC–mediated phosphorylation of vimentin. In addition, vimentin was found to be a positive regulator of EMT and regulate the expression of several migratory genes. Specifically, vimentin governs the expression of receptor tyrosine kinase Axl, which is implicated in tumour growth and metastasis. Taken together, the findings described in this thesis reveal novel aspects of the complex interplay between distinct cellular components: integrins, syndecans, and the vimentin cytoskeleton, which all contribute to the regulation of human cancer cell adhesion, migration, and invasion.

Lethal weapons : novel approaches for receptor-targeted cancer cell elimination

The currently used forms of cancer therapy are associated with drug resistance and toxicity to healthy tissues. Thus, more efficient methods are needed for cancer-specific induction of growth arrest and programmed cell death, also known as apoptosis. Therapeutic forms of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) are investigated in clinical trials due to the capability of TRAIL to trigger apoptosis specifically in cancer cells by activation of cell surface death receptors. Many tumors, however, have acquired resistance to TRAIL-induced apoptosis and sensitizing drugs for combinatorial treatments are, therefore, in high demand. This study demonstrates that lignans, natural polyphenols enriched in seeds and cereal, have a remarkable sensitizing effect on TRAIL-induced cell death at non-toxic lignan concentrations. In TRAIL-resistant and androgen-dependent prostate cancer cells we observe that lignans repress receptor tyrosine kinase (RTK) activity and downregulate cell survival signaling via the Akt pathway, which leads to increased TRAIL sensitivity. A structure-activity relationship analysis reveals that the γ-butyrolactone ring of the dibenzylbutyrolactone lignans is essential for the rapidly reversible TRAIL-sensitizing activity of these compounds. Furthermore, the lignan nortrachelogenin (NTG) is identified as the most efficient of the 27 tested lignans and norlignans in sensitization of androgen-deprived prostate cancer cells to TRAIL-induced apoptosis. While this combinatorial anticancer approach may leave normal cells unharmed, several efficient cancer drugs are too toxic, insoluble or unstable to be used in systemic therapy. To enable use of such drugs and to protect normal cells from cytotoxic effects, cancer-targeted drug delivery vehicles of nanometer scale have recently been generated. The newly developed nanoparticle system that we tested in vitro for cancer cell targeting combines the efficient drug-loading capacity of mesoporous silica to the versatile particle surface functionalization of hyperbranched poly(ethylene imine), PEI. The mesoporous hybrid silica nanoparticles (MSNs) were functionalized with folic acid to promote targeted internalization by folate receptor overexpressing cancer cells. The presented results demonstrate that the developed carrier system can be employed in vitro for cancer selective delivery of adsorbed or covalently conjugated molecules and furthermore, for selective induction of apoptotic cell death in folate receptor expressing cancer cells. The tested carrier system displays potential for simultaneous delivery of several anticancer agents specifically to cancer cells also in vivo.

Luminescence-based assay methods in drug discovery

The drug discovery process is facing new challenges in the evaluation process of the lead compounds as the number of new compounds synthesized is increasing. The potentiality of test compounds is most frequently assayed through the binding of the test compound to the target molecule or receptor, or measuring functional secondary effects caused by the test compound in the target model cells, tissues or organism. Modern homogeneous high-throughput-screening (HTS) assays for purified estrogen receptors (ER) utilize various luminescence based detection methods. Fluorescence polarization (FP) is a standard method for ER ligand binding assay. It was used to demonstrate the performance of two-photon excitation of fluorescence (TPFE) vs. the conventional one-photon excitation method. As result, the TPFE method showed improved dynamics and was found to be comparable with the conventional method. It also held potential for efficient miniaturization. Other luminescence based ER assays utilize energy transfer from a long-lifetime luminescent label e.g. lanthanide chelates (Eu, Tb) to a prompt luminescent label, the signal being read in a time-resolved mode. As an alternative to this method, a new single-label (Eu) time-resolved detection method was developed, based on the quenching of the label by a soluble quencher molecule when displaced from the receptor to the solution phase by an unlabeled competing ligand. The new method was paralleled with the standard FP method. It was shown to yield comparable results with the FP method and found to hold a significantly higher signal-tobackground ratio than FP. Cell-based functional assays for determining the extent of cell surface adhesion molecule (CAM) expression combined with microscopy analysis of the target molecules would provide improved information content, compared to an expression level assay alone. In this work, immune response was simulated by exposing endothelial cells to cytokine stimulation and the resulting increase in the level of adhesion molecule expression was analyzed on fixed cells by means of immunocytochemistry utilizing specific long-lifetime luminophore labeled antibodies against chosen adhesion molecules. Results showed that the method was capable of use in amulti-parametric assay for protein expression levels of several CAMs simultaneously, combined with analysis of the cellular localization of the chosen adhesion molecules through time-resolved luminescence microscopy inspection.

In vivo imaging of vascular adhesion protein 1 - preclinical studies with positron emission tomography

The golden standard in nuclear medicine imaging of inflammation is the use of radiolabeled leukocytes. Although their diagnostic accuracy is good, the preparation of the leukocytes is both laborious and potentially hazardous for laboratory personnel. Molecules involved in leukocyte migration could serve as targets for the development of inflammation imaging agents. An excellent target would be a molecule that is absent or expressed at low level in normal tissues, but is induced or up-regulated at the site of inflammation. Vascular adhesion protein-1 (VAP-1) is a very promising target for in vivo imaging, since it is translocated to the endothelial cell surface when inflammation occurs. VAP-1 functions as an endothelial adhesion molecule that participates in leukocyte recruitment to inflamed tissues. Besides being an adhesion molecule, VAP-1 also has enzymatic activity. In this thesis, the targeting of VAP-1 was studied by using Gallium-68 (68Ga) labeled peptides and an Iodine-124 (124I) labeled antibody. The peptides were designed based on molecular modelling and phage display library searches. The new imaging agents were preclinically tested in vitro, as well as in vivo in animal models. The most promising imaging agent appeared to be a peptide belonging to the VAP-1 leukocyte ligand, Siglec-9 peptide. The 68Ga-labeled Siglec-9 peptide was able to detect VAP-1 positive vasculature in rodent models of sterile skin inflammation and melanoma by positron emission tomography. In addition to peptides, the 124I-labeled antibody showed VAP-1 specific binding both in vitro and in vivo. However, the estimated human radiation dose was rather high, and thus further preclinical studies in disease models are needed to clarify the value of this imaging agent. Detection of VAP-1 on endothelium was demonstrated in these studies and this imaging approach could be used in the diagnosis of inflammatory conditions as well as melanoma. These studies provide a proof-of-concept for PET imaging of VAP-1 and further studies are warranted.

JNK, a versatile regulator of kinesin-1 transport and cytoskeleton dynamics

C-Jun N-terminal kinase (JNK) is traditionally recognized as a crucial factor in stress response and inducer of apoptosis upon various stimulations. Three isoforms build the JNK subfamily of MAPK; generally expressed JNK1 and JNK2 and brain specific JNK3. Degenerative potency placed JNK in the spotlight as potential pharmacological option for intervention. Unfortunately, adverse effects of potential drugs and observation that expression of only JNK2 and JNK3 are induced upon stress, restrained initial enthusiasm. Notably, JNK1 demonstrated atypical high constitutive activity in neurons that is not responsive to cellular stresses and indicated existence of physiological activity. This thesis aimed at revealing the physiological functions of JNK1 in actin homeostasis through novel effector MARCKS-Like 1 (MARCKSL1) protein, neuronal trafficking mediated by major kinesin-1 motor protein and microtubule (MT) dynamics via STMN2/SCG10. The screen for novel physiological JNK substrates revealed specific phosphorylation of C-terminal end of MARCKSL1 at S120, T148 and T183 both ex vivo and in vitro. By utilizing site-specific mutagenesis, various actin dynamics and migrations assays we were able to demonstrate that JNK1 phosphorylation specifically facilitates F-actin bundling and thus filament stabilisation. Consecutively, this molecular mechanism was proved to enhance formation of filopodia; cell surface projections that allow cell sensing surrounding environment and migrate efficiently. Our results visualize JNK dependent and MARCKSL1 executed induction of filopodia in neurons and fibroblast indicating general mechanism. Subsequently, inactivation of JNK action on MARCKSL1 shifts cellular actin machinery into lamellipodial dynamic arrangement. Tuning of actin cytoskeleton inevitably melds with cell migration. We observed that both active JNK and JNK pseudo-phosphorylated form of MARCKSL1 reduce actin turnover in intact cells leading to overall diminished cell motility. We demonstrate that tumour transformed cells from breast, prostate, lung and muscle-derived cancers upregulate MARCKSL1. We showed on the example of prostate cancer PC-3 cell line that JNK phosphorylation negatively controls MARCKSL1 ability to induce migration, which precedes cancer cell metastasis. The second round of identification of JNK physiological substrates resulted in detection of predominant motor protein kinesin-1 (Kif5). Mass spectrometry detailed analysis showed evident endogenous phosphorylation of kinesin-1 on S176 within motor domain that interacts with MT. In vitro phosphorylation of bacterially expressed kinesin heavy chain by JNK isoforms displayed higher specificity of JNK1 when compared to JNK3. Since, JNK1 is constitutively active in neurons it signified physiological aspect of kinesin-1 regulation. Subsequent biochemical examination revealed that kinesin-1, when not phosphorylated on JNK site, exhibits much higher affinity toward MTs. Expression of the JNK non-phosphorable kinesin-1 mutant in intact cells as well as in vitro single molecule imaging using total internal reflection fluorescence microscopy indicated that the mutant loses normal speed and is not able to move processively into proper cellular compartments. We identify novel kinesin-1 cargo protein STMN2/SCG10, which along with known kinesin-1 cargo BDNF is showing impaired trafficking when JNK activity is inhibited. Our data postulates that constitutive JNK activity in neurons is crucial for unperturbed physiologically relevant transport of kinesin-1 dependant cargo. Additionally, my work helps to validate another novel physiological JNK1 effector STMN2/SCG10 as determinant of axodendritic neurites dynamics in the developing brain through regulation of MT turnover. We show successively that this increased MT dynamics is crucial during developmental radial migration when brain layering occurs. Successively, we are able to show that introduction of JNK phosphorylation mimicking STMN2/SCG10 S62/73D mutant rescues completely JNK1 genetic deletion migration phenotype. We prove that STMN2/SCG10 is predominant JNK effector responsible for MT depolymerising activity and neurite length during brain development. Summarizing, this work describes identification of three novel JNK substrates MARCKSL1, kinesin-1 and STMN2/SCG10 and investigation of their roles in cytoskeleton dynamics and cargo transport. This data is of high importance to understand physiological meaning of JNK activity, which might have an adverse effect during pharmaceutical intervention aiming at blocking pathological JNK action.

From epithelial to mesenchymal: regulation of invasive cancer cell motility

Metastasis is the main cause of death among cancer patients. In order to initiate the metastatic cascade cancer cells have to undergo epithelial-to-mesenchymal transition (EMT). In EMT epithelial cells lose their cell-cell and cell-extracellular matrix (ECM) contacts and become more motile. The expression of the transcription factor Slug and of the mesenchymal intermediate filament vimentin is induced during EMT. Vimentin is often overexpressed in malignant epithelial cancers but the functional role of vimentin remains incompletely understood. In addition, kinases such as AKT and ERK are known to be involved in the regulation of EMT and cancer cell motility but the mechanisms underlining their functions are often unclear. Integrins are heterodimeric receptors that attach cells to the surrounding tissue and participate in regulating cell migration and invasion. Changes in integrin activity are linked to increased cell motility and further cancer metastasis. The aim for my PhD studies was to investigate the role of cellular signalling pathways and vimentin in the regulation of cancer cell motility and EMT. Our results revealed that in prostate cancer the downregulation of AKT1 and AKT2, but not AKT3, induces activation of cell surface 1-integrins leading to enhanced cell adhesion, migration and invasion. In addition, our findings demonstrated a reciprocal regulatory interaction between vimentin and ERK2 facilitating ERK-mediated phosphorylation of Slug at serine-87 (S87) in breast cancer. Surprisingly, Slug S87 phosphorylation is dispensable for E-cadherin repression but essential for the induction of vimentin and Axl expression in early onset of EMT. Our findings reveal previously unknown mechanistic information of how prostate and breast cancer cell motility and disease progression is regulated

Immune Evasion by Borrelia burgdorferi – With Special Reference to CD38-mediated Chemotaxis of Neutrophils and Dendritic Cells

Lyme borreliosis is a tick-transmitted infection caused by the spirochete bacterium Borrelia burgdorferi sensu lato. The tick injects bacteria into host skin, where a first line defence, mainly the complement system, neutrophils, dendritic cells and macrophages are ready to attack foreign intruders. However, in the case of Lyme borreliosis, the original immune response in the skin is untypically mild among bacterial infections. A further untypical feature is the ability of B. burgdorferi to disseminate to distant organs, where, in some patients, symptoms appear after years after the original infection. This study aimed at uncovering some of the immune evasion mechanisms utilized by B. burgdorferi against the complement system, neutrophils and dendritic cells. B. burgdorferi was shown to inhibit chemotaxis of human neutrophils towards nformyl- methyl-leucyl-phenylalanine (fMLP). Outer surface protein B (OspB) of B. burgdorferi was shown to promote resistance to the attack of the complement system and neutrophil phagocytosis at low complement concentrations. B. burgdorferi was shown to inhibit migration of dendritic cells in vitro towards CCL19 and CCL21 and also in an in vivo model. This effect was shown to be due to the absence of CD38 on the borrelia-stimulated dendritic cell surface. A defect in p38 mitogen-activated-protein-kinase (p38) signaling was linked to defective CD38 expression. A defect in CD38 expression on B. burgdorferi-stimulated neutrophils was also observed. In this study, a number of novel immune evasion strategies utilized by B burgdorferi were chracterized. However, further studies are needed as other immune evasion mechanisms await to be uncovered.

Molecular Characteristics of Neuroblastoma with Special Reference to Novel Prognostic Factors and Diagnostic Applications

Molecular Characteristics of Neuroblastoma with Special Reference to Novel Prognostic Factors and Diagnostic Applications Department of Medical Biochemistry and Genetics Annales Universitatis Turkuensis, Medica-Odontologica, 2009, Turku, Finland Painosalama Oy, Turku, Finland 2009 Background: Neuroblastoma, which is the most common and extensively studied childhood solid cancer, shows a great clinical and biological heterogeneity. Most of the neuroblastoma patients older than one year have poor prognosis despite intensive therapies. The hallmark of neuroblastoma, biological heterogeneity, has hindered the discovery of prognostic tumour markers. At present, few molecular markers, such as MYCN oncogene status, have been adopted into clinical practice. Aims: The aim of the study was to improve the current prognostic methodology of neuroblastoma, especially by taking cognizance of the biological heterogeneity of neuroblastoma. Furthermore, unravelling novel molecular characteristics which associate with neuroblastoma tumour progression and cell differentiation was an additional objective. Results: A new strictly defined selection of neuroblastoma tumour spots of highest proliferation activity, hotspots, appeared to be representative and reliable in an analysis of MYCN amplification status using a chromogenic in situ hybridization technique (CISH). Based on the hotspot tumour tissue microarray immunohistochemistry and high-resolution oligo-array-based comparative genomic hybridization, which was integrated with gene expression and in silico analysis of existing transcriptomics, a polysialylated neural cell adhesion molecule (NCAM) and poorly characterized amplicon at 12q24.31 were discovered to associate with outcome. In addition, we found that a previously considered new neuroblastoma treatment target, the mutated c-kit receptor, was not mutated in neuroblastoma samples. Conclusions: Our studies indicate polysialylated NCAM and 12q24.31 amplicon to be new molecular markers with important value in prognostic evaluation of neuroblastoma. Moreover, the presented hotspot tumour tissue microarray method together with the CISH technique of the MYCN oncogene copy number is directly applicable to clinical use. Key words: neuroblastoma, polysialic acid, neural cell adhesion molecule, MYCN, c-kit, chromogenic in situ hybridization, hotspot

Differential Regulation of c-FLIP Isoforms Through Post-translational Modifications

Cells are constantly responding to signals from the surrounding tissues and the environment. To dispose of infected and potentially dangerous cells, to ensure the optimal execution of developmental processes and to maintain tissue homeostasis, a multicellular organism needs to tightly control both the number and the quality of its cells. Apoptosis is a form of active cellular self-destruction that enables an organism to regulate its cell number by deleting damaged or potentially dangerous cells. Apoptosis can be induced by death ligands, which bind to death receptors on the cell surface. Ligation of the receptors leads to the formation of an intracellular death inducing signaling complex (DISC). One of the DISC components is caspase-8, a protease that triggers the caspase cascade and is thereby a key initiator of programmed cell death. The activation of caspase-8 is controlled by the cellular FLICE-inhibitory proteins (c-FLIPs). Consequently, sensitivity towards receptor-mediated apoptosis is determined by the amount of c-FLIP, and the c-FLIP levels are actively regulated for example during erythroid differentiation of K562 erythroleukemia cells and by hyperthermia in Jurkat leukemia cells. The aim of my thesis was to investigate how c-FLIP is regulated during these processes. We found that c-FLIP isoforms are short-lived proteins, although c-FLIPS had an even shorter half-life than c-FLIPL. In both experimental models, increased death receptor sensitivity correlated with induced ubiquitylation and consequent proteasomal degradation of c-FLIP. Furthermore, we elucidated how phosphorylation regulates the biological functions and the turnover of c-FLIP, thereby contributing to death receptor sensitivity. We mapped the first phosphorylation sites on c-FLIP and dissected how their phosphorylation affects c-FLIP. Moreover, we demonstrated that phosphorylation of serine 193, a phosphorylated residue common to all c-FLIPs, is primarily mediated by the classical PKC. Furthermore, we discovered a novel connection between the phosphorylation and ubiquitylation of c-FLIP: phosphorylation of S193 protects c-FLIP from ubiquitylation. Surprisingly, although all c-FLIP isoforms are phosphorylated on this conserved residue, the biological outcome is different for the long and short isoforms, since S193 specifically prolongs the half-lives of the short c-FLIP isoforms, but not c-FLIPL. To summarize, we show that c-FLIP proteins are modified by ubiquitylation and phosphorylation, and that the biological outcomes of these modifications are isoform-specifically determined.

Alpha2-Adrenoceptor-mediated vascular responses induced by dexmedetomidine

Alpha2-Adrenoceptors are cell-surface G protein coupled receptors that mediate many of the effects of the catecholamines noradrenaline and adrenaline. The three human α2-adrenoceptor subtypes are widely expressed in different tissues and organs, and they mediate many different physiological and pharmacological effects in the central and peripheral nervous system and as postsynaptic receptors in target organs. Previous studies have demonstrated that α2-adrenoceptors mediate both vascular constriction and dilatation in humans. Large inter-individual variation has been observed in the vascular responses to α2-adrenoceptor activation in clinical studies. All three receptor subtypes are potential drug targets. It was therefore considered important to further elucidate the details of adrenergic vascular regulation and its genetic variation, since such knowledge may help to improve the development of future cardiovascular drugs and intensive care therapies. Dexmedetomidine is the most selective and potent α2-adrenoceptor agonist currently available for clinical use. When given systemically, dexmedetomidine induces nearly complete sympatholysis already at low concentrations, and postsynaptic effects, such vasoconstriction, can be observed with increasing concentrations. Thus, local infusions of small doses of dexmedetomidine into dorsal hand veins and the application of pharmacological sympathectomy with brachial plexus block provide a means to assess drug-induced peripheral vascular responses without interference from systemic pharmacological effects and autonomic nervous system regulation. Dexmedetomidine was observed to have biphasic effects on haemodynamics, with an initial decrease in blood pressure at low concentrations followed by substantial increases in blood pressure and coronary vascular resistance at high concentrations. Plasma concentrations of dexmedetomidine that significantly exceeded the recommended therapeutic level did not reduce myocardial blood flow below the level that is observed with the usual therapeutic concentrations and did not induce any evident myocardial ischaemia in healthy subjects. Further, it was demonstrated that dexmedetomidine also had significant vasodilatory effects through activation of endothelial nitric oxide synthesis, and thus when the endothelial component of the blood vessel response to dexmedetomidine was inhibited, peripheral vasoconstriction was augmented. Hand vein constriction responses to α2-adrenoceptor activation by dexmedetomidine were only weakly associated with the constriction responses to α1-adrenoceptor activation, pointing to independent cellular regulation by these two adrenoceptor classes. Substantial inter-individual variation was noted in the venous constriction elicited by activation of α2-adrenoceptors by dexmedetomidine. In two study populations from two different continents, a single nucleotide polymorphism in the PRKCB gene was found to be associated with the dorsal hand vein constriction response to dexmedetomidine, suggesting that protein kinase C beta may have an important role in the vascular α2-adrenoceptor signalling pathways activated by dexmedetomidine.

Receptors and endocytosis of coxsackievirus A9

Coxsackievirus A9 (CV-A9) belongs to human enteroviruses within family Picornaviridae, which are the main cause of aseptic meningitis. In addition, CV-A9 causes a wide range of other clinical manifestations of acute disease including respiratory infections, myocarditis, encephalitis and severe generalized infections in newborns. In this study, the functions of integrins αVβ6 and αVβ3 in the attachment and cellular entry of CV-A9 were analyzed. Further, virus and cell surface interactions and endocytosis of CV-A9 were studied in specific cell lines. Also, a method for production of GFP-expressing CV-A9 particles by long PCR-mediated mutagenesis and in vivo transcription was developed. The results indicated that RGD-motif (arginine-glycine-asparagine) that resides in the viral capsid is important for CV-A9 infection particularly in cell lines expressing integrin αVβ6 and that this integrin serves as a high affinity attachment receptor for the virus. CV-A9 is also capable of infecting certain cell lines independently of αV-integrins by binding to the cell surface HSPA5 protein. Regardless of the attachment stage, the internalization of the virus occurs via the same entry pathway and is dependent on β2M, dynamin, and Arf6 but independent of clathrin and caveolin-1. Furthermore, the virus internalization occurs within Arf6-containing vesicles suggesting that Arf6 is central mediator of CV-A9 endocytosis. While in this study the results of CV-A9 endocytosis were based on microscopical visualization within individual fixed cells, a rapid method for generation of a virus for real-time imaging was also described.

Integrins on the move

Cancer is a leading cause of death worldwide accounting for 13% of all deaths in 2005. The spread of cancer and formation of metastases is the major cause of mortality among cancer patients. The spread of cancer is based on the cancer cell’s ability to break away from the surrounding tissue and to migrate into new areas in the body. The ability of cells to bind its surroundings and to move is controlled by the mechanical cell surface adhesion receptors called the integrins. Integrins have a critical role in cell adhesion, cell motility and tissue homeostasis. By communicating with ECM, integrins transmit signals from the surrounding environment inside the cell and modulate the function of many important signalling pathways involved in cell survival, development, gene expression, proliferation, motility and cytoskeletal organization. During cell migration integrin-matrix adhesions are formed in front of the cell while rear-adhesions are released during migration. Integrins are endocytosed from the plasma-membrane into the cytoplasm and partly recycled back to new adhesion sites in a process called integrin trafficking. Also, the cell cytoskeleton and protrusions are important in cell migration. Finger-like actin protrusions called filopodia display an interesting cancer relevant cooperation with integrins that is required for cell migration. The expression and function of integrins changes markedly as cells acquire carcinogenic properties. Changed integrin function is partly responsible for detachment of tumor cells from neighbouring cells and for providing enhanced invasive capabilities for tumor cells to disseminate. Similarly, the formation of filopodia is increased in cancer. High myosin-10 expression is related to poor outcome in breast cancer and increased cell migration. The proper function of myosin-10 induced filopodia needs association with β1 integrins. The importance of integrin trafficking and filopodia formation is becoming increasingly more recognized in cancer. This thesis focusses on the role of integrins, integrin trafficking and myosin-10 induced filopodia cancer cell migration.

Vesicular Trafficking in Osteoclasts

Osteoclasts are multinucleated bone-degrading cells that undergo large changes in their polarisation and vesicular trafficking during the bone resorption cycle. Rab proteins are small GTPases that offer both temporal and spatial regulation to the transport between membranous organelles. Previously the presence and function of only few of the currently known 60 Rab proteins in osteoclasts have been reported. In this study, the expression of 26 Rab genes in bone-resorbing osteoclasts was demonstrated with gene-specific primer pairs. The further analysis of three Rab genes during human osteoclast differentiation revealed that Rab13 gene is highly induced during osteoclastogenesis. The presence of Rab13 protein in the secretory vesicles directed towards the ruffled border and in the endocytotic or transcytotic pathways in resorbing osteoclasts was excluded. The localisation of Rab13 suggests that that it is associated with a previously unknown vesicle population travelling between the trans-Golgi network and the basolateral membrane in bone resorbing osteoclasts. Rab proteins convey their functions by binding to specific effector proteins. We found a novel Rab13 interaction with endospanins-1 and -2 that are yet poorly characterised small transmembrane proteins. The Rab13 subfamily member Rab8 also bound to endospanins, while Rab10 and unrelated Rabs did not. Rab13 and endospanin-2 co-localised in perinuclear vesicles in transfected cells, demonstrating the interaction also in vivo. The inhibition of Rab13 did not interfere with the localisation of endospanin-2 nor did it affect the cell surface expression of growth hormone receptor, as has been previously described for endospanins. The physiological role of this novel protein-protein interaction thus remains to be clarified. The analysis of the transcytotic route in bone resorbing osteoclasts revealed that multiple vesicle populations arise from the ruffled border and transport the bone degradation products for exocytosis. These vesicles are directed to the functional secretory domain that is encircled by an actin-based molecular barrier. Furthermore, the transcytotic vesicles contain abundant Helix pomatia lectin binding sites and represent lipid raft concentrates. Finally, autophagosomal compartments may also be involved in the transcytosis.