Certain organs are favoured sites for circulating cancer cells to develop metastases, and this can only arise through a permissive microenvironment in the target tissues which facilitates tumour growth. This hypothesis of "seed-and-soil" was postulated more than 100 years ago by Stephen Paget, and is being gradually justified by the identification of matrix and molecular factors, some specific to the affected organs subject to metastasis.

Comparisons of metastasis with the homing of leukocytes to specific organs suggest that chemokines expressed in the affected organs attract or aid the survival of circulating metastatic cells. Organs affected by breast cancer metastases have been found to express chemokine CXCL12, and its receptor CXCR4 is expressed in breast cancer metastatic cells. By introducing antibodies against CXCR4 in-vivo, the occurrence of metastases in all organs could be reduced. [1]

A further important area of investigation is the role of carbohydrate antigens such as sialosyl Lewisx and T-antigen that are strongly expressed in many types of cancer cells, including breast cancer. These interact strongly with animal lectins (e.g. E-selectin), which are induced in endothelium by the cytokines also released by cancer cells, and this interaction mediates adhesion of the cancer cells to the endothelium. [2] Sialosyl Lewis A and B antigens may be good indicators of metastatic potential in breast carcinoma, particularly to the lymph nodes [5]. It has been shown recently that T-antigen present on the surface of breast cancer cells is important ligand for galectin-3 expressed by endothelial cells.

Lesions in bone are the most common metastases in advanced breast cancer. There is considerable evidence that breast cancer cells express several bone matrix proteins, including bone sialoprotein (BSP). MetaBre partner University of Liège was the first to demonstrate that BSP, an osteoblastic protein, was ectopically expressed in human breast cancer. Of particular interest was that the level of expression of BSP at the primary breast lesion was significantly predictive of the risk for the patient to develop bone metastases.

Osteocalcin, osteonectin, osteopontin (OPN), transcription factors (cbfa-1) and soluble factors (TGFß, BMPs,VEGFs, CTGF) are also expressed in a similar manner to osteoblastic cells. This "osteomimicry", possibly a form of EMT, could be responsible for the preferential homing of breast cancer cells to bone matrix. However, it is not clear whether osteomimicry is a consequence of EMT at the primary site due to clonal selection in cancer cell populations of the primary tumor, or is result of induction by the bone/bone marrow microenvironment. Also uncertain is whether osteomimicry/EMT plays a causal role in determining bone homing of breast cancer cells and subsequent invasive growth in the bone microenvironment. The expression of soluble factors can be targeted for their role in acquisition of EMT/osteomimicry and therefore the bone metastatic potential and proliferation of cancer cells.

The continuous remodelling in healthy bone makes it a permissive microenvironment for bone metastases to be established, e.g. through the release of growth factors and TGFß. Osteolytic bone metastases cause damage to bone structure through stimulation of the bone-resorbing osteoclast cells with tumour-expressed factors including PTHrP and M-CSF. Some tumours may also induce upregulation of RANKL in osteogenic cells, a receptor activator that matures osteoclast precursor cells. A vicious circle results because the bone matrix as it is resorbed releases more tumour stimulating cytokines and growth factors.

Bone metastases are often found to be surrounded by osteoclasts, and tumour cells and osteoclasts stimulate each other by the production of growth factors. This signalling is mediated particularly by integrin aVß3. Its receptor is rarely found expressed in most human cells, but is strongly expressed in bone metastases cell lines from breast cancer and other malignancies.

Correlations have also been found between the level of aVß3 integrin expression, and the extent of cell proliferation, homing of malignant cells to the bone microenvironment, and the greater degree of osteotropism. In this respect, MetaBre partner INSERM has shown that aVß3 integrin can promote bone metastasis formation by increasing breast cancer cell adhesion to BSP.

Integrin aVß3 is also strongly expressed, along with other integrins, MMPs, actin and actin binding proteins and tyrosine kinase signalling machinery in the invadopodia and podosomes. These are specialised membrane structures developed in invasive cells (invadopodia in metastatic cells, podosomes in osteoclasts necessary for efficient bone resorption) that protrude into the ECM and ensure a strong adhesion for successful degradation. The molecular composition of invadopodia and podosomes remains poorly characterised as well as the overall structure of the ECM degradation sites. MetaBre partner Consorzio Mario Negri Sud has shown that the ubiquitous Dynamin 2 GTPase protein, required in endocytic membrane fission, caveolae internalisation, protein trafficking in the Golgi apparatus and the internalisation of particles during phagocytosis, is also necessary for ECM degradation by invadopodia and podosomes. MetaBre will take further the understanding of the role of Dynamin 2 in ECM degradation, as well as other cytoskeletal components of the invadopodia and podosomes, particularly actin and cortactin. This work has dual relevance for the project being applicable to both cell invasiveness and bone resorption.

Recently the use of bisphosphonates for treatment of bone metastases has become widespread. These compounds act to block bone resorption by inhibiting osteoclast activity. They also may decrease the proliferation of bone metastases through the reduction in stimulating factors released from the bone matrix or osteoclasts themselves, and it is suggested now that some bisphosphonates increase apoptosis in breast cancer cells, and inhibit MMPs [4]. All animal and clinical trials with bisphosphonates to date have shown no clear extension of life expectancy, despite 30-40% of patients experiencing reduced skeletal problems. A new bisphosphonate zoledronic acid shows promise for anti-metastatic activity in bone and visceral organs and also against angiogenesis. [5] The mechanisms are not fully elucidated and this will be a task of partner University of Liège in MetaBre. New therapeutic targets may also provide a way forward, combined with results from work on known targets, for improved treatment of metastatic disease in breast cancer patients.