In order to invade the tissues where metastases can develop, tumour cells must penetrate basement membranes and the extracellular matrix (ECM). The proteolytic enzymes responsible for normal processes of tissue remodelling in the body (for example in wound healing, bone resorption and embryo implantation) are all implicated in invasive processes of metastasis and angiogenesis. These include serine proteases (e.g. urokinase plasminogen activator), cysteine proteinases (e.g. cathepsins), and zinc dependent matrix metalloproteinases (MMPs). In many forms of human malignancy the expression and activity of the nineteen known MMPs are particularly abnormal. Recent research has suggested that MMP-2 and MMP-9 (both gelatinases) play a particularly important role in breast cancer initiation, angiogenesis, growth and metastasis. These MMPs are thought to degrade the collagen IV component of the basement membrane in ECM, and this allows malignant cells to invade organs and enter the bloodstream.

In addition, MMPs may release matrix-sequestered tumour and endothelial growth factors, further promoting the "vicious cycle" of autocrine growth and acting as an angiogenic switch. The expression of MMP-2 and MMP-9 is regulated by the key c-proto-oncogenes implicated in breast cancer, such as the c-erbB family or c-ras for increased MMP-2 and the ets gene PEA-3 for MMP-9. The membrane-type MMPs are much less well understood that the soluble MMPs but recent results based on transgenic mouse models and overexpression of different MMPs suggest that MT-MMPs are the master regulators in the invasive process, both by activating soluble MMPs and engaging in ECM degradation themselves.

As well as the basement membrane, ECM is also composed of interstitial connective tissue, and it is believed that this component is degraded by the MMPs classed as interstitial collagenases, and the stromelysins. Some of the evidence connecting MMPs to metastasis is based on the reduction in colonisation and metastasis development in vivo through introduction of MMP inhibitors, for example the endogenous TIMPs (tissue inhibitors of metalloproteinases), antisense oligodeoxynucleotides or Batimistat [1]. TIMPs are generally ruled out for therapy since they are multifunctional and can also act to stimulate cell proliferation. Synthetic MMP inhibitors are generally peptide molecules with a zinc-binding group.

Some of these MMP inhibitors have been in clinical trials for advanced cancer patients, but these trials failed because it is now clear that therapy needs to be initiated in less bulky disease. MetaBre partner CancerTek Pharmaceuticals Ltd has a new approach based on targeting phospholipase D (PLD), an important enzyme in the pathway for MMP-9 synthesis, and this will be further investigated in the MetaBre project, alongside other targets downstream of PLD. They have developed an innovative cellular assay for screening of inhibitors of MMP synthesis [patent WO0189449].

Also, the intracellular sorting, processing and secretion of MMPs is not well characterised, and research on the mechanisms and molecular machinery of MMP trafficking within MetaBre may also lead to novel targets that will provide the basis for an effective therapy.

Angiogenesis is finely tuned by a balance between activators and inhibitors. It is critical for the development of metastases and also their primary tumours, as without a sustained blood supply they cannot thrive. The dissolution of the basement membrane and surrounding ECM, followed by migration and proliferation of endothelial cells results in the establishment of the new microvessels supplying the tumour. The number and size of these microvessels is a good prognostic indicator of metastatic dissemination. It is clear that hypoxia and many oncoproteins including the c-erbB family potentiate angiogenesis and lymphangiogenesis via upregulation of vascular endothelial growth factors (VEGFs) that stimulate proliferation of endothelial cells. They also induce the expression of plasminogen activators and MMPs that allow movement of endothelial cells through the ECM, rendering these signalling pathways attractive targets for therapy.

Another pro-angiogenic factor is the soluble ectodomain of N-cadherin that is overexpressed in cancer cells. N-cadherin is also an invasion promotor and a marker for EMT. It will be of interest to characterize the enzymes responsible for the cleaving of the ectodomains in breast carcinoma. Furthermore, it is still uncertain via which receptor and which pathway isolated N-cadherin ectodomains are pro-angiogenic in endothelium. In addition to these angiogenic factors, anti-angiogenic factors (angiostatin, endostatin, etc) produced endogenously by the microenvironment through the degradation of the basement membranes have been shown to negatively regulate the growth of metastases. Their involvement in the dormancy and/or growth of bone marrow metastatic cells is, however, largely unknown.