Imagine that you are the proud owner of a successful grocery store, having found the perfect recipe for a thriving business. However, before considering expansion, you should be aware that the necessary conditions for a successful chain operation can be surprisingly different to those required for the primary store. Similarly, different properties are required for the metastasis of primary tumours to secondary sites than for primary tumorigenicity. Twelve genes have already been identified as specific suppressors of metastasis. Keller and colleagues now add RAF kinase inhibitor protein (RKIP) to this list as a suppressor of prostate cancer metastasis.

The authors had previously shown that RKIP is expressed at a lower level by the metastatic prostate cancer cell line C4-2B than by the non-metastatic cell line LNCaP from which it is derived. This result was also shown to be true for human prostate cancer samples. RKIP expression was highest for benign samples, lower for primary tumours and absent in metastases.

What is the functional relevance of RKIP expression? To answer this, C4-2B cells were stably transfected with sense RKIP to increase their level of expression of RKIP, and LNCaP cells were transfected with antisense RKIP to decrease their expression. Transfection did not alter primary tumorigenic properties of the cells, such as in vitro proliferation rate or ability to form colonies, but did affect their invasive potential. RKIP expression was found to be inversely associated with in vitro invasion.

An in vivo model was then used to confirm these findings by implanting cell clones into the prostates of mice. After 10 weeks, there were no differences in the sizes of the primary tumours, but fewer mice injected with RKIP-expressing C4-2B cells had lung metastases than mice injected with control C4-2B cells, and the average number of metastases was lower. For the few RKIP-transfected C4-2B cells that formed metastases, these secondary tumours did not express RKIP. RKIP expression was correlated with decreased vascular invasion in the primary tumour, indicating a potential mechanism by which RKIP might suppress metastasis.

Finally, the authors looked at the molecular mechanism of RKIP action. RKIP is an inhibitor of RAF-mediated signalling, so its loss of expression in metastases might lead to increased RAF signalling. This was confirmed by the observed negative correlation between the level of RKIP expression and the degree of phosphorylation (activation) of the RAF targets MEK and ERK. Inhibiting the kinase activity of MEK with the inhibitor PD-098059 decreased the invasive ability of C4-2B cells, indicating that MEK might have an important role in the regulation of metastasis through RKIP.

Just as over-ambitious expansion can be the end of many businesses, most patients with prostate cancer die from metastases rather than from the primary tumour, so this work could have important therapeutic implications. It also provides the most complete description of a metastasis suppressor so far, which might extend to other forms of cancer.