GFP-stained mouse lung metastases.

From Nature 449, 682–688 (2007).

The invasion-metastatic cascade involves a series of steps whereby tumour cells leave the primary tumour, intravasate into the circulation, extravasate at distant tissues, and establish micrometastases, some of which may grow into macroscopic secondary tumours. MicroRNAs (miRNAs) are emerging as key regulators of human tumour progression, but their specific roles at different steps of this cascade have not been yet described. In Cell, Robert Weinberg and colleagues now report that miR-31 specifically inhibits breast cancer metastasis by repressing a cohort of metastasis-promoting genes.

Having surveyed a group of ten mouse and human breast cancer-associated miRNAs, the authors identified miR-31 as being specifically attenuated in aggressive cell lines. miR-31 overexpression in MDA-MB-231 human breast cancer cells dramatically reduced invasiveness, motility and anoikis resistance in vitro. When injected into mouse mammary fat pads, MDA-MB-231 cells forced to express miR-31 did not invade surrounding tissues and formed 95% fewer lung metastases than controls, showing that miR-31 also suppresses metastasis in vivo. The authors injected the cells directly into the circulation of mice to show that miR-31 not only affected the initial steps of the metastatic cascade, but also the later stages including intraluminal viability, extravasation and survival in the lung parenchyma, and eventual colonization of the tissue.

Robert Weinberg and colleagues went on to identify targets of miR-31. Among 200 putative target mRNAs identified by algorithm. six genes — frizzled 3 (Fzd3), integrin 5 (ITGA5), myosin phosphatase-Rho interacting protein (M-RIP), matrix metallopeptidase 16 (MMP16), radixin (RDX) and RhoA — were confirmed to be inhibited by miR-31 through their 3'UTRs, as determined using a reporter assay. Furthermore, the individual expression of Fzd3, ITGA5, RDX and RhoA versions insensitive to miRNAs partly reverted the malignant phenotypes suppressed by miR-31 in vitro, showing that these four genes are functionally relevant effectors of miR-31. RhoA's role was further assessed in vivo by re-expressing it in miR-31-containing MDA-MB-231 cells. Injection of these cells into mice partially restored lung metastases, suggesting that miR-31's ability to prevent metastasis requires RhoA inhibition, but also that miR-31's total effect depends on the concomitant repression of multiple targets.

This work shows that a single miRNA — miR-31 — might prevent breast cancer metastasis through coordinated repression of a group of genes. Importantly, miR-31 expression was inversely correlated with the clinical progression of different human breast tumours, making miR-31 a good candidate marker for metastasis in a variety of breast cancer subtypes.

Author: Kim Baumann