Upon receiving the AACR Outstanding Investigator Award for Breast Cancer Research, Yibin Kang, PhD, Princeton University, Princeton, New Jersey, USA, discussed his research into the origin of metastatic traits in breast cancer (BC). Because the majority of patients with BC will develop metastatic disease, identification of the mechanisms of metastasis will provide new therapeutic targets.

Key questions include, what genes give metastatic tumor cells the ability to escape from the intrinsic constraints of the epithelium and initiate new tumors in a distinct microenvironment, and when does this occur? One level of research involves examining the natural heterogeneity of cells in tumors (intratumoral heterogeneity) to identify metastasis genes.

Another level involves looking at intertumoral heterogeneity, important because tumors of the same type from different patients behave differently. One hypothesis for this is the cell-of-origin model, in which tumors develop from different normal cells: aggressive tumors develop from stemlike cells, which become more metastatic after oncogenic events. Other tumors, derived from fully differentiated cells, are less likely to metastasize.

Another hypothesis, the oncogenic driver mutation model, suggests that poor-prognosis tumors result from oncogenic driver events during tumor initiation. Tumors may come from the same cell of origin but, through different oncogenic driver events, give rise to tumor cells that have little chance of metastasizing; another oncogenic driver event may lead to formation of highly aggressive metastatic tumors.

One amplicon associated with poor prognosis in breast cancer was identified by a computational biology in 8q22 [Hu G et al. Cancer Cell. 2009]. This small region contains about a dozen genes, half of which are highly differentially expressed. These were cloned and overexpressed in tumor cells to determine which could drive lung metastases in a mouse model, leading to the identification of one gene, metadherin (MTDH), a poor prognosis marker and functional driver of breast cancer metastasis.

Knockout of MTDH does not affect embryonic or postnatal development in mice; in a model of a highly aggressive tumor, it slows tumor progression, reduces total tumor burden, and eliminates metastases. Tumor formation in MTDH knockouts in an in vivo model of tumor initiation is shown in Table 1.

Staphylococcal nuclease domain containing 1 (SND1) is associated with poor prognosis and interacts with MTDH. Knockdown of SND1 reduces lung metastases and sensitizes cells to chemotherapy-induced apoptosis [Blanco MA et al. J Biol Chem. 2011; Wan L et al. Cancer Cell. 2014]. Dr Kang’s group has used crystal structural analysis to determine the sites of interaction between MTDH and SND1 [Guo F et al. Cell Reports. 2014]. Mutation at either 1 of 2 tryptophans eliminates the binding of MTDH to SND1 and the ability of MTDH to rescue tumor-initiating functions. The group is now screening small molecular compounds to interfere with the cooperative activity of MTDH and SND1 in tumor initiation.

Genes that confer survival advantage to primary tumor cells during tumor initiation may play crucial roles in metastasis via several mechanisms, including promotion of the development of tumor-initiating cells or by mediation of crosstalk between tumor cells and their environment. Therapeutic targeting of metastasis genes may have a far-reaching impact on the prevention and treatment of metastatic diseases.