Learning therapeutic lessons from metastasis suppressor proteins

Abstract

Metastasis suppressor proteins regulate multiple steps in the metastatic cascade, including cancer cell invasion, survival in the vascular and lymphatic circulation, and colonization of distant organ sites. Understanding the biology of metastasis suppressors provides valuable mechanistic insights that may translate to therapeutic opportunities. Several reports have explored novel strategies for restoring metastasis suppressor function, including gene transfer, induction of previously suppressed gene expression and exogenous administration of gene product. Pathways activated downstream of metastasis suppressor loss can also be targeted. Although none of these strategies are yet in routine clinical use, several are being tested preclinically and in clinical trials.

Figure 1. Metastasis suppressor genes and steps in the metastatic cascade in human cancer

The biological process of metastasis is a complex cascade with multiple steps in which suppressor activity may prevent clinically apparent metastasis. Given the prevalence and mortality of human breast cancer, we focus here on this tissue type, although these principles apply to most solid malignancies. In the primary tumour, deregulation of oncogenes and tumour suppressor genes mediates the conversion of normal cells to a neoplastic phenotype. By definition, metastasis suppressor genes do not prevent these steps, but must function subsequently in the cascade as shown. Invasion of the basement membrane, stroma and vasculature constitutes one key, negative prognostic turning point in the natural history of breast and other cancer types. Many metastasis suppressor genes, including NM23 (REF. 40), DLC1 (REF. 115), KAI1 (REF. 116) and NDRG1 (REF. 70) have been shown to function in in vitro and in vivo surrogates of invasion in breast cancer cells. Nodal metastasis is also a key prognostic stage for breast cancer patients, now aggressively managed clinically. Loss of the metastasis suppressors Raf kinase inhibitor protein (RKIP) and RhoGTPase dissociation inhibitor 2 (RHOGDI2) have been shown to be associated with nodal metastasis in breast cancer. Survival during circulatory dissemination is another step of the metastatic cascade that has been studied in less detail; however, several reports have suggested that resistance to anoikis is important for metastasis and that the metastasis suppressor gene BRMS1 (breast cancer metastasis suppressor 1) increases anoikis,. Because increasing data suggest that tumours are widely disseminated on an individual cellular basis even at the time of diagnosis of localized disease, the role of metastasis suppressors in preventing outgrowth of isolated single or cellular clusters (micrometastases), known as the ‘colonization stage’ of metastasis, is compelling. NM23 (REF. 29), KAI1 (REF. 54), RHOGDI2 (REF. 94) and KISS1 (REFs 82,120) have been shown to function at this stage.

Figure 2. Strategies for restoring metastasis suppressor function

Given the key roles shown for metastasis suppressors in vivo and the association of their loss with negative outcomes in patients, re-expression of these proteins would seem a rational therapeutic strategy. This has been accomplished by strategies that are both specific — for example, re-induction of endogenous NM23 expression by medroxyprogesterone acetate (MPA) (a) — and general — for example, the use of chromatin-modifying drugs,, (b), including trichostatin A, which induces repressed gene expression by inhibiting histone deacetylases (HDACs). Exogenous viral gene therapy for re-expression of metastasis suppressor protein has also been reported (c) and has shown promise in preclinical models,,. Non-viral vectors for suppressor re-expression, including naked plasmid and cationic liposomes, have also been reported. Direct administration of the metastasis suppressor protein is another strategy (d). In at least two cases, the metastasis suppressor protein itself is a soluble, secreted factor, amenable to direct therapeutic administration, in a fashion analogous to that of insulin. In 2001, Ohtaki et al. reported the use of osmotic pumps to administer KISS1 to prevent the development of metastasis. This may proceed through induction of dormancy through KISS1R or another receptor and, for KISS1, small molecule mimetics are also under development.

Figure 3. Targeting key genes downstream of metastasis suppressors

In the case of two metastasis suppressor genes, NM23 (REFs 40,100) and RhoGTPase dissociation inhibitor 2 (RHOGDI2),, investigators have found by microarray expression profiling that loss of metastasis suppressor expression results in transcriptional deregulation. a. By re-expressing RHOGDI2 in metastatic bladder cancer cells and comparing transcripts repressed by the suppressor to genes overexpressed in invasive bladder tumours, Titus et al. discovered both endothelin 1 (ET1), which is druggable by an inhibitor of its receptor, atrasentan, and neuromedin U, which is a potent pro-metastatic gene that cannot yet be targeted therapeutically. b. By microarray profiling breast cancer cells re-expressing NM23 and mutants lacking suppressor activity, then screening differentially expressed transcripts for appropriate correlations to NM23 expression levels in human tumours, Horak et al. uncovered LPAR1, a lysophosphatidic acid receptor, among other genes, as a key downstream target potentiating motility and metastasis in breast cancer cells. An antagonist of LPAR1, Ki16425, has been reported and is being evaluated for therapeutic activity.

Figure 4. Targeting metastasis suppressor signatures through the Connectivity Map and COXen (Coexpression extrapolation)

Advanced informatic technologies including the Connectivity Map allow for screening an entire ‘metastasis suppressor-expressing cell’-type gene expression signature against gene expression changes induced by compounds, potentially systematically identifying molecules capable of inducing non-metastatic gene expression and phenotype. The list of hits is then further filtered using the COXEN algorithm, which provides a list of those suppressor signature-inducing agents that additionally have therapeutic efficacy in the NCI-60 cell line system and, most importantly, predicts function in patient tumours. Together, this integrated approach could deliver therapeutics based on metastasis suppressor biology to the clinic with a high likelihood of efficacy.

Timeline

Key advances in the metastasis suppressor field