Research highlights
Metastasis: Attack of the mutant p53
Nature Reviews Cancer 9, 6 (01 June 2009) | doi:10.1038/nrc2658
Standfirst
The molecular details of how transforming growth factor-
(TGF) can both inhibit tumorigenesis and promote metastasis are unclear. Stefano Piccolo and colleagues describe a new pathway that suggests opposing roles for mutant p53 and wild-type p63 in TGF-induced metastasis and uncover a gene signature that could be clinically useful for predicting breast cancer prognosis.
Stable reconstitution of TP53-null H1299 lung cancer cells with the p53 R175H hotspot mutation, combined with TGF1 treatment, induced a mesenchymal phenotype and in vitro migration. These effects were not observed with p53R175H or TGF1 alone. In MDA-MB-231 breast cancer cells, which express endogenous mutant p53 (p53R280K), TGF1 induced migration that was dependent on mutant p53 expression and canonical Smad signalling, as assessed by short hairpin RNA knockdown of either p53R280K or the common mediator SMAD4. When injected into the mammary fat pad of immunocompromised mice, these cells formed primary tumours with similar rates and sizes, but knockdown of mutant p53 reduced the number of lymph node metastases (these form in response to autocrine TGF). Mutant p53 was also required for lung colonization following intravenous injection of MDA-MB-231 cells.
How does mutant p53 induce cell migration and invasion? A previous study showed that loss of one Trp63 allele in Trp53+/- tumours induced a metastatic phenotype similar to that induced by mutant p53 expression, suggesting a possible role for p63. Expression of either the TA or 
Np63
isoform of p63 inhibited TGF1-induced in vitro migration of MDA-MB-231 cells, and knockdown of p63 in H1299 cells increased migration. Loss of p63 in MDA-MB-231 cells also rescued TGF1-induced migration and lung colonization in cells lacking p53, placing p63 downstream of mutant p53 in the pathway.
As mutant p53 cannot complex with Np63, the p63 isoform most commonly expressed in epithelium, the authors hypothesized that Smads, activated downstream of TGF, might link p53 and p63; this was confirmed by co-immunoprecipitation experiments. In vitro binding assays indicated that receptor-regulated Smads (R-Smads) serve as a bridge between the domain of Np63 and the TA domain of mutant p53 (notably the TA domain is lacking in Np63). Complex formation was also dependent on oncogenic Ras, which has previously been shown to be required for TGF-induced metastasis.
How does this ternary mutant p53–R-Smad–Np63 complex increase metastasis? In the presence of TGF1, mutant p53 blocks both the binding of p63 to known p63 DNA-binding sites and p63-dependent transcription. Surprisingly, only five genes were found to be co-regulated by mutant p53 and TGF1 in MDA-MB-231 cells, and the authors chose to focus their analysis on two negatively regulated genes (CCNG2, which encodes cyclin G2, and the basic helix–loop–helix family member BHLHE41, also known as SHARP1) that are therefore likely metastasis suppressors. Loss of p63 or TGF1 treatment inhibited expression of CCNG2 and SHARP1, and knockdown of either gene restored TGF1-induced migration in cells lacking mutant p53. The molecular mechanisms by which cyclin G2 and SHARP1 restrain metastasis remain to be determined.
Finally, the authors examined the clinical relevance of this pathway. Using accessible gene expression datasets from primary breast tumours, they found that low expression of CCNG2 and SHARP1 correlated with a higher probability of recurrence compared with tumours that had high expression of these genes. Furthermore, this two-gene 'minimal signature' was able to stratify patients with stage 2 tumours of uncertain prognosis into those with good or bad prognoses, and performed comparably to the 70-gene breast cancer signature (used in the MammaPrint test) in predicting clinical outcome.
