MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression by targeting mRNAs, and are often deregulated in tumours. But miRNAs have hundreds of targets, so it is challenging to understand the relevance of a specific miRNA–target interaction to tumorigenesis, especially in mammals, in which it is expected that many interactions need to be disrupted to obtain a tumorigenic phenotype. Now David Bartel and colleagues show that disrupting the miRNA regulation of a single mammalian gene leads to a tumorigenic phenotype.

The authors examined the gene that encodes the high mobility group A2 (HMGA2) protein, which regulates gene expression by altering chromatin structure, and is expressed early during development and in many different cancers. In tumours, HMGA2 is often the target of chromosomal rearrangements that cause the loss of the C terminus of the protein and the 3′ untranslated region (3′ UTR) of the mRNA. Interestingly, the mouse Hmga2 3′ UTR contains seven conserved sites complementary to the let-7 miRNA, which is expressed late during development, suggesting that let-7 might control Hmga2 expression.

The authors showed that the introduction of let-7 into F9 mouse embryonic carcinoma cells was able to repress HMGA2 expression, whereas inhibiting endogenous let-7 increased HMGA2 expression in NIH3T3 cells. They then generated mutants of the Hmga2 3′ UTR, disrupting two, four or all seven let-7 complementary sites, and tested them in F9 cells. Strikingly, the level of Hmga2 downregulation induced by let-7 co-transfection correlated with the number of intact sites, and repression was restored by the co-transfection of a let-7 mutant that was able to bind the mutated 3′ UTR.

So, let-7 is able to directly repress Hmga2, but is disrupting this interaction sufficient to transform cells? The stable expression of a vector containing the wild-type Hmga2 open reading frame but mutated let-7 sites led to the anchorage-independent growth of NIH3T3 cells. These cells also generated tumours in immunocompromised mice, indicating that let-7 miRNA functions as a tumour suppressor through the direct repression of an oncogenic gene.

These findings suggest that loss of miRNA-mediated gene regulation is likely to be a common mechanism of tumorigenesis, and should be considered when investigating cancer-associated mutations.