Oncogenic HMGA2: short or small?

The High Mobility Group A (HMGA) family of proteins are architectural transcription factors that bind to DNA and introduce structural alterations in chromatin. The HMGA family consists of two members, HMGA1 and HMGA2, which are encoded by separate genes. As one of the major nonhistone chromosomal proteins, HMGA proteins are multifunctional and are involved in many fundamental cellular processes, including gene regulation, cell cycle, differentiation, and viral integration (Reeves 2001). Not surprisingly, therefore, HMGA mutations contribute to many common diseases, including benign and malignant tumors (Sgarra et al. 2004), obesity (Anand and Chada 2000), diabetes (Foti et al. 2005), and atherosclerosis (Schlueter et al. 2005). HMGA proteins are relatively abundant in the early embryo, where cells are proliferating rapidly, whereas they are undetectable in terminally differentiated cells (Sgarra et al. 2004). hmga2-deficient mouse embryonic fibroblasts (MEFs) grow more slowly than wild-type MEFs (Zhou et al. 1995), suggesting that HMGA2 confers a growth advantage. Consistent with this observation, HMGA is indeed oncogenic and is up-regulated in many tumors (see below). However, it was recently shown that HMGA proteins are also involved in cellular senescence, a state of “permanent” cell cycle arrest, adding further diversity to the many facets of the HMGA proteins (Narita et al. 2006). HMGA proteins are primarily, but not exclusively, upregulated in tumors of mesenchymal origin. Up-regulation of the HMGA gene products often results from genetic alterations such as gene amplification and translocation. Rearrangement of the HMGA2 gene by reciprocal translocations in particular, is frequently observed in benign tumors such as lipomas, lung hamartomas, uterine leiomyomas, endothelial polyps, fibroadenomas, and adenolipomas of the breast (Ashar et al. 1995; Kazmierczak et al. 1995, 1996; Schoenmakers et al. 1995). HMGA is also involved in malignant tumors such as liposarcomas and osteosarcomas, acute lymphoblastic leukemia, as well as many lung carcinomas (Berner et al. 1997; Xu et al. 2004; Sarhadi et al. 2006). The HMGA2 gene, which is located in the q13-15 segment of chromosome 12, consists of five exons. Each of the first three exons encodes for an AT-hook domain, so called because they bind to the minor groove of AT-rich stretches of DNA. Exons 4 and 5 encode the linker and acidic tail regions, respectively. In most HMGA2 translocations, the breakpoint occurs within the ∼140-kb third intron, giving rise to chimeric transcripts that are a fusion of HMGA2’s three AT-hooks with various other genes (Goodwin 1998). It was therefore presumed that the fusion proteins themselves were the cause of the tumorigenicity, as exemplified by the classical oncogenic translocations PMLRARa and BCR-ABL. Surprisingly, however, a truncated form of HMGA2, containing only the three AT-hook domains, was sufficient for transformation in vitro. This finding raises the possibility that the loss of the acidic tail, rather than the generation of chimeric proteins, is the critical step in causing HMGA oncogenicity (Fedele et al. 1998). To date, multiple hmga2 transgenic mouse models have been generated, which constitutively express either the full-length or the truncated form of the protein (Arlotta et al. 2000; Fedele et al. 2002). Although these mice all develop benign tumors, their tumor spectrum is slightly different; the truncated form of HMGA2 gives rise to gigantism associated with lipomatosis, while full-length HMGA2 leads to pituitary tumors with milder lipomatosis. Recently, new transgenic mouse models were generated, in which either a full-length or a truncated from of HMGA2 was expressed under a promoter specific for differentiated mesenchymal tissues (Zaidi et al. 2006). Both transgenic models developed various benign mesenchymal tumors, yet the severity of the tumor phenotype was not affected by the presence or absence of the C-terminal acidic tail. These findings, in contrast to the original hmga2 transgenic models that ubiquitously and constitutively express HMGA2, suggest that the misexpression of HMGA2 itself is sufficient for tumorigenesis, at least in the mesenchymal lineage. Thus, how translocations in the HMGA2 locus lead to tumor development is a matter of debate. Now, Lee and Dutta (2007) in this issue of Genes & Development, together with others (Hebert et al. 2007; Lee and Dutta 2007; Mayr et al. 2007; Wang et al. 2007), provide new insight into this long-standing debate, indicating that the chromosomal translocations within HMGA2 may contribute to tumorigenesis by removing HMGA2’s 3 untranslated region (UTR). These reports identify HMGA2 as a target of the let-7 family of microRNAs (miRNAs). miRNAs encode a class of small, regulatory noncoding RNAs, which have come to prominence as a key mechanism of post-transcriptional reguCorresponding author. E-MAIL [email protected]; FAX 44-1223-404208. Article is online at http://www.genesdev.org/cgi/doi/10.1101/gad.1554707.