Abstract
The transcription factor B-Myb plays a critical role in regulating gene expression and is implicated in controlling carcinogenesis and cellular senescence. Transcription of the B-Myb gene is regulated by retinoblastoma proteins acting directly on the B-Myb promoter. Recently, we found that microRNAs also control the abundance of B-Myb mRNA during senescence, adding another level of complexity to B-Myb regulation. This review focuses on the importance of B-Myb in cancer and senescence, with an emphasis on the regulation of B-Myb expression and activity. Cancer Res; 71(16); 5370–3. ©2011 AACR.
Introduction
The v-Myb myeloblastosis viral oncogene homolog (avian)–like 2 (MYBL2; B-Myb) encodes a transcription factor that regulates the expression of numerous genes during cell cycle progression (1). The Myb gene family of transcription factors is present in all vertebrates (2). c-Myb is the homolog of the v-Myb oncogene, which is present in avian retroviruses that cause acute leukemia (3, 4). The other 2 family members, A-Myb and B-Myb, were cloned on the basis of homology to c-Myb (5). In mammals, c-Myb and A-Myb expression is restricted to specific cell types and stages of development, whereas B-Myb is expressed in virtually all proliferating cells (2, 5). In cultured cells, B-Myb expression is highest in the S-phase of the cell cycle, and changes in B-Myb expression have been linked to growth arrest, apoptosis, carcinogenesis, and senescence (6, 7). Thus, it is essential to determine the mechanisms that regulate expression of this important gene.
B-Myb Regulates Genes Important in Cell Proliferation and Survival
All 3 Myb proteins bind to consensus Myb-binding sites (MBS) in DNA (2), but the phenotypes of knockout mice that lack individual Myb family members are strikingly different, with mice lacking B-Myb dying very early in development because the blastocyst inner cell mass does not form (8). These diverse phenotypes show that the Myb proteins carry out different biological functions. Divergent function of the Myb proteins was also suggested by expression profiling, which revealed that ectopic expression of each of the 3 Myb proteins activates different sets of genes (9).
B-Myb stimulates transcription of genes that promote entry into the S- and M-phases of the cell cycle (1, 6). For several of these genes, such as DNA topoisomerase IIα and c-Myc, B-Myb acts by binding directly to MBSs in their promoters (see refs. 10, 11). B-Myb activates the expression of other genes that lack MBSs, including B-Myb itself, by interacting with transcription factors that bind to the promoters of these genes (12, 13). In conjunction with E2F proteins, B-Myb also stimulates the expression of genes required for the G2 to M-phase of the cell cycle (14). In addition, binding of B-Myb to the multiprotein LINC/DREAM complex regulates the ability of this complex to affect gene expression (15–18). B-Myb can also negatively repress gene expression, perhaps by competing with other transcription factors whose binding sites overlap with MBSs in target promoters (see refs. 19, 20). Finally, B-Myb associates directly with clathrin and filamin, components of the mitotic spindle (21). The absence of B-Myb reduces the amount of clathrin in the spindle and causes mitotic arrest.
The transcriptional activity of B-Myb is regulated by posttranslational modifications and by interactions with other proteins (1). B-Myb is activated by phosphorylation and acetylation, which release it from transcriptional corepressors (22, 23). However, phosphorylation also leads to ubiquitylation of the B-Myb protein (24), decreasing its half-life and limiting its activity to S-phase. B-Myb activity is also regulated by binding to transcriptional coactivators, corepressors, and other proteins (reviewed in ref. 6).
B-Myb in Cancer and Senescence
Given the provenance of c-Myb as a cellular proto-oncogene and the ability of B-Myb to regulate the expression of cell cycle genes, it is not surprising that B-Myb is involved in cell proliferation and carcinogenesis. B-Myb expression is required for entry into S-phase and can overcome growth inhibitory signals (25–27). Cytogenetic analysis of several types of cancers revealed amplification of chromosome 20q13, where B-Myb is located (see ref. 28). In addition, B-Myb overexpression occurs in several cancers and has been linked to aggressive tumor growth and poor outcomes in neuroblastomas and other tumors (28–30). Conversely, B-Myb repression can inhibit the proliferation of normal and tumor cells (11, 26, 31, 32). The ability of B-Myb to increase the expression of anti-apoptotic genes such as Bcl2, survivin, and clusterin may also contribute to cancer progression (17, 33–35). Interestingly, certain inherited sequence variants of B-Myb are associated with altered cancer risk (30, 36).
Cellular senescence, a form of irreversible growth arrest, seems to be an important obstacle that cells must bypass during carcinogenesis (37). Because B-Myb expression is strongly repressed during senescence, it seems reasonable that loss of B-Myb expression may play an important role in senescence. Several findings support this hypothesis. Repression of B-Myb inhibits proliferation of mouse BALB/c 3T3 fibroblasts, whereas constitutive B-Myb expression allows the cells to grow with reduced growth factors (32). In primary mouse embryonic fibroblasts, overexpression of the ras oncogene induces premature senescence, but coexpression of B-Myb abrogates this response (20). Our group previously showed that inhibition of B-Myb expression by short hairpin RNAs (shRNA) induces senescence in primary human fibroblasts and HeLa cervical cancer cells (38). In human embryonic lung fibroblasts, B-Myb is a direct transcriptional repressor of the cyclin-dependent kinase inhibitor p16INK4a, which is involved in the induction of senescence (39). Overexpression of B-Myb in these cells increases their in vitro lifespan, and B-Myb repression induces premature senescence. Taken together, these data suggest that B-Myb plays a central role in controlling senescence.
B-Myb Is Regulated at the Posttranscriptional Level by MicroRNAs during Senescence
B-Myb expression is low in quiescent cells because of retinoblastoma (Rb)-mediated repression (7). Repressive complexes between members of the E2F family and the Rb family (specifically E2F4/p107 and E2F4/p130) bind to E2F sites in the B-Myb promoter in G0 cells and repress its activity (40). Recently, our group discovered that B-Myb is also repressed at the posttranscriptional level by microRNAs (41).
Small noncoding microRNAs regulate gene expression by base-pairing with specific mRNA targets and affecting their translation or stability. We found that the expression of approximately 50 cellular microRNAs changes in HeLa cells undergoing Rb-induced senescence. Several members of the miR-29 and miR-30 families are upregulated in senescent HeLa cells and primary human foreskin fibroblasts. This upregulation is Rb dependent and likely to involve the oncogene c-Myc. It has been shown that c-Myc binds directly to the miR-29 and miR-30 promoters and acts as a transcriptional repressor of these genes (42). However, c-Myc itself is an E2F-responsive gene repressed by the Rb pathway during senescence (43). Taken together, these data suggest that miR-29 and miR-30 upregulation during senescence is due to Rb-driven inhibition of c-Myc expression and the consequent loss of c-Myc–mediated repression of the promoters controlling transcription of these microRNAs.
By using reporter constructs, mutational analysis of microRNA binding sites, and ectopic expression or inhibition of members of the miR-29 and miR-30 families, we showed that these microRNAs bind to the 3′ untranslated region (UTR) of B-Myb mRNA during senescence and reduce the amount of this mRNA, presumably by reducing its stability. Furthermore, overexpression of miR-29 and miR-30 inhibits the expression of endogenous B-Myb and reduces DNA synthesis. The inhibition of DNA synthesis by these microRNAs was partially rescued by exogenous expression of B-Myb, suggesting that B-Myb repression is responsible, at least in part, for growth inhibition in cells overexpressing miR-29 and miR-30. Finally, antagonizing the activity of miR-29 and miR-30 allows a population of cells to escape Rb-induced senescence, showing the importance of miR-29 and miR-30 in this process.
These experiments show the complexity of B-Myb regulation during Rb-induced senescence (Fig. 1). Activation of the Rb pathway mobilizes p107 and p130, which, in complex with E2F4, bind to the B-Myb promoter and inhibit transcription. In addition, Rb activation increases the expression of miR-29 and miR-30, which bind to B-Myb mRNA and destabilize it. As noted above, B-Myb is also regulated at the protein level by posttranslational modification and association with other cellular proteins. Presumably, this 3-tiered regulation of B-Myb evolved to ensure profound inhibition of B-Myb activity and complete blockade of cell cycle progression.
Three-tiered control of B-Myb. Activation of the Rb pathway during senescence restricts B-Myb action at multiple levels. At the DNA level, activating E2F complexes at the B-Myb promoter are replaced by inhibitory complexes between E2F4 and p107 and p130. At the RNA level, elevated levels of miR-29 and miR-30 base-pair to the 3′ UTR of B-Myb mRNA, reducing its abundance. At the protein level, reduced expression of B-Myb–responsive genes lowers cyclin-cdk activity and the extent of activating phosphorylation of B-Myb. In addition, ubiquitylation and degradation of B-Myb, as well as preferential association with corepressors, may further inhibit its ability to activate target genes.
Three-tiered control of B-Myb. Activation of the Rb pathway during senescence restricts B-Myb action at multiple levels. At the DNA level, activating E2F complexes at the B-Myb promoter are replaced by inhibitory complexes between E2F4 and p107 and p130. At the RNA level, elevated levels of miR-29 and miR-30 base-pair to the 3′ UTR of B-Myb mRNA, reducing its abundance. At the protein level, reduced expression of B-Myb–responsive genes lowers cyclin-cdk activity and the extent of activating phosphorylation of B-Myb. In addition, ubiquitylation and degradation of B-Myb, as well as preferential association with corepressors, may further inhibit its ability to activate target genes.
miR-29 and miR-30 affect the expression of numerous genes in addition to B-Myb, and many of these targets may also play a role in senescence or carcinogenesis. For example, Fabbri and colleagues have identified DNA methyltransferase 3A and 3B as targets of miR-29 (44). Similarly, in addition to miR-29 and miR-30, numerous microRNAs are upregulated or downregulated during senescence (41). These microRNAs may also regulate the expression of genes involved in senescence.
Conclusions and Perspectives
B-Myb exerts powerful effects on cell behavior. The multitude of B-Myb protein partners and target genes allows it to regulate many important cellular processes, including proliferation, senescence, apoptosis, and mitosis. On the basis of this spectrum of activities and the association of B-Myb with human cancer, it may be possible to exploit B-Myb as an important prognostic, diagnostic, and therapeutic tool. Do mutations that affect the binding of Rb proteins to the B-Myb promoter or the binding of microRNAs to B-Myb mRNA affect cancer risk by regulating levels of B-Myb? Is the presence of such mutations or the level of B-Myb itself an informative biomarker? Is it possible to influence the activity of B-Myb and affect cellular behavior by modulating its expression or its ability to bind to DNA or its protein partners, including components of the LINC/DREAM complex and the mitotic spindle? Can as-yet-unrecognized layers of complexity in the regulation of B-Myb be exploited therapeutically? The central role played by B-Myb in cellular proliferation and related processes suggests that the factors that regulate the expression and activity of B-Myb deserve special scrutiny in attempts to develop rational approaches to control cancer.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Acknowledgments
We thank Joan Steitz and members of the DiMaio laboratory for critical reading of the manuscript.
Grant Support
Work conducted in the authors' laboratory was supported by a grant from the National Cancer Institute (CA016038).