Striking it Rich by Data Mining

detected in breast cancer, ovarian tumors, and colo-rectal tumors (reviewed in Zahnow, 2002). In contrast, C/EBP␤ null mice are refractory to Ras-mediated skin tumorigenesis (Zhu et al., 2002). Thus, disruption of sig-In this issue of Cell, Lamb et al. have used a combina-naling through C/EBP␤ appears to contribute to malig-tion of molecular genetics, DNA microarray, and data nant transformation. mining of human tumor expression databases to iden-To provide independent evidence that supports the tify a cdk-independent mechanism by which the in-role of C/EBP␤ in the mechanism of cyclin D1 action, teraction of cyclin D1 with the transcription factor Lamb et al. (2003) conducted a functional analysis of C/EBP␤ may mediate tumorigenesis. target gene promoters using conventional transfection assays. Cyclin D1-responsive regions were mapped in Although the protooncogene cyclin D1 is overexpressed the promoters of six genes, including HSP70-2, which in approximately 45% of breast cancers (Hui et al., 1996) the original microarray analysis showed to be induced as well as a broad range of other human tumor types, with similar kinetics by both wild-type and mutant cyclin the mechanisms through which cyclin D1 mediates its D1. Consensus C/EBP␤ binding sites were present in effects on tumorigenesis have not yet been elucidated. each of these promoters, and mutation of these sites The D-type cyclins represent a fundamental link be-resulted in a higher basal activity. Surprisingly, these tween signal transduction pathways and the cell cycle results suggest that C/EBP␤ acts as a constitutive re-machinery; thus, it has been suspected that cyclin/cdk pressor of cyclin D1 targets and that cyclin D1 antago-complexes act not only as essential regulators of cell nizes this repressor function. Additional support was division, but are also responsible for the oncogenic ef-provided through the use of a dominant-negative iso-fect of cyclin D1. However, several recent studies, inform of C/EBP␤, LIP, which is translated from the third cluding the analysis of cyclin D1-deficient mice, have in-frame AUG and therefore lacks most of the amino-suggested that cyclin D1 may indeed have other activi-terminal transactivation domain, but still contains the ties separate and distinct from its role as a cdk regula-DNA binding and leucine zipper dimerization domains. tory subunit and regulator of Rb (reviewed in Coqueret, LIP can recapitulate the effects of cyclin D1, suggesting 2002). Lamb et al. (2003) have now taken a molecular an alternative mechanism of regulation of target gene genetic approach to identify alternative cyclin D1 activi-expression. The …