ABSTRACT
The fork-head box transcription factor FoxMl is essential for hepatocellular carcinoma (HCC) development and its overexpression coincides with poor prognosis. Here, we show that the mechanisms by which FoxM1 drives HCC progression involve overcoming the inhibitory effects of the liver differentiation gene FoxA2. First, the expression patterns of FoxM1 and FoxA2 in human HCC are opposite. We show that FoxM1 represses expression of FoxA2 in G1 phase, a phase in the cell cycle in which cells can undergo differentiation. Repression of FoxA2 in G1 phase is important, as it is capable of inhibiting expression of the pluripotency genes that are expressed mainly in S/G2 phases. Using a transgenic mouse model for oncogenic Ras-driven HCC, we provide genetic evidence for a repression of FoxA2 by FoxM1. Conversely, FoxA2 inhibits expression of FoxM1, and inhibits FoxM1-induced tumorigenicity of HCC cells. Moreover, expression of FoxA2 in mouse liver expressing activated Ras inhibits FoxM1 expression and inhibits HCC progression. The observations provide strong genetic evidence for an opposing role of FoxM1 and FoxA2 in HCC progression.
AUTHOR SUMMARY Liver cancer remains untreatable because it is diagnosed at a stage when the cancer is aggressive and resistant to therapeutics. The mechanism that drives aggressive liver cancer is poorly understood. These cancers are made up of poorly differentiated cancer cells. Interestingly, the FoxM1 gene is overexpressed in the aggressive liver cancers. Although FoxM1 is important for expression of the proliferation genes, it does not explain why it is overexpressed mainly in the undifferentiated cancers. The current study addresses this puzzle. Our previous studies demonstrated that FoxM1 increases expression of the pluripotency genes that are expressed mainly in the stem-like cells. In the current manuscript we show that, in addition to activating the pluripotency genes, FoxM1 inhibits expression of the liver differentiation gene FoxA2. Overexpression of FoxM1 is important for this inhibition function, as it involves the retinoblastoma family of proteins, which are often inactivated in cancer cells, and thus, are of low-abundance. Moreover, the inhibition of FoxA2 is significant because FoxA2 could inhibit expression of the pluripotency genes as well as FoxM1. The observations provide new insights into how FoxM1 drives progression of aggressive liver cancer.