BMP activated Smad signaling strongly promotes migration and invasion of hepatocellular carcinoma cells.

Several of the different bone morphogenetic proteins (BMPs) are involved in development and progression of specific tumors. For hepatocellular carcinoma (HCC) only BMP4 and BMP6 are described to be important for carcinogenesis. However, up to now neither the influence of other BMPs on tumor progression, nor the responsible signaling pathways to mediate target gene expression in HCC are known. In order to characterize BMP expression pattern in HCC cell lines, we performed RT-PCR analysis and revealed enhanced expression levels of several BMPs (BMP4, 6, 7, 8, 9, 10, 11, 13 and 15) in HCC. Thus, we treated HCC cells with the general BMP inhibitors chordin and noggin to determine the functional relevance of BMP overexpression and observed decreased migration and invasion of HCC cells. A cDNA microarray of noggin treated HCC cells was performed to analyze downstream targets of BMPs mediating these oncogenic functions. Subsequent analysis identified collagen XVI as 'Smad signaling specific' and nidogen-2 as 'MAPK/ERK signaling specific' BMP-target genes. To examine which signaling pathway is mainly responsible for the oncogenic role of BMPs in HCC, we treated HCC cells with dorsomorphin to determine the influence of BMP activated Smad signaling. Interestingly, also migratory and invasive behavior of dorsomorphin treated HCC cells was diminished. In summary, our findings demonstrate enhanced expression levels of several BMPs in HCC supporting enhanced migratory and invasive phenotype of HCC cells mainly via activation of Smad signaling.

Downregulation of hemojuvelin prevents inhibitory effects of bone morphogenetic proteins on iron metabolism in hepatocellular carcinoma.

Recently, we revealed that bone morphogenetic protein (BMP) 4 is increased in hepatocellular carcinoma (HCC). Furthermore, latest reports described BMPs, in particular BMP6, as important regulators of hepcidin expression in iron homeostasis. Therefore, we aimed to unravel why enhanced BMP expression in HCC patients does not lead to severe changes in iron metabolism. Initial analysis of the BMP4 and BMP6 expression patterns revealed enhanced expression on mRNA and protein level in HCC cell lines and tissue samples compared with primary human hepatocytes (PHHs) and normal liver tissues. However and interestingly, hepcidin expression was reduced in HCC cell lines and tissues. Analysis of BMP6 receptor expression revealed loss of BMP6-specific receptor subunit in HCC. To identify a possible regulatory mechanism causing lack of reaction to BMP4 we analyzed the expression of hemojuvelin (HJV), which is involved in iron metabolism as BMP co-receptor. HJV expression was markedly decreased in HCC cell lines and tissues. HJV promoter analysis revealed potential HNF-1α and snail-binding sites, but functional analysis ruled out that these transcriptional regulators or promoter methylation are the cause of HJV downregulation in HCC. However, we identified AU-rich elements in the HJV 3'-untranslated region and revealed significantly faster decay of HJV mRNA in HCC cells as compared with PHH indicating decreased mRNA-stability as the reason for the loss of HJV expression in HCC.

Iron-induced expression of bone morphogenic protein 6 in intestinal cells is the main regulator of hepatic hepcidin expression in vivo.

BACKGROUND & AIMS: Recent studies identified bone morphogenic protein 6 (BMP6) as a key regulator of hepatic hepcidin expression and iron metabolism, but the cellular source of BMP6 and the reason for its specific effect on hepatocytes are unknown. METHODS: BMP and hepcidin expression upon iron sensing were analyzed in vivo in BMP6(-/-) and BMP6(+/+) mice and ex vivo in tissue and in vitro in cells of the liver and the small intestine. RESULTS: BMP6(-/-) mice developed severe hepatic iron accumulation and reduced hepcidin expression with increasing age. This phenotype could be triggered in younger BMP6(-/-) mice by dietary or parenteral iron application. Furthermore, both treatments induced a marked up-regulation of BMP6 expression in the small intestine of BMP6(+/+) mice. Ex vivo treatment of intestinal tissue of BMP6(+/+) mice with iron sulfate or holo-transferrin confirmed epithelial cells as an inducible source of BMP6. In contrast, iron overload did not promote a striking induction of BMP6 expression in hepatocytes or macrophages. Furthermore, iron-supplemented diet induced a compensatory up-regulation of BMP2, BMP4, and BMP9 in the small intestine of BMP6(-/-) mice that was apparently not sufficient to assure iron homeostasis. As a potential explanation, analysis of hepatocytes revealed an expression pattern of BMP receptor subunits preferentially used by BMP6, and treatment of hepatocytes with different recombinant BMPs identified BMP6 as the most potent stimulator of hepcidin expression. CONCLUSIONS: Epithelial cells of the small intestine are the predominant cellular source of BMP6 upon iron sensing. Our findings reveal a previously unknown mechanism in which the small intestine controls iron homeostasis.

Evidence for a lack of a direct transcriptional suppression of the iron regulatory peptide hepcidin by hypoxia-inducible factors.

BACKGROUND: Hepcidin is a major regulator of iron metabolism and plays a key role in anemia of chronic disease, reducing intestinal iron uptake and release from body iron stores. Hypoxia and chemical stabilizers of the hypoxia-inducible transcription factor (HIF) have been shown to suppress hepcidin expression. We therefore investigated the role of HIF in hepcidin regulation. METHODOLOGY/PRINCIPAL FINDINGS: Hepcidin mRNA was down-regulated in hepatoma cells by chemical HIF stabilizers and iron chelators, respectively. In contrast, the response to hypoxia was variable. The decrease in hepcidin mRNA was not reversed by HIF-1alpha or HIF-2alpha knock-down or by depletion of the HIF and iron regulatory protein (IRP) target transferrin receptor 1 (TfR1). However, the response of hepcidin to hypoxia and chemical HIF inducers paralleled the regulation of transferrin receptor 2 (TfR2), one of the genes critical to hepcidin expression. Hepcidin expression was also markedly and rapidly decreased by serum deprivation, independent of transferrin-bound iron, and by the phosphatidylinositol 3 (PI3) kinase inhibitor LY294002, indicating that growth factors are required for hepcidin expression in vitro. Hepcidin promoter constructs mirrored the response of mRNA levels to interleukin-6 and bone morphogenetic proteins, but not consistently to hypoxia or HIF stabilizers, and deletion of the putative HIF binding motifs did not alter the response to different hypoxic stimuli. In mice exposed to carbon monoxide, hypoxia or the chemical HIF inducer N-oxalylglycine, liver hepcidin 1 mRNA was elevated rather than decreased. CONCLUSIONS/SIGNIFICANCE: Taken together, these data indicate that hepcidin is neither a direct target of HIF, nor indirectly regulated by HIF through induction of TfR1 expression. Hepcidin mRNA expression in vitro is highly sensitive to the presence of serum factors and PI3 kinase inhibition and parallels TfR2 expression.

Bone morphogenetic protein 4 is induced in hepatocellular carcinoma by hypoxia and promotes tumour progression.

Striking similarities exist between molecular mechanisms driving embryonic liver development and progression of hepatocellular carcinoma (HCC). Bone morphogenetic proteins (BMPs), particularly BMP4, have been proposed to regulate embryonic hepatic development. BMP expression has been observed in neoplasia but the expression and biological role of BMP4 in human HCC are unknown. We found increased BMP4 mRNA and protein in HCC cell lines and tissue samples compared to primary human hepatocytes and corresponding non-tumourous tissue. Hypoxia further induced BMP4 expression in HCC cells, which was abolished by transfection of a dominant negative form of HIF-1 alpha (dnHIF-1 alpha). However, gel shift assays revealed only minor binding activity in nuclear extracts from (hypoxic) HCC cells to a putative hypoxia-response element in the BMP4 promoter. Sequence analysis of the BMP4 promoter revealed two Ets-1 binding sites, and Ets-1 activity was increased in HCC cells under hypoxic conditions. Transfection of dnHIF-1 alpha completely abrogated hypoxia-induced Ets-1 activity as well as BMP4 expression. Overexpression of Ets-1 markedly enhanced BMP4 promoter activity, while antisense Ets-1 almost completely abolished basal as well as hypoxia-induced BMP4 expression. These data demonstrate that Ets-1 activity contributes to baseline expression of the BMP4 gene and is the predominant mediator of the HIF-dependent BMP4 induction under hypoxic conditions. To determine the functional relevance of BMP4 expression, HCC cell lines were treated with antisense BMP4 constructs or siRNA against BMP4. BMP4 suppression resulted in a strong reduction of the migratory and invasive potential and anchorage-independent growth. Furthermore, tube formation assays indicated that BMP4 expressed by HCC cells promotes vasculogenesis. Our findings demonstrate that BMP4 is increased in HCC and promotes HCC progression. Therefore, BMP4 expression may have clinical relevance, and interfering with BMP4 signalling appears as an attractive therapeutic target for this highly aggressive tumour.

GLUT1 expression is increased in hepatocellular carcinoma and promotes tumorigenesis.

Accelerated glycolysis is one of the biochemical characteristics of cancer cells. The glucose transporter isoform 1 (GLUT1) gene encodes a key rate-limiting factor in glucose transport into cancer cells. However, its expression level and functional significance in hepatocellular cancer (HCC) are still disputed. Therefore, we aimed to analyze the expression and function of the GLUT1 gene in cases of HCC. We found significantly higher GLUT1 mRNA expression levels in HCC tissues and cell lines compared with primary human hepatocytes and matched nontumor tissue. Immunohistochemical analysis of a tissue microarray of 152 HCC cases revealed a significant correlation between Glut1 protein expression levels and a higher Ki-67 labeling index, advanced tumor stages, and poor differentiation. Accordingly, suppression of GLUT1 expression by siRNA significantly impaired both the growth and migratory potential of HCC cells. Furthermore, inhibition of GLUT1 expression reduced both glucose uptake and lactate secretion. Hypoxic conditions further increased GLUT1 expression levels in HCC cells, and this induction was dependent on the activation of the transcription factor hypoxia-inducible factor-1alpha. In summary, our findings suggest that increased GLUT1 expression levels in HCC cells functionally affect tumorigenicity, and thus, we propose GLUT1 as an innovative therapeutic target for this highly aggressive tumor.