Purpose: Hepatocellular Carcinoma (HCC) is one of the most malignant cancers worldwide. Due to the male prevalence of HCC, androgen signals have been suspected to be involved in the disease process. Although over the past fifty years there have been studies showing that androgens plays positive roles in HCC progression, controversial results with antiandrogens on patients’ prognosis led to puzzles in the past decade. An increasing number of studies pointed out the distinct functions between androgens and the androgen receptor (AR) in cellular and molecular regard. Therefore, approaches to dissect AR roles in hepatocyte transformation and HCC progression are necessary to understand the disease. Methods: Using conditional knockout of AR specifically in the hepatocyte (L-ARKO), without changing the serum testosterone level, we characterized the pathophysiological function of AR in HCC. Our previous works illustrated the oncogenic role of AR to promote hepatocarcinogenesis in the early stage of cancer development. Here we demonstrated the AR function in the later stage of cancer progression and delineated the possible mechanisms. Results: To our surprise, the L-ARKO mice had poorer survival, higher lung metastasis rate, and HCC tumors of more malignant cancer variants, less collagen deposition, and higher cellular proteolysis activity compared to wild type mice. To explain the invasive cancer phenotype observed in the L-ARKO, we found AR suppressed cancer cells invasion through ablation of NFκB activation, and indirectly inhibited MMP9 expression. Except for the suppression effect of AR on cell invasion, we found cell focal adhesion and anoikis might be one of the potential mechanisms to explain the pathological transition of AR on cancer progression. This is the first demonstration that AR promotes cancer growth and death in different conditions. We found AR promotes HCC cell growth through expression of the integrin family, ECM (extracellular matrix) receptor, and enhanced cell focal adhesion, ultimately enhanced cell anti-apoptotic capacity in the adhered environment. On the other hand, AR also enhances cell anoikis (detached cell death) through facilitated amorphosis (actin filament destruction). Conclusion: This report explains that the failure of antiandrogens in patients could be due to differential performance of AR at different stages of HCC. This report provides another evidence for “dependence receptor” hypothesis to explain cancer metastasis underlying the AR effect. In addition, diverging from the conventional biological functions of AR in anti-apoptosis, we are the first to demonstrate detached cell death in the HCC model. Finally, this report not only explains the failure of antiandrogens, but also suggests the combinatorial therapy potential of antiandrogens (or anti-AR) and anti-metastatic agents in the clinical practice.

99th AACR Annual Meeting-- Apr 12-16, 2008; San Diego, CA