Search

Purpose Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related death worldwide with. lack of effective systemic chemotherapy. In this study, we aimed to evaluate the value of ATPase family AAA domain-containing protein 2 (ATAD2) as a biomarker and potential therapeutic target for HCC. Methods The expression of ATAD2 was tested in different HCC patient cohorts by immunohistochemistry and comparative transcriptional analysis. The co-expression of ATAD2 and proliferation markers was compared during liver regeneration and malignancy with different bioinformatics tools. The cellular effects of ATAD2 inactivation in liver malignancy was tested on cell cycle, apoptosis and colony formation ability as well as tumor formation using RNA interference. The genes affected by ATAD2 inactivation in three different HCC cell lines were identified by global gene expression profiling and bioinformatics tools. Results ATAD2 is overexpression is closely correlated with HCC tumor stage. There was gradual increase from dysplasia, well differentiated and poorly differentiated HCC, respectively. We also observed transient upregulation of ATAD2 expression during rat liver regeneration in parallel to changes in Ki-67 expression. ATAD2 knockdown resulted in apoptosis and decreased cell survival in vitro and decreased tumor formation in some HCC cell lines. However, three other HCC cell lines tested where not affected. Similarly, gene expression response to ATAD2 inactivation in different HCC cell lines was highly heterogeneous. Conclusions ATAD2 is a potential proliferation marker for liver regeneration and HCC. It may also serve as a therapeutic target despite heterogeneous response of malignant cells.

BACKGROUND: Breast cancer is a remarkably heterogeneous disease. Luminal, basal-like, "normal-like", and ERBB2+ subgroups were identified and were shown to have different prognoses. The mechanisms underlying this heterogeneity are poorly understood. In our study, we explored the role of cellular differentiation and senescence as a potential cause of heterogeneity. METHODOLOGY/PRINCIPAL FINDINGS: A panel of breast cancer cell lines, isogenic clones, and breast tumors were used. Based on their ability to generate senescent progeny under low-density clonogenic conditions, we classified breast cancer cell lines as senescent cell progenitor (SCP) and immortal cell progenitor (ICP) subtypes. All SCP cell lines expressed estrogen receptor (ER). Loss of ER expression combined with the accumulation of p21(Cip1) correlated with senescence in these cell lines. p21(Cip1) knockdown, estrogen-mediated ER activation or ectopic ER overexpression protected cells against senescence. In contrast, tamoxifen triggered a robust senescence response. As ER expression has been linked to luminal differentiation, we compared the differentiation status of SCP and ICP cell lines using stem/progenitor, luminal, and myoepithelial markers. The SCP cells produced CD24+ or ER+ luminal-like and ASMA+ myoepithelial-like progeny, in addition to CD44+ stem/progenitor-like cells. In contrast, ICP cell lines acted as differentiation-defective stem/progenitor cells. Some ICP cell lines generated only CD44+/CD24-/ER-/ASMA- progenitor/stem-like cells, and others also produced CD24+/ER- luminal-like, but not ASMA+ myoepithelial-like cells. Furthermore, gene expression profiles clustered SCP cell lines with luminal A and "normal-like" tumors, and ICP cell lines with luminal B and basal-like tumors. The ICP cells displayed higher tumorigenicity in immunodeficient mice. CONCLUSIONS/SIGNIFICANCE: Luminal A and "normal-like" breast cancer cell lines were able to generate luminal-like and myoepithelial-like progeny undergoing senescence arrest. In contrast, luminal B/basal-like cell lines acted as stem/progenitor cells with defective differentiation capacities. Our findings suggest that the malignancy of breast tumors is directly correlated with stem/progenitor phenotypes and poor differentiation potential.