CONTRIBUTIONS OF EPIGENETICS TO DRUG RESISTANCE IN CANCER STEM CELLS
Several mechanisms have been hypothesized as a basis for drug resistance in solid tumors, including drug inactivation, quiescence, and increased DNA repair in chemoresistant tumor cells. The classical drug-resistance model invokes drug selection of clones possessing mutations (i.e., increased DNA repair, quiescence, etc.) that confer a survival advantage from various drug insults (e.g., platinum damage of DNA); each cell within the tumor would have an equal probability of gaining such mutations. However, the above-mentioned phenotypes also describe tissue stem cells. These normal progenitor cells are generally long-lived (thus quiescent or slowly proliferating) and therefore require survival-conferring characteristics such as enhanced DNA repair and the expression of membrane transporters that allow increased resistance to environmental toxins. As a consequence of the cancer stem cell theory, it is believed that malignancies that initially undergo complete remission but subsequently relapse to a completely refractory state (e.g., ovarian cancer and anti-hormone-refractory breast and prostate cancers) are more likely to possess tumor stem cells than are cancers that do not respond well to primary therapy (e.g., pancreatic cancer). In that model, chemotherapeutics preferentially target transit-amplifying cells, causing tumor regression, but fail to eradicate drug-resistant CSCs. Following «complete» remission of detectable disease, CSCs proliferate and probably confer the drug-resistant phenotype to their progeny, resulting in a tumor that is now fully refractory to further treatment, as occurs in patients suffering from several advanced-stage malignancies. Consequently, therapies are greatly needed that specifically target the small percentage of tumor progenitors in addition to the nontumorigenic progeny that comprise the bulk of the tumor mass. As epigenetic modifications are the primary driving force for differentiation in normal somatic cells, it is highly likely that these also play a role in maintaining differentiation states in tumors (see below), necessitating a comprehensive study of such modifications (and agents capable of their reversal) in malignant progenitors. As the Polycomb repressor EZH2 is widely implicated in aggressive malignancies, it is likely that this enzyme plays a significant role in CSC self-renewal and multipotency. In this regard, it was recently demonstrated that in vivo delivery of a small inhibitory RNA against E2H2 could abolish human prostate xenograft bone metastases in athymic mice. Moreover, as histone deacetylases (HDACs) and DNA methyltransferases (DNMTs) may associate with the PRC2 complex it is feasible that inhibition of these enzymes could disrupt PRC2 activity and, consequently, stem cell establishment. In support of such a hypothesis, it was demonstrated that treatment with HDAC inhibitors could deplete EZH2 and other PRC2 component proteins from acute leukemia cells. We and others have also shown that HDAC and DNMT inhibitors, alone and in combination with conventional chemotherapies, are strongly antiproliferative to drug-resistant human cancer cells, suggesting that these therapies may hold the potential to directly target tumor stem cells.