1 Modulation of Apoptosis to Reverse Chemoresistance Gil Mor, Michele K. Montagna, and Ayesha B. Alvero Summary Interference with the innate apoptotic activity is a hallmark of neoplastic transformation and tumor formation. Modulation of the apoptotic cascade has been proposed as a new approach for the treatment of cancer. In this chapter, we discuss the role of apoptosis in ovarian cancer and the use of phenoxodiol as a model for the regulation of apoptosis and potential use as chemosensitizer for chemoresistant ovarian cancer cells. Key Words: Apoptosis; ovarian cancer; X-linked inhibitor of apoptosis protein (XIAP); phenoxodiol; caspases. 1. Introduction Ovarian cancer is the fourth leading cause of cancer-related deaths in women and is the most lethal of the gynecological malignancies (1) . One in 70 women will develop ovarian cancer and one out of 100 women will die from it. The high mortality rate is due in part to the lack of means to detect early disease such that approximately 80% of patients are initially diagnosed in advanced- staged disease. In these patients, 80–90% will initially respond to chemotherapy; however, less than 10–15% will remain in remission because of the subsequent development of chemoresistance. Treatment advances have led to improved 5-year survival, approaching 45%; however, no advances have been made in the overall survival. One way to improve survival may be by influencing the apoptotic pathways. From: Methods in Molecular Biology, vol. 414: Apoptosis and Cancer Edited by: G. Mor and A. B. Alvero © Humana Press Inc., Totowa, NJ 1 2 Mor, Montagna, and Alvero Chemotherapy in the treatment of cancer was introduced into the clinical practice more than 50 years ago. Although this form of therapy has been successful in the treatment of some forms of cancer, it has not been the case for the majority of epithelial tumors of the breast, colon, lung, and ovary. Initially, the development of chemotherapeutic agents was based on the obser- vation that tumor cells proliferate faster than normal cells. Therefore, the original strategy was to interfere with DNA replication or cellular metabolism. A better understanding of the molecular mechanisms of apoptosis and its function in normal physiology has resulted in a better understanding of the effect of chemotherapy and the mechanisms of chemoresistance. Current under- standing suggests that the induction of apoptosis in target cells is a key mechanism for most anti-tumor therapies, including chemotherapy,  -radiation, immunotherapy, and cytokines (2) . Therefore, defects in apoptosis may cause resistance . The realization that apoptosis is a key factor that contributes to the anti-tumor activity of chemotherapeutic drugs has allowed us to understand how drug resistance may arise and to look for new approaches for the treatment of cancer. 2. The Apoptotic Cascade Apoptosis is characterized by morphological changes including cell shrinkage, membrane blebbing, chromatin condensation, and nuclear fragmen- tation (3) . All these changes are the result of the activation of a cascade of intracellular factors known as caspases. Caspases are highly specific proteases synthesized as zymogens and activated by cleavage at aspartate, which generates the large and small subunits of the mature enzyme (3 , 4) . These aspartate cleavage sites are themselves caspase sites. Therefore, caspases can collaborate in the proteolytic cascade by activating themselves and each other (4 , 5) . Within these cascades, caspases can be divided into “initiator” caspases and downstream “effectors” of apoptosis. Initiator caspases, such as caspase-8 and caspase-9, mediate their oligomerization and auto-activation in response to specific upstream signals. The best-documented pathway of caspase activation is the assembly of the death-induced signaling complex (DISC) induced by the binding of the members of the death receptor family (Fas, TNF-related apoptosis inducing ligand (TRAIL), and tumor necrosis factor (TNFR1)) to its ligand (reviewed in refs 6 – 10 ). Caspase-8 is recruited to the DISC, undergoes spontaneous auto-activation, and activates downstream caspase-3 (11) ( see Fig. 1 ). Another well-described pathway is the mitochondrial pathway, which is initiated by the release of cytochrome c from the mitochondria to the cytosol, where it assembles with Apaf-1 and the caspase-9 holoenzyme, leading Modulation of Apoptosis 3 TYP E I TYP E I I Pro-Caspase 8 Caspase 8 Caspase 3 Caspase 6 and 7 Apoptosis FADD DD FA S Cytochrome c Pro - Caspase 9 Caspase 9 Pro-Caspase 8 Caspase 8 Bid FLIP XIAP Bcl 2 Fig. 1. The apoptotic cascade. Type I represents the death receptor pathway mediated by caspase-8 and Type II represents the mitochondrial pathway regulated by Bcl2 family members. to the activation of caspase-9, which also activates downstream caspase-3 (reviewed in refs 4 , 12 – 14 ). In contrast to caspase-8 and caspase-9, there are conflicting depictions for the role of caspase-2. Caspase-2 has been shown to be an important initiator caspase of the mitochondrial pathway, acting upstream of caspase-9 (15–17 ) . However, other studies show that caspase-2 activation occurs downstream of the mitochondria and caspase-9, and therefore, caspase-2 is an effector and not an initiator caspase (18) . Thus, the placement of caspase-2 in the apoptotic cascade is controversial. In addition, its mode of activation and regulation has not been characterized. The limited data available on caspase-2 are brought in part by earlier findings showing the absence of severe phenotype in caspase-2-deficient mice (19) , leading to the speculation that it does not play a significant role in apoptosis. Recent studies, however, demonstrated caspase-2’s central role in cytotoxic stress-induced apoptosis in some human cell lines (20 , 21) . In addition, our preliminary data in epithelial ovarian cancer cells (EOC) cells show that caspase-2 is one of the earliest caspase activated in response to chemotherapy and that it is essential for the induction of the 4 Mor, Montagna, and Alvero apoptotic cascade. Moreover, we have identified caspase-2 activation as one of the steps occurring in chemosensitive EOC cells in response to chemotherapy, but not in chemoresistant cells. This suggests that alterations in the regulation of caspase-2 may be one of the ways EOC cells develop chemoresistance. The effector caspases include caspase-3, caspase-6, and caspase-7, which cleave cellular substrates and precipitate apoptotic death. To date, more than 280 caspase targets have been identified (22) . These include multiple proteins involved in cell adhesion, cell cycle regulation, DNA synthesis, cleavage and repair, RNA synthesis and splicing, protein synthesis, and those that make up the cytoskeletal and nuclear structures. Our preliminary results show that two of these proteins, vimentin and keratin, are differentially expressed in cells responding to therapy and in those that are not responding. Some of these proteins may “leak” to the circulation and be detected in the serum. Thus, one of our hypotheses is that downstream targets of the apoptotic cascade are potential markers of chemoresponse that warrant further investigation. Currently, the best marker available in the clinics to detect chemoresponse is cancer antigen 125 (CA125). In first-line treatment, CA125 has an estab- lished role in monitoring the efficacy of treatment (23 , 24) . Serial changes in its concentrations are well correlated with response and survival (25) . However, in second-line treatment, where chemoresistance is widely seen, several studies identify patients in whom CA125 levels are in discordance with the radiographic evaluation of changes in tumor load (26 – 28) . The identification of sensitive and specific markers of chemoresponse will provide a way to immediately monitor response to treatment and will aid in the tailoring of therapy in ovarian cancer patients, especially those with recurrent disease and in their second line of treatment. 3. Inhibitors of Apoptosis and Chemoresistance Each step in the apoptotic cascade is delicately controlled by intracel- lular factors that can block the apoptotic pathway either at the “initiator” or “effector” level. There are three groups of proteins that are known to inhibit apoptosis ( see Fig. 1 ): (i) the bcl2 family of proteins, which stabilize the mitochondrial membranes and prevent cytochrome c release (reviewed in refs 12 , 29 , 30 ); (ii) the FADD-like interleukin-1 converting enzyme (FLICE)- inhibitory proteins (FLIP), which are caspase-8-like proteins and interfere with caspase-8 for binding to the DISC, thus preventing caspase-8 oligomerization and auto-activation (reviewed in refs 31 – 33 ); and (iii) the inhibitors of apoptosis proteins (IAP), which constitute a family of evolutionarily conserved apoptotic suppressors (reviewed in refs 34 – 36 ). All these inhibitory factors are, in the Modulation of Apoptosis 5 normal cell, inactivated in response to apoptotic signals such as hormone withdrawal, DNA damage, or activation of death receptors. In chemoresistant EOC cells, however, these blockers have been shown to remain active even after drug treatment. XIAP, which is the prototype of the IAP family of proteins, is a unique regulatory protein because of its capacity to block the apoptotic cascade both at the initiation and effector levels. It is capable of binding to pro- caspase-9, thus preventing the activation of the mitochondrial pathway, and it also binds and inhibits the effector caspases, caspase-3 and caspase-7 (37) ( see Fig. 1 ). Several studies, including our own, have shown that the intracellular blockers of apoptosis may contribute to drug resistance (34 , 35 , 38) . Indeed, we have previously shown that XIAP is highly expressed in all tested ovarian cancer samples (39) . 4. Regulation of XIAP The role of XIAP in chemoresistance is already well established. Numerous studies have demonstrated that its up-regulation confers resistance to chemotherapy and its down-regulation confers sensitivity (34 , 40 – 43) . The mechanisms of XIAP regulation, which ultimately leads to its inactivation and hence apoptosis in sensitive cells, are, however, not clearly understood. Similarly, the mechanisms that prevent its inactivation in resistant cells are unknown. Three negative regulators of XIAP have been identified: (i) XIAP- associated factor 1 (XAF-1) is a nuclear protein that binds XIAP and antago- nizes its ability to suppress the caspases (44 , 45) ; (ii) Smac/direct inhibitor of apoptosis-binding protein with low pl (DIABLO) is a mitochondrial protein that also binds and inhibits XIAP (46 , 47) ; and (iii) Omi/HtrA2 is another mitochon- drial protein that can bind XIAP and in addition, through its protease activity, can induce XIAP cleavage and inactivation (48 – 52) . Recently, the relationship between XIAP and Akt has been characterized. Dan et al. (53) showed that XIAP regulates Akt activity in response to cisplatin. Specifically, they showed that down-regulation of XIAP by cisplatin induces Akt cleavage and that XIAP over-expression induces Akt phosphorylation. Another study demonstrated that XIAP is a substrate of Akt and that phospho- rylation of XIAP by Akt stabilizes XIAP and prevents its degradation by the ubquitin-proteasome pathway (54) . Although all of these individual mechanisms may be significant, the mode of regulation and the sequence of events of XIAP inactivation, which is responsible for chemosensitivity and hence is possibly altered in chemoresistant cells, remain to be identified and characterized. We hypothesize that the cleavage of XIAP by Omi/HtrA2 is required for the full activation of the apoptotic cascade