Multicellular organisms often need to eliminate cells for normal development and homeostasis through a process now known as apoptosis. Apoptosis is also known as programmed cell death and leads to a highly regulated destruction of a cell. It counters cell division and ensures proper function in a variety of tissues, including the reproductive organs and the immune system (Osborne, 1996; Tilly, 1996). Loss of apoptosis regulation will lead to a wide variety of diseases and developmental defects. Apoptosis may be induced in order to remove abnormal cells. It is a fundamental requirement for maintaining healthy ovarian function (Tilly, 1996; Morita and Tilly, 1999). The mechanisms of apoptosis have complicated pathways for regulation that require several families of proteins.
The beginning of the apoptosis process requires the generation of a death signal. Messenger molecules (cytokines, TNF-α, FASL, etc.) often initiate the death signal under normal circumstances. Radiation and some chemical and pharmacological agents will also trigger the signal cascade, which induces apoptosis. The death signal needs to be interpreted by the cell and molecular events determine the gravity of the signal against factors that promote cellular survival (Hengartner, 2000).
Deregulation of the apoptosis signal control can result in various disease states. One of the major diseases associated with the loss of apoptosis regulation is autoimmune disease. An example of this was reported to occur in infants and children, when a lack of the tissue necrosis factor receptor (TNFR) family member CD95 (Fas) resulted in lymphadenopathy, splenomegally and a high level of circulating antibodies (Fisher et al., 1995; Le Deist et al., 1996). Furthermore, overexpression of the anti-apoptotic protein Bcl-2, CD95, and TNF-α were reported to be common markers of systemic lupus erythemaosus in 51 patients with the disease (Miret et al., 2001). These reports indicate the importance of apoptosis regulation in the regulation of normal immune function.
Organ and graft rejection is mediated by cytokines and cytotoxic T-lymphocyte permeation of the donor tissue (Heusel et al., 1994). It has also been reported that apoptosis may have a critical role in neuodegenerative diseases (Gschwind and Huber, 1995; Roy et al., 1995; Ham et al., 2000).
Cancer can result from a lack of proper apoptotic regulation. It has been reported that initial development of ovarian tumors is accompanied by high expression of CD95, CD95 ligand, and Bcl-2 proteins; also, the common lymphoid malignancy follicular lymphoma results from the overexpression of Bcl-2 (Tsujimoto et al., 1985; Ben-Hur et al., 1999). It is well known that many ovarian carcinomas have a defect in the p53 gene or its regulation (Marks et al., 1991). It has also been reported that p53 alterations correlate significantly with resistance to chemotherapy, early relapse, and shortened survival rates in ovarian cancer patients (Reles et al., 2001). P53 is a tumor suppressor gene that can regulate pre-cancerous cells in two ways: (1) arrest of the cell cycle to allow for the repair of the genomic DNA, (2) generation of a signal that induces apoptosis. Inheriting a mutated allele of p53 causes the individual to be highly susceptible to cancer (Malkin et al., 1990; Srivastava et al., 1990; Donehower et al., 1992).
APOPTOSIS VERSUS NECROSIS
Necrosis is another form of cell death that is caused by a range of noxious chemicals, biological agents, or physical damage. Typically, necrosis occurs in groups of cells that have been damaged in the same way. It is characterized by the rupturing of cellular membranes, swelling of the cells, and the destruction of the cellular structures. Inflammation is associated with necrosis due to the release of cytosolic proteins to the intercellular space.
Apoptosis generally occurs in individual cells independent of neighboring cells and/or the intracellular matrix. An apoptotic cell loses volume and shrinks as opposed to the swelling associated with necrosis. Importantly, apoptosis is a coordinated process that requires energy. Caspase activity characterizes apoptosis. The activity of caspases inactivates DNA repair proteins, destroys the cellular matrix, and activates proteins that enhance the apoptosis signal. The outer membrane is left intact, while the nuclear structure breaks down releasing chromatin, nucleic acids, and other proteins associated with the interior of the nucleus to the cytosol. Neighboring cells or macrophages will phagocytize apoptotic cells preventing an inflammatory response (Savill and Fadok, 2000).
In general, the two separate methods of cell death can be identified by molecular means. The common characteristics of these two forms of cell death are summarized in Table 1. It is important to note that proof of apoptosis often requires more than one method; this is due to the fact that not all cell death can be easily classified as either necrosis or apoptosis. There are some cells that undergo apoptosis and yet will not have the characteristic shrinkage and chromatin loss and necrosis may display some of the characteristics of apoptosis (Fady et al., 1995; Meneses-Acosta et al., 2001).
Table 1. Apoptosis versus necrosis
Large groups of cells
No energy requirement
Organized directed destruction of cellular proteins and structures
Random disintegration of cellular structures
Cell shrinks in size/nuclear collapse
Cells swell in size/nucleus intact
Nuclear envelope breached/outer membrane intact
Rupture and blebbing of outer membrane
Chromatin is released to the cytosol
Cytosolic proteins are released to intracellular space
p53 MEDIATED GROWTH ARREST AND APOPTOSIS
P53 has been reported to be involved in differentiation, senescence, development, inhibition of angiogenesis, and apoptosis (Yonish-Rouach et al., 1991; Shaw et al., 1992; Clarke et al., 1993; Dameron et al., 1994; Aloni-Grinstein et al., 1995; Atadja et al., 1995; Sah et al., 1995). The main function of activated p53 is growth suppression at the G1 phase of the cell cycle through the upregulation of p21WAF1/CIP1. Stress response, such as DNA damage, will be mediated by p53 dependent growth arrest and the initiation of repair or if the damage is severe, induction of apoptosis. DNA damage is detected directly by p53 through the binding of single strand DNA to the carboxyl-terminal domain of the p53 protein. Cells with mutated or deleted p53, such as most carcinomas, are resistant to DNA damage induced apoptosis and cell cycle control. Although cell cycle arrest is often reported to occur at the G1/S phase during apoptosis it is not a requirement for apoptosis. It has been reported that p53 can induce apoptosis through both transactivation-independent and transactivation-dependent pathways (Yonish-Rouach et al., 1991; Caelles et al., 1994; Wagner et al., 1994; Attardi et al., 1996). However, there is no clear pathway described that p53 induces to initiate apoptosis independent of transcription. P53 has been reported to upregulate the pro-apoptotic Bax protein and downregulate the anti-apoptotic Bcl-2 protein (Miyashita et al., 1994; Selvakumaran et al., 1994; Miyashita and Reed, 1995; Findley et al., 1997). The apoptotic signal death receptor CD95 (Fas) is induced by p53 increasing sensitivity to apoptosis signals (Morimoto et al., 1993; Owen-Schaub et al., 1995; Muller et al., 1998; Waring and Mullbacher, 1999). The majority of evidence indicates that transcriptional regulation of proteins associated with apoptosis is the primary way p53 induces apoptosis. It has been reported that many viruses associated with cancer, encode proteins to inactivate p53, preventing upregulation of apoptosis proteins. Examples of such viral proteins are the E1B protein of adenovirus, the E6 protein of human papilloma virus, and the large T antigen of SV 40 virus. It has also been proposed that the level of p53 may determine whether growth arrest (low levels) or apoptosis (high levels) pathways are induced (Kagawa et al., 1997; Wrone-Smith et al., 1999; Zhao et al., 2000b; Pyrzynska et al., 2002). These findings hint at the need for higher levels of p53 in order to transactivate pro-apoptotic genes. Although the decision process for apoptosis or growth arrest after DNA damage is not fully realized, it is clear that p53 has an important role to play in this process.
DEATH RECEPTOR PROTEINS
External apoptotic signals are transmitted through cell death adaptor proteins associated with transmembrane receptor proteins (Fig. 1). These adaptor proteins contain a motif known as the death effector domain (DED) or just death domain (DD). It is this DD that will recruit and cause the activation of the execution proteins known as caspases.
The important ligands for the TNF-R transmembrane family of receptors are CD95L (Fas ligand), TNF-α, and TRAIL (Schulze-Osthoff et al., 1998; Ashkenazi and Dixit, 1999; Walczak and Krammer, 2000; Daniel et al., 2001; Mitsiades et al., 2001). It has been reported that the receptors and the ligands form a trimeric structure (Karpusas et al., 1995; Gruss, 1996; Schneider et al., 1997; Hymowitz et al., 1999). Interestingly the ligands are expressed as transmembrane proteins but soluble forms have been identified for all of them (Schneider and Tschopp, 2000). The soluble forms are generated through the activity of metalloproteinases (Kayagaki et al., 1995; Chandler et al., 1997; Black et al., 1997; Mariani and Krammer, 1998; Itai et al., 2001). It has not been determined if there are specific metalloproteases for each ligand.
TNF-R1 and CD95 are the most studied members of the death receptors. Through these proteins a multicellular organism can identify and eliminate senescent, injured or mutated cells by induction of apoptosis. These receptors have adapter proteins associated with the cytosolic domain of the protein. Proteins, such as Fas-associated death domain (FADD) and TNF-receptor associated death domain (TRADD), link caspases, in particular caspase-8, via their death domain to the DD of receptors for external apoptosis signaling. Apoptosis regulation is extremely important for the regulation of autoimmunity through the ability to eliminate any lymphoid cells that could be characterized as autoreactive.
Some tissues need to be free of immune cell monitoring, such as brain, pancreas, ovary, testis, eye, and the placenta during pregnancy. These tissues are called immune privileged tissues (Watanabe-Fukunaga et al., 1992; Leithauser et al., 1993; Hiramatsu et al., 1994; Galle et al., 1995; Xerri et al., 1997; Guller and LaChapelle, 1999; Bernstorff et al., 2002; Niederkorn, 2002). Immune privilege occurs because these tissues express CD95 ligand normally on their surface and normal monitoring immune cells carry the death receptors resulting in a clearing of the immune cells via apoptosis (Krammer, 1998). Improper expression of the CD95 ligand on the surface of these tissues results in a number of diseases (Zipp et al., 1999; Contreras et al., 2000).
CERAMIDE AND SPHINGOMYELINASE APOPTOTIC SIGNALING
Sphingomyelin is a membrane phospholipid that is usually located in the outer portion of cellular membranes. Sphingomyelinase breaks down sphingomyelin into a number of components including ceramide. Sphingomyelinase is activated by diacylglycerol (DAG) through the accumulation of apoptotic signals via the DD of receptors such as CD95 and TNF-R1. Ceramide, from the breakdown of sphingomyelin, is the major part of one of the most conserved receptor-mediated signal transduction pathways. Ceramides mediate many different cell responses such as proliferation, differentiation, growth arrest, and apoptosis (Jayadev et al., 1995; Obeid and Hannun, 1995; Pushkareva et al., 1995; Mathias et al., 1998; Claus et al., 2000; Gallardo et al., 2000). A number of different pathways are activated by ceramide through the action of ceramide-activated protein kinase (CAPK). CAPK will phosphorylate important MAP kinase signal pathway members Raf-1, MEK, PLA2, and ERK-2 (Maziere et al., 2001; Willaime et al., 2001). Ceramide is a catalyst for the activation of stress response proteins MEKK1, SEK1, SAPK, and c-Jun (Smith et al., 1997; Basu and Kolesnick, 1998; Ruvolo, 2001). Interestingly, ceramide can activate PP2A (Ruvolo, 2001). PP2A is a phosphatase involved in a diverse number of pathways, such as cell cycle control. There is a cytosolic sphingomyelinase that becomes active hours after death receptor crosslinking that is responsible for a continued and persistent accumulation of ceramide (Cifone et al., 1994; Brenner et al., 1998; Chen et al., 1998). There is speculation that ceramide may contribute to amplification of the apoptotic signal by association with the mitochondrial membrane and initiating a loss of membrane potential.
Mitochondria represent the central hub for the regulation of the apoptotic signal as to whether a cell should commit suicide through apoptosis or survive (Kroemer and Reed, 2000). Mitochondria sequester a number of pro-apoptotic proteins (apoptotic protease-activating factor-1 (APAF-1), cytochrome c, and caspase-9) that can initiate apoptosis upon release to the cytoplasm. The release of these proteins occurs after the loss of membrane potential by the mitochondria. The amplification of the apoptotic signal into a clear death signal occurs when mitochondria have a loss of membrane potential. The Bcl-2 family of proteins consists of both pro- and anti-apoptotic members that can control the loss of membrane potential (Tsujimoto, 1998; Costantini et al., 2000; Lohmann et al., 2000; Raisova et al., 2001).
Mitochondria transfer apoptotic signals to the adapter protein APAF-1 through the release of cytochrome c (Solary, 1998; Perkins et al., 2000; Cao et al., 2002). It is speculated that cytochrome c binding to APAF-1 causes a conformational change in APAF-1 that exposes a caspase recruitment domain for procaspase-9 (Vaughn et al., 1999; Zhou et al., 1999). The complex of APAF-1 and cytochrome c activates the procaspase-9. Active caspase-9 will activate caspase-3 and -7, the major caspases for cell death and apoptosis signal amplification.
Bcl-2 PROTEIN FAMILY
The Bcl-2 family is directly associated with the mitochondrial membrane and can control mitochondrial permeability and the release of APAF-1, cytochrome c, and caspase-9 (Marsden et al., 2002). The Bcl-2 family includes both pro- and anti-apoptotic members. The proportions of these pro- and anti-apoptotic proteins can control the amplification of apoptosis signals (Chao and Korsmeyer, 1998; Raisova et al., 2001). The anti-apoptotic members include Bcl-2, Bcl-X(L), Bcl-w, A1 and Mcl-1.
The anti-apoptotic Bcl-2 family of proteins associates with nuclear membranes, mitochondrial membranes, and endoplasmic reticulum. It has been proposed that these proteins associate with transmembrane pores to control permeability transitions (Lee et al., 1999; Vander Heiden and Thompson, 1999; Hoetelmans et al., 2000; Ferri and Kroemer, 2001). It has also been proposed that anti-apoptotic members will bind to adaptor proteins, such as APAF-1, inhibiting the activation of the caspase cascade (Hu et al., 1998; Inohara et al., 1998).
The pro-apoptotic members include Bax, Bcl-X(S), Bak, Bad, Bik, Bim, and Bid. This group is proposed to either bind directly to the anti-apoptotic members preventing their function or through competition for a common binding site (Tsujimoto, 1998; Shimizu et al., 2000; Sugiyama et al., 2002).
The Bcl-2 family of proteins can be circumvented for the activation of apoptosis. It has been reported that an accumulation of apoptosis signal proteins upstream of mitochondrial depolarization could initiate apoptosis without involvement of the Bcl-2 family of proteins (Tartier et al., 2000). For example, caspase-3 can be activated directly by caspase-8 after association with the DD of a cell surface receptor activates caspase-8.
Caspases are cysteine proteases that have a central role in apoptotic cell death. Caspases were first discovered in the Caenorhabditis elegans and soon after sequence homology studies revealed that many multicellular organisms have caspases. They are constitutively expressed as latent zymogens known as procaspases. Caspase activation occurs at the onset of apoptosis and inhibition of caspase activity curtails apoptosis. Active caspases systematically destroy the cellular matrix through the cleavage of key cellular proteins after aspartate residues.
At present, 14 caspases (Table 2) have been identified. All caspases share a common structure as determined by X-ray crystallography and sequence analysis. The carboxy-terminal contains one or two aspartate residues and is the site of cleavage for the activation of caspases. Cleavage separates the protein into small and large subunits, which form a heterodimer. The active protease is reported to act as a homodimer of heterodimers consisting of two large and two small subunits. Although the active caspases are present in this tetrameric form, each heterodimer has an active site for protease activity. All caspases cleave after aspartate residues exclusively.
Table 2. Caspase characteristics
X, unrestricted amino acid specificity; NA, not applicable; M, only characterized in mouse.
Caspases can be divided into three groups: (1) apoptotic initiator caspases (caspase-2, -8, -9, -10, -12) are activated by the initial signal for apoptosis and promote the activation of downstream caspases through direct association and cleavage. For example, caspase-12 has been characterized as an activator of caspase-3 during stress or calcium flux of the endoplasmic reticulum (Nakagawa et al., 2000; Yoneda et al., 2001). (2) Apoptotic executioner caspases (caspase-3, -6, -7) are activated by apoptotic initiator caspases and are responsible for the cleavage of key proteins that are the targets of apoptosis. (3) Pro-inflammatory caspases (caspase-1, -4, -5, -11, -13, -14) are less well characterized as to their specific function, although, caspase-11 may aid in the activation of caspase-1 by a non-cleavage interaction (Kang et al., 2000). These inflammatory caspases may play a role when there is a need to involve an immune system response, such as defense against viral infections.
It has been reported that activation of caspases may occur by autoactivation, transactivation or proteolysis by other proteases and caspases. It is speculated that localized concentrations of caspases will promote (Kang et al., 2000) procaspsae-3. Caspases, once active, transactivate other caspases amplifying the apoptotic signal, providing a positive feedback loop (Fig. 1). It is important to note that a single caspase cannot activate all of the members of the caspase family directly. This provides a layer of control linked to the specific pattern of caspase expression by a given cell.
Another mechanism, for the activation of caspases, is through non-caspase proteases (Fig. 1). Granzyme-B, calpain, cathepsin D and others have been reported to be responsible for the activation of caspases during apoptosis but it is not known if activation is initiated by direct interaction with the caspases. Interestingly, cathepsin G activates caspase-7 by cleavage at glutamine-194 indicating that cleavage at the aspartate is not always necessary for activation (Zhou and Salvesen, 1997).
Activated caspases will commonly target pro- and anti-apoptotic proteins initially. The elimination of anti-apoptotic proteins will prevent the blockage of apoptosis. An example of this is active caspase-3, which cleaves Bcl-2 and Bcl-X(L) thereby eliminating these pro-survival proteins (Bossy-Wetzel and Green, 1999; Zhang et al., 2002). On the other hand, caspase-8 will cleave the pro-apoptotic protein BID. The carboxy-terminal fragment of BID is able to induce the release of cytochrome c from the mitochondria initiating the activation of caspase-9 (Bossy-Wetzel and Green, 1999).
Another target of active caspases are proteins that will enhance the signal for apoptosis. These proteins are often effectors of apoptosis or kinases that can enhance the signal for apoptosis such as inhibitor of caspase-3 activated DNAse (ICAD), p21Cdc42/Rac-activated kinase 2 (PAK2), protein kinase c (PKCδ) and MAP ERK kinase kinase 1 (MEKK1). ICAD is cleaved releasing the active caspase-3 activated DNAse (CAD) (Charrier et al., 2002). CAD will directly cleave DNA resulting in the characteristic DNA laddering associated with apoptosis (Zhao et al., 1999b; Meng et al., 2000). Another example is cleavage of MEKK1 by caspases enhances the activation of other caspases. This suggests that kinase activation by caspases may act to enhance the apoptosis signal (Widmann et al., 1998).
In summary, activation of caspases occurs through two distinctly different pathways involved in the induction of apoptosis: the extrinsic pathway and the intrinsic pathway (Fig. 2; Nunez et al., 1998). The extrinsic pathway involves the activation of the death receptors after binding their specific ligand. The death receptors will recruit initiator caspases and initiate their activation. The intrinsic pathway involves the depolarization of the mitochondria due to cellular stress signals causing the release of cytochrome c and subsequent activation of the initiator caspase, caspase-9. Initiator caspases activated by either the extrinsic or intrinsic pathways will activate the executioner caspases. The apoptotic signal induces the activation of caspases, which in turn amplify the signal for cell death and amplify caspase activation. Active caspases can then target cytoskeletal and structural proteins associated with nuclear integrity. This would enhance the condensation of the cell and nucleus. Furthermore, caspases will target proteins responsible for DNA repair, cell signaling, and protein building proteins increasing the total collapse of the cell into a null body ready for phagocytosis.
It is important that caspase activation occurs in a controlled and well-regulated manner. An inability to express or activate appropriate caspases could result in malignant tumors since these cells cannot be removed from the tissue. The opposite is just as devastating, the increase signaling for apoptosis by the constitutive activation of caspases. This would lead to many degenerative diseases such as Huntington's disease or Alzheimer's disease (Eldadah and Faden, 2000).
Physiological mechanisms for the removal of abnormal cells are a fundamental requirement for maintaining healthy ovarian function (Tilly, 1996; Morita and Tilly, 1999). In theory, failure of the molecular mechanisms responsible for the physiological induction of apoptosis may result in tumorogenisis of ovarian cells (Ghahremani et al., 1999; Morita and Tilly, 1999). It may be possible to restore or induce the disrupted apoptotic mechanisms within the ovarian tumor cells through the use of chemotherapeutic reagents.
Retinoids are a group of natural and synthetic analogs of vitamin A that have been shown to play an important role in cell differentiation and proliferation (Kastner et al., 1995; Means and Gudas, 1995). Since retinoid treatment inhibits the growth of a variety of epithelial cancers, such as ovarian carcinomas, retinoids have great promise in the area of cancer chemotherapy and chemoprevention (De Luca, 1991; Gudas et al., 1994; Moon et al., 1994; Wu et al., 1997a,b, 1998b; Zhang et al., 2000).
Natural retinoids, such as all-trans-retinoic acid (ATRA), for the most part, do not induce apoptosis. However, a number of synthetic retinoids have been reported to inhibit the growth of both ATRA-sensitive and ATRA-resistant ovarian tumor cell lines. Investigation of the effects of these synthetic retinoids on cell growth has suggested that the mechanism of growth suppression by these retinoids involves induction of apoptosis (Wu et al., 1998a; Holmes et al., 2000).
We have studied two synthetic retinoids, 6-[3-(1-Admantyl)]-4-hydroxyphenyl]-2-naphthalene carboxylic acid (CD437) and fenretinide N-[4-hydroxyphenyl] retinamide (4-HPR), that are reported as potent inducers of apoptosis. In our studies we have demonstrated that treatments with either CD437 or 4-HPR could growth inhibit and induce apoptosis in both ATRA-sensitive (CA-OV-3) and ATRA-resistant (SK-OV-3) ovarian carcinoma cell lines (Fig. 3; Holmes et al., 2002). The CA-OV-3 and SK-OV-3 cell lines are reported to have mutated p53. As discussed previously, p53 protein is responsible for the decision between growth arrest and DNA repair or the induction of apoptosis in response to DNA damage. Treatment with CD437 or 4-HPR may be able to override or circumvent the block in the normal path for induction of apoptosis. Since p53 and other proteins which control apoptosis and growth arrest are often mutated in the majority of tumor cells, it is important to understand the molecular mechanisms induced by these retinoids to further the ability to create chemotherapeutic agents for the treatment and prevention of cancer.
CD437 induced apoptosis is meditated in part by retinoid receptors
CD437 has been demonstrated to inhibit the growth of both ATRA-sensitive (CA-OV3) and ATRA-resistant (SK-OV3) ovarian tumor cell lines as well as to induce apoptosis (Wu et al., 1998a; Holmes et al., 2000). Since a number of studies have demonstrated the importance of nuclear receptors (RARs and RXRs) in mediating cellular responses to retinoids (Wu et al., 1997b, 1998b), we wished to determine the role of RARs in mediating the response of ovarian carcinoma cells to CD437 treatment. We have previously examined the role of RARs in general and RAR-γ in particular in mediating the induction of apoptosis by the RAR-γ selective synthetic retinoid, CD437, in ovarian tumor cell lines (Holmes et al., 2000). We modulated RAR level and function by overexpressing either wild type RAR-γ or a pan dominant negative mutant of all RAR subtypes called RAR-β (R269Q) (Tairis et al., 1995), or through the use of an RAR-γ antagonist, MM11253. We found that inhibition of RAR function reduced but did not eliminate induction of apoptosis in both CA-OV3 and SK-OV3 cells by CD437. Likewise, overexpression of wild type RAR-γ was found to increase apoptosis after treatment with CD437. Our results suggested that in ovarian carcinomas, CD437 induced apoptosis is mediated at least in part via an RAR pathway (Holmes et al., 2000).
CD437 and 4-HPR induced apoptosis requires the activation of caspases
Caspase activity is a hallmark of the induction of apoptosis. In light of this fact we assayed for the activation of the caspase cascade to determine if active caspases were essential for CD437 and 4-HPR induced apoptosis. Caspase-7 is almost identical in structure and function to caspase-3 and there is not yet an adequate means to differentiate the two, so all assays for caspase-3 will be referred to as caspase-3 but should be understood as representing both caspase-3 and caspase-7. Caspase-3 activity is indicative of the final stages of apoptosis induction via the caspase cascade. The synthetic retinoids, CD437 and 4-HPR induced caspase-3 activity in both the CA-OV-3 and the SK-OV-3 ovarian carcinoma cell lines. Using caspase-3 inhibitors, we determined that caspase-3 activation was essential for the induction of apoptosis by both CD437 and 4-HPR in both CA-OV-3 and SK-OV-3 cell lines (Holmes et al., 2002, 2003a).
As discussed, caspase activation occurs through the activation of a cascade of events transferring the death signal from initiator caspases (caspase-8 and caspase-9) to effector caspases (caspase-3 and caspase-7). We did not detect any appreciable caspase-8 activity and determined that caspase-8 was not necessary for the induction of apoptosis by either CD437 or 4-HPR.
Caspase-9 is an initiator caspase associated with mitochondria that has been characterized as an activator of caspase-3. Determinations of caspase-9 activity were positive in response to treatments with either CD437 or 4-HPR in ovarian carcinoma cell lines. Caspase-9 activation was determined to be essential for the induction of apoptosis by both of these synthetic retinoids. Furthermore, we were able to determine that the CD437 and 4-HPR activation of caspase-3 were directly dependent on caspase-9 activation through the use of a caspase-9 inhibitor (LEHD-cmk). This finding indicated that the pathway for the induction of apoptosis must involve the depolarization of the mitochondrial membrane, since mitochondrial membrane depolarization results in the release and activation of caspase-9 (Holmes et al., 2002, 2003a).
The depolarization of mitochondria in response to apoptotic inducing agents, such as CD437 and 4-HPR, has been reported to be necessary for the activation of caspase-9 and caspase-3 (Marchetti et al., 1999; Mologni et al., 1999; Suzuki et al., 1999; Costantini et al., 2000; Eldadah and Faden, 2000; Tartier et al., 2000). The depolarization of the mitochondrial membrane allows for the release of the pro-apoptotic proteins cytochrome c, procaspase-9, and APAF-1. These three proteins are able to associate causing the cleavage of procaspase-9 to active caspase-9. Active caspase-9 is then able to cleave procaspase-3 to active caspase-3.
We determined that mitochondrial membrane depolarization was required for the induction of apoptosis by both CD437 and 4-HPR through the use of two chemical reagents, Bongkreikic acid and Betulinic acid. Bongkreikic acid inhibits the depolarization of mitochondrial membranes, whereas, Betulinic acid induces the depolarization of mitochondrial membranes inducing apoptosis through the activation of caspase-9 and caspase-3. Using these chemicals, we determined that CD437 and 4-HPR required the depolarization of the mitochondrial membrane to induce apoptosis, activate caspase-9, and activate caspase-3 (Holmes et al., 2002, 2003a,b). This allows us to map the order of events in the late stages of apoptosis induction by CD437 and 4-HPR: mitochondrial depolarization, caspase-9 activation, subsequent activation of capase-3.
Mitochondria are targets of synthetic retinoid induced apoptosis
In order to further characterize the mechanism involved in induction of apoptosis by CD437 we explored the response of a CD437 resistant cell line (CA-CD437R) to other agents that induce apoptosis. The parental CA-OV-3 ovarian tumor cell line is sensitive to apoptosis induced by treatment with CD437, 4-HPR, and TNF-α. Surprisingly, the CA-CD437R cell line was not responsive to the effects of TNF-α treatment but was just as sensitive as the CA-OV-3 cell line to 4-HPR treatments (Holmes et al., 2002). This suggests a common mechanism for the induction of apoptosis by TNF-α and CD437 treatments. Further, the ability of 4-HPR to induce apoptosis in CA-CD437R cells indicates a clear divergence in the pathway of apoptosis induction by these two synthetic retinoids.
We have reported that the CA-CD437R cell line is resistant to apoptosis induced by TNF-α (Holmes et al., 2002). This suggests that CD437 induced apoptosis utilizes a similar mechanism as TNF-α. Cross-resistance of ovarian tumor cell lines to multiple apoptotic agents has been reported previously. Kumar et al. (2001) found that a cell line selected for resistance to paclitaxel was also resistant to CD437. A recent report suggests that resistance to CD437 of a leukemia cell line mapped upstream of cytochrome c release and activation of caspases (Ponzanelli et al., 2000). The release of cytochrome c is dependent on release from the mitochondria often through a depolarization of the mitochondrial membrane (Marchetti et al., 1999; Costantini et al., 2000). DIOC(6) staining demonstrated that mitochondria were depolarized in CA-OV-3 cells but not CA-CD437R cells after CD437 treatment. CMXRos staining confirmed the results obtained with DIOC(6) (Holmes et al., 2002). Likewise, cytochrome c was found to be released in CD437 treated CA-OV-3 cells but not in CA-CD437R cells (Holmes et al., 2002). However, in contrast, 4-HPR treatment of CA-CD437R cells did result in mitochondrial depolarization and release of cytochrome c. Thus, this step in the apoptotic pathway cannot be defective and thus not responsible for CD437 resistance in CA-CD437R cells (Holmes et al., 2002).
CD437 activates the MAP kinase signal pathway to induce apoptosis
TNF-α has been reported to activate tyrosine kinases of cellular stress-induced pathways consequently leading to activation of c-JUN N-terminal Kinase (JNK) and p38 MAPk (Nahas et al., 1996; Avdi et al., 2000; Liu et al., 2000). Recent reports suggest that activated p38 MAPk is associated with apoptosis and regulates the release of cytochrome c from the mitochondria (Assefa et al., 2000; Hatai et al., 2000; Zhuang et al., 2000). Sakaue et al. (2001) reported that CD437 treatment of H460 lung carcinoma cells was able to activate the MAP kinase cascade, suggesting that the MAP kinase cascade may be a common link for apoptosis induced by TNF-α and CD437. Only some inhibitors of upstream components of the MAP kinase signal pathway inhibited CD437 induced apoptosis. Interestingly, PD98059, an inhibitor of MEK, and SB203580, an inhibitor of p38, blocked the induction of apoptosis by CD437 but the inhibitors of RAS and c-RAF, FTP Inhibitor III and ZM336372, respectively, did not (Holmes et al., 2003b). Using a p38 activity assay we were able to determine that p38 activity occurs after CD437 treatment but not after 4-HPR treatment. This indicates that the MAP kinase cascade is involved in the induction of apoptosis by CD437.
MEK is an upstream activator of p38 that must be phosphorylated to become active. Using an antibody specific for the active form of MEK we were able to determine that MEK is activated after treatment of CA-OV-3 and CA-CD437R cells with CD437. Further, we were able to determine that the activation of MEK was responsible for the activation of p38 MAP kinase by CD437 treatment through the use of a MEK specific inhibitor (PD98059). These results further our understanding of the early molecular mechanism responsible for the induction of apoptosis by CD437 in ovarian carcinoma cells and map the activation of p38 directly to the activation of MEK (Holmes et al., 2003b). We have already determined that c-RAF is not the upstream activator of MEK but recent reports have shown MEKK1, an upstream activator of MEK, to be necessary for apoptosis induced by cisplatin and TRAIL (tumor necrosis factor (TNF)-related apoptosis inducing ligand) so it is possible that activation of MEK by CD437 may be through the MEKK1/SEK1 pathway (Gibson et al., 2000; Bild et al., 2002; Sanchez-Perez et al., 2002).
One of the important consequences of activation of the MAP kinase signal pathway is the activation of factors responsible for the transcriptional regulation of response genes. MEF2 is a transcription factor that has been shown to be activated by p38 MAP kinase through phosphorylation (Yang et al., 1999; Zhao et al., 1999a, 2000a; Han and Molkentin, 2000). MEF2 is reported to regulate the transcription of the orphan receptor TR3 (nur77) (Youn et al., 1999, 2000; Kasler et al., 2000; Youn and Liu, 2000; Liu et al., 2001). We have found that there is a direct link between the activation of MEK and subsequent activation of p38 with the phosphorylation of MEF2 (Holmes et al., 2003b). Although we did not measure MEF2 activation by 4-HPR, we were not able to detect any of the upstream events responsible for MEF2 activation such as p38 MAP kinase activity. Furthermore, inhibitors of the MAP kinase cascade did not inhibit the mitochondrial depolarization, the activation of caspases, or the induction of apoptosis by 4-HPR in ovarian carcinoma cell lines. These studies indicate that the induction of apoptosis by CD437 and 4-HPR utilizes separate early molecular pathways that converge at the depolarization of the mitochondrial membrane.
TR3 is an orphan receptor of the steroid/thyroid receptor superfamily that is present in the nucleus as a transcription factor but has been shown to translocate to the cytoplasm and cause the depolarization of the mitochondrial membrane (Li et al., 2000; Dawson et al., 2001; Holmes et al., 2002; Zhang, 2002). This translocation and association has been reported to be a mechanism for the induction of mitochondrial depolarization and subsequent apoptosis in response to apoptosis inducing agents, such as CD437 (Li et al., 1998; Zhang, 2002). We have found that GFP-TR3 is sequestered in the nucleus before treatment with CD437 and after CD437 treatment TR3 translocates to the cytosol and associates with mitochondria in ovarian carcinoma cell lines (Holmes et al., 2003b). Furthermore, 4-HPR treatments of ovarian carcinoma cells do not induce the translocation of TR3. This indicates that the depolarization of the mitochondrial membrane associated with the induction of apoptosis by 4-HPR does not require TR3 translocation, which is in contrast to the requirement of TR3 for the induction of apoptosis by CD437.
Our findings suggest that the pathways that CD437 and 4-HPR utilize to induce apoptosis is consistent with the findings by Appierto et al. (2001) who developed an ovarian carcinoma cell line which was resistant to 4-HPR but not resistant to ATRA or CD437. Moreover, we believe that demonstration of the fact that CD437 and 4-HPR induce apoptosis via different pathways will have potential therapeutic implications in that ovarian tumors resistant to one retinoid may not be resistant to the other.
The synthetic retinoids, CD437 and 4-HPR can induce growth arrest and apoptosis in both ATRA-sensitive (CA-OV-3) and ATRA-resistant (SK-OV-3) ovarian carcinoma cell lines. We determined that CD437 activated the MAP kinase signal pathway while 4-HPR and ATRA did not, indicating a divergence in the molecular mechanisms employed by these retinoids to induce growth arrest and apoptosis. CD437 treatment is able to induce the activation of MEK. The activation of MEK is able to induce a cascade of events that include p38 MAP kinase and MEF2 activation resulting in the translocation of TR3 to the mitochondrial membrane inducing its depolarization (Fig. 4).
Although we were not able to determine the early events induced by 4-HPR treatments, we did determine that the late stages of apoptosis induction by both CD437 and 4-HPR were through the same pathway. Treatment of ovarian carcinoma cell lines with either retinoid induced apoptosis through the depolarization of the mitochondrial membrane, activation of caspase-9 and subsequent activation of caspase-3.
Elucidation of the molecular pathways induced by CD437 and 4-HPR should help in the design of new chemotherapeutic reagents for the treatment of ovarian cancer. Defects or blocks in the cascade of events for the induction of apoptosis can cause malignant cells to escape immune monitoring or resist the normal physiological induction of apoptosis. Understanding the relationship between the structures of these conformationally restricted synthetic retinoids and the molecular mechanisms activated by treatment with CD437 and 4-HPR to induce the intrinsic pathway for apoptosis will allow for the creation of other retinoids with similar structures for the treatment of resistant carcinomas. Importantly, activation of separate pathways that result in the induction of apoptosis indicates that combination treatments with structurally differing and unique synthetic retinoids should prove to be more efficacious for the treatment of all ovarian tumors.
We are grateful to Uwe Reichert, PhD, from Galderma for providing the synthetic retinoid CD437 and to Edward L. Tolman, PhD, and James W. Oldham, PhD, from R.W. Johnson for providing the synthetic retinoid 4-HPR. We are also very grateful to Xiao-kun Zhang, PhD from the Burnham Institute, for providing the pGFP-TR3 construct. This work was supported by grants from the National Institutes of Health (CA 64945 and DE 13139) to KJS. WFH was funded in part by a training grant from the National Institutes of Health to the Department of Microbiology and Immunology, Temple University School of Medicine (AI 07101).