Combining proteasome inhibition with TNF-related apoptosis-inducing ligand (Apo2L/TRAIL) for cancer therapy

Abstract

Apoptosis has an essential role in embryogenesis, adult tissue homeostasis and cellular responses to stressful stimuli. Therefore, increased apoptosis is involved in the pathogenesis of various ischaemic, degenerative and immune disorders. Conversely, genetic aberration that results in a reduction or abolition of apoptosis can promote tumorigenesis and underlie the resistance of cancer cells to various genotoxic anticancer agents. Therefore, a detailed knowledge of the control of apoptotic pathways could aid in the rational design of effective therapeutics for a variety of human diseases including cancer. One major way to promote apoptosis involves signaling through members of the tumor necrosis factor (TNF) superfamily. On binding to their appropriate receptors, some TNF family members can promote caspase activation and apoptosis. Early studies on TNF indicated that a limited number of tumor cell lines could be induced to undergo apoptosis on exposure to TNF. Another member of the TNF family Fas ligand (FasL) is also known to induce apoptosis in a variety of tumor cells. Although TNF and FasL can efficiently induce apoptosis in a limited number of tumor cells, administration of either of these agents is associated with extreme toxicity. This toxicity has precluded further development of either TNF or FasL for cancer therapy. However, within the last 8 years another member of the TNF family, TNF-related apoptosis-inducing ligand (Apo2L/TRAIL) has been characterized, which induces apoptosis of a wider range of cancer cells than either TNF or FasL. Surprisingly, most normal non-transformed cells are quite resistant to the apoptotic effects of Apo2L/TRAIL. This selective toxicity for cancer cells is the basis for the current enthusiasm for Apo2L/TRAIL as a potential novel anticancer therapy. In this symposium report, we provide a brief overview of Apo2L/TRAIL, its receptors and their signaling pathways. We discuss findings on the antitumor effects of Apo2L/TRAIL alone or in combination with radiotherapy or chemotherapy. In addition, we present recent information from our groups concerning the possible therapeutic benefits of combining Apo2L/TRAIL with the proteasome inhibitor bortezomib.

Apo2L/TRAIL signaling and apoptosis

TRAIL was discovered primarily through homology cloning [1]. Independently, TRAIL was isolated by another group, who termed it Apo2 ligand (Apo2L) [2]. Several lines of evidence, including the resolution of Apo2L/TRAIL’s crystal structure, indicate that Apo2L/TRAIL forms homotrimers [3] and this multimerization seems essential for the induction of apoptosis. A variety of Apo2L/TRAIL preparations have been engineered to enhance the biological activity of Apo2L/TRAIL by promoting aggregation. These include a polyhistidine-tagged form (amino acids 114–281) [2, 4] a Flag epitope-tagged version (aa 95–281) [1] and a Apo2L/TRAIL/leucine zipper fusion protein [5]. In addition, a recombinant soluble form devoid of foreign sequence has been developed. This version requires a central zinc ion to promote homotrimerization [6]. Mutations in residues which affect zinc binding to Apo2L/TRAIL result in loss of both trimer formation and apoptotic activity.

Shortly, after the discovery of Apo2L/TRAIL, a similar strategy of searching EST databases resulted in identification of four Apo2L/TRAIL receptors [7]. Two of these, death receptor 4 (DR4 or TRAIL-R1) and death receptor 5 (DR5 , TRAIL-R2, KILLER), contained death domains (DD) motifs and promoted apoptosis when expressed in cells. Both DR4 and DR5 can form heterotrimeric complexes. Two of the other receptors DcR1 (TRAIL-R3) and DcR2 (TRAIL-R4) are thought to act as decoys. DcR1 lacked both a cytoplasmic and transmembrane domain and DcR2 only had a partial death domain, so both were unable to trigger apoptotic death. Overexpression of DcR1 and DcR2 provided protection from Apo2L/TRAIL-induced apoptosis. This led to the proposal that these decoy receptors could account for resistance to Apo2L/TRAIL. Indeed, expression of DcR1 and DcR2 was more prevalent in normal tissues than in most cancer cells. However, a number of studies did not find a good correlation between decoy receptor expression and Apo2L/TRAIL sensitivity, suggesting that additional mechanisms are also involved in the regulation of Apo2L/TRAIL-mediated apoptosis [8–10]. The secreted protein osteoprotegrin (OPG) was also identified as a protein that could bind to Apo2L/TRAIL. Whether or not OPG plays a significant role in Apo2L/TRAIL-mediated cell death remains to be determined.

In an analogous manner to Fas and TNF receptors, DR4 and DR5 use specific adaptor proteins for death signal transduction [7, 11]. On interaction with ligand, these receptors recruit specific adapter molecules such as Fas-associated DD (FADD). The binding of adapter proteins such as FADD to the receptor complex then promotes the recruitment of procaspase 8 through homotypic interactions between a second protein motif, the death effector domain (DED). This complex of proteins has been called the death-inducing signaling complex (DISC) [12]. Both FADD and procaspase 8 contain these DED, and it is likely that aggregation of procaspase 8 molecules results in enzymatic conversion to active caspase 8 and further activation of the caspase cascade of enzymes [13]. In humans, the highly homologous caspase 10 can also be recruited to the DISC and promote apoptosis; whereas, caspase 10 is not found in rodents. The recruitment of caspase 8 can be inhibited by another DED containing protein called FADD-like IL-1β-converting enzyme (FLICE) inhibitory protein (c-FLIP). FLIP was originally described as a viral protein (v-FLIP), which prevented death receptor-induced apoptosis [14]. The cellular homologue c-FLIP can also inhibit apoptosis by preventing the recruitment and activation of proximal caspases (Fig. 1). High levels of endogenous c-FLIP often correlate with resistance of cells to Apo2L/TRAIL-mediated apoptosis. However, in a similar manner to receptor heterogeneity, a direct correlation between high c-FLIP levels and Apo2L/TRAIL resistance does not always apply [9].

Fig. 1.

Control of TRAIL signaling. The extrinsic cell death pathway is triggered by binding of death-receptor ligands that causes recruitment of adapter proteins such as FADD/Mort1 to the receptor. Subsequent binding and oligomerization of caspase 8 results in proximity-induced autoproteolytic activation. Activation is regulated by decoy receptors or dominant-negative pseudo-caspases such as c-FLIP which prevent recruitment of the caspase-8 proenzyme into the receptor complex. (Figure adapted from D. Nicholson, Nature 2000, 407:812)

If a productive signal occurs from the DISC, much data suggests that the induction of caspase activity is crucial for apoptosis to occur in most cells. For Fas-mediated signaling, caspase 8 can induce apoptosis in the absence of mitochondrial involvement by directly activating effector caspases, the so-called extrinsic apoptosis-signaling pathway. Cells exhibiting this phenotype have been designated as type I cells [15]. Alternatively in other cells, cleavage of the proapoptotic Bcl-2 family protein Bid by caspase 8 is required for apoptosis [16]. Activated truncated Bid (tBid) causes mitochodrial damage and therefore acts as an amplifying signal linking the extrinsic to the intrinsic apoptotic death pathways (Fig. 2). Cells in which mitochondrial damage is essential for the apoptotic response to FasL have been termed type II cells. It seems that a similar type I and type II designation can be applied concerning the sensitivity of various tumor cells to Apo2L/TRAIL apoptosis [17]. The answer as to which of these pathways are more important for Apo2L/TRAIL-mediated apoptosis most likely involves cell-type specific differences in Apo2L/TRAIL-mediated apoptotic signaling.

Fig. 2.

TRAIL-mediated apoptosis. In some cells (Type I), the death receptor pathway involves direct activation of the caspase cascade by activation of proximal caspases such as caspase 8. Alternatively in other cells (Type II), there may be involvement of Bcl-2 family members, whereby caspase cleavage of proapoptotic Bid (probably also involving interaction with Bax and/or Bak) results in a linkage to the stress pathway, mitochondrial perturbation and subsequent apoptosis. These pathways are controlled by inhibitors c-FLIP, anti-apoptotic Bcl-2 family members or inhibitors of apoptosis proteins (IAPs)

In addition to the controls modifying, the apoptotic signal at the DISC or mitochondria, there are further controls on apopotisis distal to mitochondrial damage. The inhibitor of apoptosis (IAP) family of proteins bind specifically to the initiator caspase 9 or executioner caspases 3 and 7 inhibiting these caspases, and acting as a downstream gatekeeper in the final phases of the apoptotic process [18]. This block on caspase activity is relieved when the mitochondrial proteins Smac/DIABLO and OMI/HtrA2 are released into the cytoplasm in response to mitochondrial damage and interact with the IAPs, thus releasing them from the caspases. Peptides that cross the cell membrane as well as small molecule mimics of Smac can potentiate Apo2L/TRAIL-mediated cell death due to their ability to release caspases from inhibition by IAPs [19–21].

In addition to its ability to initiate cell death, Apo2L/TRAIL has also shown to activate the nuclear factor-kappa-B (NF−_kB), which is known to suppress cytokine production and induce apoptosis-suppressing genes. Apo2L/TRAIL receptors DR4, DR5 as well as DcR2 can activate NF−_kB under certain conditions [22–24]. Although activation of NF−_kB is often associated with cell survival signals, under some specific circumstances NF−_kB can trigger cell death such as in focal cerebral ischemia [25] or p53-mediated apoptosis [26]. Differential use of the adapter proteins FADD and TNF-receptor-associated death domain protein (TRADD) may be the molecular switch that determines whether proximal caspases (i.e. apoptosis and death) or NF−_kB (i.e. cell survival) pathways become activated. Activation of the c-Jun N-terminal kinase (JNK) by Apo2L/TRAIL has also been demonstrated. However, JNK activity alone appears to be insufficient to elicit apoptosis, but may contribute to the induction of cell death following cellular stress by activating the expression of cell death promoting proteins.

Anticancer properties of Apo2L/TRAIL

The fact that Apo2L/TRAIL potently induces rapid apoptosis in a wide spectrum of tumor cell lines; whereas, lacks cytotoxicity towards many normal cells suggests it could have application as a novel biological anticancer therapeutic [27, 28]. Furthermore, death receptors can trigger tumor cell apoptosis in cancer cells independent of their p53 status. Therefore, Apo2L/TRAIL may still induce apoptosis in tumor cells in which p53 has been inactivated, resulting in their reduced sensitivity to chemotherapy or radiation. In immunodeficient mice carrying human tumor xenografts, Apo2L/TRAIL treatment substantially inhibited tumor progression in a variety of human tumors including breast and colon carcinomas, gliomas and multiple myeloma [5, 27]. More recent studies suggest that Apo2L/TRAIL is also effective at inducing apoptosis in primary tumor samples from patients with multiple myeloma and colon carcinoma [29, 30]. Nonetheless, a number of studies have questioned the resistance of some normal cells to Apo2L/TRAIL-mediated apoptosis. Toxicity of Apo2L/TRAIL on human hepatocytes and neuronal cells has been reported. However, it must be borne in mind that there are a number of different Apo2L/TRAIL preparations being used in these studies. The recombinant version of the ligand that contains amino acids 114–281 of human Apo2L/TRAIL and has been optimized by the addition of reducing agents and Zn to the cell culture media and the extraction buffers has been described. This version is preferred for clinical application since it contains no exogenous sequences and is therefore unlikely to be immunogenic [11]. Many tumor cells are still sensitive to this nontagged Zn-optimized version of Apo2L/TRAIL. Indeed hepatocytes and keratinocytes are resistant to this optimized ligand version despite showing sensitivity to nonoptimized or antibody cross-linked versions of Apo2L/TRAIL. One potential explanation for these differences is that apoptosis induction in normal cells may require higher order multimerization of the DR4 and DR5 receptors, and there may be significantly more aggregates in nonoptimized versions of Apo2L/TRAIL. Initial studies in nonhuman primates show that the nontagged Zn-bound Apo2L/TRAIL is well tolerated even at high doses [31], and phase 1 clinical trials of this preparation of Apo2L/TRAIL produced by Genentech and Amgen recently commenced.

Although Apo2L/TRAIL ligand can induce apoptosis in a variety of human cancer cells, some tumors may be protected by the fact that they express high levels of the decoy receptors. In an attempt to circumvent this, several groups have generated agonist monoclonal antibodies directed against the DR4 and DR5 death receptors [32–34]. These antibodies have been successfully used to promote tumor cell apoptosis in vitro and in vivo using xenogeneic human tumors in immunodeficient mice. Indeed these antibodies may offer significant advantages for clinical application over the ligand since they are specific for the death receptors, do not bind to decoys, and they have a significantly increased half life in vivo compared to the ligand. In addition, administration of antibodies to cancer patients as therapeutics has been recently been validated based on the clinical success of antibodies to VEGF and EGFR. As is the case with Apo2L/TRAIL, there are legitimate concerns that there may be toxicity associated with administration of these antibodies. However, this may be due to the nature of the antibody used such as the epitope it is directed against, its affinity, isotype etc. Nonetheless, Human Genome Sciences in association with Cambridge Antibody Technologies are entering phase 2 trials with an anti-DR4 antibody (HGS-ETR1) and have initiated phase 1 trials with an anti-DR5 monoclonal (HGS-ETR2). Also, a second anti-DR5 antibody (HGS-TR2J), developed by Human Genome Sciences in collaboration with the pharmaceutical division of the Kirin Brewery Company, is in phase 1 clinical trials. To date no major toxicities have been observed at least in response the doses administered during these initial clinical trials. Interestingly, a thorough evaluation of tumor and normal tissues using both DR4 and DR5 antibodies for immunohistochemistry has revealed an increased intensity of staining in many tumors as compared to the surrounding normal tissue. Also, it was reported at recent meetings that most normal tissues examined exhibited only weak staining with these antibodies [35, 36].

Since Apo2L/TRAIL alone has exhibited significant therapeutic benefit as a single agent in various animal tumor models, it has been also used in combination with commonly used therapies such as chemotherapy or radiation. Ionizing radiation can sensitize breast cancer cells and leukamias to Apo2L/TRAIL-induced apoptosis [37, 38]. This synergistic effect was shown to be p53 dependent and may involve radiation-induced upregulation of the death receptor DR5. Therefore, further the studies on combinations of radiotherapy and Apo2L/TRAIL are required. The combination of Apo2L/TRAIL with chemotherapeutics such as doxorubicin, 5-fluorouracil, etoposide, CPT-11, actinomycin D and camptothecin, has also been shown to significantly enhance Apo2L/TRAIL-mediated apoptosis in a variety of human tumor cell lines [11, 30]. However, the synergistic proapoptotic affects of Apo2L/TRAIL when combined with chemotherapy and radiation seems to require a functional p53. There are contrasting findings concerning the molecular basis by which chemotherapeutic drugs enhance Apo2L/TRAIL-mediated apoptosis of cancer cells. It has been reported using colon cancer cell lines that chemotherapy sensitizes cells by increasing concentrations of the proapoptotic protein Bak [39]. Higher levels of Bak would increase the apoptotic signal by amplifying the intrinsic apoptotic pathway involving the mitochondria. In contrast in other studies, using the same colon cancer cells, it was argued that in response to chemotherapy, the subsequent increase in p53 amplified the extrinsic apoptotic pathway by increasing levels of the death receptor DR5 [40]. As a result, these colon cancer cells were converted from a type II cell (dependent on mitochondrial participation in the apoptotic process) to a type I cell where apoptosis can proceed independent of the mitochondrial pathway. If chemotherapy could indeed convert tumor cells from a type II to a type I phenotype, this would be encouraging, since many tumors circumvent the mitochondrial apoptotic pathway due to alterations in levels of Bcl-2 family members. Nonetheless, if wild-type p53 is vital for synergy between chemotherapy and Apo2L/TRAIL, this combination would not be expected to prove very efficacious in the many human cancers where p53 is mutated or non-functional.

Combination of Apo2L/TRAIL with proteasome inhibition

The proteasome inhibitor bortezomib (Velcade, formerly PS-341) has shown much initial promise as an anticancer agent both in vitro and in vivo [41–43]. Treatment of tumor cells with bortezomib results in multiple biological effects including inhibition of the cell cycle, inhibition of NF-κB activation, changes in cell adherence and increased apoptosis [41–43]. Indeed, bortezomib was recently approved by the FDA for the therapy of human multiple myeloma. Due to the pivotal role the proteasome plays in apoptosis, inhibitors of this enzyme, such as bortezomib, provide an opportunity for exploring synergy between proteasome inhibition and other apoptosis-inducing agents. Indeed, bortezomib in combination with Apo2L/TRAIL further enhances apoptosis of some human multiple myeloma cells in vitro [29]. One of the major effects of proteasome inhibition in many cells is the inhibition of NF-κB activation. Since Apo2L/TRAIL binding to its receptors activates NF-κB (which can result in the induction of apoptosis suppressing genes), it is tempting to speculate that proteasome inhibition may enhance Apo2L/TRAIL-mediated apoptosis principally by blocking NF-κB. Blocking of NF-κB using either degradation resistant IκB constructs [44, 45] or proteasome inhibition [45, 46] has been demonstrated to increase the sensitivity of a variety of tumor cells to Apo2L/TRAIL-mediated apoptosis. However, NF-κB activation does not protect all cells from Apo2L/TRAIL-mediated apoptosis [47–49]. We have observed that bortezomib can sensitize a mouse myeloid leukemia (C1498) to Apo2L/TRAIL apoptosis independent of any effects on NF-κB activation [50]. Sensitization of these leukemia cells by bortezomib appeared to involve decreases in levels of the antiapoptotic protein c-FLIP, while levels of Bcl-2 family members and various IAP’s did not change. However, further studies on C1498 cells also revealed that bortezomib treatment also increased cell surface levels of mouse DR5. Increases in DR5 expression in response to proteasome inhibitors and chemotherapy have been reported, suggesting this could also be a crucial component in sensitization of cells to Apo2L/TRAIL [51, 52]. While studying a number of human renal carcinoma cells in vitro, we have noted that some (but not all) tumor cells were sensitized to Apo2L/TRAIL apoptotis by bortezomib treatment. In two of three sensitized renal carcinomas, c-FLIP levels were decreased in the absence of major effects on Apo2L/TRAIL receptors. In the remaining renal carcinoma, c-FLIP levels remained the same while surface expression of DR4 and DR5 increased significantly following bortezomib treatment (unpublished observations). Therefore, it seems likely that the relative importance of decreases in c-FLIP or increases in DR 4 and 5 following proteasome inhibition may be cell specific, and in some tumor cells both decreases in c-FLIP and increases in DR4 and 5 may cooperate to amplify the apoptotic signal. It should also be borne in mind that even though the blocking of NF-κB by proteasome inhibition is not essential for the sensitization of all tumor cells to Apo2L/TRAIL, NF-κB inhibition may still be an important event contributing to the sensitization of some tumors [53]. Other studies have noted that proteasome inhibition can modulate the balance between pro and antiapoptotic members of the Bcl-2 family, resulting in accumulation of the proapoptotic protein Bik [54]. Additionally, proteasome inhibition is also reported to promote activation of proapoptotic caspases such as caspase 3 [55]. Interestingly, there are also reports that proteasome inhibitors can reduce levels of XIAP in normal keratinocytes thus sensitizing them to Apo2L/TRAIL-mediated apoptosis [56]. Clearly proteasome inhibition may sensitize cells to Apo2L/TRAIL apoptosis by modification of both NF-κB-dependent and independent pathways. Since many tumor cells are only moderately sensitive or are completely resistant to the apoptotic effects of Apo2L/TRAIL, a combination of Apo2L/TRAIL with enhancers of its activity (such as proteasome inhibitors) may be required for optimal therapeutic effects in vivo.

Apo2L/TRAIL as a tumor suppressor

Although exogenous administration of Apo2L/TRAIL has shown some convincing antitumor effects, the physiological role of Apo2L/TRAIL is still unclear. A number of recent studies using neutralizing antibodies to Apo2L/TRAIL or Apo2L/TRAIL deficient mice demonstrated that Apo2L/TRAIL could be involved in limiting the development of experimental tumors [57, 58]. In these mouse tumor models, NK cells were the major source of Apo2L/TRAIL, and NK cells in the liver expressed high endogenous levels when compared to NK isolated from other tissue sites. These NK cells were responsible for natural protection against liver metastases of Apo2L/TRAIL-sensitive tumors [59–61]. Also, Apo2L/TRAIL expression was increased in the presence of interferon gamma (IFN-γ) [62]. The anti-metastatic effects of biological responses modifiers such as IL12 or α-GalCer in vivo seemed to be due, at least in part, to increases in IFN-γ and Apo2L/TRAIL occurring in these mice following these biological therapies [62, 63]. Interestingly, Apo2L/TRAIL can be expressed by human peripheral blood cells following treatment with anti-CD3 and both type 1 and II interferons [64]. Type I and II interferons induce Apo2L/TRAIL expression on human NK cells, monocytes, DCs and neutrophils accounting for their cytotoxicity against Apo2L/TRAIL-senstive tumor cell lines in vitro [65–67]. The role that Apo2L/TRAIL induction plays in the anti-tumor effects of IFNs is unknown. However, an association between the success of BCG therapy of bladder cancer with the IFN-γ mediated increase in Apo2L/TRAIL expression on local neutrophils has recently been described [68]. In certain tumor cell lines Apo2L/TRAIL expression can be induced on the tumor cells themselves by agents such as IFN alpha or all trans-retinoic acid (ATRA), which may result in Apo2L/TRAIL-mediated tumor cell apoptosis in a paracrine manner [69]. Apo2L/TRAIL may also play a role as an effector molecule in the beneficial effects of GVT following allogeneic bone marrow transplantation, yet may not be involved in the GVHD that is a major side effect of this procedure [70]. Proteasome inhibition using bortezomib has been recently evaluated in allogeneic bone marrow transplantation (BMT) models to ascertain effects on GVHD/GVT. Inhibition of GVHD by early administration of bortezomib was observed in BMT recipients [71] with increased donor T cell apoptosis being reported. Importantly, preservation of GVT was also observed indicating that proteasome inhibition may be of use in BMT to augment antitumor effects. However, consistent with published reports by us and others [50, 72] that proteasome inhibition can augment apoptosis in response to death ligands, we have recently observed that later administration of bortezomib at times in which acute GVHD was actively occurring results in markedly opposite effects in which GVHD-toxicities were augmented by bortezomib (Sun et al. manuscript submitted). Thus, the use of proteasome inhibition can potentially sensitize both the host cells as well as the tumor to immune-mediated attack. The immune lytic molecules responsible for both the GVHD and GVT effects following bortezomib remain to be elucidated and possibly exploited. In conclusion, Apo2L/TRAIL may be an important contributor to the antitumor effects observed following a number of biological therapies of cancer.

Future directions

Administration of Apo2L/TRAIL, agonist anti-DR4 or DR5 antibodies or agents that induce Apo2L/TRAIL expression in vivo is a promising strategy to complement and enhance the efficacy of traditional chemotherapy or radiation. Although much has been learned concerning the complex molecular events necessary for apoptotic signaling by Apo2L/TRAIL, further studies are clearly necessary. The mechanism of Apo2L/TRAIL resistance in normal cells, how oncogenic transformation makes them more sensitive to Apo2L/TRAIL, and whether cells respond to Apo2L/TRAIL in ways other than cell death remains to be completely defined. The identification of crucial signaling molecules controlling apoptotic pathways in cancer cells may allow for the design of more specific molecular targeting of these molecules, thus allowing proapototic signals to reach the critical strength necessary to result in cell death. Furthermore, an increased understanding of crucial signaling molecules could eventually allow for molecular profiling of tumor cells in advance of treatment, to determine if tumor cells of a specific individual can be sensitized to Apo2L/TRAIL therapy or not. On the other hand, the ability of agents such as bortezomib to influence multiple components of the apoptotic pathway may be critical in sensitizing many tumor cells to Apo2L/TRAIL. It still remains a concern that the administration of various agents to increase the sensitivity to Apo2L/TRAIL may also be associated with the amplification of undesired toxicities. Therefore, further studies in tumor-bearing animals are clearly required. Recent exciting data in mice using an agonist antibody to mouse DR5 (MD5-1) are of particular interest [73]. In these studies, the agonist antibody to DR5 induced apoptosis of Apo2L/TRAIL-sensitive tumor cells in vivo when cross-linked by FcR-expressing innate immune cells. There were no signs of toxicity accompanying this treatment. However, more importantly, the anti-DR5-mediated tumor rejection subsequently evoked a tumor-specific T cell immunity. Although most reports indicate that signaling via the Apo2L/TRAIL receptors promote apoptosis, there are instances when cellular necrosis occurs in response to Apo2L/TRAIL [74]. Since necrotic death can promote inflammatory responses [75], the relative importance of the apoptosis or necrosis of tumor cells in promoting T cell immunity in vivo is certainly worthy of more intensive investigation. Interestingly, the antitumor T cells generated following anti-DR5 treatment could reject Apo2L/TRAIL-resistant variants of the tumor in a perforin and/or FasL dependent manner. Current methods for dramatically reducing tumor burden, such as chemotherapy or radiation, are associated with potent immunosuppression. It is possible that the new generation of targeted, tumor-specific agents such as Apo2L/TRAIL may either have little effect on, or even actually enhance, immune responses to cancer. This may allow for the efficient instigation of immune responses to cancer in the context of a low residual tumor burden, in an individual whose immune system has not been severely compromised. This hypothesis would seem to warrant further investigation in the future.

Abbreviations

Apo2L/TRAIL

Tumor necrosis factor-related apoptosis-inducing ligand

IAP

Inhibitor of apoptosis protein

NF-κB

Nuclear factor of kappa-B

SMAC/DIABLO

Second mitochondrial activator of caspases/direct IAP binding protein with low pI

GVT

Graft-versus-tumor

GVHD

Graft-versus-host disease