| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Tumor Immunology Program, German Cancer Research Center (DKFZ), Heidelberg, Germany
Correspondence: Prof. Dr. Peter H. Krammer, Tumor Immunology Program, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. E-mail: P.Krammer{at}dkfz-heidelberg.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT TUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
Key Words: CD95L • cytotoxic T cells • CD95 • death receptor • immunosuppression
|
TUMORS AND THE IMMUNE SYSTEM |
|---|
|
TOPABSTRACTTUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
Numerous tumor antigens have been identified that can be recognized by T cells [3 4 5 ]. Some of these antigens are expressed exclusively by tumors and thus are called tumor-specific antigens. These antigens arise from mutations or translocations of normal cellular genes (e.g., ß catenin, cdk4, ras). The mutations may be involved directly in carcinogenesis. Another group of antigens are the tumor-associated antigens that are not only expressed by tumor cells, but are also expressed by other cells of the body. Cancer-testes antigens are expressed on a variety of epithelial tumors as well as on testis and placental tissue (e.g., MAGE, BAGE, GAGE). Overexpressed nonmutated proteins (e.g., p53, Her2/neu) may also serve as tumor antigens for T cells. In addition, tumor-infiltrating lymphocytes have been identified that are reactive against differentiation antigens present on normal melanocytes as well as melanomas (e.g., MART-1/Melan-A, tyrosinase, gp100). Moreover, antigens from tumorigenic viruses are presented on tumor cells. The expression of tumor antigens may be heterogeneous within a tumor, and a single patient can develop immune reactions to multiple antigens [6 , 7 ].
Two different models for the immune response to tumors have been proposed: the concept of immunosurveillance and the danger model. According to the immunosurveillance hypothesis, tumors expressing antigens are regarded as "nonself" by the immune system, and a major function of the immune system is to survey the body for the development of malignancy and to eliminate tumor cells as they arise [8 ]. To detect "danger," the immune system uses professional antigen-presenting cells (APC) as sentinels of tissue damage. In the presence of danger signals, APC—such as dendritic cells, activated macrophages, and B cells—stimulate the T cell response. The danger model proposes that cancer cells do not appear dangerous to the immune system, so that the response of T cells to tumors is not initiated [9 ].
Natural killer (NK) cells of the innate immune system also play an important role in immune surveillance of tumors [1 ]. NK cells kill MHC class I-deficient cells—a phenomenon that is part of the "missing self" hypothesis [10 , 11 ]. The activity of NK cells is controlled by a balance of positive and negative signals. Engagement of inhibitory receptors by MHC class I molecules blocks activation signals. Two families of inhibitory receptors have been identified in humans: the immunoglobulin-like killer cell inhibitory receptors and the lectin-like CD94-NKG2 receptors. Stimulatory receptors comprise receptors (e.g., CD16, CD94-NKG2C, natural cytotoxicity receptors) that are supposed to bind to constitutively expressed ligands [12 ] and NKG2d receptors, which bind to molecules that are induced by cellular stress [13 14 15 ]. Ligands for NKG2d receptors are the MHC class I chain-related (MIC) glycoproteins MICA and MICB in humans and the minor histocompatibility antigen H60 and the retionic acid early inducible (Rae-1) family in mice.
Additional cells of the innate immune system involved in immunity against tumors are macrophages and neutrophils [16 17 18 19 ]. They can reject tumors by direct killing of the tumor cells, by destruction of tumor vessels and matrix, and by inhibition of angiogenesis. Moreover, they display tumor antigens and can stimulate other immune cells such as CTL, NK cells, or APC. In contrast, inflammatory cells may also contribute to tumor progression by production of tumor growth factors and stimulation of angiogenesis [20 ]. Macrophages and neutrophils are recruited to the tumor site by expression of adhesion molecules on endothelial cells and by chemotactic proteins.
|
TUMOR IMMUNE ESCAPE MECHANISMS |
|---|
|
TOPABSTRACTTUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
Growth of antigenic tumors in the presence of potent immune cells cannot be explained by immunological ignorance alone. A major goal of cancer immunotherapy is to generate an anti-tumor immune response, e.g., by vaccination with cancer cells fused with APC or by transfer of anti-tumor T cells. However, many of these approaches do not efficiently stimulate immunity, or the tumors continue to grow despite the presence of an immune reaction [1 , 3 , 29 ]. Multiple mechanisms have been identified that tumors use to escape from rejection (Table 1 ) [30 31 32 ]. It is likely that malignant escape variants are selected by the immune system.
|
Another strategy that tumors use to escape from immune-mediated rejection is the expression of immunosuppressive factors [18 , 52 ]. These factors may be expressed by the malignant cells themselves or by noncancerous cells present at the tumor site, such as immune, epithelial, or stromal cells. The most prominent of these factors is transforming growth factor ß (TGF-ß) [53 54 55 56 57 58 ]. TGF-ß is a cytokine that affects proliferation, activation, and differentiation of cells of innate and adaptive immunity and thus inhibits the anti-tumor immune response. Moreover, vascular endothelial growth factor (VEGF) is produced by many tumor cells [59 ]. Besides its angiogenic properties, it inhibits the differentiation of progenitors into dendritic cells. Further immunosuppressive factors expressed by malignant cells are prostaglandins [60 61 62 ], interleukin (IL)-10 [63 ], macrophage-colony stimulating factor [64 , 65 ], and soluble tumor gangliosides [66 ]. The membrane protein RCAS1 (receptor-binding cancer antigen expressed on SiSo cells) inhibits proliferation and induces apoptosis in T cells in vitro, suggesting a role of this molecule in immune evasion of tumors [67 ].
The lack of a T cell response against tumor-associated antigens that are also expressed by other cells of the body or during development may be explained by tolerance [30 , 68 ]. In healthy organisms, self-reactive T cells are tolerized mainly by deletion in the thymus, a process known as central tolerance. The mechanisms of peripheral tolerance induction prevent autoimmunity by tolerizing T cells that have escaped the process of central tolerance. Peripheral tolerance induction is a complex multistep process [69 , 70 ], but in principle, four major mechanisms can be distinguished. One mechanism is the induction of anergy. T cell activation requires two signals, binding of a peptide-MHC complex to the TCR and binding of costimulatory molecules (e.g., B7) to their ligands (CD28) on the T cell surface [71 , 72 ]. If a T cell binds via its TCR to a peptide-MHC complex on the target cell without sufficient costimulation, the T cell is rendered anergic and does not become activated when restimulated with antigen. Many tumor cells do not express costimulatory molecules and thus may induce anergy in anti-tumor lymphocytes [27 , 73 74 75 ]. Another process of tolerance induction that tumors exploit is immune deviation. In this process, the immune response is driven toward a Th2 humoral response away from a Th1 response required for efficient tumor rejection by cytotoxic T cells. The mechanism of immune deviation is not exactly understood, but it may depend on secretion of TGF-ß and IL-10 [76 ] or on the presentation of the tumor antigen by B cells to CD4+ Th cells [77 ]. Tumors can also induce the generation of regulatory T cells [78 ]. Although the molecular mechanism is not clear, a subset of CD4+ T cells seems to suppress the response of cytotoxic T cells against tumors in some settings [79 , 80 ]. A further mechanism to establish peripheral tolerance to self-antigens is T cell deletion. Repetitive stimulation of T cells with the antigen induces apoptosis, a process referred to as activation-induced cell death (AICD). Thus, administration of antigens, such as superantigens, peptides, or allogeneic cells, or direct restimulation of the TCR by anti-CD3 antibodies has been shown to induce tolerance by T cell deletion [81 82 83 84 ]. This process is mainly mediated via the CD95/CD95L death system [85 86 87 88 ]. Costimulation via CD28 can rescue T cells from AICD [89 ], implicating another important role for expression of B7 on tumor cells. Tumors have been shown to induce tolerance by deleting anti-tumor T cells [90 , 91 ]. AICD, as a result of chronic stimulation with the tumor antigen may contribute to immune escape, yet the significance of this process has not directly been shown.
Two further strategies used by tumors to evade rejection by the immune system are related to apoptosis. First, malignant cells have changes in the expression of molecules involved in apoptosis signaling, resulting in resistance of the tumor to the killing mechanisms of the immune system. Second, tumors may adopt a killing mechanism from cytotoxic immune cells to delete the attacking anti-tumor lymphocytes, a concept called "tumor counterattack." These apoptosis-related immune-escape mechanisms will be discussed in detail below.
|
KILLING MECHANISMS OF THE IMMUNE SYSTEM |
|---|
|
TOPABSTRACTTUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
m drops. In the cytosol, cytochrome c forms a complex with Apaf-1, adenosine 5'-triphosphate, and procaspase-9. This complex is called apoptosome. Within the apoptosome, the initiator caspase-9 is activated.
|
Various proteins regulate the apoptotic pathway at different levels. FLIPs (FLICE-inhibitory proteins) interfere with the initiation of apoptosis directly at the level of death receptors [112 ]. Two splice varians have been identified in human cells, a long form (cFLIPL) and a short form (cFLIPS). They share structural homology with procaspase-8, but lack its catalytic site. This structure enables them to bind to the DISC, thereby inhibiting the processing and activation of the initiator caspases.
A major class of regulatory proteins are members of the Bcl-2 family that regulate apoptosis at the mitochondrial level [109 , 110 ]. According to their function, Bcl-2 family members can be divided into antiapoptotic and proapoptotic proteins. Bcl-2 family proteins influence the permeability of the mitochondrial membrane; however, the biochemical mechanism of their function is not entirely clear.
A third class of regulatory proteins are the IAPs (inhibitor of apoptosis proteins), which bind to and inhibit caspases [113 ]. They may also function as ubiquitin ligases, promoting the degradation of the caspases bound. However, not all IAPs have been shown to suppress apoptosis, and some of them may also have functions other than caspase inhibition. IAPs themselves are inhibited by a protein, Smac/DIABLO [114 , 115 ], which is released from mitochondria along with cytochrome c during apoptosis and promotes caspase activation by binding to IAPs.
In the calcium-dependent granule exocytosis pathway, lymphocytes secrete perforin and granzymes from cytotoxic granules toward the target cell. In the presence of calcium, perforin polymerizes and initiates as yet ill-defined changes in the target cell membrane, which allow granzymes to pass into the cell [116 117 118 ]. Granzymes are neutral serine proteases that can activate caspases in the target cell [119 120 121 ]. In addition, granzyme B may directly cleave the Bcl-2 family member Bid, initiating the mitochondrial death pathway [122 ]. Perforin-deficient mice are very susceptible to some carcinogens and oncogenic viruses [123 ] and develop spontaneous lymphomas with age [124 ]. Although it is clear that granzymes are indispensable effector molecules in a granule exocytosis-mediated host defense against viral pathogens [125 ], their contribution to tumor rejection remains controversial [126 ]. Thus, mice deficient for granzymes A and B are capable of rejecting tumors in an efficient way.
Macrophages and neutrophils use totally different killing mechanisms. In principle, they use four kinds of cytotoxic molecules [127 ]. The first effector mechanism is the oxidative burst consisting of the release of reactive oxygen species (superoxide anion, hydrogen peroxide, and derivatives) produced by the phagocytic reduced nicotinamide adenine dinucleotide phosphate oxidase (NADPH) [128 129 130 ]. Reaction of peroxide via the myeloperoxidase pathway can yield hypochlorous acid and chloramines. A further cytotoxicity mechanism is the production of nitric oxide (NO) by the inducible NO synthase [129 , 131 ]. The toxicity of NO is enhanced greatly by reacting with superoxide to form peroxynitrite. The molecular targets of reactive oxygen species, NO, and derivatives inside the target cells are not fully defined yet, but may include enzymes essential for cellular survival. Phagocytes also release antimicrobial peptides, defensins and cathelicidins, which have affinity for bacterial and eukaryotic membranes and may lyse cells by disrupting the phospholipid bilayer [132 ]. Other mechanisms include the production and release of a variety of proteases including elastase, proteinase 3, and metalloproteases [127 ]. These proteases degrade extracellular matrix components and other proteins.
|
RESISTANCE TO APOPTOSIS AND IMMUNE ESCAPE |
|---|
|
TOPABSTRACTTUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
Although many mechanisms of tumor resistance to apoptosis have been identified (Fig. 1) , only a few of them have directly been shown to be involved in immune escape. One strategy tumors use to acquire apoptosis resistance is the overexpression of antiapoptotic molecules. The antiapoptotic proteins FLIPL,S interfere with apoptosis induction at the level of death receptors, but do not prevent apoptosis by perforin/granzyme [112 , 135 , 136 ]. Nevertheless, overexpression of FLIP mediates the immune escape of tumors in mouse models. Tumors with high expression levels of FLIPL were shown to escape from T cell-mediated immunity in vivo despite the presence of the perforin/granzyme pathway [137 ]. In vivo tumor cells were selected for elevated FLIPL levels. FLIPL overepxression also prevents rejection of tumors by perforin-deficient NK cells [138 ]. Human melanomas were shown to express high levels of FLIP [135 , 139 , 140 ]. Moreover, in Epstein-Barr virus-positive Burkitt’s lymphoma cell lines, an increased ratio of FLIP to caspase-8 correlated with resistance to CD95-mediated apoptosis [141 ]. Viral analogs of FLIP, viral FLIPs (v-FLIPs), are encoded by some tumorigenic viruses, such as Kaposi sarcoma-associated herpesvirus (KSHV) [142 143 144 ]. KSHV-FLIP promotes tumor establishment and growth in immunocompetent mice by prevention of death receptor-induced apoptosis triggered by T cells [145 ]. Therefore, v-FLIPs may contribute to immune escape of v-FLIP-encoding viruses.
Further antiapoptotic proteins are involved in apoptosis resistance of tumors, although the significance for immune escape is less clear. Enhanced Bcl-2 expression is found in follicular B-cell lymphoma (Bcl) with the chromosomal translocation t(14;18) that couples the Bcl-2 gene to the immunoglobulin heavy chain locus [146
147
148
149
]. In cooperation with the oncogenes c-Myc or promyelocytic leukemia retinoic acid receptor 
, Bcl-2 contributes to tumorigenesis [150
151
152
153
]. In some studies, high Bcl-2 expression correlates with the grade of malignancy of human tumors [154
155
156
]. Moreover, it has been shown in in vitro and in vivo models that Bcl-2 expression confers resistance to CD95L and other apoptosis stimuli [154
, 157
158
159
]. In some types of tumors, high Bcl-2 expression appears to be predictive of a poor disease-free survival [155
, 156
, 160
]. The tumor-associated viruses Epstein-Barr virus and human KSHV encode proteins that are homologs of Bcl-2. Both proteins, BHRF1 or KSbcl-2 (vBcl-2), respectively, have an antiapoptotic function and enhance survival of the infected cells [161
162
163
164
]. Thus, they may contribute to apoptosis resistance of virus-induced tumors. In addition, the antiapoptotic Bcl-2 family members Bcl-xL and Mcl-1 are up-regulated in tumors and can confer resistance to multiple apoptosis-inducing pathways [159
, 165
166
167
168
].
The IAP family member Survivin is expressed in a highly tumor-specific manner [169 ]. It is found in the vast majority of human tumors, but not in normal adult tissues [170 ]. In neuroblastomas, expression correlates with a more aggressive and unfavorable disease [171 ]. Besides its antiapoptotic activity, Survivin also has an apparent role in the cell cycle [169 ]. In transgenic mice expressing Survivin in the skin, its antiapoptotic function was more prominent than its role in cell division; Survivin, however, did not affect CD95-induced cell death [172 ]. Expression of a mutant of Survivin that could not be phosphorylated induces cytochrome c release and cell death. In xenograft tumor models, this mutant suppressed tumor growth and reduced intraperitoneal tumor dissemination [173 , 174 ]. The tumor-suppressing activity has been observed in immune-deficient mice and thus does not seem to depend on the immune system but on other apoptosis stimuli. Another IAP family member, cIAP2, is affected by the translocation t(11;18)(q21;q21) that is found in about 50% of marginal cell lymphomas of the mucosa-associated lymphoid tissue (MALT) [175 ]. This suggests a role of cIAP2 in the oncogenesis of MALT lymphoma.
Tumor cells cannot only resist killing by cytotoxic lymphocytes through mechanisms blocking the death receptor pathway, but also through direct interference with the perforin/granzyme pathway. The serine protease inhibitor PI-9/SPI-6 that inhibits granzyme B is expressed in a variety of human and murine tumors. Overexpressions result in resistance of tumor cells to cytotoxic lymphocytes and to immune escape [176 177 178 ].
A different strategy of tumors to escape from the effector mechanisms of the immune system may be to neutralize or impair the death-inducing stimuli. Thus, the expression of soluble receptors that act as decoys for death ligands may interfere with apoptosis induction via death receptors. To date, two distinct soluble receptors, soluble CD95 (sCD95) and decoy receptor 3 (DcR3), have been shown to inhibit CD95 signaling competitively. sCD95 is expressed in various malignancies, and elevated levels can be found in the sera of cancer patients [179 180 181 182 ]. High sCD95 serum levels were associated with poor prognosis in melanoma patients. DcR3 is amplified genetically in a number of lung and colon carcinomas and is overepxressed in several adenocarcinomas, glioma cell lines, and glioblastomas [101 , 102 , 183 , 184 ]. Ectopic expression of DcR3 in a rat gliosarcoma model resulted in a decrease of immune cell infiltration, suggesting an involvement of DcR3 in immune evasion of malignant glioma [184 ]. Moreover, a mechanism for resistance toward the perforin/granzyme pathway may be that binding of perforin to the tumor cell membrane is impaired [185 ]. Acute myeloid leukemia cells that failed to bind perforin on their surface were completely resistant toward NK cell-mediated cytotoxicity.
Tumors can also acquire apoptosis resistance by down-regulation or inactivation of proapoptotic molecules. In comparison to their normal counterparts, some tumor cells show a decreased expression of the death receptor CD95. This has been demonstrated for hepatocellular carcinomas, neoplastic colon epithelium, melanomas, and other tumors [186 187 188 189 190 ]. Loss of CD95 may contribute to chemoresistance and immune evasion. Transcriptional mechanisms may account for this. Oncogenic Ras seems to down-regulate CD95 [191 ], and in hepatocellular carcinomas, loss of CD95 expression was accompanied by p53 aberrations [190 ]. Several CD95 gene mutations have been demonstrated in primary samples of myeloma and T cell leukemia [192 193 194 ]. The mutations include point mutations in the cytoplasmic death domain of CD95 and a deletion leading to a truncated form of the death receptor. These mutated forms of CD95 may interfere in a dominant-negative way with apoptosis induction via CD95. In families with germline CD95 mutations, usually resulting in autoimmune lymphoproliferative syndrome, the risk for the development of lymphomas is increased [195 ]. Deletions and mutations of the death receptors TRAIL-R1 and TRAIL-R2 have also been observed in tumors [196 197 198 199 ]. The frequent deletion of the chromosomal region 8p21-22 in head and neck cancer and non-small cell lung cancers affects the TRAIL-R2 gene. Mutations have been found in the ectodomain or the death domain of TRAIL-R1 or TRAIL-R2. Further mutations result in a truncated form or other antiapoptotic forms. Down-regulation or mutation of death receptors may impair tumor immunosurveillance by NK and T cells.
In neuroblastomas with amplification of the oncogene MYCN, the gene for the initiator caspase-8 is frequently inactivated by DNA methylation and gene deletion [200 ]. Caspase-8-deficient neuroblastoma cells are resistant to death receptor- and doxorubicin-mediated apoptosis.
In certain types of cancer, the proapoptotic Bcl-2 family member Bax is mutated [201 202 203 ]. Common mutations comprise frameshift mutations leading to loss of expression and mutations in the BH domains resulting in loss of function. Tumor cell lines with frameshift mutations are more resistant to apoptosis. Reduced Bax expression is associated with a poor response rate to chemotherapy and shorter survival in some situations [204 ]. Several mouse studies confirmed the function of Bax as a tumor suppressor [205 206 207 ]. However, Bax may be relevant primarily for apoptosis stimuli such as chemotherapy or p53 and not for apoptosis triggered by the immune system.
Metastatic melanomas often do not express Apaf-1, which forms an integral part of the apoptosome [208 ]. A high rate of allelic loss of the Apaf-1 locus can be observed. The remaining allele is transcriptionally inactivated by gene methylation. Apaf-1-negative melanomas fail to respond to chemotherapy, a situation found commonly in this type of tumor.
Taken together, tumor cells use many mechanisms to acquire apoptosis resistance. Although a direct role of these mechanisms for immune escape has only been shown in a few studies, it is likely that apoptosis resistance is not only relevant for tumorigenesis and resistance to chemotherapy, but also influences immunosurveillance and immunotherapy.
|
TUMOR COUNTERATTACK |
|---|
|
TOPABSTRACTTUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
Many publications have supported the idea of tumor counterattack as an immune escape mechanism. CD95/CD95L interactions are discussed as being an important mechanism for the maintenance of immune privilege. CD95L is expressed in immune-privileged sites, e.g., the eye and the testis [212 213 214 ]. Thus, in the eye a high percentage of human corneal transplants are accepted without tissue matching or immunosuppressive therapy. In a mouse model, CD95L-positive corneal allografts were accepted at a rate of about 45%, whereas all grafts from mice with a mutated, nonfunctional CD95L (gld) were rejected [214 ]. Moreover, inflammatory cells entering the anterior chamber of the eye in response to viral infection underwent apoptosis dependent on CD95/CD95L interactions and did not produce any tissue damage [213 ]. In contrast, viral infection in gld mice, which lack functional CD95L, resulted in inflammation and invasion of ocular tissue by cells without signs of apoptosis. CD95-mediated apoptosis of lymphoid cells was necessary for tolerance induction following antigen injection into the anterior chamber of the eye [215 ].
Data along similar lines were also found with murine testis grafts. Grafts expressing wild-type CD95L survived indefinitely when transplanted under the kidney capsule of allogeneic animals, whereas testis grafts from gld mice were rejected [216 ]. However, attempts to reproduce these observations by another group were not successful [217 ]. Because neurons and astrocytes also express CD95L [212 , 218 , 219 ], immune privilege in the central nervous system may also involve CD95L.
These findings suggest that CD95L may be used to render a transplanted tissue an immune privileged site. Indeed, it was demonstrated that cotransplantation of syngeneic CD95L-transfected myoblasts protected allogeneic islets of Langerhans grafts from immune rejection [220 ]. However, other authors have found exactly opposite results using a similar technology [221 ]. Numerous studies have shown that tumor counterattack may be a relevant immune escape mechanism of tumors. Many CD95-resistant tumors express CD95L constitutively or after induction by chemotherapy [211 , 218 , 222 223 224 225 226 227 ]. Such tumor cells killed CD95-positive, apoptosis-sensitive cells in vitro. Moreover, apoptosis of tumor-infiltrating lymphocytes has been found in situ within CD95L-expressing human tumors [186 , 224 ]. In esophageal cancer, the number of infiltrated lymphocytes was reduced concomitantly with increased lymphocyte apoptosis within CD95L-expressing areas of the tumors [223 ].
Various animal models have been used to demonstrate the ability of CD95L expressed on tumors to down-regulate anti-tumor immune responses. Tumor growth of a subcutaneously injected CD95L-positive murine melanoma cell line was slightly faster in wild-type or gld mice than in mice with a mutated or down-regulated CD95 receptor (lpr mice) [224 ]. Another study in syngeneic mice showed that growth of tumors of murine CD95L-transfected cells was significantly better than that of control cells when implanted under the kidney capsule [228 ]. Immunosuppression in vivo was directly demonstrated in allogeneic mice injected with CD95L-transfected colon carcinoma cells. Alloantibodies were virtually completely abolished and allospecific cytotoxic T lymphocytes and helper T cells were reduced [229 ].
Despite the wealth of data accumulated in support of the tumor counterattack hypothesis, many contradictory studies have also been published [230 ]. In contrast to the above findings, it has been shown that CD95L expression by grafts or by tumor cells targeted the cells for rapid destruction by neutrophils. CD95L expression on pancreatic islets transplanted into allogeneic hosts resulted in a massive infiltration of neutrophils and in islet destruction. Similarly, transgenic mice expressing CD95L in pancreatic ß cells developed a massive infiltration of neutrophils and diabetes [231 ]. CD95L-expressing islet ß cells transplanted under the kidney capsule of syngeneic or allogeneic animals were not protected from rejection [232 , 233 ]. In addition, CD95L-expressing hearts from transgenic mice transplanted into sygneneic and allogeneic recipients were more rapidly rejected than control grafts and showed massive neutrophil infiltration as early as 1 day after transplantation [234 ].
Various studies have shown that the overexpression of CD95L in murine tumor cells resistant to CD95-mediated apoptosis did not affect growth in vitro, but caused rejection by neutrophils in vivo [235 236 237 238 239 ]. CD8+ T cell-mediated protective immunity against subsequent challenge with the parental tumor cells was elicited. Rejection of the CD95L-expressing tumor has even been observed in the study mentioned above, demonstrating immunosuppression by the tumor in an allogeneic mouse model [229 ]. In addition, when control tumor cells were co-implanted with the CD95L-expressing cells at the same site, a "bystander rejection" of the CD95L-negative cells was found [236 237 ]. Infection of a subcutaneously growing CD95-negative tumor by an adenoviral vector encoding CD95L resulted in rapid elimination of the tumor [240 ]. These studies indicate that CD95L has a proinflammatory function and that gene transfer of CD95L may be used in tumor eradication.
Several mechanisms for the recruitment of the graft- or tumor-rejecting neutrophils have been proposed. Two studies suggested that soluble CD95L is directly chemotactic for neutrophils implicating that soluble CD95L promotes rejection of CD95L-expressing grafts [241 , 242 ]. However, others have not found a chemotactic activity of soluble CD95L in vitro or in vivo [235 , 243 ]. Moreover, tumor cells expressing only soluble CD95L did not elicit a neutrophilic response [243 , 244 ]. Conversely, tumor cells expressing a noncleavable membrane-bound form of CD95L were rapidly rejected [243 244 245 ]. The extent of inflammation induced by the various transfectants seemed to correlate with the cytotoxic activity of CD95L. Alternatively, it has been suggested that CD95L acts on surrounding cells to induce the production of granulocyte chemoattractants. Thus, it has been demonstrated that CD95L induces the processing and release of IL-1ß, which may be responsible for the infiltration by neutrophils [246 ]. CD95L may act on resident macrophages, leading to increased production of IL-1ß and macrophage inflammatory proteins [247 ]. Moreover, engagement of CD95 on dendritic cells may induce the secretion of proinflammatory cytokines [248 ]. These data support an indirect mechanism for the inflammatory effect of CD95L. Nevertheless, the exact mechanism of how neutrophils are recruited to CD95L-expressing tumors and how CD95L-expressing tumors are rejected is still unclear.
Many studies of tumor counterattack have been published, but the results are contradictory, and therefore do not clarify whether tumor counterattack is a relevant immune escape mechanism in vivo. No study has demonstrated conclusively that a tumor (or graft), by CD95L expression, deleted anti-tumor-specific lymphocytes, escaped the immune response, and thus had a growth advantage in vivo.
Additional factors may influence the outcome of CD95L expression on tumors in vivo and thus may account for the controversial situation. Level and time-point of CD95L expression may be particularly relevant. In most animal experiments, CD95L-transfected tumor cell lines have been used. These cells may express different, probably higher levels of CD95L than naturally occurring tumors. Overexpression of CD95L may lead to rejection by neutrophils, whereas physiological levels may not induce neutrophilic infiltration but may still suffice to delete anti-tumor lymphocytes. Moreover, tumors growing in situ may express CD95L at late stages of tumorigenesis, e.g., after induction by chemotherapy [249 , 250 ].
Therefore, the time-point of CD95L expression may directly influence tumor counterattack. Furthermore, the sensitivity of T cells to CD95-mediated apoptosis varies considerably with respect to the activation status of the T cells [84 , 89 , 251 252 253 ]. Therefore, it may be crucial when T cells encounter CD95L and how the T cells have been activated and costimulated. The consequences of CD95L expression may also depend on the microenvironment. It has been suggested that TGF-ß is necessary as a cofactor for promoting immunologic tolerance [254 ]. CD95L-expressing tumor cells were rejected by neutrophils when injected subcutaneously. When TGF-ß was simultaneously provided to the subcutaneous sites, protection against tumor rejection was observed. However, it is not clear whether in vivo TGF-ß and CD95L together are necessary for immune escape and whether CD95L has an immunosuppressive effect on T cells in the presence of TGF-ß.
Despite the controversial situation concerning CD95L and tumor counterattack, other molecules have also been implicated in T cell deletion by tumors. Thus, the death ligand TRAIL has been suggested to suppress tumor-specific T cell responses in a similar manner as CD95L [255 ]. Moreover, the membrane protein RCAS1 [67 ] not only inhibits proliferation, but also induces apoptosis in T cells in vitro. Chemokine production by tumor cells may also sensitize T cells to apoptosis [256 ].
|
CONCLUSIONS |
|---|
|
TOPABSTRACTTUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
It is the present aim of research in this area to understand these events further and to use this insight to resensitize tumor cells to apoptosis. Recently, it has been shown that low doses of chemotherapy or irradiation sensitized resistant cells to TRAIL-induced apoptosis in vitro and in vivo [257 , 258 ]. Therefore, these treatments might enhance the susceptibility of a tumor to immune attack. A future therapeutic strategy may also involve down-regulation of antiapoptotic molecules such as FLIP or Bcl-2 by antisense oligonucleotides or dsRNA interference [259 , 260 ]. However, a macroscopic tumor is heterogeneous, and different cells within the tumor may also use different immune escape mechanisms including apoptosis resistance, impaired antigen presentation, secretion of immunosuppressive factors, and other strategies (Table 1) . Moreover, multiple mechanisms may develop in a single tumor cell. Therefore, it is questionable whether a single, predominant immune-escape mechanism can be identified in a tumor and whether therapeutic targeting of one mechanism alone is promising. Deeper insight into the molecular mechanisms underlying tumor immune escape may finally lead to novel therapeutic approaches that will be used for the benefit of cancer patients.
|
ACKNOWLEDGEMENTS |
|---|
Received February 3, 2002; revised March 18, 2002; accepted March 20, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTTUMORS AND THE IMMUNE...TUMOR IMMUNE ESCAPE MECHANISMSKILLING MECHANISMS OF THE...RESISTANCE TO APOPTOSIS AND...TUMOR COUNTERATTACKCONCLUSIONSREFERENCES |
|---|
