Immunological Control of Neoplasia - Chapter 2

Author: Christine Ichim

Chapter 2 Components of Immune Response

Overview of the Antitumour Response

The immune system, according to Janis Kuby is "a remarkably adaptive defense system that has evolved in vertebrates to protect them from invading microorganisms and cancer" (156). Although a comprehensive review of the immune response to tumours is beyond the scope of this work, the typical immune response to tumours will be delineated. For reviews of this field the reader is referred to the following excellent papers (157-165).

The initiation of an immune response requires a signal to alert the host that there exists a situation in the body dangerous to its overall being. Whether this signal comes from detection of a foreign protein, a damaged cell, or a live parasite, the initiation of the response commences with a "danger" signal (166,167). Cells that detect these "danger" signals belong to the front line defenses, called "innate" immunity. These cells activate the specialized immune cells to respond to the danger. These specialized immune cells steer the profile of the response, attempting to make it a protective response, then activate the effector mechanisms in the immune response. Thus, from a simplistic perspective the immune response can be divided into three phases: the sensory phase, the synthesis phase and the effector phase.

Sensory Phase

Cells involved in the sensory phase include dendritic cells, macrophages, neutrophils, gamma delta T cells and natural killer cells. These cells first encounter the danger signal and then initiate activation of cells in the synthesis phase. One example of the sensory phase is the ability of dendritic cells to recognize tumour cells and present the antigens to the synthesis phase T-cells via MHC class 2 T cell receptor interaction (168). Does the dendritic cell present only tumour antigens or does it present all antigens and have the T cell “select” which one it will be activated by? The answer is both, T cells that react to self antigens are usually induced to undergo apoptosis in the thymus (169) while dendritic cells generally have a higher tendency to present foreign antigens as opposed to self antigens. In a normal immune response to tumour cells, the dendritic cell engulfs tumour debris from damaged or aged tumour cells and processes the tumour-derived proteins intracellularly (168). The dendritic cell, now activated, migrates to the draining lymph node of the tumour, it is here where the synthesis phase of the immune system gets activated.

Synthesis Phase:

During the synthesis phase a decision is made regarding the type of immune response that will ensue. This phase is critical. To confer protection, not only is it important to initiate an immune response but the response initiated must be of the proper type. An inappropriate response can have disastrous consequences. This is exemplified in leprosy where a cell-mediated response will yield protection whereas an antibody-mediated response will result in wasting (170). The synthesis phase of the immune response occurs in the lymph nodes. Little is known about the mechanisms involved in decision making at the synthesis phase of the immune response. However, work done at a reductionist level demonstrates that the interaction between host genetics, pathogen secreted substances, and area of infection all are involved in the "thinking" process required before launching the immunological attack. The decision reached at the synthesis phase then is transmitted to the effector phase, usually through soluble molecules such as cytokines. Examples of synthesis phase decisions are the Th1 and Th2 polarization of responses.

Effector Phase:

The effector phase of the immune response can be crudely grouped into two categories: the antibody mediated effectors, usually driven by Th2 cytokines and the cell mediated effectors, usually driven by Th1 cytokines (171).

Generally the antibody immune response is the protective response against extracellular pathogens and parasites. Upon antigen stimulation and in concert with T cell help, activated B cells differentiate into antibody producing plasma cells. Antibodies are extracellular molecules that can recognize exogenous antigen by noncovalently binding to it. Antibodies cannot recognize intracellular antigens since they cannot pass through the cell membrane. Antibodies mediate their effector functions through one of the following means: 1. Neutralization of antigen. This is the case with antibodies directed toward bacterial LPS, they bind onto LPS's active site and prevent it from inducing its toxic biological effects (172). 2. Agglutination of antigen. This occurs in bacterial infections of the small intestine where inflammatory conditions would have disastrous consequences for the host. By agglutination, the bacteria can easily be discarded by the cilia. In contrast, this clearance does not happen to unagglutinated bacteria (156). 3. Opsonization. This occurs when an antibody, or opsonin, binds onto the antigen on the target pathogen and makes the pathogen more palatable for phagocytosis by neutrophils (173). 4. Antibody mediated cellular cytotoxicity (ADCC). When an antibody binds an antigen sometimes the Fc portion of the antibody will link the target cell to an effector cell such, as an NK cell, and promote its lysis (174). The NK cell, monocytes (175), neutrophils (176) and eosinophils (177) all possess receptors for the Fc portion of the antibody molecule. 5. Induction of complement. Antibodies binding to antigen induce an alteration on their Fc portion that activates a cascade of plasma dwelling proteins called complement. The complement cascade culminates in lysis of the cells that have bound the activated antibody. The cytotoxic effects of complement cascade activation result from the formation of a nine-protein polymer called the "membrane attack complex" that inserts itself into the target cells and causes lysis (178,179).

Cell mediated immunity (CMI) generally aims to eradicate intracellular pathogens such as viruses, intracellular bacteria and cancer. The most widely studied mechanism behind CMI is the activity of the cytotoxic T lymphocyte (CTL). CTLs are CD8+ T cells that circulate in the blood as precursors (pCTL) (180). A pCTL will differentiate into a CTL if two signals are present: the first is a signal from the target cell and the second is from a T helper cell (181). The signal from the target cell usually occurs due to an intracellular abnormality, such as a mutated self protein (182-185). CTLs can recognize intracellular abnormalities because they are continuously patrolling host cells for the peptides presented to them by MHC class one. The peptides presented on MHC class 1 by the host cell are usually intracellular self proteins. Since the immune system deletes autoimmune pCTL in the thymus (169), then hypothetically all pCTLs circulating in the body are specific toward foreign proteins. Therefore, when a cell mutates, or expresses viral proteins on its class 1 MHC, the pCTL will change into a CTL and abolish the abnormality if the secondary signals are present. In absence of the second signal, the pCTL will either become anergic or undergo apoptosis (186). The second signal comes in the form of cytokine support from the T helper cell. The T helper cell is the command post of the immune system since it carries out a significant portion of the synthesis phase of immune activation. For a T helper cell to get activated, the antigen must be presented on MHC class 2 (187). While almost every cell in the body expresses MHC class 1, only antigen presenting cells express MHC class 2. Antigen presenting cells are usually phagocytes that engulf antigen and express it in a way that induces a T helper response. The often asked question is how do tumour cells activate CD4 + helper cells since the tumour is not an antigen presenting cell? The answer lies in the fact that dead cancer cells are phagocytosed by antigen presenting cells that subsequently activate CD4+ helper cells.

Immune Response to Tumours

Tumour specific antigens generally initiate an immune response by activation of the sensory component of the immune system. Tumour characteristics recognized by the immune system could be; mutated intracellular peptide presented on MHC 1 such as mutant Ras (188), abnormal mucins (189,190), downregulation of MHC molecules (191) or heat shock protein expression (192-194). These activation signals are not mutually exclusive; it is very likely that tumours with mutant ras also express heat shock proteins and have downregulated MHC. Little work has been done, to the author's knowledge, regarding how these multiple signals converge into the initiation of immune response or its subsequent outcome. It is interesting that the immune activation signals on the tumour are targeted toward different branches of the immune response. For example, the mutated ras would activate a T cell (188), which is considered part of the synthesis phase of the immune response. T cells are also activated by MUC antigens that express different glycosylation patterns on cancer cells as opposed to normal cells (195). Downregulation of MHC usually activates NK cells (191) or potentially even monocytes (196); both are parts of the sensory phase of the immune response. Heat shock proteins can stimulate NK cells (192-194), and gamma delta T cells (197,198), both of which belong to the sensory component of the immune response. Cancer antigens can therefore be viewed as the antigens needed to initiate the immune response (the antigens that activate sensory immune cells) and the antigens needed to maintain a response (the antigens that activate the synthesis phase of the immune system). The fact that cancer antigens exist that activate different parts of the immune system should be examined with the hope of elucidating which combination of antigens will lead to protective immunity. Only by dissecting the different pathways in which the cancer cell stimulates the immune response, can we hope to devise proper vaccination strategies. The remainder of this chapter will examine immune cells involved in neoplasia with the hope of illustrating the multidimensional interactions occurring during antitumour responses.

Dendritic Cells:

One of the first cells to encounter antigen and initiate the immune response is the dendritic cell (DC); therefore, we call them sensory cells. These are tissue-fixed cells that are interspersed throughout the body. When DC are activated through contacting an antigen, they migrated to the local lymph node where they activate the synthesis part of the immune system, i.e., the T cells (199). Evidence of the importance of dendritic cells in the initiation of an immune response came from experiments that demonstrated a vital need for these cells in the generation of in vitro anti-sheep red blood cell responses (200), mixed lymphocyte reaction (201), and in vitro anti-TNP T cell responses (202). Dendritic cells express very high levels of MHC class 2, this allows them to present antigen to T-cells at a very efficient rate. This antigen presentation step is where the sensory phase of the immune response encounters the synthesis phase. At this phase not only is antigen presented to T cells, but several secondary signals must also be present. These signals may be cytokines, surface molecules or other soluble mediators. The signals given to the T cell from the antigen presenting cell assist the T cell in determining the type of immune response that will ensue. For example, dendritic cells are potent secretors of the signaling cytokine IL-12 (203), this cytokine is needed for the initiation of a Th1 immune response (204-206). However, in some immune responses the dendritic cells present antigen but do not provide IL-12, this results in anergy, an antigen specific tolerance characterized by the death or inactivation of the T cells specific for that antigen (207). An unanswered question is "through what mechanisms do the dendritic cells decide which signals to give to the T cell when it presents it with antigen?" Do these signals depend on the structural features of the antigen? Are there molecularly conserved features that are recognized as "danger" and therefore, the dendritic cell knows to provide the signals that will ensue the initiation of a strong immune response? An interesting hypothesis is that the dendritic cell itself is an area of information synthesis and decision making, just like the T cell. The signals influencing the dendritic cell will effect what signal the dendritic cell will send to the T cell. The area of dendritic cell manipulation is a rapidly growing field largely due to the recent ability to grow respectable numbers of these cells in vitro (208,209).

Since these dendritic cells seem to play such an important role in immune response initiation, the question arises whether the tumour cell can modulate host dendritic cells so that protective responses will not be initiated. An experiment addressing this question was conducted by Steinbrink et al when his group cultured dendritic cells with IL-10 (a Th2 cytokine secreted by tumours). In contrast to controls, IL-10 cultured tumours lost the ability to stimulate in vitro T cell responses. These IL-10 "deactivated" dendritic cells actually induced a state of anergy in the T cells specific for the antigen being presented (210). Another example of tumour-released soluble mediators deactivating dendritic cells was demonstrated by Gabrilovich et al (213) who showed that culturing dendritic cell progenitors with vascular endothelial growth factor (VEGF), a tumour secreted angiogenic factor, prevents dendritic cell maturation and inhibits the ability of these cells to activate T cells. Hypothetically deactivation of the dendritic cell’s immunoinitiating abilities, as seen in the aforementioned example, could occur to the dendritic cells that are in close proximity to the tumour. These deactivated dendritic cells would then present tumour antigens but simultaneously send inhibitory signals to the T cells that recognize the tumour antigen, thereby diminishing the host’s T cell response against the tumour. Such an explanation could account for observations that cancer patients do not always suffer from systemic immunosuppression: the immunosuppression may be restricted only to tumour antigens.

Although a limited study, the clinical importance of dendritic cells can be seen in an experiment by Enk et al (211) where the dendritic cells of melanoma patients who were either responding to treatment and those from patients who were not responding to treatment were compared. Dendritic cells purified from the tumours of responding individuals had a five fold greater ability than the dendritic cells of non-responding patients to stimulate a mixed lymphocyte reaction. Also, the dendritic cells from non-responding patients possessed less B-7.2 costimulatory molecule than dendritic cells from responding patients. In addition, the cytokines secreted from the responding patients’ dendritic cells where of the Th1 family (IL-2, IL-12, IFN-(), whereas those secreted by the dendritic cells of the non-responding patients were of the Th2 family (IL-10). The ability of dendritic cells to assist tumour growth was also assessed by Chaux et al (212). Utilizing a rat colon carcinoma they demonstrated the existence of tumour associated dendritic-like cells. These cells were presumed to be involved in tumour evasion since they possessed protolerogenic characteristics such as a tumour secreted angiogenic factor, lack of costimulatory molecules but expression of MHC class 2. The concept of tolerogenic dendritic cells is supported by findings in the field of transplantation immunology where passenger leukocytes assist in graft acceptance. These passenger leukocytes are donor-derived dendritic cells that present the alloantigen to recipient T cells. However, the presentation takes place so that instead of entering a state of activation, the recipient T cells enter a state of anergy. The field of tolerogenic dendritic cells is reviewed in the following articles with respect to cancer (214) and transplantation (215,216).

Since dendritic cells possess strong antigen presentation ability, a question many immunologists were asking was whether dendritic cells could be used as an adjuvant. This can be accomplished by either transfecting the dendritic cell with the gene encoding the antigen to be immunization with (217-220), or by pulsing the dendritic cell with antigen ex vivo followed by infusion (221-223). In both situations the immunostimulating activity of dendritic cells as adjuvant is superior to other immunization procedures such as pulsing macrophages with antigen, or immunizing with classical adjuvants such as Freunds. In human clinical trials, peptide pulsed dendritic cells have been used in prostate cancer. A group of those patients developed cytotoxic T cell responses to the prostate cancer antigen immunized for, as well, a subset of this group demonstrated clinical response (224,225). In summary, the dendritic cell is an interesting sensory cell of the immune system that we will definitely hear more about in the future.

The Macrophage

Macrophages are a heterogenous population of phagocytic cells that provide a front line defense mechanism against invading pathogens or transformed cells. They belong to the sensory part of the immune system. However, the macrophages can also play an effector role when specific immunity is activated and B-cells start producing antibodies. The macrophage’s effector functions are both antibody dependent and antibody independent. The macrophage functions as an antibody independent effector when it becomes activated and by T cell secreted IFN-(, this endows them with cytotoxic activity toward tumour and virally-infected targets. The antibody dependent macrophage cytotoxicity occurs when the macrophage Fc receptor binds to the Fc portion of antibody/antigen complex. This binding makes the macrophage phagocytose the antigen-antibody complex (opsonization), or if the antigen is on the surface of another cell, the macrophage can kill it (antibody dependent cellular cytotoxicity).

Macrophages are very heterogenous cells. The two basic types are the exudate and the resident. Also, resident macrophages can be subdivided into a number of subtypes depending on their anatomical location and function (226). Generally, exudate macrophages are one of the first cells to enter a site of inflammation. Exudate macrophages are derived from blood born monocytes and cannot proliferate. In contrast, resident macrophages proliferate in response to a variety of stimuli (227). Macrophage recognition of target cells and pathogens occurs through mechanisms that for the most part are unknown. Known mechanisms of macrophage activation include recognition of cells lacking MHC 1 molecules (196), recognition of bacterially conserved DNA motifs (228) and binding of LPS or LPS-LBP complexes to macrophage receptors (229). From these presumed mechanism, only the first applies to neoplasia since cancer cells generally express low or absent levels of MHC 1, this is partly to escape detection by T cells. T cells recognize target cells by binding foreign or mutated self peptides that are presented on MHC 1 molecules on the surface of the target cell. Therefore if the target cells downregulates expression of MHC 1 then they can escape T cell attack. Subsequently, the macrophage attacks these cells. T cell activation is generally a late event (effector stage) in the immune response, since it is known macrophages are one of the first cells the begin infiltration of tumors, we know the "macrophage recognition of cancer by downregulated MHC 1" cannot be the whole story.

Macrophages are called to the site of tissue injury by chemoattractants such as complement components and factors released from damaged cells. This initial leakage of material from dead cells is what Matzinger describes as "danger signals.” When these macrophages recognize the concentration gradient of a chemoattractant, they move through diapedesis toward the source of highest concentration. The macrophages that primarily enter the sites of inflammation are exudate macrophages that derive from the blood-borne monocyte. These macrophages phagocytose material on their way up the concentration gradient. Having reached the area of inflammation they begin to secrete factors that recruits in more cells--primarily T helper cells, but also other cells of the synthesis phase of the immune response (230). Most helper T cells, however, get activated in the draining lymph node for the area of inflammation. Besides T cells, a very important cell called into the site of tissue damage by macrophage-released mediators is the dendritic cell (231), which is key in activating T cells present in the lymph node. Components of the clotting system are also involved in the attraction of macrophages to the site of injury. This is because injury naturally activates the clotting cascades through the extrinsic pathway that involves tissue factor that activate prothrominases to cleave prothrombin into thrombin (232). Thrombin is a serine protease that acts as a chemotactic agent for macrophages (233,234), and induces production of molecules with chemotactic ability (235).

The role of the macrophage in the immune response to tumours is very interesting. Macrophages infiltrate tumours but in some cases they assist in tumor growth. Macrophage infiltration of tumours is widely reported in the clinical literature (see Ref 236-239 for reviews) although no consistent correlation between prognosis and infiltration is evident. For example, in breast cancer, macrophage infiltration was believed to be associated with better prognosis according to a review of the 1970s literature by Underwood (240). However newer studies argue such infiltration suggests a poor prognosis (241) or bears no prognostic weight (242). Confusion regarding the prognostic significance of macrophage infiltration is seen in other types of cancers that possess such infiltrates, these include soft tissue sarcoma, ovarian cancer, and hepatoma.

One would expect that the different tumour responses induced by macrophages would depend on the type of macrophage infiltrating, and the type of tumour being infiltrated. For example, tumour secreted products, such as IL-10, have the ability to deactivate macrophages and render them nontumourocidal. Tumours are heterogenous with regards to IL-10 secretion, thus if a macrophage bound on to a tumour not expressing IL-10, the macrophage may kill the tumour. It is important that the macrophage in contact with the tumour is activated since unactivated macrophages that contact tumour can stimulate tumour proliferation (243). Tumour heterogeneity can also effect the level of macrophage recruitment. Tumours attract macrophages in part because of the extensive cell damage and death that occurs in the microenvironment of the tumour. This tissue damage usually results in macrophage chemotaxis and the activation of clotting cascades. In fact, some tumour cells express procoagulant tissue factor on their surface (244,245), this activates clotting and is partly responsible for the fibrin coating encapsulating many tumours (246). Since the amount of tissue damage induced by the tumor and the level of tumour bound procoagulant activity is heterogenous with respect to the tumour type, stage of tumour development and tumour vascularization, it should not be a surprise that tumours vary in the amount of macrophages they attract.

From the tumour’s perspective, macrophages should be avoided since they have the potential to eradicate the tumor. Therefore, it would be expected that tumours secrete agents which prevent the infiltration of macrophages. It is therefore surprising that tumours secrete macrophage chemoattractants such has the macrophage chemoattractant protein-1 (MCP-1) (247-249). Even more surprising is the observation that transfecting tumour cells with large levels of MCP-1 can increase their immunogenicity, allowing them to act as tumour vaccines (250). Thus, it appears that tumours secrete just enough MCP-1 to attract macrophages, but not enough to endow them with tumouricidal properties. The reason tumours may desire macrophage recruitment, is to use them to serve the purposes of the tumour (i.e., to provide growth factor and angiogenic support). An in vivo example of how macrophages are used by tumour cells to facilitate this augmentation of tumour growth and metastasis is the tumour challenged osteopetrotic mouse (op/op mouse). These mice have a genetic abnormality that prevents them from secreting the cytokine monocyte colony stimulating factor (M-CSF). Since M-CSF is a differentiation factor for the formation of macrophages, the op/op mice almost has an absolute absence of macrophages. When op/op mice are challenged with syngeneic Lewis Lung Cancer (LLC), the tumours grow at a reduced rate compared to wild type mice challenged with LLC. To demonstrate that the impaired growth is due to the lack of macrophages and not an intrinsic anticancer effect associated with the op/op phenotype, the authors administered M-CSF to op/op mice and then challenged the mice with LLC. The macrophage containing op/op mice now developed tumors that grew at the same rate as the wild-type mice. Most interestingly, the growth of LLC in op/op mice was characterized by a low mitotic index (probably meaning that the macrophages secret growth factors that increase cancer proliferation) and by decreased angiogenesis (implying that the macrophage contributes to tumour angiogenesis) (251).

Macrophage augmentation of tumour growth has been shown to occur by the following mechanisms: inhibiting activation of Th1 cells and subsequent T cytotoxic responses, secretion of tumour stimulatory growth factors, and induction of angiogenesis.

The first mechanism involves having the macrophages take on a "suppressor phenotype", this can be induced in vitro by culturing them with tumour derived immunosuppressants such as IL-10. Macrophages cultured under such conditions become insensitive to a variety of activatory signals (252), secrete suppressed levels of inflammatory cytokines such has TNF, IL-1, IL-6 (253) and downregulate costimulatory signals necessary for proper T cell activation such has B-7 (254). Suppression of these functions is detrimental for the host because direct killing of transformed cells will not occur, and more importantly, it activates the synthesis phase of the immune response without the necessary activation signals. If the synthesis phase is activated through a suppressed signal then the response will be molded against what the immune system has perceived to be a weaker danger and therefore the response from the effector phase will be smaller and probably ineffective.


IL-10 induced suppression of B-7 is used by a variety of pathogens to improperly activate the synthesis phase of the immune response. Since the interaction of the macrophage (an antigen presenting cell) with a T cell represents the sensory-synthesis communication, it is essential that the macrophage provide the T cell with both an immunogenic antigen, and the necessary costimulatory signals to instruct the T cells in terms of the profile of the immune response to launch. Pathogen downregulation of costimulatory molecules such as B-7 on antigen presenting cells occurs so that the T cell will not be activated to induce the response needed for eradication of the antigen. One means of downregulating B-7 is to increase the concentration of IL-10 in the microenvironment surrounding antigen presentation. An example of this phenomenon is seen in murine schistosomiasis induced granulomas, where the parasite egg somehow triggers the macrophages in the granuloma to secrete IL-10. This IL-10, through an autocrine fashion, inhibits expression of the costimulatory molecule B-7. Since B-7 is not expressed, the macrophages in the granuloma will not be able to activate the T cell response properly and will result in an ineffective response or anergy. If antibodies to IL-10 are given to the granuloma cells surrounding the egg, B-7 is expressed and the T cell becomes properly activated (255).

Macrophages can also be immunosuppressory by secreting soluble mediators that inhibit T cell activation. Since macrophages are one of the first immune cells to enter the tumour microenvironment, secretion of immunosuppressory factors by these cells is likely to inhibit the function of more specific effector cells that enter the tumour at later time points such as the T cell. Should infiltrated macrophages interfere with proper T cell activation and effector function, the tumour may have "devised" a very clever way of using cells of the immune system to protect itself from immune-mediated destruction. Prostaglandin E-2 (PGE-2) is an eicosanoid product of arachinonic acid metabolism via the cyclooxygenase pathway. PGE-2 is associated with inflammatory reactions but also possesses immunomodulatory activity. PGE-2 is secreted by a variety of tumours such as liver (256), lung (257), breast (258), and corectal cancer (259). In addition, tumour associated macrophages also secrete PGE-2 (260,261). Heterogeneity in PGE-2 expression has been observed in macrophages, this may be attributed to the different needs of inflammatory mediators at different anatomical sites. For example, microglia cells are resident macrophages of the brain, while these cells perform an excellent job of antigen phagocytosis, they secrete little PGE-2 upon stimulation. In contrast, the exudate macrophages that invade inflamed tissue have the capability of secreting large levels of PGE-2, this allows for inflammation to ensue. Macrophages invading tumour tissue are generally of the exudate subtype, since it is these cells that respond to the chemotactic signal produced by the tumour. PGE-2 inhibits T cell immunity primarily by suppressing the ability of T cells to proliferate, this was demonstrated in head and neck cancer (262), melanoma (263) and colon cancer (264). In addition, PGE-2 has been shown to inhibit the generation of T cytotoxic cells, important effector cells in antitumour immunity (265). In vivo, PGE-2 is involved in tumour progression since inhibiting PGE-2 pharmacologically by administering indomethacin significantly impedes metastatic progession in several animal s (266-268).

The second method in which the macrophage assists tumour growth is through secreting mitogenic substances. It is important to note that substances present at biological doses that are not mitogenic to nontransformed cells may be mitogenic to neoplastic cells due to typical features of transformed cells such as increased receptor number, decreased receptor turnover, and increased signal transduction ability. Physiologically, the macrophage is involved at the initiation, effector, and resolution of inflammatory responses. Macrophages recognizing LPS or other bacterially derived substances become activated, calling in other inflammatory cells and producing inflammatory mediators. Macrophage effector functions during inflammation typically consist of engulfing cellular debris either through phagocytosis or by virtue of receptor mediated endocytosis. At the end of inflammation the macrophages also are active assisting in tissue repair (269). In order for macrophages to facilitate wound healing certain agents must be secreted that accelerate the proliferation of the cells that were lost, or induce formation of new blood vessels so proliferation of new cells may begin. In cancer both these processes are induced at inappropriate times and as a result the tumour obtains a survival advantage. Growth factors secreted by activated macrophages during normal wound healing include epidermal growth factor (EGF) (270,271), platelet derived growth factor (PDGF) (272), transforming growth factor (TGF) (273), and fibroblast growth factor (FGF)(274). To determine if macrophage released growth factors are associated with increased tumour growth, researchers are assaying macrophages at tumour sites for growth factor secretion. Early studies in this field were initiated before cloning of growth factors and had to rely on macrophages extracted from the tumour. In 1981 Currie demonstrated that coculturing of syngeneic macrophages with murine fibrosarcoma cells enhanced proliferation in vitro through the secretion of a soluble factor (275). A decade later, clinic studies demonstrated that tumour associated macrophages in nasal polyps (276) and lung cancer (277) secrete PDGF at doses that are mitogenic in vitro. Future approaches may be aimed at administering cytokines to diminish macrophage secretion of PDGF. One such cytokine is IFN-(, which has been shown to suppress secretion of PDGF by activated macrophage (278). Tumour associated macrophages have also been shown to secrete EGF at mitogenic concentrations in human breast cancer (279), although the factors inducing EGF were not investigated. FGF secretion by tumour associated macrophages is important in macrophage induced tumour angiogenesis (280) and tumour proliferation (281,282). Similarly, TGF-B can be secreted by macrophages at concentrations that are mitogenic (283).

The third mechanism in which macrophages assist tumour progression is through the induction of angiogenesis. In order for tumours to grow larger then 3 mm cube, new blood vessels must be formed so that it can recieve oxygen and nutrients. The process of new blood vessel formation, called angiogenesis, normally occurs in development and wound healing (284). Angiogenesis is a multistep process that can be initiated through several means. Tumour hypoxia generally induces expression of VEGF (285-288), insulin-like growth factor (289), and IL-8 (290), all of which promote the angiogenic process. It is also interesting that all of the aforementioned agents have immunomodulatory properties. VEGF, for example, is able to alter the antigen presenting ability of dendritic cells by restricting their normal differentiation process (213, 291). Insulin-like growth factor increases monocyte production of IL-1 (292), IL-1 being a potent stimulator of the immunosuppressive mediator PGE-2 (293,294). IL-8 is a monocyte and neutrophil chemoattractant (295). Although tumour cells generally initiate their own angiogenesis, this process is greatly assisted by macrophages that secrete contributing factors. One example of this is the ability of the macrophage to degrade matrix. In order for angiogenesis to proceed, tumour matrix must be degraded so endothelial cells can arrive and colonize inside the tumour. Macrophages can degrade tumour matrix by activation of plasmin from plasminogen. Plasminogen is the inactive form of the proteolytic enzyme plasmin, which can cleave matrix proteins such as collagen, fibronectin and laminin (296). Macrophages activate plasmin from plasminogen by forming a catalyst when the macrophage bound urokinase-type plasminogen activator receptor binds serum floating or membrane bound urokinase-type plasminogen activator (297). One way tumour cells can induce macrophages to initiate the angiogenic process is by increasing expression of the of urokinase-type plasminogen activator on macrophages. Tumours can modulate macrophage levels of this pro-angiogenic protein by secreting TGF-B. Both in vitro and in vivo TGF-B was shown to possess this tumour promoting property (298). Many other pathways that the macrophage could use to induce angiogenesis are likely to exist. Evidence for this was previously discussed, with the example of suppressed angiogenesis in mice lacking macrophages (251), and increased survival with decreased angiogenesis in breast cancer patients lacking tumour associated macrophages (241).

It appears that macrophages can act as a "two edged sword" in terms of anticancer immunity. On one hand macrophages have been shown repeatedly to possess anticancer properties, while on the other hand they also assist tumour growth and metastasis (since metastasis can't occur without angiogenesis). The critical questions in the authors’ opinion are 1. Can markers be found that identify tumour-promoting macrophages from tumour suppressing ones? 2. Are tumour promoting macrophages the same cells as tumour suppressing ones except for the fact that they have been "reprogrammed" by the neoplasia? This would also leave the possibility that tumour-suppressing macrophages are the same cells as the tumour promoting ones except one cell type is at a different stage of differentiation than the other. 3. What tumour secreted factors can induce the changes in macrophages? 4. Can the tumour promoting phenotype of a macrophage be changed to a tumour suppressing phenotype by cytokine manipulation?

The Neutrophil

Granulocytes are a family of inflammatory cells called polymorphonuclear leukocytes (PMN) which includes neutrophils, eosinophils and basophils. This work will only examine neutrophils due to space limitations, although eosinophils have been shown by some groups to be important in some types of antitumour responses. The reader is referred to the following articles for the role of eosinophils in host antitumour defense (299-304). Neutrophils resemble macrophages in that both are able to phagocytose and kill antigens or antigen coated cells. Neutrophils like macrophages are also considered part of the sensory phase of immune response and part of the effector phase. Since neutrophils are called to the site of inflammation immediately after tissue damage, the immunomodulatory environment formed by the invading neutrophils is very important to how the synthesis phase of the immune response will assess the antigenic threat. Once the synthesis phase of the immune response instructs the initiation of the effector phase, antibody secretion will allow the neutrophils to perform opsonization (305) and antibody dependent cellular cytotoxicity (306).

Neutrophils possess a multilobed nucleus and granulated cytoplasm in which various bioactive compounds are stored. Neutrophils are short-lived cells that circulate for 7-10 hr and then enter the tissue where they live for another 3 days before undergoing apoptosis (307). The life of neutrophils can be extended by administration of cytokines such as GM-CSF (308), TNF-" (309) or interaction with activated platelets (310). Neutrophils are nonrecycling phagocytic cells in that they release an oxidative burst post-phagocytosing the antigen. This destroys the target but also the neutrophil. Neutrophil cytotoxicity occurs through generation of several free radicals (ie. O-, H2O2, OH) and halide species (ie. hypochlorous acid) both through myeloperoxidase dependent and independent mechanisms. The free radicals and halides are directly cytotoxic to tumour and nontumour targets partly through induction of membrane damage (311).

In order for neutrophils to bear any significance on tumour growth, a mechanism of neutrophil homing must exist that guides them to the site of neoplasia. Since tumours are often associated with hypoxia, there is often cell death in various parts of the tumour, this is what elicits inflammatory-like reactions such as infiltration of immune cells. Generally, neutrophils nonspecifically enter tissue in response to inflammatory stimuli. Tissue damage releases histamine and other vasoactive compounds which "activate" the endothelium. Neutrophils possess L-selectin, a molecule that allows them to roll on activated endothelium through interactions with sialyl Lewis X. Rolling neutrophils then are in closer proximity to accept signals from the endothelium such as the chemotactic signal of IL-8. IL-8 signals neutrophils to express beta 2 integrins that bind on with strong affinity to ICAM-1 on activated endothelium. The beta 2 integrin-ICAM-1 binding allows the neutrophil to stop its flow in the blood vessel and to begin transendothelial migration (312). It is not completely understood what mechanisms the neutrophil uses to move through the extracellular matrix (ECM) although the concentration gradient of the chemotactic signal is the guiding mechanism (313). Neutrophils secrete cytokines that have been shown to influence the development of immune response. An early study by Lichtenstein et al in 1985 (314) demonstrated the ability of neutrophils to alter the course of antitumour response. By interperitoneally injecting C. parvum bacteria into ovarian cancer bearing mice, a macrophage mediated tumour reduction was achieved. Surprisingly, neutrophilic infiltrates were essential for the accumulation of tumourolytic macrophages. In addition, adoptive transfer of neutrophils from C parvum injected mice into tumour bearing mice was able elicit a monocyte mediated antitumour response. Today it is known that neutrophils are able to activate macrophages through secretion of cytokines such as interferon-gamma (315), thus molecularly validating the results of Lichtenstein. Neutrophils are also able to influence the type of immune response that will arise by giving specific signals to the synthesis phase. IL-12 is a cytokine secreted by antigen presenting cells such as dendritic cells and macrophages which acts as a costimulatory signal, telling the T helper cell what type of immune response is needed. When the dendritic cell does not secrete IL-12 but presents antigen to the T cell, the T cell will orchestrate a different immune response than if IL-12 was present. Specifically, presentation in conjunction with IL-12 shifts naive T cells to secrete interferon-gamma and aquire characteristics of Th1 cells. Th1 cells generate an immune response that is geared toward eradication of intracellular pathogens such as listeria. Since listeria resides intracellularly, antibody responses are useless since antibodies can not enter the cell. Effective immune response against listeria requires activation of the components of the immune system that can "sense" intracellular abnormalities such as NK cells and T cytotoxic cells. The importance of IL-12 in inducing cellular immunity is demonstrated by experiments indicating the most effective method of inducing immunity to listeria is by coinjecting listeria vaccines with IL-12 (316). Neutrophils play an important role in modulating whether an immune response will be effective against intracellular or extracellular antigens by means of the immunomodulatory signals they secrete. In murine candidiasis an intracellular immune response is protective, and as in listeria, initiation of this response is dependent on IL-12. Neutrophils have been shown to be the greatest secretors of IL-12 in these infections and depletion of neutrophils before candida infection will not allow the protective response to occur (317). Additional studies by the same group demonstrated neutrophil dependent IL-12 protection could be substituted by exogenous injection of the cytokine in absence of neutrophil. Further, mice recovering from candida had elevated IL-12 and suppressed IL-10 neutrophil production compared to neutrophils from mice succumbing to disease which secreted low IL-12 and high IL-10 (318).

Neutrophils can produce several other cytokines upon stimulation and these include IL-1 (319), IL-1 receptor antagonist (320), IL-8 (321), TGF-B (322), TNF (323), MIP-1 (324), IL-3, GM-CSF (325), IFN-a (326) and VEGF. Unfortunately little research has been performed at deciphering the conditions under which neutrophils release different types of cytokines and how this is related to development of immune response. Nevertheless, neutrophil modulation of immune response will likely be an intense area of investigation for the years to come.


The Natural Killer Cell

Natural killer (NK) cells are large granular lymphocytes, which have the ability to kill tumour and virally infected target cells without prior sensitization. These cells are nonphagocytic and kill through a variety of mechanisms: secretion of cytotoxic compounds found in their membranes such as granzymes, induction of apoptosis in target cells through fas, and making holes in target cells by releasing perforin. NK cells belong to both the sensory phase and the effector phase of the immune response since they can kill targets upon direct recognition or they can kill via ADCC. The surface phenotype of NK cells is CD3-, CD56+, and CD16+ (328). NK cells are important in controlling tumour metastasis since mice lacking NK cytotoxicity (beige mice) have an increase in cancer metastasis compared to controls (329-334). There is data which suggests beige mice possess increased incidences of spontaneous tumours (335), this would indicate a role for the NK cell as the "policemen" of the immune system, guarding the body against development of neoplasia. In fact, NK cells can specifically lyse oncogene transfected fibroblast cell lines while sparing the untransfected control (336). If NK cells are indeed the frontline defence against neoplasia, patients lacking their activity should have increases in cancer incidence. A Japanese study by Kobayashi demonstrated a 200 fold increased risk of developing malignant neoplasia in humans with Chediak-Higashi, a disease comparable to the NK deficiency in beige mice (337). Cancer patients generally possess suppressed NK activity (338-363) although there is argument as to whether the NK suppression contributes to the causation of cancer or whether it is an effect, since cancer cells secrete soluble factors that are suppressory for NK cells in vitro (364-367).


In some types of cancer, NK activity is positively correlated with improved prognosis. This would agree with in vitro studies that show NK cells have some importance in host response to neoplasia. In order to increase antitumour responses, several investigators have tried to increase NK activity in cancer patients by administration of immunomodulatory compounds such as IL-2. As discussed in the previous chapter, systemic IL-2 administration results in dose limiting toxicities that prevent its routine usage at concentrations needed to properly stimulate NK functions. In the trials were IL-2 was able to increase NK activity, the effect of the increase on the tumour mass was minimal. Since anticancer immune response requires a cooperation between various facets of immune response, increasing NK activity alone may not address all the immune abnormalities in the cancer patient. It should therefore not be surprising that IL-2 therapy has not lived up to the expectations placed upon it (84-86). Investigators therefore have searched and presently continue to search for augmenters of NK function that lack toxicity and possess antineoplastic activity. Besides IL-2, a plethora of biological agents increase NK activity, these include 1.) Hormones, such as prolactin (379), growth hormone (380), melatonin (381) and insulin-like growth factor (382). 2) Cytokines, including IL-7 (383), IL-12 (384), IL-15 (385), and IL-18 (386). 3) Neuropeptides such as Met5 enkalphin (387), beta-endorphin (388) and substance P (389). Understanding interactions between these agents may one day lead to discovery of synergies which can be therapeutically efficacious ways of increasing the NK activity of cancer patients. An interesting example of developing these synergies is the work of Lissoni et al who is combining pineal hormones such as melatonin with IL-2 therapy in order to decrease toxicity but to increase NK activity (90,390-397). These methods are presently in clinical trials and the phase 2 data seems promising (396,397).

Although evidence exists for the importance of NK cells in antitumour immunity, it is still unclear the mechanism through which NK cells mediate this function. NK cells possess antitumour cytotoxic ability in in vitro assays, however the ratio of NK to target cells needed to obtain toxicity is usually 10-100 (398), which is unlikely to occur in vivo considering that NK cells are infiltrating a tumour mass, not the tumour cell infiltrating a NK mass. Realistically the NK to target ratio should be at maximum one to one if the NK cell is touching only the cell at the outermost part of the tumour. Taking this into account we must first speculate that NK cells are more efficacious at combating tumour cells in circulation (because of the higher effector-target ratio) and second, there must exist other antitumour mechanisms besides direct cytotoxicity by which they can induce regression of cancer. Such mechanisms include a recently described proteolytic activity were NK cells can disrupt the three dimensional shape of the tumour without actually killing the tumour cells (398a). Such a disruption would allow other immune cells to enter the tumour, or on the negative side, could promote tumour metastasis. A more accepted role of NK cells is their ability to induce the activation of other immune cells. The actions of NK cells at the synthesis phase of immune response are seen in their ability to influence a variety of immunocytes, these will be listed below:

The Impact of NK Cells on Macrophages:

As stated previously, macrophages possess tumour suppressing and tumour enhancing ability, in part through the secretion of soluble mediators and also through direct induction of tumour cell lysis. While unactivated macrophages can increase tumour proliferation (243), IFN-( activated macrophages are potent tumouricidal effectors (399). Interestingly, NK cells upon activation start secreting IFN-( (400,401) at concentrations that are able to activate macrophages in vitro. Such an interaction between macrophages and NK cells is seen in the murine tumour of Abe et al in which NK cells are needed for macrophage cytotoxic activity towards the tumour. In agreement with the recognized role of IFN-( in macrophage activation, antibodies to IFN-( decreased the ability NK cells had to activate macrophages (402). By virtue of IFN-( secretion NK cells may inhibit the ability of tumour promoting macrophages to secrete mitogenic factors such as PDGF since IFN-( treatment of such macrophages inhibits secretion of the mitogen (278). Interestingly PDGF is a suppressor of NK activity (403). In another system, macrophages are needed for NK cells to secrete IFN-( and therefore activate the macrophages (404,405). A possible explanation for the importance of macrophages in NK activation is that NK cells respond to the macrophage bound T-cell costimulating molecules B-7.1 (CD80) and B7.2 (CD86) (406-409). Interestingly, in other interactions, tumour associated macrophages produce PGE-2 that suppressed NK activation and subsequent tumour lysis (410). It appears that the NK interaction with tumour associated macrophages is complex and depends on variables such as the type of tumour, the type of macrophages infiltrating the tumour, and the local cytokine milieux. Studies using the macrophage deficient op/op mouse will assist in elucidating these interactions.

The Impact of NK cells on T cells

T helper cells are important in immune responses against pathogens since they provide a coordination of the type of immune response that will be mounted. One of the mechanisms by which T helper cells determine the immune response necessary is through synthesizing the information given to them by the sensory arm of the immune response, part of which is the NK cell. As previously stated NK cells are secretors of IFN-( upon activation, this cytokine influences the T cell to become a Th1 cell which will promote immune responses against intracellular pathogens and cancer (411). In addition, NK cell secreted IFN-( will increase expression of MHC class 2 (412) as well as levels of the B-7 costimulatory molecule (413) on macrophages. Both of these will influence the way the T helper cell will shape the immune response. Specifically, B-7 costimulatory molecules on antigen presenting cells increase the probability of a Th1 response occurring (414,415) whereas increased MHC 2 expression seems to drive the response towards Th2 (416). More work needs to be performed in this field to determine what molecular events decide the course the response will take.

Another area where NK cells effect T cells is in the mixed lymphocyte reaction (MLR). This is an in vitro assay that was used extensively in the 1970s to determine T cell proliferation. MLR is performed by culturing two populations of allogeneic T cells in vitro in the presence of accessory cells such as macrophages and allowing them to proliferate. The T cells proliferate since they are recognizing the different MHC antigens on each other. Interestingly, adding purified populations of NK cells to MLR will increase the amount of T cell proliferation (417,418). Thus a role for NK cells in T cell proliferation seems to exist. In another study, Kos and Engleman describe the ability of NK cells to substitute for T helper cells in the induction of cytotoxic T cells generated through the MLR (419). This is very relevant to tumour immunology because, as will be discussed in the next section, T cells that can kill the tumour are of the cytotoxic type. Cytotoxic T cells recognize intracellular antigen only when it is presented on MHC class 1. Tumour cells can express MHC class 1 but don’t express MHC class 2. In order for the T cytotoxic cells to kill tumours, they need a signal from the tumour expressing the antigen on MHC class 1, and a second signal from the T helper cell. T helper cells can only be activated by the recognition of antigen on MHC class 2. Since tumour expression of MHC class 2 is absent or very small, it was always a mystery how the tumour specific cytotoxic cells can get activated. The generally accepted explanation (which contains many weaknesses) is that dead tumour cells are phagocytosed by antigen presenting cells which express MHC class 2 and these in turn activate the T helper cells which can send the second signal needed for the T cytotoxic cell to get activated. A very interesting hypothesis that requires testing is whether the NK cells that are recognizing the tumour can substitute for T helper cells at the generation of T cytotoxic cells.

The Impact of NK cells on B cells

Upon activation, B cells can mature into antibody producing cells called plasma cells. In terms of immunity to tumours, the type of antibody secreted by plasma cells can be crucial whether the response will be protective or nonprotective. As stated previously, some types of antibody responses to tumours actually protect the tumour from the effective Th1 response by immunological enhancement (127-131). Other types of antibodies mediate ADCC and can induce tumour lysis by activation of complement. The type of antibody response that will ensue is dependent on the cytokine microenvironment, as well as costimulatory signals provided by T cells and NK cells. Activated NK cells have been shown to skew antibody responses towards the antitumour Th1 type that is characterized by secretion of the immunoglobulin isotype IgG2a (420).


The Impact of NK Cells on Natural Suppressor (NS) Cells

Besides the ability of NK cells to possess antitumour activity, there exists a subtype of NK cells that suppress immune function and under some circumstances promote tumour growth. Present day immunological literature does not usually make mention of this cell type since the interests of immunologist have shifted in areas deemed more worthy of investigation. For more information, the reader is referred to the following papers (421-425).


The Significance of NK Cells in Autoimmunity

The importance of NK cells in maintaining health of the host may not only be limited to prevention of neoplastic disease but also at protecting the host from autoimmunity. In patients with arthritis (425a), systemic lupus erythematosus (425b)

T cells

T cells are called as such because they originate in the thymus, although those of the gamma delta variety can develop in other sites. T cells can be classified into four categories: T helper, T cytotoxic, gamma delta T cells and T suppressor cells. The helper T cell is the synthesis part of the immune system, it draws information about the pathogen from the sensory cells of the immune response and subsequently decides what response needs to be mounted. The T cytotoxic cell is part of the effector arm of the immune system since it kills target cells that are expressing antigens on MHC class 1. Gamma delta T cells are a strange population of T cells, which contribute to tumour immunity in that they can, under some circumstances spontaneously lyse tumours in absence of prior stimulation. Although the field of gamma delta T cells is rapidly expanding, a review of it is outside the scope of this paper and the reader is referred to the following reviews (426-430). T suppressor cells are a controversial type of T cell which is responsible for antigen specifically downregulating immune response, or secreting agents which do so. The importance of this cell population in cancer is that some groups have generated data showing the activation of these cells in neoplasia. Due to the conflicting reports and the modern belief that T suppressor cells are helper T cells with altered cytokine secretion, T suppressor cells will not be discussed in this section. The reader is referred to the following reviews (480-483).



T helper cells (Th)

Although there have been several allusions to the role of Th cells in mounting immune responses, this section will examine this role in detail. “Immune response” according to the classical definition can not occur in the absence of Th cells since these are needed specific clonal expansion of T and B effector cells and for induction of immunological memory (431). Other responses such as NK cells spontaneously killing tumour targets are not classified as immune responses since there is no memory thought to be involved. The tumouricidal ability of activated macrophages, neutrophils, and NK cells is said to be part of the “innate response”. The memory or “aquired” responses are manifested only by the Th cells, T cytotoxic cells and the B cells. All of these cells possess the capability of clonally expanding in response to antigen. By “clonally expanding” we mean that each of these cells possesses a different receptor for each antigen that they may encounter, upon contact with the antigen and the appropriate secondary signals, only the cell which possesses the receptor that binds to the antigen will proliferate. For example, T cells (both cytotoxic and helper) possess a receptor called the T cell receptor (TCR), which binds to antigen. Amazingly, the TCR on each T cell, through a process of gene shuffling and recombination, is made differently so that each individual T cell recognizes a specific antigen. In total, the human immune system possesses 108 different T cells that recognize 108 different antigens (432). When an antigen activates one of these T cells it proliferates, thus increasing the percentage of T cells in the body which can recognize the antigen. Increasing the number of antigen specific T cells in response to an antigenic stimulation is called clonal expansion.

The process of Th activation by antigen is a complex one, we will first describe it briefly and then mention details effecting the activation and events effected by the activation. Generally, the Th recognizes antigen only when an antigen-presenting cell with MHC 2 molecules derived from the host presents it. The antigen presenting cells are dendritic cells, macrophages, and B cells since these cells have the ability to ingest antigen, process it intracellularly, and present it on MHC 2. MHC 2 is a molecule that intracellularly associates with peptide sequences derived from proteins ingested by the antigen presenting cells. The TCR on the surface of the Th cell binds to the MHC 2 and to the antigen, this results in generation of the primary signal needed for T cell activation. The second signal needed is that of the costimulatory molecules such as CD80 and CD86, which bind on to receptors on the T cells such as CD28. When the Th cell receives both the primary signal and the secondary signal, it begins clonally expanding and activating other cells. If the Th cell receives only the primary signal, then instead of becoming activated, it will enter a state of perpetual nonresponsiveness (even if activation is performed by both the primary and secondary signal thereafter) called anergy (433,434). Anergy is a very interesting phenomena since it antigen specifically inactivates Th clones. This could be physiologically beneficial to the host in preventing the induction of autoimmunity. For example, during fetal life, most autoreactive T cell clones are deleted in the thymus by a process called thymic education (435), however some autoreactive Th cells remain in circulation even in life. Should such Th cells be activated, autoimmunity will result, therefore a build-in safeguard against activation of such autoreactive cells is that two signals need to be presented to the Th cell for activation. Autoantigens can be presented on MHC 2 to the autoreactive Th cell since APCs present most material that they phagocytose from their environment, including dead cells belonging to self. However, self reactive Th cells usually are not activated since APCs don’t provide the needed secondary signal to the T cell should they be presented with self antigens. Since Th cells activated by the primary but not secondary signal enter a state of anergy, these autoreactive Th cells will be silenced and not threaten the body with autoimmunity. The reasons why APC do not provide the secondary signal when presenting self antigen is not completely understood, but it is likely to do with intrinsic characteristics of APC to differentiate ingested proteins on the degree of foreignness. For example, macrophages that ingest LPS containing pathogens intrinsically associate LPS with “foreignness” and therefore increase expression of the costimulatory molecule CD80 that will allow for activation of Th cells specific for antigens on the LPS containing pathogen (436). It is intriguing that forcing expression of costimulatory molecules can predispose the host to autoimmunity, this is seen in the clinical correlation between certain infectious agents and initiation or relapse of autoimmune disease. Patients with rheumatoid arthritis relapse after infection with LPS containing, Gram-negative bacteria (437), this is likely because LPS nonspecifically increases expression of costimulatory molecules that will provide a second signal for the activation of the already clonally expanded autoreactive Th cells.

Once Th cells become activated, they can choose to promote various types of immune responses, depending on the sensory input during the activation process. According to the Th1/Th2 paradigm originally proposed by Mossman and Coffman (438), Th cells begin as Th0 and then subsequently differentiate into Th1 cells if the effective response to pathogen calls for cell mediated immunity or Th2 if the response calls for antibody production or protection against parasites. Typically, when the pathogen is intracellular, the effective response is of the Th1 type, whereas if the pathogen is extracellular, clearance of the antigen will require antibody production and thus a Th2 response. This Th1/Th2 paradigm is in agreement with the examples we have spoken of previously for leprosy, schistosoma, and listeria where a Th1 response will clear the pathogen whereas a Th2 response will exacerbate the disease. Typically, Th1 responses are associated with the production of IL-2, IL-12, IL-15, IL-18 and interferon-( (439- 443), whereas Th2 responses produce TGF-B, IL-4, IL-10, IL-13 and IL-14 (444-448). The level of interactions between these cytokines is very complex and illustrates several feedback mechanisms that are only beginning to be elucidated. Thus while we call IL-4 a product of Th2 cells, the IL-4 knockout mice cannot generate Th-1 responses, this demonstrating a need for a Th2 product in initiation of Th1 response (449). Another example of cooperation between Th1 and Th2 cytokines is exemplified in a study that shows that priming macrophages with IL-13 (Th2 cytokine) increases their ability to secrete IL-12, a Th1 cytokine (450).

Besides Th0 cells becoming Th1 or Th2, some investigators have coined the term Th3 for Th cells secreting TGF-B but not IL-10 (451). Since Th cells can have distinct patterns of cytokine secretions based on disease, host, and other factors it will be interesting to see if other subsets of Th cells will be identified such as Th4, Th5, etc. The importance of different types of differentiated Th cells for specific conditions lies in the fact that these each disease condition will require a specific immune response in combating the pathogen. Thus while the work of Mossman and Coffman was monumental in elucidating two broad cytokine profiles associated with two types of immune responses, the real situation is much more disease specific. This manuscript will however concentrate on the broad generalization of Th1 and Th2 since data is not available yet to comment on the other types of responses.

It will be interesting when molecular markers are found on Th cells that identify their cytokine profile. So far the only such marker is CD30 that is preferentially expressed on Th2 cells (452). Patients with various types of neoplasia have been noted to increase their numbers of CD30 positive Th cells, congruent with the shift toward the ineffective Th2 response. As we increase knowledge of different types of immune responses, markers will be found for the Th mediating the effective response in each case (such as Th3, Th4, etc.). Identification of such markers will allow for ex vivo growing of the effective immune cells needed to combat various diseases and provide a powerful diagnostic tool.

The key to understanding of immune responses appears to be in the forces that direct the Th0 cell to differentiate into Th1 or Th2. The first part of this differentiation process occurs through activation of Th0 by the primary and secondary signals received from the APC. These signals are accompanied by much other input from the microenvironment that can additionally skew the response to Th1 or Th2. Additionally, “Th1" or “Th2" can have varying degrees of intensity in that some Th1 responses will only mediate a weak activation of the cellular effector mechanisms whereas other Th1 responses will activate cell mediated immunity very strongly. This different level of Th cell activation can be paralleled to the different levels of anergy that Th cells can enter upon restimulation with only a primary signal from the APC. In other words, after stimulating a Th cell with a primary signal several times, pulling that Th cell out of anergy would be more difficult, than if activating the Th cell only once by the primary signal induced the anergy (453). The Th0 cell can be manipulated by several factors to become Th1/Th2 that includes: 1. Level of MHC 2 on APC (416). 2. Type of MHC 2 alleles (454). 3. Molecular structure of the antigen presented on MHC 2 (455). 4. The amount of antigen injected (455a). 5. To what extend costimulatory molecules are expressed (414,415). 6. Type of costimulatory molecule expressed (456). 7. Oxidative stress in the microenvironment (457). 8. Presence of antigen-antibody complexes (458). 9. Presence of cytokines in the microenvironment (459). 10. Location of antigen injection (459a). 11. Amount of inflammatory mediators such as histamine (459b). 12. Presence of hormones in the microenvironment (459c) 13. Adrenergic stimulation (459d). All these factors are present in cancer and will be used in the next chapter to explain the mechanism of Th1 suppression during neoplasia.

T Cytotoxic (Tc) Cell

Tc cells belong more to the effector and sensory arm of the immune response than the synthesis arm. The sensory ability of Tc cells is epitomized by screening cells of the body for mutations and viral infections. When a cell is altered, the Tc cell with the activation signal from the Th cell will lyse the abnormal cell. Therefore many cancer researchers have concentrated on identifying and ex vivo culturing the Tc cells that can bind and kill cancer cells. The mechanism through which Tc cells identify abnormal cells is by binding their Tcr to the target cell’s MHC and antigen complex. All cells in the body except the brain, testis and eye lens contain MHC 1. MHC 1 binds onto degraded intracellular proteins and presents them to the Tc cell. The Tc cell will become activated if it recognizes mutated or foreign proteins in the cells belonging to the body. Thus, the analogy is often made between the police patroller and the Tc cell.

Tc cells can recognize many types of tumour cells, but they have been most widely studied in melanoma (460-465). In fact, melanoma specific antigens have been discovered and molecularly cloned. These antigens plus MHC 1, are the targets of melanoma infiltrating Tc cells (466-468). Evidence of the immunogenicity of melanoma from mouse s and clinical trials (469-471), and the increased survival of melanoma patients that have Tc cell infiltrating melanomas (472), has provided the rational for conducting clinical trials using autologous TILs expanded ex vivo to treat melanoma patients. These trials had minimal success compared to standard chemotherapeutic regimes. However, some important observations were made. Namely, the TILs possessed the ability to traffic to the melanoma from which they were derived, and in the patients where the TILs did not traffic, no response was present (473). Important work needs to be done at increasing efficacy of Tc cell responses against tumours.

Tc cells mediate destruction of transformed or infected cells through induction of both apoptosis and necrosis. Tc cells induce necrosis by secreting perforin, a polymeric protein that embeds itself in the target cell’s membrane and forming a hole. Necrosis results because the osmolytic difference between the inside and the outside of the cell causes the cell wall to burst (474). As an aside, perforin knockout mice have a diminished ability to reject tumours (475), this indicating the in vivo importance of Tc cell killing through necrosis. Tc cell induction of apoptosis occurs through expression of apoptotic inducing proteins such as the TNF family member TRAIL/Apo2 ligand (476), secretion of specific types of enzymes that activate intracellular apoptotic mechanisms such as the granzymes (477), and by the expression of an apoptotic inducing membrane bound protein called fas-ligand (CD95L) (478). Some clones of Tc cells have been shown to utilize one of the above mechanisms of inducing target cell death preferentially. An area of little research that deserves attention is identifying the functional importance of the heterogeneity of Tc cell target killing. For example, elucidating conditions that stimulate expansion of specific Tc cells that kill primarily through CD95L would enable researchers to activate only the part of the immune response needed for killing of certain targets and thus spare the host of unnecessary toxicity.

Like Th and B cells, Tc cells can become memory cells after activation. This is important since it keeps the host prepared from further attacks by the pathogen. Conceptually, memory responses may be important in the control of neoplasia. One reason is that once a type of cancer has acquired the necessary mutations to become metastatic and invasive (lets say 8 mutations) then the cancer, for the sake of argument, also becomes antigenic. If an immune response can eradicate the fully developed cancer, there will still exist a population of cells that possess 7 mutations. These cells with 7 mutations are not dangerous to the host but upon gathering one mutation they will be. The memory Tc cells could function by destroying the cells that do get the 8th mutation, not allowing them to expand into a mass that would be dangerous to the host. This is just the author’s speculation and future experiments will definitely shed light on this issue.

Memory Tc cells upon activation secrete cytokines belonging to both the Th1 and Th2 family (479). These cytokines can activate other cells that do not have memory on their own but can become involved in the effector stage of the memory response such as NK cells, macrophages, and neutrophils. An example of the interplay between memory Tc cells and these nonmemory effectors can be illustrated in the following situation: An immunogenic tumour arises which possesses antigens recognizable by the memory Tc cell. Upon activation, the Tc cell kills the tumour cell but also starts releasing IFN-( (imagining the response takes a Th1 direction). IFN-( activates macrophages, increasing their capability to both phagocytose tumours, and present tumour antigens through the MHC 2 pathway. The increase in antigen presentation through MHC 2 will allow for activation of Th cells. These Th cells will not only be activated by the antigens that the Tc cell has memory for, but also by new antigens that the tumour may have accumulated due to its high rate of mutation. In turn, these newly activated Th cells will activate Tc cells that previously did not recognize the new tumour antigens. Therefore, the immune response will “learn” new tumour associated antigens through the interaction between the memory and effector parts of the immune response. Another effect that the memory Tc cell could have is downregulate the immune response. Since Tc cells exist that secrete immune inhibitory cytokines such as IL-10, the interaction between such a Tc cell and a tumour will decrease expression of costimulatory molecules on the macrophages surrounding the tumour, this will not allow for them to activate Th cells but instead will put them in a state of anergy and subsequently, blunt the response.

B cells

B cells are lymphoid cells that develop in the bone marrow of humans and mice. Upon proper activation by antigen these cells turn into anti-body secreting plasma cells. In contrast to T cells, B cells can recognize antigen in the absence of MHC, this allows B cells to respond to many blood born antigens. Antigen recognition is mediated through the B cell receptor (Bcr) which has a unique specificity for every B cell. The Bcr is made by the same process of random gene rearrangement that allows for the generation of diversity of the Tcr. In fact, the enzyme that initiates recombination of the Bcr and Tcr genes, RAG, is essential for development of these two cell lineages. Mice that lack the RAG gene develop a condition called severe combined immunodeficency (SCID), characterized by disfunctional B and T lymphocytes (486). Analogous to the two-signal required for T cells activation, B cell activation also requires two signals. The primary signal comes from an antigen binding onto the Bcr; the secondary signal comes from an activated Th cell through the form of a cytokine (487) or the ligation of CD40 (488) to the CD40 ligand. Once the B cell is activated it begins proliferation, transforming into either a memory cell, or an antibody secreting plasma cell. The antibodies generated recognize only the antigen that originally activated the B cell. However, as the plasma cells secrete antibodies, the isotype of the antibody, (i.e. the structural type of antibody secreted), can be changed by the conditions regulating the immune response. This allows the immune response to secrete antibodies that will best deal with the disease. For example, during a neonatal intestinal infection, the host will not want to initiate an inflammatory reaction against the pathogen since such inflammation will cause tissue damage and even death. Therefore, the immune system, by virtue of the Th cell-secreted cytokines will instruct the plasma cell to secrete antibodies that do not cause inflammation but will neutralize the pathogen by agglutination. Such noninflammatory antibodies are usually of the IgA family (489). If the protective response needed requires activation of inflammatory causing actions such as complement or ADCC, the plasma cells can release antibodies that can facilitate these. The ability of IgG2a to activate complement is better than IgG1a (490), consequently, immune responses initiated with the help of Th1 cells generally are associated with higher titers of IgG2a (491). The cytokines involved in determining the type of antibody the plasma cells will secrete are known for several s of disease and neoplasia. These cytokines are in line with the idea that Th1/Th2 has a crucial role in coordinating the synthesis phase of the immune response. Examples occur in which the antibody secreted against the pathogen, instead of eradicating it, protects it from the proper immune response. This usually applies to noncomplement fixing antibodies and to asymmetrical antibodies. Both of which have been found in pregnancy (492-495), cancer (496-498), and transplantation (499,500), and have been used to explain lack of destructive responses in each of the respective situations.

Although reports exist of B cells infiltrating tumours (484,485), the fact that B cells are noncytotoxic explains the lack of research in B cell manipulation for cancer therapy. This is not to say that B cells play no role in the cancer-immune interaction. Antitumour responses mediated by tumour specific monoclonal antibodies have been reported in both murine s (501-503) and in the clinical setting (504-506). The drawbacks of antibody mediated therapies include ineffective antibody penetration (507), inefficient activation of cytotoxic mechanisms (508), and antigenic modulation of the tumour (509). Recent attempts to improve on these limitations include the construction of chimeric antibodies that cross-link tumour cells with immune cells to induce tumour killing (510-512), usage of antibodies to which the nonantigen binding portion is molecularly attached to a toxin (immunotoxins)(513-515), and the use of single chain antibodies for greater penetration into the tumour (516). Clinical results are not available or still inconclusive.

An interesting type of B cell is the B-1 or CD5 positive cell. Although morphologically indistinguishable from conventional B cells, B-1 cells traffic to areas such as the peritoneum where the former cells do not (517-519). B-1 cells secrete antibodies called “natural antibodies” against self proteins and are believed to be part of the mechanism through which the immune system recognizes itself, otherwise called the immunological humonculus (520). Interestingly B-1 cells are the cells that secrete antibodies in several autoimmune diseases (521-523) and the cells that undergo neoplastic transformation in B-cell leukemia (524). The cytokine primarily secreted by B-1 cells is IL-10, and upon certain culture condition, B-1 cells can differentiate into macrophages (525). The biological role of these B-1 derived macrophages is unknown, however they are resistant to IL-10 treatment, a cytokine that is generally inhibitory to macrophage proliferation (525). A hypothesis that the author proposes is that B-1 derived macrophages are the tumour enhancing macrophages described earlier on in this chapter. Support for the hypothesis lies in that the levels of IL-10 secreted by several macrophage-infiltrated tumours theoretically should not allow for normal macrophages to proliferate or even invade the tumour, yet they do. Further, the conditions used to initiate B-1 differentiation into macrophages mimics the tumour microenvironment in certain ways ie expression of various myelopoietic growth factors and Th2 cytokines.

Chapter 1 - Introduction to The Immunotherapy of Cancer
Chapter 2 - Components of Immune Response
Chapter 3 - Immune Suppression in Cancer
Chapter 4 - Immune Interactions with Chronic Myeloid Leukemia