6
Animal Models for Use in Detecting Immunotoxic Potential And Determining Mechanisms of Action

This chapter deals with approaches to and methods for using experimental animals to identify xenobiotic substances that produce injury to the immune system and to determine the mechanisms by which injury occurs. The contributions and limitations of animal models are discussed in relationship to extrapolation to humans. Specific approaches and methods are outlined and discussed. As discussed in Chapter 3, immunogenicity of a xenobiotic substance in experimental animal models should provide a flag to indicate the potential of the xenobiotic to produce a hypersensitivity response in humans.

Although every effort should be made to assess the impact of environmental exposure on the human immune system in human populations, experimental animals are the best surrogates for detecting harmful xenobiotic substances and for determining their mechanism of action. With some exceptions, the immune and metabolic systems of humans and experimental animals are similar enough that animal models can provide the conduit to detect immunotoxic chemicals. Most agents that suppress the immune system in humans produce similar results in rodents, and the mechanisms for immunosuppressive action are similar in experimental animals and humans. Data obtained from animal immunotoxicity studies are an important component of the evaluation of health risks associated with environmental chemical exposure.

ANIMAL IMMUNOTOXICITY BIOASSAYS

Much attention has been given to the need to verify immunotoxicity bioassays that appraise immune function after exposure to xenobiotics. Such bioassays should meet several criteria:

  1. They should be reproducible within a laboratory and between laboratories.

  2. They should be specific to the part of the immune system to be assessed.

  3. They should be sensitive enough to measure normal and abnormal immune function.

  4. They should be able to measure alterations in immune function caused by exposure to known immunotoxicants.

Control compounds known to alter immune



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 83
Biologic Markers in Immunotoxicology 6 Animal Models for Use in Detecting Immunotoxic Potential And Determining Mechanisms of Action This chapter deals with approaches to and methods for using experimental animals to identify xenobiotic substances that produce injury to the immune system and to determine the mechanisms by which injury occurs. The contributions and limitations of animal models are discussed in relationship to extrapolation to humans. Specific approaches and methods are outlined and discussed. As discussed in Chapter 3, immunogenicity of a xenobiotic substance in experimental animal models should provide a flag to indicate the potential of the xenobiotic to produce a hypersensitivity response in humans. Although every effort should be made to assess the impact of environmental exposure on the human immune system in human populations, experimental animals are the best surrogates for detecting harmful xenobiotic substances and for determining their mechanism of action. With some exceptions, the immune and metabolic systems of humans and experimental animals are similar enough that animal models can provide the conduit to detect immunotoxic chemicals. Most agents that suppress the immune system in humans produce similar results in rodents, and the mechanisms for immunosuppressive action are similar in experimental animals and humans. Data obtained from animal immunotoxicity studies are an important component of the evaluation of health risks associated with environmental chemical exposure. ANIMAL IMMUNOTOXICITY BIOASSAYS Much attention has been given to the need to verify immunotoxicity bioassays that appraise immune function after exposure to xenobiotics. Such bioassays should meet several criteria: They should be reproducible within a laboratory and between laboratories. They should be specific to the part of the immune system to be assessed. They should be sensitive enough to measure normal and abnormal immune function. They should be able to measure alterations in immune function caused by exposure to known immunotoxicants. Control compounds known to alter immune

OCR for page 83
Biologic Markers in Immunotoxicology function, such as cyclophosphamide, should be included in experiment design until the investigators have confidence in the reproducibility and usefulness of the assay in their laboratory. The ability of the assay to predict the potential incidence of human disease must be considered if the assay is to assist in regulatory decisions. The ability of the assay to forecast the effects in humans also should be considered and should be discussed by the investigators when presenting their results. The development of assays to determine the immunosuppressive potential of chemicals has relied mainly on the use of the mouse, whereas the potential to induce hypersensitivity is most often studied in the guinea pig. Although the mouse is not used routinely in initial toxicologic evaluations, except in the determination of carcinogenic potential, the mouse immune system is well characterized, and most of the necessary reagents for immunotoxicity testing are specific for the mouse. Because the rat is the animal used most often in initial toxicologic evaluations of chemicals, most of the pharmacokinetic and toxicologic data available are for this species. An effort has been made to use the rat to optimize and verify methods that assess immune function and immunotoxic effects and to develop the reagents necessary to assess immunotoxicity in the rat. Several assays have been optimized and found to be sensitive and reproducible in the rat. One laboratory has optimized a system to examine several areas of immunotoxicity simultaneously in the rat; however, each assay also can be assessed individually (Koller and Exon, 1985; Exon et al., 1984, 1986). All the caveats given here for mouse assays apply to assessments of rat immune-system function. Recent studies have examined cytomegalovirus infection in the rat, but no host-resistance models are currently considered feasible for routine immunotoxicologic evaluation. Table 6-1 lists several series evaluations that can be used in succession for hazard evaluation. That is, if results of several of the assays in the first group are positive, the investigator should proceed with selected assays in the following groups. Mechanistic procedures are included in Table 6-1. Elucidation of the mechanism by which a chemical affects the immune system provides important information for hazard evaluation. Some of these procedures are performed in vivo, but the majority are conducted in vitro. Since the immune system is extremely complex and is regulated by other body systems, application of immunotoxicologic data for risk-assessment purposes must be based on data obtained from in vivo exposures. However, both in vivo and in vitro immune procedures that have been verified are appropriate to evaluate chemical-induced immune dysfunction. Several reviews discuss the various tests used to assess immune function in mice and rats after exposure to potentially immunotoxic compounds (Vos, 1977, 1980; Speirs and Speirs, 1979; Dean et al., 1982; Luster et al., 1982, 1988; Bick et al., 1985; Koller and Exon, 1985; Exon et al., 1986). Some assays are preferred because of their ease of incorporation into current toxicologic examinations. However, many of these procedures are sensitive to modulation by xenobiotic levels that do not produce adverse effects in other organ systems. Other immunotoxicity assays are preferred because they respond to modulation by toxicants. These assays are more difficult to incorporate into routine toxicologic evaluations. The advantages and disadvantages of each assay are discussed below. Table 6-2 summarizes the sensitivity and predictivity of these assays. Pathologic Evaluation Pathologic evaluation as an initial screen for immunotoxic chemicals can be useful because it is incorporated into the standard toxicologic assessment of new chemicals

OCR for page 83
Biologic Markers in Immunotoxicology TABLE 6-1 Approaches to Animal Immuntoxicity Testing Rapid Screen Pathologic evaluation of lymphoid organs T-cell-dependent antibody response Enzyme-linked immunosorbent assay Plaque-forming-cell assay Further identification of immunotoxicity Cell-mediated immunity  Lymphoproliferation Mixed-lymphocyte reaction, phytohemagglutinin, concanavalin A, lipopolysaccharide, anti T-cell receptor complex, anti-immunoglobulin + interleukin-4 Delayed hypersensitivity response Cytotoxic T-cell response Nonspecific Natural-killer-cell cytotoxicity Macrophrage bactericidal activity and interleukin-1 tumor necrosis factor activity Interleukin-2 activity Immune-cell surface markers Colony-forming units (spleen); or (granulocyte/monocyte) Host-resistance models Listeria monocytogenes Streptococcus pneumoniae or pyogenes Influenza virus B16F10 melanoma tumor PYB6 tumor Mechanistic studies Interleukin-2 receptor expression Ia receptor expression Transferrin receptor expression Mac 1 and Mac 2 receptor expression F4/80 receptor expression mRNA for cytokines Complement components Antibody-dependent cell-mediated cytotoxicity Respiratory burst (macrophages, polymorphonuclear leukocytes) Antigen presentation (Vos, 1977). If the pathology serves as a basis for assessing potential immunotoxic effects, then it is not necessary to use more animals than those required in the standard toxicologic bioassay. The areas studied by immunopathology include hematologic values, as well as the weight, histopathology, and cellularity of lymphoid organs. Hemograms

OCR for page 83
Biologic Markers in Immunotoxicology TABLE 6-2 Validated Rodent Immunoassays   Sensitivea Predictiveb Pathology (M,R)c   ? Humoral immunity Plaque-forming cell (M, R) × × Enzyme-linked immunosorbent assay or radioimmune assay (R) × × B-cell markers, immunoglobulin (M, R) ×   Cell-mediated immunity Delayed-type hypersensitivity (M, R) × × Mixed-lymphocyte reaction (M) × × Cytotoxic T cell (M) × × T-cell markers (M, R) ×   CD3, CD4, CD8, Thyl (M) ×   Nonspecific Natural-killer-cell cytotoxicity (M, R) × × Macrophage phagocytosis (M, R)   × Macrophage bactericidal activity (M, R) × × Macrophage antitumor activity × × Interleukin-2 (M, R) × × Interleukin-1 (M, R) × × Prostaglandin E-2 (M, R) ×?   Bone marrow CFU-S (M) ×   CFU-C ×   Host resistanced Streptococcus (M) × × Listeria (M, R) × × Influenza virus (M, R) × × Herpes simplex virus (M)   × Plasmodium (M) ×   PYB6 tumor (M) × × B16F10 melanoma (M) × × a Sensitive: procedures with inherently small variation or sensitive to a small degree of immune modulation. b Predictive: procedures corollary to immune function. c M, mouse; R, rat. d Sensitive for detection of pathogenesis but not immune function. should include a complete blood-cell count; enumeration of the number of white blood cells; and a differential count of the lymphocytes, polymorphonuclearneutrophils, basophils, eosinophils, and monocytes. The lymphoid organs that should be examined are the thymus, spleen, lymph nodes, and bone marrow. Bone marrow is sensitive to

OCR for page 83
Biologic Markers in Immunotoxicology compounds that block division of rapidly proliferating cells. Until now, very few data have been accumulated on the ability of the changes in these measures of immunopathology to predict disease states, and they generally are thought to be ineffective biologic markers of immunotoxicity. Functional alterations of markers related to immune dysfunction are not likely to correlate with changes in histopathologic markers. Humoral Immunity The humoral immune response results in the production of antibodies by differentiated B cells that recognize the immunizing antigen (Ehrich and Harris, 1945; Raidt et al., 1968), and initial studies can be used to determine the number of splenic or peripheral-blood B cells. However, as with histopathologic evaluation, an alteration in the number of splenic or peripheral B cells is not a sensitive indicator of chemical insult to the immune system. Therefore, a determination of immune function should be conducted. The generation of a primary humoral immune response requires the interaction of macrophages, regulatory T cells, and B cells (Mosier, 1967; Gorczynski et al., 1971; Miller et al., 1971; Jakway et al., 1975). Because of this complexity, the ability of immunocytes to generate a primary immune response is a sensitive indicator of immune dysfunction caused by immunotoxicants. The generation of humoral immune responses can be performed both in vivo and in vitro. After in vivo immunization by intravenous injection of the T-dependent antigen (sheep red blood cells), the response can be measured either by quantitation of the serum antibody titer to the antigen (measured by immunoassay or hemagglutination) or by counting the cells that produce antigen-specific antibodies (through determining the number of plaque-forming cells) (Jerne and Nordin, 1963; Cunningham, 1965). The humoral immune response also can be assessed by the ability of immunocytes to proliferate in response to a mitogen, such as lipopolysaccharide (Andersson et al., 1972). However, some compounds that cause changes in serum antibody titer or in the number of B cells that produce specific antibody do not alter mitogenic responses (Sikorski et al., 1989; Cao et al., 1990). A decrease in the resistance of mice to influenza virus is associated with suppression in the number of plaque-forming cells and in the mitogenic responses of B cells (Luster et al., 1988). The production of antibodies that opsonize (alter bacteria to enable more efficient phagocytization) and fix complement is involved in the elimination of streptococcal infection. In addition, a decrease in humoral immune responses results in increased parasitemia after infection with Plasmodium yoelli (Luster et al., 1988). Cellular Immunity A cellular immune response results in the generation of effector cells that phagocytize or lyse invading antigens. For example, in an immune response to an alloantigen (an antigen responsible for transplant rejection), the effector cell is a cytolytic T cell (Lindahl and Bach, 1976). Therefore, initial studies could involve a differential count of T cells by cytofluorometry. The same count should be performed in both the spleen and the thymus. However, as discussed above for B cells, simple counting of T cells is not as accurate an indicator of immune-system responses as is the measurement of effector function. Two assay systems have been used to measure the effect of environmental toxicants on the in vivo generation of cell-mediated immune responses. One is the generation of a delayed hypersensitivity response to antigens, such as keyhole limpet hemocyanin or sheep red blood cells. This response has been measured by the area of induration

OCR for page 83
Biologic Markers in Immunotoxicology formed, the ability of radiolabeled monocytes to migrate and become macrophages, or the amount of radiolabeled albumin that infiltrates the area, all as a result of antigen challenge (Lefford, 1974; Holsapple et al., 1984). The second measure that can be made after either in vivo or in vitro exposure to antigen is the generation of cytotoxic T lymphocytes to alloantigen (Cantor and Boyse, 1975a,b). The level of response is assessed by the ability of immunized cells to lyse target cells that have the same major histocompatibility locus as that of the immunizing antigen. Both these assay systems involve complex interactions of many cell types. Another assay used to determine the effects of environmental toxicants on the cellular arm of the immune system assesses the ability of immunocytes to proliferate in response to alloantigen. This assay is called a mixed-leukocyte response (MLR). Although MLR does not measure the ability of the effector cell to eliminate antigen, it is sensitive to perturbation by chemicals known to affect cellular immunity. MLR is generally more sensitive to changes than are the proliferative responses to mitogens (Bach and Voynow, 1966; Harmon et al., 1982). The mitogens used to stimulate the proliferation of T cells are the plant lectins phytohemagglutinin and concanavalin A (Andersson et al., 1972). Some studies have shown that the concentration of mitogens most likely to stimulate T-cell proliferation is also optimal for the generation of suppressor cells (Dutton, 1972; Rich and Pierce, 1973; Redelman et al., 1976). A change in the ability of the host to resist influenza virus, herpes simplex virus, Listeria monocytogenes, and Plasmodium yoelli infections and to eliminate the tumor PYB6 has been shown to correlate with alterations in MLR (Harmon et al., 1982; Luster et al., 1988). Changes in T-cell responses to mitogen and cell-mediated immune responses, such as a delayed hypersensitivity response, also are correlated with alterations in the ability of the mouse to eliminate Listeria monocytogenes, herpes simplex virus, PYB6, and Plasmodium yoelli. Nonspecific Immunity The immune system responds to and eliminates antigens before the generation of a specific immune response through nonspecific mechanisms. Natural killer cells can kill sensitive tumor targets (such as YAC-1 for mouse and K562 for human) upon their initial exposure (Reynolds and Herberman, 1981). Although the mechanism of cytolysis could be slightly different, natural killer cells eliminate sensitive tumor cells by a method very similar to that of cytotoxic T lymphocytes. A decrease in natural-killer-cell activity is associated with the reduced ability of an animal to respond to cytomegalovirus and to eliminate PYB6 and B16F10 tumors (Luster et al., 1985). Standard assays for nonspecific leukocyte function include quantitation of peritoneal macrophage number (basal and in response to in vivo stimuli), quantitation of polymorphonuclear cells, and leukocyte phagocytic ability (basal and stimulated). Although these measurements are easy to perform, they are not sensitive to chemical perturbation. Additional estimates of macrophage function suggested are the quantitation of ectoenzymes, bactericidal activity, and tumoricidal activity. Alterations in polymorphonuclear-cell function and leukocyte phagocytic ability have been correlated with changes in resistance to Streptococcus pyogenes infection. Resolution of an infection with Listeria monocytogenes requires the appropriate function of macrophages. In addition, modulation of macrophage function has been correlated with changes in the elimination of the B16F10 tumor and Plasmodium yoelli. Soluble mediators of immune responses also are measured to assess immune function.

OCR for page 83
Biologic Markers in Immunotoxicology Interferon and complement serve to eliminate nonspecific complement pathogens. Alterations in interferon levels have been correlated with modulation in resistance to influenza virus. In addition, changes in complement activity have been correlated with alterations in the host defense against Streptococcus pyogenes infection. Bone Marrow The reservoir of stem cells that replenishes erythroid and immune cells is found in the bone marrow. Because this organ contains many highly proliferating cells, it is sensitive to toxic agents that modulate cellular proliferation (such as antineoplastic agents). Therefore, a change in the cellularity of bone marrow could be a useful indicator of a general toxicity, but is not necessarily specific to the immune system. However, upon stress of the immune system, when it could be necessary to call on the bone marrow reserve, alterations in bone marrow cellularity will lead to immunotoxicity. Two assays are used to assess the effects of xenobiotic substances on the stem cell activity of the bone marrow. One is a determination of the number of cells able to form colonies in the spleen (CFU-S). Additionally, the CFU-GM assay used often in immunotoxicology is based on the ability of bone marrow cells to form colonies of granulocytes and monocytes in vitro in response to the appropriate hormones. Host Resistance The models of host resistance now used to assess the integrity of the mouse immune system after exposure to xenobiotics were discussed above. These assays are expensive to run, require special housing to isolate infected animals, and require special facilities to grow the pathogens. Most laboratories undertaking immunotoxicity studies for hazard evaluation will have these models available, but they may not be feasible for individual researchers to undertake. Because the assays are expensive and use a great number of animals, they should not be used for screening. However, once an alteration in one area of immune function is noted, the appropriate model of host resistance could be used to determine the effect of a xenobiotic on the whole animal's response to disease. These host-resistance models are the best available models for illustrating the link between immunosuppression and clinical manifestation of disease end points. Wholeanimal responses could be the best way to predict alterations in immune function, although they might not be sensitive to minimal xenobiotic modulation (Bradley, 1985). Mechanistic Studies If the cellular or subcellular site of action in animals is known, the opportunity to determine whether the effect will occur in humans is increased. Often, mechanistic studies can be performed in human cells, and a specific response in an exposed population can be investigated. Indirectly acting chemicals are now being identified. Agents that alter immune function as a consequence of effects on other tissues are only beginning to be investigated. Agents can produce tissue damage, releasing acute-phase reactive proteins, which then modify the immune system. The classical example is casein, which will decrease immune-system activity by causing the release of serum amyloid protein. Agents that alter nervous-system function, kidney function, and liver function are also known to alter immune status. Because the immune response can be generated in vitro, the actions of a chemical can be investigated at all stages of cellular activity. Several current investigations are directed at determining the biochemical basis for immune-system alterations. Luster et al.

OCR for page 83
Biologic Markers in Immunotoxicology (1985) have shown that benzidine affects the leukotriene pathway in immunocytes. Dean et al. (1985b) reported that dimethylbenzanthracene (DMBA) inhibits interleukin-2 (IL-2) production. Efforts are being made to understand more about pharmacologic receptors on immunocytes and the transduction events that are responsible for immune regulation. Studies by Sanders and Munson (1984a,b) and Fuchs et al. (1988) showed that the β-adrenoceptor activation enhances primary antibody response. Studies by Tucker et al. (1986) and Kerkvliet and Brauner (1987) implicated the Ah (aromatic hydrocarbon) receptor in immunosuppression caused by dioxin. Ah mediates the induction of AAH (aryl hydrocarbon hydrolase) when dioxin and the receptor bind. Dioxins and polychlorinated aromatic compounds most likely act directly by means of a receptor mechanism. In contrast, cyclophosphamide must be metabolically activated, because its metabolites are the reactive agents. Likewise, the polyaromatic compounds and nitrosamines are metabolized to active chemicals. It is likely that halogenated methanes require activation also. ASSAYS OF PULMONARY IMMUNOCOMPETENCE The assays described in the previous section are applicable to situations in which the test chemical can be given orally. Special assays are needed to determine harm from exposure to air pollutants. They should be meaningful in predicting the susceptibility to, severity of, or recovery from disease. They should detect insults to both the humoral and the cell-mediated immune systems, evaluate local pulmonary versus systemic immunity, and provide data that are useful for comparing interspecies sensitivity for risk assessment (G.R. Burleson, EPA, personal commun., 1990). Several pulmonary immune functions are important in cell-mediated immunity against viral disease and are proposed for use in evaluating immunocompetence: interferon production, alveolar macrophage function, natural-killer-cell activity, and cytotoxic-T-lymphocyte (CTL) activity (Burleson, 1987, and personal commun., 1990). These immune functions increase in the lungs of rats infected with rat-adapted influenza virus. Interferon is measured in the bronchoalveolar lavage fluid (BALF), and alveolar macrophage function is assessed in cells from BALF. Bronchoalveolar lavage is a relatively noninvasive procedure that allows direct comparisons of immune function in animals and humans. Natural-killer-cell and CTL activities are measured in whole-lung homogenate. CTL activity also is detected in the lung-associated lymph nodes (LALNs), spleen, and peripheral blood after viral infection. This pulmonary model for cell-mediated immunity has been used to evaluate the immunotoxicity of inhaled compounds (Burleson and Keyes, 1989; Burleson et al., 1989; Ehrlich et al., 1989; Ehrlich and Burleson, in press). Pulmonary humoral-mediated immunity also should be evaluated after exposure to inhaled pollutants. The important considerations for evaluation of pulmonary humoral-mediated immunocompetence are reviewed by Bice and Shopp (1988). Changes in antigen elimination: Inhaled pollutants can damage epithelial cells and alter lymphatic clearance of antigens from the lung to the LALNs (Schnizlein et al., 1982; Hillam et al., 1983). Changes in the number or function of alveolar macrophages or neutrophils entering the lung also could alter clearance of antigen (Harmsen et al., 1985, 1987). Changes in the function of LALNs: Lymphocyte subpopulations and numbers in the LALNs can be altered as a result of inhaled particulate matter eliminated by the LALNs. Insoluble particles remain in the LALNs for long periods (Snipes et al., 1983), and these toxic materials can alter antigen-handling cells and other immune-system functions (Bice et al., 1985, 1987a). Changes in the recruitment of immune cells

OCR for page 83
Biologic Markers in Immunotoxicology into the lungs: Changes in vascular permeability are important in the recruitment of lymphoid cells and antibody-forming cells into the lung (Bice et al., 1982). Thus, inflammation and lung damage as a result of exposure to pollution could alter the recruitment of immune cells and antibody from the blood into the lung. Changes in antibody production of the lung: Large numbers of antibody-forming cells are present in BALF from immunized lung lobes (Bice et al., 1980a,b, 1982). Plasma cells in the alveoli and interstitial tissues are at risk from inhaled pollutants (Bice et al., 1987b). Damage to these cells could reduce the production of local antibodies. These are important in pulmonary defense against pathogens. Cells that produce antibody in immunized lung lobes have been reported 3 years after the last antigen exposure (Bice and Muggenberg, 1989). The cells responsible for long-term maintenance of antibody levels in the lung could be important in preventing recurrent pulmonary infections. The cells responsible for long-term antibody production are at risk of damage caused by inhaled pollutants. Changes in the function of immune-memory cells in the lung: Immune-memory cells are recruited or produced locally after lung immunization (Mason et al., 1985; Bice and Muggenberg, 1988). These cells provide local immune memory against inhaled or aspirated pathogens. Damage to immune-memory cells or to the cells responsible for antigen presentation could result in a loss of localized lung immune memory and lead to recurrent pulmonary infections. BALF contains numerous biochemical indicators of lung damage, including antioxidants, proteins, enzymes, cytokines, growth factors, arachidonic acid metabolites, and cells. BALF analysis has been used to assess the pulmonary toxicity of numerous inhaled pollutants (Henderson et al., 1979a,b, 1985a,b, 1986, 1988; Henderson, 1984, 1988, 1989). It thus provides biologic markers that allow direct extrapolation of results from animal studies to human subjects. Most of the cells in BALF consist of alveolar macrophages. However, neutrophils provide an indicator of inflammatory response to infection and to inhaled xenobiotics. The above categories of pulmonary immunologic response can provide information to assist in evaluating immunotoxicity of inhaled chemicals and in comparing interspecies sensitivity for use in risk assessment. The severity of change of pulmonary immune functions for humoral and cell-mediated immunity to assess immunotoxicity can now be used to obtain results for the extrapolation of data from experimental animals to humans. ASSAYS REQUIRING ADDITIONAL DEVELOPMENT The previous sections listed assays that are generally acceptable for immunotoxicity testing in animals. This section is devoted primarily to animal bioassays that are still in development. Some of these procedures have been used in immunotoxicology testing protocols but require additional testing and confirmation before they can be accepted as validated. Some of the more advanced procedures could require little additional testing to determine their reproducibility; others, used in basic immunology, could require modification or additional development for adaptation to immunotoxicology. Some of these procedures could be inherently insensitive; others might not test for, or correlate with, immune function. Immunotoxicity assays and correlated immune-system biologic markers should be investigated in both humans and animals. The example in a following section relating the effects of ultraviolet light and suppression of Th/Ts ratios is an example of the type of cross-discipline activity that needs encouragement. Subpopulations of immunocytes can be identified and enumerated by specific surface antigens that are peculiar to an individual cell type, but studies in animals suggest that immunomodulatory chemicals and drugs may not target the same subpopulation of immunocytes

OCR for page 83
Biologic Markers in Immunotoxicology that is affected in humans. Thus, enumeration of immunocyte surface markers as correlates of immune function has not been of much value. Nevertheless, the ease of these procedures and the remote possibility that a correlation could occur suggest that further comparative investigation is in order. In addition, markers of cellular activation that are modulated on the basis of the differentiation status of the cell, such as Tal or class II MHC proteins, may be useful in detecting immunomodulation by xenobiotics. Lymphokine production and activity have been studied on a limited basis for selected species. Additional information is needed to ascertain the effect of xenobiotics on lymphokine production and activity in animal species and to elucidate the ability of such alterations to correlate with changes in host resistance. Assays are available for the 10 known animal interleukins and several of the cytokines. Animals also produce interferon and other cytokines that are similar to those in humans. Those that require additional investigation for applicability in immunotoxicity testing are tumor necrosis factor and the factor that stimulates CFU-GM in bone marrow (Aggarwal et al., 1985; Pestka et al., 1985; Nathan, 1987; Malkovsky et al., 1988). As new lymphokines are discovered, measurement of immunohormones to detect xenobiotic-induced immune dysfunction deserves attention. Further exploration of the macrophage as an indicator of xenobiotic-induced immunotoxicity could be of considerable value. Evaluation of cell-surface markers, such as the Mac-1 (type 3 complement receptor) (Beller et al., 1982; Garner and Elgert, 1986), Mac-2 (activation marker) (Ho and Springer, 1984; Garner and Elgert, 1986), F4/80 (activation marker) (Austyn and Gordon, 1981), and Ia antigen (Beller et al., 1980; Bhattacharya et al., 1981), could prove a valuable asset in assessing immune dysfunction. Other properties of the macrophage, such as respiratory burst, antigen presentation, monokine secretion, and activity, cytostasis, cytotoxicity, protease activity, and size heterogeneity, require additional investigation to assess their ability to reflect modulation by xenobiotics (Nathan, 1987). Other markers that require additional investigation in animals are the antibody-dependent cytotoxic cells (Cordier et al., 1976), transferrin receptor (Neckers and Cossman, 1983), IL-2 receptor (Cantrell et al., 1988), Ia receptor, and complement activity (Ehlenberger and Nussenzweig, 1977). Some of these have been used, and a few appear promising as sensitive indicators of immunomodulation by xenobiotics, but all require additional investigation before they can be considered as validated assays for immunotoxicology. Other procedures that deserve additional attention are detection of mRNA levels (Lindsten et al., 1989) and analysis of bone marrow factors, such as colony forming unit-granulocyte (CFU-G), colony forming unit-basophil (CFU-B), and stromal-cell culture. USE OF IMMUNOTOXICITY BIOASSAYS The animal assays used to study the effects of xenobiotic exposure on the immune function of rodents have proved to produce consistent results in several laboratories and are correlated with a change in host resistance. Some assays are of less value, because they lack the ability to detect changes in the immune functions at low levels. Table 6-1 summarizes the status of these assays. Of all animal studies conducted thus far, the generation of a primary immune response to antigen seems to be the most responsive to modulation by xenobiotic substances. Considerations in the Design of Immunotoxicity Testing The first choice of an experimental animal

OCR for page 83
Biologic Markers in Immunotoxicology for toxicologic investigation is the one whose toxicokinetics of the test chemical are similar to those of the human. If the metabolism of the chemical is not known, the best experimental animal is usually the one best for measuring the response. The immune systems of most commonly used animals present similar targets and operate in a similar manner. Although the mouse was the experimental animal of choice because much of what is known about the immune system was discovered in the mouse, the rat is equally useful for most aspects of immunotoxicity assessment. The dog is an important experimental animal in toxicology, but has not yet been used widely in studies of immunotoxicity. Primates are now beginning to be used to determine how chemicals target their immune systems. Many of the reagents available for human immune assessment can be used in cynomolgus and rhesus monkeys. The duration of exposure also is an important consideration. For most studies aimed at determining target-organ toxicity, 90 days of exposure is believed to be adequate to produce and demonstrate most toxicities. Exceptions are made for studies of developmental, reproductive, and genetic effects. For target organs not closely aligned with metabolism and excretion (liver and kidney) of a chemical, 90 days of exposure can mask an effect. The long exposure allows induction of tolerance mechanisms to take place. This could be an important consideration for the immune system, which, for the most part, is not in the mainstream of chemical metabolism and excretion. The cells of the immune system are activated by the introduction of an antigen and, if the level of the chemical in the lymphoid tissue is decreased because of enhanced metabolism or excretion, an effect that could occur with short exposure could be missed with a longer exposure. An exposure less than 90 days is justified in immunotoxicology because of the fairly rapid turnover of most of the immunocompetent cells once they have been stimulated by an antigen. Except for memory cells, most cells of the immune system perform their function in 3-14 days. Thus, an exposure that maximizes the levels of the chemical in the lymphoid tissues for this period should provide the best chance of showing the effect on the immune system. Exposure of 14-30 days is usually adequate to demonstrate immunotoxicity. This minimal time requirement for hazard evaluation constitutes a significant advantage over other types of assays that often require much longer periods. However, the prediction of disease end-point incidences is complicated by the necessity of challenges with infectious or neoplastic agents during the duration of immune suppression. Reversibility of effects on the immune system is related to the chemical and the age of the experimental animal. Most of the effects of immunosuppressive antimetabolites, such as azathioprine and methotrexate, are readily reversed. Most immune responses are restored within 2 weeks when alkylating agents, such as cyclophosphamide, are used. However, high doses of alkylating agents can injure pluripotential stem cells that might not be regenerated. In animals, the immune responses could be intact when the animals are challenged, but, if the stem cells are again insulted by administration of antimetabolites, a more severe and prolonged depressed response will occur. The reversibility of the effect on the immune system is important information for hazard evaluation. As do other systems of the organism, the immune system has many compensatory mechanisms, and in the normal individual a large reserve exists. The immune system has several levels of protection, and innate immunity has several components, as does acquired immunity. Even when innate immunity is sufficient to handle a particular pathogenic insult, the acquired immune system is activated to provide future protection. There is overlap between humoral and cell-mediated immune responses to pathogens. Although one system can predominate in protecting against a given pathogen, the other often will back up or synergize with

OCR for page 83
Biologic Markers in Immunotoxicology the primary system. For example, neutrophil bactericidal activity can be sufficient to protect from a streptococcal insult, but IgM and IgG antibodies afford future protection. When neutrophil activity is inadequate, IgG antibody and complement might be sufficient to remove the invader. In addition, antibody-coated bacteria are more efficiently recognized and killed by tissue macrophages. Similar examples can be shown for cooperation between T-and B-cell-mediated protective activity. Because dose is a function of the amount of the xenobiotic received and the duration of exposure, because most agents have several target organs leading to both quantitative and qualitative changes in the dose-response curve, and because the immune-system response can be influenced by other target-organ toxicities, care must be taken in selecting doses for immunotoxicity assays, To use these assays, one must be aware of the temporal consideration, especially important to ensure system responses. The dose should not overwhelm the functional capacity of other systems or the metabolic capability of the test animals, and the apparent immune changes should not be the result of influences by other systems. Minimal modulation of the nervous and endocrine systems can result in magnification of the effect in the immune system. Contrary to general belief, the immune system can remain intact when other toxic manifestations are great. In a series of experiments with tetraethyl lead and chlordane, animals surviving a median lethal dose (LD) had intact immune systems. Animals that are moribund because of chemical exposure often can respond normally to antigens. Immune alteration can be secondary to other effects. Acute-phase reactive proteins, produced and released in some hepatotoxic events, can lead to immunosuppression. It is important to know whether the immune system is the most sensitive target, so that appropriate exposure levels can be established and at-risk groups identified. Immunotoxicity as a Basis for Risk Assessment Immune function can be defined as the normal, special, or proper action of immunocytes and their secretory products. Action may be expressed if a macrophage phagocytizes foreign debris; B lymphocytes produce antibody that ''neutralizes" foreign antigen; T lymphocytes, natural killer cells, or lymphokine-activated killer cells kill tumor cells; or cytokines regulate the immune network. Several immune procedures are available to assess xenobiotic-induced dysfunction in animals. Fewer assays are available to evaluate immune dysfunction in animals. Measuring immune function is essential and critical in interpreting and translating data from animals to humans and in understanding the relationship of immunocompetence to disease resistence. In using immunotoxicity data for risk analysis, one must be cautious to ensure that the data are a correlate of immune function, i.e., disease resistance. The immune system can be a sensitive target organ, compared with conventional toxicologic procedures used to evaluate the toxicity of drugs and chemicals. Although it is not possible to predict the outcome of exposure and human disease incidence from animal immunotoxicity bioassays, they are useful in the hazard evaluation phase of risk assessment. Additionally, immunotoxicity data may clarify the mechanisms whereby agents induce adverse effects that are manifested as other end points, such as cancer. Chemical suppression of humoral immunity: The immunotoxicity of trans -1,2-dichloroethylene (DCE) was assessed in a 90-day study by Shopp et al. (1985). In the cell-mediated immunity assays, no dose-related decreases were seen in either sex. Humoral immune status was assessed by measuring quantitation of spleen IgM antibody-forming cells (AFCs), hemagglutination titers to sheep erythrocytes (SRBCs), and spleen response to the B-cell mutagen lipopolysaccharide

OCR for page 83
Biologic Markers in Immunotoxicology (LPS). Cell-mediated immunity was assayed by delayed-type hypersensitivity (DTH) to SRBCs, popiteal lymph-node proliferation in response to SRBCs, and spleen-cell response to challenge with concanavalin A. In males on day 4 after treatment, there was a significant decrease in AFCs at 17, 175, and 387 mg/kg per day, but the decreases were significant at 175 and 387 mg/kg per day only when the data were calculated on the basis of spleen cells. No dose-dependent effects were noted in the DCE-treated female mice. In both sexes, there were no changes in hemagglutination titers after DCE treatment and spleen responsiveness to LPS was unaltered in male mice. However, females exposed at 452 mg/kg per day did have an enhanced response to LPS. In addition, lymphocyte responsiveness in the absence of the mitogen (LPS) was decreased at 224 and 452 mg/kg per day in females only. This suggests a no-observed-adverse-effect level (NOAEL) of 22 mg/kg per day in female mice. The impact of the decrease in spleen AFCs in male mice at all three dose levels is uncertain, but the decrease is significant only at the 175 and 387 mg/kg per day on the basis of AFCs. Therefore, the NOAEL in male mice is taken to be 17 mg/kg per day. This study illustrates three areas of uncertainty in the use of immunotoxicity assays. The first, in translation of animal data to human risk, is the difficulty in measuring a given indicator of adverse effect in the same compartment in animals and humans. Rodent immunotoxicity studies most often center on effects in the spleen, as in this study, or in lymph node, thymus, or cells from the target site, such as alveolar macrophages. For the most part, these compartments are the sites of immune function. Thus, the results give the best prediction of the potential actions of a xenobiotic substance on the immune system. In contrast, the major source of immunocompetent cells available for determination of human immune status is peripheral blood. The immune compartments and stages of differentiation of cells in the peripheral blood and in the lymphoid tissue are different. Correlations must be made from knowledge of the action of immune cells in each of the tissues. Unless there is a major breakdown in immune-system function or the challenge by the secondary agent occurs at the right time, an immune-system alteration induced by a xenobiotic is not likely to be detected in specific disease end points. The second uncertainty of extrapolation of animal data to human risk is the variability in measurement of the indicator of toxicity. Not only are fewer human assays available, but the variability in the human population is considerable. In most immunotoxicity studies, inbred animals are used to decrease variation between individuals. Thus, animal studies are more likely to detect an immunosuppressive xenobiotic substance. Because of genetic differences among humans and because of the many factors that can influence the response of the immune system, it is difficult to detect immunosuppression at low exposure levels. Likewise, immunosuppression can lead to infectious diseases caused by opportunistic microorganisms if simultaneous exposures (i.e., to pathogenic microorganisms and xenobiotics) occur. The third uncertainty relates to the temporal separation of cause and effect with respect to immune-system suppression. There is a continuum of consequences of immunosuppression that is related to the degree of suppression. The most serious consequence is the time required to recover from an infection. Major immunosuppression is readily detected, and the hazard is easily described. Moderate to minor immunosuppression is less readily detected, but the consequences can be considerable. Prolongation of an infection or development of the disease can occur when moderate immunosuppression occurs. Assuming that this variation in immune-system activity represents an indication of the potential for disease, uncertainty factors of 10 for variability

OCR for page 83
Biologic Markers in Immunotoxicology within the test species (in this case mice), 10 for variation between animals and humans, and 10 if one applies this information derived from a short-term study to a life-time exposure could seem reasonable. Using these uncertainty factors suggests that an intake of 0.0175 mg/kg per day over the lifetime of humans would probably not produce adverse effects. A role for mechanistic studies in the management of risk: Understanding of the mechanisms of immunosuppression, most frequently developed from animal experiments, can facilitate our understanding of risk and play an important role in the selection of alternatives used to decrease risk. Light (ultraviolet-B, 290-320 nm) suppresses cell-mediated immune responses in the skin (De Fabo and Noonan, 1983). In mice, the suppression occurs regardless of whether the sensitizer (antigen) is applied locally or at a site distant from the site of UV exposure (Noonan and De Fabo, 1990). This depression of the immune system decreases immune responses to conact sensitizers (Noonan et al., 1981) and to alloantigens (Williams et al., 1990) and is critical to the growth of UV-induced tumors (Noonan and De Fabo, 1990). Urocanic acid (UCA) is a major UV-absorbing component of the skin (stratum corneum). UV light converts the trans isomer of UCA to the cis-UCA isomer, an immunosuppressive agent (Noonan et al., 1988). In mice, the immunosuppression is affected by the generation of suppressor T cells (Ross et al., 1987) and defective in antigen-presenting cells (Noonan et al., 1988). Frentz et al. (1988) have shown that in patients (with nonmelanoma skin cancers) with a history of UV exposure, the ratio of circulating helper T lymphocytes (Th) to suppressor T lymphocytes (Ts) was abnormally low (p < 0.010) compared with that in patients with a history of x-ray exposure or controls. The low Th:Ts ratio (CD4:CD8) was associated with an increase in the absolute number of Ts cells. The immunodeficiency in the skin would enhance the survival of initiated tumor cells, increasing the likelihood of progression of these cells to skin carcinomas. This information is important in the understanding of the risk associated with changes in stratospheric ozone (De Fabo et al., 1990). On the basis of the information presented above and the report by Reeve et al. (1989) that topically applied UCA enhanced the number of skin tumors and the malignancy rate in hairless mice, several cosmetic companies have removed UCA from some products, such as moisturizing creams. Disruption of normal modulation in the immune system: Malathion is a widely used organophosphate pesticide that acts through inhibition of acetylcholinesterase. Inhibition of acetylcholinesterase of red blood cells or pseudocholinesterase in the serum is the standard by which organophosphate toxicity is measured (i.e., this is thought to be the most sensitive indicator of organophosphate toxicity). Studies have shown that administration of doses of malathion that inhibit acetylcholinesterase suppresses the humoral immune response to antigen (Casale et al., 1983). However, recent studies showed that administration of noncholinergic doses of malathion enhanced the immune response to antigen (Rodgers et al., 1986). In this strain of mice, using purified malathion (>99.9% pure), administration of malathion at 715 and 900 mg/kg through oral lavage led to 0% and 10% inhibition of serum pseudocholinesterase, respectively. Further studies showed that enhanced macrophage function contributed to this enhanced immune responsiveness after malathion administration (Rodgers and Ellefson, 1990). Most recent studies were conducted to determine the NOAEL for malathion on the mouse immune system. For this, the respiratory burst of peritoneal cells was used as a measure (Rodgers and Ellefson, in press). The respiratory burst of macrophages is instrumental in the bactericidal and tumoricidal activity of leukocytes. On the other hand, an enhanced respiratory burst of leukocytes has been associated with tissue and DNA damage. An enhanced respiratory burst of

OCR for page 83
Biologic Markers in Immunotoxicology leukocytes is thought to be involved in the pathogenesis of rheumatoid arthritis, ischemia-reperfusion injury, and arthrogenesis. Therefore, an alteration in the level of respiratory burst activity, either increased or decreased, can be detrimental to the host. For this measure of immune function, an unusual dose-response curve was noted. There are dose-dependent changes after malathion at above 150-300 mg/kg (administered orally) and below 1 mg/kg. Between these doses of malathion, there is an elevation in the respiratory burst activity of peritoneal cells, compared with vehicle or naive controls, but not relative to one another (i.e., plateau noted). In this study, lowest-observed-adverse-effect level (LOAEL) and NOAEL for oral administration of a single dose of purified malathion for enhanced respiratory bursts of peritoneal cells were 0.25 (120% of control) and 0.1 mg/kg, respectively. Thus, an uncertainty factor of 10 for variability within species and 10 for rodent to human extrapolation would lead to a relatively safe dose of 0.001 mg/kg. This seems appropriate for occasional exposures to malathion. SUMMARY A large body of data exists on animal models, some of which have been verified and many more of which are in development. Over the last decade, several important animal models have been developed to detect the immune potential of chemical agents and to elucidate immunologic mechanisms of injury. The following are two of the most productive areas of development: Animal models are essential for detecting immunosuppressive potential. They provide the means whereby dose-response and mechanistic studies can be performed to provide a basis for hazard evaluation as related to human risk. When possible, the experimental design for animal studies should be applicable to humans with respect to the following: Ability to measure the status of the immune system. Use of a route of administration that emulates human exposure. Matching of the toxicokinetics (i.e., absorption, distribution, and metabolism of the xenobiotic in the experimental animal and humans). Experimental animals provide an important and necessary means for detecting immunotoxic compounds and for determining their mechanisms of action in the induction of disease. Many of the methods described in this chapter can be used to detect immunotoxic chemicals. They have been validated in laboratories, and several of them can be used for in-depth mechanistic investigations. The principles of toxicology that are applied to assessing immunotoxic chemicals cover the selection of the following: experimental animals; route and duration of exposure; sensitivity, reproductibility, and interpretability of methods; and ability of these bioassays to be applied to hazard evaluation. RECOMMENDATIONS The Subcommittee on Immunotoxicology recommends that studies on the toxicity of xenobiotic substances to human immune function include the generation of primary cellular and humoral immune responses to an antigen, such as keyhole limpet hemocyanin. Inclusion of this assay in studies of alterations in immune responsiveness to xenobiotics should vastly enhance their usefulness. It will be necessary to develop models that detect potential immunosuppressive chemicals in species other than rodents. Dogs and nonhuman primates are widely used in toxicity studies, and attention should

OCR for page 83
Biologic Markers in Immunotoxicology be given to their use in immunotoxicity investigations. Models need to be developed for closer extrapolation to humans by using severe combined immunodeficiency and transgenic and congenic mouse and human cell lines. The relationship between indicators of immune status in peripheral blood and lymph nodes (spleen) should be established. Because the business end of the immune system is in the immune system's organs and the available marker medium in humans is blood, the relationship of these components should be well defined with agents that are known to alter the immune system. The mechanisms by which classes of chemicals produce immune alterations should be determined. Mechanisms classified according to chemical family will assist the regulatory establishment in making decisions related to risk assessment. Knowledge of the mechanism of immune-system injury will enhance extrapolation to humans and give a rational basis to remediation. There should be investigation of the role of immunocompetent cells in metabolizing chemicals to reactive agents that are either immunosuppressive or that are recognized as nonself and elicit a hypersensitivity reaction. In vitro immunologic assays that can be used for rapid detection of potential immunosuppressive agents need to be developed to determine cellular and subcellular sites of action. There needs to be investigation of the relationships between specific chemically induced immunosuppression and quantitative and qualitative changes in host resistance to microorganisms, parasites, and neoplastic diseases. In host-resistance assays, research should be directed to measure more sensitive end points than mortality. Some suggested approaches include measurement of the generation of an immune factor that is known to eliminate a pathogen after exposure. For example, measurement of viral-induced natural-killer-cell function might be more sensitive and relevant to human disease than is death of a rodent.