Animal Model Approach to Testing

Immunosuppression. To draw meaningful conclusions from host resistance studies, the animal model must fulfill certain criteria. The criteria are that (1) the model must simulate a relevant disease or infection with known adverse health effects, (2) the immune effector mechanism active in the disease or infection must be the same in humans and the test animal, (3) microorganisms must be administered by the natural route, (4) the concentration of the challenge agent should not overwhelm the host defense system, and (5) relative resistance or susceptibility to microorganisms and viruses is genetically determined (Bradley, 1985). Based on these criteria, the major considerations in the design of host defense or tumor models are the genetic background, the dose, and the number of animals used in the study.

The genetic background has long been known to control resistance or susceptibility to disease. In mice, influenza-induced macrophage antiviral activity is controlled by the presence of the Mr gene. Viruses can replicate in macrophages lacking the Mr gene even when high concentrations of interferon are present. In contrast, the virus does not replicate in cells carrying the Mr gene (Haller et al., 1980).

Resistance to murine cytomegalovirus (MCMV) is carried in the H-2k MHC

locus. Mice expressing this marker have augmented NK cell activity and are ten times more resistant to the virus than are mouse strains lacking the H-21. Therefore, MCMV host resistance studies use C3H/HeJ or B6C3F1 mouse strains.

The response in host defense assays depends, in part, on the dose of the organism used in the study. Clearly, a high dose of a virulent organism will overwhelm the host and kill most of the animals. In contrast, low concentrations will fail to detect changes in immunocompetence.

Two different dose regimens can be used in host resistance studies. The LD20-LD30 is the most common dose used in host defense studies. This dose will kill 20-30%of the animals in the control group. In the second regimen, a dose slightly below the dose inducing effects in the control group (ED0) is administered. The determination of the LD20 or the ED0 is difficult and depends on the test species. In practice, extreme accuracy in administration of the dose is required (Anonymous, 1997).

Statistically significant changes in host defense are a reflection of the number of animals in the test groups treated with xenobiotics (Table 3). For example, when 15 animals in each test group are immunized with an LD20-LD 30 inoculum, 13 of 15 animals must be affected before changes are statistically significant. If the number of animals per group is increased to 30, only 18/30 must be

Tabie3, Effect of Dose of Bacteria (ED) on the Number of Animals Required for a Host Defense Study

ed30 ed30 EDo

No. affected

No. affected

No. affected

(n = 15)

p value*

(n = 30)

p value*

(n = 15)

p value*

Control

3/15

6/30

0115

Test

1

4/15

0.532

8/30

0.428

1/15

0.516

2

5/15

0.411

10/30

0.273

2/15

0.274

3

6/15

0.312

12/30

0.165

3/15

0.150

4

7/15

0.233

14/30

0.096

4/15

0.084

5

8/15

0.173

16/30

0.055

5/15

0.048**

6

9/15

0.128

18/30

0.031**

8

10/15 11/15

0.094 0.069

9

12/15

0.051

10

13/15

0.038**

Modified from International Program on Chemical Safety, Principles and methodsfor assessing direct immunotox-icity associated with exposure to chemicals, Anonymous, 1995a. Geneva. *Chi-square one tailed test. **Statistically significant^ < 0.05.

Modified from International Program on Chemical Safety, Principles and methodsfor assessing direct immunotox-icity associated with exposure to chemicals, Anonymous, 1995a. Geneva. *Chi-square one tailed test. **Statistically significant^ < 0.05.

Table 4. Biologically Relevant Immunotoxicity Induced by Chemicals in Host Defense Models

Model

Chemical

Reference

Listeria

A'-Tetrahydrocannabinol

Morahan et al. (1979)

Streptococcus

TCDD

White et al. (1986)

Cytomegalovirus

Organotin

Garssen et al. (1993)

Influenza virus

Benzo[a]pyrene

Munson and White (1990)

Benzo [e] pyrene

Methyl isocyanate

Luster et al. (1986)

TCCD

House et al. (1990)

Dimethylnitrosamine

Thomas et al. (1985)

Trichinella

Organotin

Vos et al. (1990)

Acyclovir

Stahlmann et al. (1992)

Plasmodium

4,4'-Thiobis (6-r-butyl-m-cresol)

Holsapple et al. (1988)

Fish oil supplement

Blok et al. (1992)

Styrene

Dogra et al. (1992)

affected to achieve statistical significance. Only a limited number of chemicals have been tested in host defense (Table 4) or tumor models (Table 5).

The dose of the infectious agents also influences the number of animals needed in host defense assays. When the subclinical dose (ED0) is used in the assay, only 5 of 15 animals must be affected for statistical significance (Anonymous, 1997).

Data from the host resistance assays can be analyzed by Fisher's exact test. The use of this test requires several basic assumptions. First, it is assumed that death is the result of direct or indirect action of the test chemical. Second, it is assumed that death is the result of viral, bacterial, or fungal infections. Third, it is assumed that death from infection is the event of interest.

Table5. Biologically Relevant Immunotoxicity Induced by Chemicals in Tumor Models

Model

Chemical

Reference

B16F10 melanoma

Phorbol myristate acetate

Murray et al. (1985)

Gallium arsenide

Sikorski et al. (1989)

4,4'-Thiobis(6-r-butyl-m-cresol)

Holsapple et al. (1988)

PYB6 fibrosarcoma

Aroclor

Luster et al. (1986)

Dimethylbenzanthracene

Dean et al. (1986)

Benzene

Rosenthal and Snyder (1987)

MAD106

Nickel chloride

Smialowicz et al. (1987)

Often overlooked in the statistical analyses of host defense data is the power of the results. This measurement determines the frequency that the expected result would occur. For example, if the results were expected 9 of 10 times, the power to detect would be 90%. Assuming approximately 20% mortality in the control group and three dose levels, a mortality rate of 63% in the highest dose group would yield a power of 90%.

Listeria monocytogenes Model. The immune response to intracellular listeria involves several different immune effector mechanisms. An initial response is mediated by macrophages activated by T-cell-derived IFN-y. The activated macrophages attempt to kill the extracellular organisms by phagocytosis (Van Loveren et al., 1987). After 2 or 3 days, CD8 cytotoxic cells lyse infected phagocytic cells releasing bacterial protein antigens. CD4 Thl cells respond to the proteins when presented in context with MHCII. A cell-mediated response is elicited resulting in granuloma formation in the liver and spleen. In mice, additional IFN-y production causes the B cells to synthesize IgG2a that opsonizes the bacteria and activates complement.

In the typical listeria challenge study, three groups of mice (12 mice per group) are injected intravenously with bacterial numbers that produce 20,50, and 80% mortality. The intracellular bacteria begin to grow in the liver and spleen within 24 to 72 hr. Mortality is observed at 5 to 14 days, depending on the dose. Several endpoints can be measured: (1) the number of microorganisms retrieved from organs relative to the controls (Reynolds and Thomson, 1973), (2) the percent mortality relative to controls, (3) histopathology caused by listeria. Lesions are characteristic lymphocyte foci with histiocytic cells. Both the severity and the duration of the lesion can be altered by immunotoxicants.

Streptococcus pneumoniae. In the early phase of infection, C3 is deposited on the bacterial surface and the alternate complement pathway is activated. Opsonization bacteria are ingested by monocytes and other phagocytic cells. During the late phase of the infection, streptococcal polysaccharide (a T-cell-independent antigen) elicits production of antibodies. IgM antibody provides protection against circulating bacteria and prevents dissemination. Tissue invasion is prevented by the production of IgG antibodies. Activation of the classical complement pathway, opsonization, and phagocytosis terminate the infection (Winkelstein, 1981).

In summary, host defense against S. pneumoniae requires several effector mechanisms including antibody response to a T-cell-independent antigen (Mox-on, 1981), phagocytosis by monocytes and PMNs, and activation of complement (Schiffman, 1983).

Depending on whether the early or late phase of the infection is being probed, two strains of streptococcus are used in the defense model. Smooth, encapsulated S. pneumoniae, which causes rapid death, is used to probe the complement activation observed early in the disease. Another strain of streptococcus is used to probe the late-phase immune response. S. zooepidemicus tests for antibody-mediated resistance during the late phase of the infection. S. zooepidemicus animals have a much longer time to death (Fugmann etal., 1983).

The effect of immunotoxicants on the complement system can be defined in acute-phase response by using different strains of S. pneumoniae. For example, type 14 S. pneumoniae ATCC 6314 activates both pathways whereas S. pneumoniae type 25 ATCC 6325 activates only the alternative pathway.

In the assay, bacteria are usually administered intravenously in large doses at 20, 50, and 80% of the lethal dose. Deaths occurring within the first 48 hr are related to a complement defect. During the 2- to 4-day period, deaths can be attributed to decreases in phagocytosis. Defects in antibody production result in deaths at 5-8 days.

Cytomegalovirus (CMV) Model. Infection with CMV elicits a number of different immune responses. Interferon is synthesized during the first 48 hr. This is followed by a vigorous NK cell response that peaks at day 6 (Butowski et al., 1986). Virus-specific cytotoxic T cells appear in large numbers by day 10. Neutralizing antibodies are also produced late in the response. A role for macrophages in CMV infection is doubtful. However, the CMV may infect macrophages that serve as reservoirs for future infections (Boos, 1980). Termination of the infection is primarily brought about by NK cells, cytotoxic T cells, and ADCC (Selgrade etal., 1982a).

Two routes of administration have been used in the model. Mice are generally immunized by the intraperitoneal or intratracheal route. Intraperitoneal administration initiates a systemic infection. The highest virus concentration is found in the salivary glands followed by the liver, spleen, and lung. When the virus is administered by the intratracheal route, the virus usually localizes in the salivary glands and death is a rare event. Generally, the viral replication is slow with peak viral loads at 15-20days. Mortality or the concentration of virus in tissue are common endpoints (Bruggeman et al., 1983, 1985).

With intratracheal administration in the rat, there are species-specific differences in virus localization (Bruggeman et al., 1983). In PVG rats, the highest viral load is in the salivary glands. In other rat strains such as Lewis or LN, the viral load is higher in the kidney than in the salivary gland (Bruning, 1985).

In the assay system, pathogen-free, 3-week-old CD-I mice are the strain of choice. This strain is susceptible to the virus. Other strains are resistant. Resistance or susceptibility to CMV is genetically determined. Mouse strains carrying the H-2k haplotype are 10-20 times more resistant to the virus than are strains carrying the H-2b or H-2d haplotype (Chalmers etal., 1977). Resistance may be related to the fact that strains with the H-2k haplotype respond more vigorously with an NK cell response (Bancroft et al., 1981).

The CMV model is used as a surrogate for human opportunistic infections in immunosuppressed subjects and chemicals that suppress NK cell function. There is a significant correlation between immunosuppression and the appearance of CMV infections (Rubin et al., 1981). CMV infections are commonly observed in patients purposefully (e.g., transplantation), accidentally (AIDS), or naturally (e.g., the elderly) immunosuppressed (Rubin, 1990). In addition, there is a correlation between chemically induced suppression of NK cells and infections with CMV.

Influenza Virus Model (A/Port Chalmers/1/72/H3N2). Host response to influenza infection is a cascade of effector mechanisms that include (1) interferons, (2) interleukins, (3) alveolar macrophages, and (4) NK cells, cytotoxic T cells, and antibody production (Burleson etal., 1987). However, resistance to the virus is associated with antibody production (Vireligier, 1975) and interferon (Hoshino etal., 1983). Mortality is commonly measured at 14 days. However, virus concentration in the lung may also be determined by the virus plaque assay.

In theory, mice are inoculated with the LD20 virus concentration. In practice, the LD20 is difficult to define accurately. Continued viral passage in mice increases the viral virulence. Increased virulence reduces the number of viruses necessary to achieve an LD20 and increases the variability around the LD20. To abrogate the variability, three different viral concentrations near the 0.2 logi0 challenge dose are usually administered (Burleson, 1995).

The model has limitations. It is a nasal instillation model with effects in the lung. Frequently, the lung acts as an autonomous immunological organ that requires no interdiction from the peripheral blood immune system. Therefore, it is conceivable that chemicals lowering systemic antibody levels will have no effect on the host defense against in the inhalation model.

Trichinella spiralis Model. T spiralis has a complicated life cycle. Following ingestion of the encysted helminth larvae, the larvae excyst in the acid environment of the stomach. They enter the small intestine where they mature within 3-4 days. After copulation, the gravid females penetrate the intestinal mucosa and produce offspring larvae. This event results in an inflammatory response consisting of mast cells and eosinophils. Over a 3-week period, viviparous larvae enter the lymphatic system and migrate to striated muscle. In the muscle, the larvae become encapsulated and remain viable for months or, perhaps, years. The encysted larvae evoke a vigorous T-cell-dependent inflammatory response.

The adult worms resident in the gut are expelled approximately 6 days after infection. Expulsion is mediated by a specific inflammatory response consisting of gut mast cell activation, goblet cell hyperplasia, and accumulations of eosinophils, PMNs, lymphocytes, and plasma cells. Evidence suggests that the lymphocytic response is T cell dependent (Vos etal., 1983). Antibodies (IgM, IgG, and IgA) are produced and directed toward surface antigens. The IgG-coated parasites are lysed via ADCC mediated by neutrophils and eosinophils (Ruitenberg et al., 1983). Complement and enzymes released from granulocytes may also play a role in destruction of the parasites in the early stages.

This model is unique in that different effector mechanisms are involved in host resistance to life cycle changes of the parasite. T cells are involved in the expulsion of parasites during the first infection. T, B cells, and eosinophils are involved in preventing parasite migration and limiting parasite reproduction (Van Loveren etal., 1995).

A number of different endpoints can be measured in the trichinella model. An increased worm load in the gut and larvae in the muscle are hallmarks of immunosuppression. Histopathology can be used to compare the inflammatory responses to encysted larvae in treated and control animals. Serum antibody titers are also reduced in compromised animals (Ruitenberg et al., 1983).

Malaria Host Defense Model. Resistance to this organism involves antibody production, macrophage activation, and T-cell functions (Luster et al., 1988). Multiple effector cells are involved because malaria has a complicated life cycle using several different host organs and tissues. Sporozoites released from the mosquito salivary gland enter the bloodstream and ultimately infect hepato-cytes. After sexual reproduction in the liver, merozoites are released and infect circulating red blood cells. The merozoites mature in the red cell becoming tropozoites. When the red blood cell ruptures, the merozoites infect other red cells. Some merozoites mature in gametocytes transmitted to a feeding mosquito. Sexual reproduction of the gametocytes occurs in the mosquito.

Several different strains of Plasmodium have been used in model systems. Lethal (Pyl7L) and nonlethal (Pyl7NL) strains of Plasmodiumyoelii are commonly used in model mouse systems. Susceptibility to these organisms is influenced by genes within and outside of the MHC complex. C57BL/6 mice are susceptible to infection and DBA/2 mice are resistant (Sayles and Wassom, 1988).

Less frequently used in assays are Plasmodium chabaudi and P. adami that produce nonlethal infections in some strains of mice. R chabaudi is lethal for BALB/c and C3H/HeJ mice but C57BL/6, C57L, DBA/2, and B 10.A are resistant to infection (Stevenson et al., 1982).

Plasmodium berghei initiates a self-limiting disease in most mouse species. However, peak parasitemia, anemia, and other hallmarks of infection differ with the mouse strain (Eling etal., 1977). Host resistance to this strain ofplasmodium is a result of T and B cells and macrophage activation (Bradley and Morahan, 1982).

Usually, 106 parasitized red cells are inoculated in naive hosts via intra-venousor intraperitoneal routes.Bloodsamplesare takenat days 10, 12, and 14. In most mouse strains, peak parasitemia occurs at day 12. Endpoints include the number of parasites in the blood, the number of parasitized red cells, and the number of red cells in the circulation (Dockrell and Playfair, 1983). These measurements are made manually or by flow cytometry (Whaun et al., 1983).

B16F10 Melanoma. Host defense to melanoma is mediated by NK cells, macrophages, and some T cells (Parhar and Lala, 1987). Using an artificial tumor metastasis model in syngeneic C57BL/6 mice, mice are challenged with different number oftumor cells (Fidler etal., 1978). Because cells are injected into the tail vein, the cells lodge in the lung, which is the first capillary bed encountered. Two endpoints are usually measured. The proliferation rate of cells in the lung can be measured by injecting fluorodeoxyuridine, which blocks de novo nucleotide synthesis, followed by [125I]iododeoxyuridine. The latter nucleotide is incorporated into the DNA of newly divided cells (White, 1992). As a second endpoint, the number of tumor nodules per lung can be determined. Tumor cells form characteristic black nodules on a background of white or yellow lung tissue. Up to 200-250 nodules can be enumerated in each lung section.

Immune responses to the melanoma tumor are time dependent. IL-2 production and an increased NK cell cytolysis occur early in the tumor response (Makovic and Murasko, 1991). The intermediate response consists of macrophages and T cells. Macrophages release TNF that kills tumor cells and inhibits tumor growth (Loveless and Munson, 1981). T cells release lymphokines that initiate proliferation of cytotoxic T cells. Cytotoxic T cells participate in the late responses to the tumor.

PYB6 Fibrosarcoma. This model was originally developed as a screening tool for immunosuppression. The tumor produces a vigorous NK and cytotoxic T cell response (Urban et al., 1982). In the assay, tumor cells are introduced subcutaneously in the thigh and observed for 42 days. Endpoints include the incidence of tumors, time to tumor development, and tumor size.

MAD106. This mammary adenocarcinoma, which is syngeneic in the Fischer 344 rat, is the only rat tumor model. The host response consists of NK and LAK cells (Barlozzari etal., 1985). Following the intravenous administration of tumor cells, the time to death is usually 3 weeks.

How To Prevent Skin Cancer

How To Prevent Skin Cancer

Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.

Get My Free Ebook


Post a comment