Testing Human Biologicals in Animal Host Resistance Models

Abstract

The purpose of immunotoxicity testing is to obtain data that is meaningful for safety assessment. Host resistance assays are the best measure of a toxicant's effect on the overall ability to mount an effective immune response and protect the host from infectious disease. An outline is presented for immunotoxicological evaluation using host resistance assays. The influenza virus host resistance model is useful to evaluate the overall health of the immune system and is one of the most thoroughly characterized host resistance models. Viral clearance requires all aspects of the immune system to work together and is the ultimate measure of the health of the immune system in this model. Mechanistic immune functions may be included while measuring viral clearance and include: cytokines, macrophage activity, natural killer (NK) cell activity, cytotoxic T-lymphocyte (CTL) activity, and influenza-specific IgM and IgG. Measurement of these immunological functions provides an evaluation of innate immunity (macrophage or NK activity), an evaluation of cell-mediated immunity (CMI) (CTL activity), and an evaluation of humoral-mediated immunity (HMI) (influenza-specific IgM or IgG). Measurement of influenza-specific IgM or IgG also provides a measurement of T-dependent antibody response (TDAR) since influenza is a T-dependent antigen. There are several targeted host resistance models that may be used to answer specific questions. Should a defect in neutrophil and/or macrophage function be suspected, Streptococcus pneumoniae, Pseudomonas aeruginosa, or Listeria monocytogenes host resistance models are useful. Anti-inflammatory pharmaceuticals or therapeutics for rheumatoid arthritis or Crohn's disease that target TNFα may also be evaluated for immunotoxicity using the S. pneumoniae intranasal host resistance assay. Marginal zone B (MZB) cells are required for production of antibody to T-independent antigens such as the polysaccharide capsule of the encapsulated bacteria that are so prominent in causing blood-borne infections and pneumonia. Intravenous infection with Streptococcus pneumoniae, an encapsulated bacterium, results in a blood-borne infection that requires MZB cells for clearance. The systemic S. pneumoniae host resistance assay evaluates whether a therapeutic test article exerts immunotoxicity on MZB cells and measures the T-independent antibody response (TIAR). Suppression of CMI or in some cases HMI may result in reactivation of latent virus that may result in a fatal disease such as progressive multifocal leukoencephalopathy (PML). The murine cytomegalovirus (MCMV) reactivation model may be used to evaluate a pharmaceutical agent to determine if suppression of CMI or HMI results in reactivation of latent virus. Candida albicans is another host resistance model to test potential immunotoxicity. Host resistance assays have been the ultimate measure of immunotoxicity testing for environmental chemicals and pharmaceutical small molecules. Human biologicals are now an important component of the drug development armamentarium for biotech and pharmaceutical companies. Many human biologicals are fusions of IgG, and/or target immune mediators, immunological receptors, adhesion molecules, and/or are indicated for diseases that have immune components. It is therefore necessary to thoroughly evaluate human biological therapeutics for immunotoxicity. Numerous biologicals that are pharmacologically active in rodents can be evaluated using well-characterized rodent host resistance assays. However, biologicals not active in rodents may use surrogate biologicals for testing in rodent host resistance assays, or may use host resistance assays in genetically engineered mice that mimic the effect of the human biological pharmacological agent.

INTRODUCTION

In immunotoxicity safety testing, the major objective is to determine the significance of an immunotoxic effect with respect to increased susceptibility to infectious or neoplastic disease (Gleichmann et al., 1989). Screening assays include data normally obtained in standard toxicology tests, as well as splenic cellularity, splenic differentials, and immunophenotyping to determine the population size and/or numbers of B-lymphocyte, T-lymphocyte, T-lymphocyte subsets, NK cells, and monocytes. Screening assays are useful in allowing elimination/selection of candidate compounds early in development. However, standard immunophenotyping performed with toxicology studies do not usually include marginal zone B (MZB) cell markers unless something is already known that indicates further investigation (i.e., drug target) and then standard immunophenotyping would be customized. Histopathological evaluation may also detect loss of MZB cells, especially if the drug target suggests a possible effect.

Functional assays should be considered surrogates. These assays measure not whether the cell type is present or absent, but whether the cell type is functional. Functional assays measure but one component of the total immunological response and include: (1) non-specific or innate immunological determinations such as NK cell activity, and macrophage activity (phagocytosis and respiratory burst); and, (2) evaluation of specific acquired immunity including (a) T-dependent assays to quantify antibody responses to keyhole limpet hemocyanin (KLH), sheep red blood cells (SRBC), or influenza virus to evaluate humoral immunity. The CMI response has been measured by (a) CTL activity assay or (b) delayed-type hypersensitivity (DTH). It should be noted that although DTH is a cell-mediated immune response, DTH and CTL are not synonymous (Yap and Ada, 1978). Class I-restricted DTH-mediating T-lymphocytes generated in mice infected with virus convey protection (Leung and Ada, 1981). However, Class II-restricted DTH-mediating T-lymphocytes generated by immunization with virus augments disease (Leung et al., 1980). Therefore, T-lymphocytes exhibiting DTH in viral disease measures both T-lymphocytes that are beneficial and those associated with immunopathology and a decreased DTH may either be adverse or beneficial.

The major function of the immune system is protection from infectious or neoplastic disease and most immunotoxicologists regard host resistance assays to be the most relevant for both validating the usefulness of other detection methods and for extrapolating the potential of a substance, drug or chemical, to alter host susceptibility in the human population (Germolec, 2004). It is not known what degree of change in a screening or functional assay constitutes an adverse result, i.e., increased susceptibility to infectious or neoplastic disease. There is no formula to determine how large a decrease in a functional assay is reflected in an effect on bacterial or viral clearance. This uncertainty complicates assessment of immunotoxicity. However, dosing with a test article and a resultant decrease in clearance of an infectious agent in a host resistance model is unambiguously adverse and by definition the test article is immunotoxic. The major objective of immunotoxicity safety testing is to determine the significance of an immunotoxic effect with respect to increased susceptibility to infectious or neoplastic disease (Gleichmann et al., 1989). Furthermore, host resistance models provide the only means to directly assess the functional reserve of the immune system.

Host Resistance Models in Rodents

Well-characterized host resistance models exist for viral, bacterial, fungal, parasitic, and neoplastic diseases. Immunologists have used these infectious or neoplastic disease models to study immunological functions associated with clearance of the challenge agent; by immunopharmacologists to evaluate therapeutic or prophylactic anti-viral, anti-parasitic, anti-bacterial agents, anti-fungal, anti-tumor agents and biological response modifiers; and by immunotoxicologists to evaluate the immunotoxicity of a test article by quantifying the adverse effect on disease susceptibility.

Immunological clearance of the infectious challenge agent is a more sensitive and meaningful measure of immunological function (Lebrec and Burleson, 1994; Burleson, 1995; Selgrade et al., 1998) than mortality. The number of infectious particles per organ or per gram of organ is quantified. Challenging the immune system with an extremely virulent or with an extremely high titer of infectious agent may overwhelm the immune system with death occurring before development of the cascade of immunological responses required for clearance. Challenge with a highly virulent or with a high titer of infectious agent may reflect a model of sepsis or result in a “cytokine storm.” Titer does not necessarily correlate with mortality; that is, similar titers of virus were reported in the lungs of mice infected with either the mouse-adapted lethal influenza A/Hong Kong/8/68 virus or the mouse-adapted nonlethal influenza A/Port Chalmers/1/73 virus (Lebrec and Burleson, 1996). Viral titers also did not correlate with mortality in studies evaluating the immunotoxicity of TCDD (Burleson et al., 1996).

GENERAL HOST RESISTANCE MODELS

Influenza Virus Host Resistance Model

Host resistance assays may be employed as a first approach or used as a second approach if immunotoxicity is observed in screening and/or immune function assays. However, host resistance assay is recommended if the test article pharmacological target is designed to alter a component of the immune system. The most thoroughly characterized host resistance model is the mouse or rat influenza host resistance model (Table 1).

TABLE 1 Host resistance model to test the overall health of the immune system. Mechanistic endpoints may or may not be included

Influenza Virus Host Resistance Model

▪ Cytokines

▪ Interferon Activity

▪ Macrophage Activity

▪ NK cell activity

▪ CTL activity

▪ Influenza-specific IgM, IgG (IgG1 and IgG2a) – TDAR

▪ Immunophenotyping

▪ Histopathology

Measurement of viral clearance is a measure of the overall health of the immune system since clearance of virus requires an intact and functional immune system that incorporates a cascade of immune responses including: innate immunity (cytokine production, NK cell activity, and macrophage activity) as well as acquired or adaptive immunity (CTL activity and antibody production) (Burleson, 1995; Burleson and Burleson, 2007). Immunotoxicity caused by a test compound will be reflected in an impaired clearance of the infectious agent. Numerous influenza viruses have been used for immunotoxicity testing and include Influenza A/PR8/34 (H1N1), Influenza A/Taiwan/1/64 (H2N2), Influenza A/Aichi (H3N2), Influenza/A HKx31 (H3N2), Influenza A/Hong Kong/8/68 (H3N2)and Influenza A/Port Chalmers/1/73 (H3N2) (reviewed by Burleson and Burleson, 2007). Mouse-adapted influenza A/Port Chalmers/1/73 (H3N2) (Mouse-adapted Influenza Virus -MAIV) has been used for numerous immunotoxicity studies (Burleson, 1995; Burleson and Burleson, 2007).

The typical influenza virus host resistance assay involves a 28-d repeat dose study. Mice are dosed with the test agent 28 consecutive days (7 days pre-infection and 21 days post-infection). Groups of mice are sacrificed and infectious virus measured for viral clearance on Days 2, 6, 8, 10, and 21. Dexamethasone is used as a positive immunomodulatory control. Immunological function assays are monitored at selected times after infection. Influenza virus is a T-dependent antigen and formation of antibody to influenza virus requires functional T-lymphocytes, B-lymphocytes, and macrophage antigen processing and presentation activity, and is thus a T-dependent antibody response (TDAR). The influenza model can be used to evaluate clearance as an overall measure of immunological health and mechanistic immune evaluations may or may not be included. The mouse and rat influenza models have been used to evaluate the immunotoxicity of numerous pharmaceuticals with the following representative publications (Cowan et al., 2002; Steele et al., 2005; Olivier et al., 2007; Miller et al., 2007; Zhu et al., 2007).

Targeted Host Resistance Models

The influenza model in mice or rats is best used to evaluate the overall health of the immune system, i.e., how the numerous components of the functional immune system work together to clear an infection. Targeted host resistance assays are also available to evaluate specific immunotoxicity questions (Table 2). These host resistance models answer specific questions concerning the immune system.

1. Streptococus pneumoniae Pulmonary Host Resistance Model

   • Therapeutics affecting neutrophils/macrophages may be evaluated for immunotoxicity using the Streptococcus pneumoniae pulmonary host resistance assay. This assay has been used in mice and rats to produce a pulmonary infection following intranasal infection (Gilmour et al., 1993; Gilmour and Selgrade, 1993; Burleson, personal communication). S. pneumoniae or pneumococcal pneumonia is the most common cause of bacterial infections acquired outside of hospitals and pneumococcal pneumonia accounts for 25–30% of all community-acquired pneumonia, and causes an estimated 40,000 deaths annually (Centers for Disease Control and Prevention, 1997). Rodent models for bacterial pneumonia can be used to evaluate immunotoxicity that may predispose to bacterial pneumonia. Macrophages were demonstrated to be important in the clearance of streptococci from the lungs of mice (Gilmour et al., 1993) and in mice and rats (Gilmour et al., 1993; Gilmour and Selgrade, 1993). Further studies by Gilmour and Selgrade (1993) demonstrated the importance of neutrophils in pulmonary streptococcal disease in rats by pretreatment with anti-neutrophil antibody. Cytokines may be measured in the lung as well as in the serum. Bacterial titers and bacterial clearance are quantified as the number of colony forming units (CFU) per organ or per gram of tissue. Bacterial titers are determined at early timepoints (1, 4, and 24 hr) and evaluate early innate immunity before the specific acquired immune responses have developed. Macrophage and/or neutrophil function assays could be measured as a mechanistic probe if an effect on bacterial clearance is observed. However, the conclusive observation is bacterial clearance. The S. pneumoniae host resistance model in mice has been used in immunotoxicity evaluations and was reported as one of a battery of three host resistance assays to evaluate a small molecule therapeutic targeted for splenic tyrosine kinase (Syk) (Zhu et al., 2007). The Streptococcal host resistance model in rats has also been used for immunotoxicity evaluation (Steele et al., 2005).

   • Model for Evaluating Anti-Inflammatory Agents. The S. pneumoniae host resistance model has been well characterized in mice and rats. Animals are infected intranasally and bacterial clearance measured. Bacterial clearance is evaluated by determining the number of CFU per gram of lung tissue or per lung. Dexamethasone is used as a positive immuno-modulatory control as it has a suppressive effect on innate immunity and delays bacterial clearance. Komocsar et al. (2007) used the S. pneumoniae pulmonary host resistance model in Lewis rats to assess the effects of anti-inflammatory agents on innate immunity. The model was able to predict potential drug suppression of the innate immune response to S. pneumoniae. The ability to rank order the severity of innate immune suppression with multiple test articles in the same study demonstrates the usefulness of this model for screening potential drug candidates.

   • Model for Therapeutics Targeting TNFα. The S. pneumoniae pulmonary host resistance model is also valuable for evaluating the importance of macrophage cytokines on bacterial host resistance. Pharmaceuticals targeting inhibition of TNFα have been used to treat inflammatory diseases such as rheumatoid arthritis, ankylosing spondylitis, psoriasis, and Crohn's disease. Decreased TNFα as a result of treatment with monoclonal antibody (MAb) to TNFα has an effect on several biomarkers of infection. TNFα plays an essential role in preventing reactivation of persistent tuberculosis (Mohan et al., 2001) as well as being important in clearance of Pseudomonas aeruginosa (Gosselin et al., 1995). Studies have reported that treatment of mice with MAb to TNFα results in a decreased level of TNFα in the lungs and serum, decreased level of neutrophils, and an increased bacterial titer with decreased survival in mice infected intranasally with S. pneumoniae (Takashima et al., 1997; van der Poll et al., 1997; Benton et al., 1998; O'Brien et al., 1999). The streptococcal pulmonary host resistance model is thus an important means to assess the functional immunological capacity of macrophages and neutrophils as well as macrophage cytokines and has been used in both mice and rats to assess immunotoxicity of pharmacological agents (Burleson, personal communication). This host resistance assay may be used to choose a lead compound causing the least degree of immunosuppression among candidates with similar anti-inflammatory potency.

2. Marginal Zone B (MZB) Cell Host Resistance Model

   Bacteria encapsulated with a polysaccharide capsule such as S. pneumoniae or Haemophilus influenzae are important blood-borne pathogens that present a different challenge to the immune system. Capsular polysaccharide antigens are thymus-independent type 2 antigens (TI-2) [Mond et al., 1995) and effective immune responses are highly dependent on the presence of a functional marginal zone (Amlot et al., 1985; Harms et al., 1996; Guinamard et al., 2000). Capsular antigens therefore stimulate a T-independent antibody response (TIAR). The MZB cell model in mice or rats measures bacterial clearance, hematology, cytokine production, and antibody production in a kinetic fashion over a 14-d period after intravenous infection to create a blood-borne infection. It is clear that MZB cells in both humans and rodents are considered a critical host defense mechanism directed against encapsulated blood-borne pathogenic microorganisms. Therefore, immunotoxicity directed against MZB cells not only decreases protection against blood-borne pathogens but also results in a depletion of immunological memory. In summary, T-independent antibody responses (TIAR) are decreased or ablated as a result of MZB cell immunotoxicity.

   Histopathology will detect defects in the splenic marginal zone and special immunophenotyping markers can be included to detect alteration in the number of MZB cells. Should an effect on MZB cells be observed, the pharmaceutical agent should be evaluated in the Streptococus pneumoniae systemic MZB host resistance model for encapsulated bacteria.

3. Pseudomonas aeruginosa Host Resistance Model

   Pseudomonas aeruginosa is a Gram-negative bacillus that is a human pathogen and primarily causes diseases of the urinary tract, septicemia, abscesses, corneal infections, meningitis, bronchopneumonia, subacute bacterial endocarditis, and in burn patients. Pneumonia is one of the ten leading causes of death including community-acquired pneumonia and hospital-acquired pneumonia. Treatment often fails and the mortality rate in Pseudomonas septicemia has been reported to be greater than 80%. P. aeruginosa is used as a pulmonary bacterial host resistance model to evaluate the immunotoxicity of therapeutics when an immunotoxic effect is suspected in neutrophils, macrophages, and/or TNFα (Gosselin et al., 1995). TNFα also is important in bacterial clearance of Streptococcus pneumoniae, and plays an essential role in preventing reactivation of persistent tuberculosis (Mohan et al., 2001).

4. Listeria monocytogenes Host Resistance Model

   Listeria monocytogenes, a Gram-positive bacillus, is a host resistance assay that previously measured mortality (Bradley, 1995) or bacterial clearance in the liver and spleen (Burleson, personal communication). The Listeria monocytogenes infection is useful primarily to determine xenobiotic effects on neutrophils and Kupfer cells of the liver and splenic macro-phages and neutrophils. NK cells and T-lymphocytes also play a role in bacterial clearance. The Listeria monocytogenes host resistance model has been used to evaluate MAbs directed against an adhesion molecule resulting in decreased neutrophils to determine if the pharmaceutical would enhance disease susceptibility to Listeria and therefore predict whether this anti-inflammatory therapeutic approach would enhance susceptibility to opportunistic infections in humans. Treatment to inhibit the diapedesis of monocytes, macrophages, and neutrophils into the inflammatory lesions in the liver and spleen resulted in an otherwise sublethal listeria inoculum that grew unrestricted within the spleen and liver and caused death in 3 d (Rosen et al., 1989; Conlon and North, 1991) and decreased clearance of bacteria in the liver and spleen (Burleson, personal communication).

5. Candida albicans Host Resistance Model

   Candida albicans has been used to test for immunotoxicity when immunosuppression is suspected of affecting resistance to a fungal infection. Candida albicans is used to infect mice intravenously and fungal clearance is monitored by quantitatively measuring the number of colony forming units (CFU) in the liver and spleen.

6. Latent Viral Reactivation Host Resistance Model

   Murine cytomegalovirus (MCMV) and rat cytomegalovirus (RCMV) are well-characterized models for CMV disease in humans. MCMV and RCMV have been used as host resistance models for evaluating immunotoxicity by measuring the effect on clearance of the virus in a primary viral infection (Goettsch et al., 1994; Garssen et al., 1995; Selgrade et al., 1995, 1998; Van Loveren, 1995; Ross et al., 1996, 1997). These viruses cause a primary infection with infectious virus detectable in a variety of organs (lung, liver, spleen and salivary gland). The virus is cleared by immunological mechanisms following primary viral replication.

   The virus can remain in a nonproductive latent state for a lifetime (Jordon et al., 1977; Shanley et al., 1979) unless viral reactivation or recrudescence occurs following intended immunosuppression in transplant patients or unintended immunosuppression in individuals receiving immunotoxic therapeutics, or in individuals immunosuppressed by infection with HIV or idiopathic immunosuppression. Reactivation of latent viral infection is a serious problem for immunosuppressed individuals and may involve latent CMV, herpes simplex virus (HSV), or either of the human polyoma viruses termed BK virus and JC virus. BK virus and JC virus were isolated in 1971, BK virus from the urine of a renal allograft patient undergoing immunosuppressive therapy and JC virus from brain tissue of a patient with PML. The majority of individuals are infected in childhood with BK virus, JC virus, CMV and HSV (Alford, 1990; Eckhart, 1990; Whitley, 1990) and those with latent viral infections are susceptible to reactivation upon immunosuppression.

   The MCMV latent viral model is an excellent model to assess reactivation of latent viral disease as a result of immunosuppression. There are many similarities between the viruses responsible for latent /reactivated viral disease. CMV and HSV belong to the Herpesviridae virus family while BK virus and JC virus belong to the Papovaviridae virus family. All these viruses have double stranded DNA (the human polyoma viruses are circular); all are ubiquitous in the human population; all cause mild primary infections followed by a latent viral infection; and immunosuppression, especially a suppressed CMI results in reactivation of latent viral infection.

   The MCMV reactivation model may be used for a pharmaceutical agent that causes immunosuppression of CMI or HMI that may result in reactivation of latent virus, which may cause severe or fatal disease such as PML. Lymphocyte depletion studies revealed a hierarchy of immune control functions of CD8⁺, NK, and CD4⁺ cells. Reactivation was rare if only one cell type was depleted, but was evident after concurrent depletion of another cell type (Polić, 1998). This study demonstrates immunological reserve since depleting one immunological function such as NK activity reactivates virus in 5.6% of the mice, depletion of either CD4 and NK or CD8 and NK activity reactivates either 25 or 80% of the mice, while abrogation of CD4, CD8, and NK activity reactivates 100% of the mice (Jonjić, 1994; Polić, 1998).

   The use of B-lymphocyte-deficient mice for this model increases the sensitivity of detecting reactivated virus by 100 to 1,000-fold (Polić, 1998). Suppression of HMI results in an increased sensitivity to reactivated virus disease (Polić, 1998) and perhaps acts in a primary or secondary manner in the induction of reactivation disease. The United States FDA issued an alert concerning spontaneous fatal PML due to viral reactivation of JC polyoma virus following reports of two patients with systemic lupus erythematosus (SLE) who had received treatment with Rituximab,® a monoclonal antibody directed against CD20 on B-lymphocytes (Fox, 2007).

   Natalizumab, a monoclonal antibody against α 4 integrins, has also been associated with the development of PML from reactivation of latent JC virus infection (Yousry et al., 2006). It is important to test therapeutics that suppress CMI using a viral reactivation model. Therapeutics that suppress the HMI may also have the potential to cause recrudescence of latent viral disease (reviewed by Burleson and Burleson, In Press).

TABLE 2 Targeted host resistance models for evaluation of immunotoxicity

Targeted Host Resistance Models:
1) Streptococcus pneumoniae pulmonary model for innate immunity:
Evaluation of therapeutics affecting neutrophils and/or macrophages
Anti-inflammatory therapeutics: Evaluate in Streptococcus pneumoniae pulmonary model for innate immunity
Therapeutics targeting TNFα : Evaluate in S. pneumoniae pulmonary model for innate immunity
2) Marginal Zone B (MZB) cell model:
Systemic Streptococcus pneumoniae host resistance model to evaluate MZB cells
3) Pseudomonas aeruginosa pulmonary host resistance model:
Neutrophils
4) Listeria monocytogenes systemic host resistance model:
Intracellular bacterial model for evaluation of liver and splenic macrophages and neutrophils
5) Candida albicans host resistance model:
Fungal host resistance model
6) Murine cytomegalovirus (MCMV) host resistance model:
Primary viral clearance host resistance model
Latent virus reactivation model

HUMAN BIOLOGICALS AND HOST RESISTANCE MODELS

Human Biologicals with Activity in Rodent Models

Human biologicals with activity in rodent models can be evaluated for immunotoxicity using host resistance assays. One example is TACI, a transmembrane activator, calcium-modulator and cyclophilin ligand-interactor. Recombinant TACI-Ig is produced as a soluble human receptor IgG₁-F_c fusion protein receptor antagonist of soluble and membrane bound BLyS and APRIL. BLyS is a B-lymphocyte stimulator and APRIL is a proliferating-inducing ligand. Treatment with TACI-Ig reduces circulating B-lymphocytes in mice and non-human primates and is in clinical evaluation for treatment of various autoimmune diseases and B-lymphocyte malignancies. TACI-Ig has been evaluated using the mouse influenza host resistance model (Roque et al., 2006). TACI-Ig treatment resulted in a dose-dependent decrease in spleen weight and influenza-specific IgM and IgG in both the lungs and serum compared to vehicle-treated animals.

Therefore, the T-dependent antibody response (TDAR) to influenza was suppressed. Flow cytometric analysis showed a decrease in B-lymphocytes, but not T-lymphocytes, in peripheral blood as a result of TACI-Ig treatment. However, there was no effect on viral clearance. This study demonstrates the importance of immunological reserve, since the TDAR response was suppressed, as expected based on the immunopharmacology of the biological therapeutic, but the remaining immunological functions were intact and sufficient for viral clearance.

Surrogate Biologicals

For monoclonal antibodies (MAbs) or human biologicals that do not have biological activity in rodents, parallel development of the human biological in rats or mice will facilitate toxicology evaluations, including immunotoxicological evaluations, using well-established and validated methodologies with sufficient numbers of test subjects required for the statistical power to detect immunotoxicity. Integrins are cell-surface receptors that are important in the trafficking, activation, and retention of lymphocytes, monocytes, and granulocytes within inflamed tissues (Springer, 1994). Each integrin consists of an α -subunit and a β -subunit. The α₁β₁ integrin (VLA-1 or very late antigen-1) is involved in inflammation and is involved in extracellular matrix (ECM)-binding as well as leukocyte attachment to endothelial cells and extravasation of cells into tissues. Parallel development of a surrogate protein to the human MAb that targets inhibition of the murine α₁β₁ integrin was used in immunotoxicity testing. The surrogate hamster anti-mouse α₁β₁ integrin monoclonal antibody (HA31/8) was tested in the mouse influenza host resistance model (Olivier et al., 2007). HA31/8 had no effect on influenza viral clearance and was therefore not immunotoxic in this host resistance model. HA31/8 also had no effect on influenza-specific IgG TDAR response in this host resistance model.

The Listeria monocytogenes host resistance model has been used to evaluate MAbs directed against CD11b to determine whether inhibition of this adhesion molecule would enhance disease susceptibility. CD11b/CD18 (Mac-1) is a leukocyte integrin that plays a critical role in neutrophil adhesion and the initiation of acute inflammatory processes and is therefore a therapeutic anti-inflammatory target. CD11b (α M integrin) complexes with CD18 (β 2 integrin) to form complement receptor type 3 (CR3) heterodimer. Treatment of mice with either of two surrogate biological MAbs designated NIMP-R10 or 5C6, both directed against CD11b resulted in decreased clearance of listeria in the liver and spleen with increased mortality. Neutrophils and monocytes were decreased and mice were unable to control the infectious intracellular bacterial disease (Rosen et al., 1989; Conlon and North, 1992; Burleson, personal communication).

Regulation of osteoclast differentiation is mediated by the receptor activator of NF-κ B (RANK) ligand (RANKL; a member of the TNF superfamily of ligands) and two receptors, osteoprotegrin (OPG) and RANK (reviewed by Miller et al., 2007). RANK-Fc was evaluated for immunotoxicity using the influenza host resistance model with dexamethasone as a positive immunomodulatory control. rRANK-F_c is a fusion protein containing aa 1-213 of the murine RANK extracellular domain with the C terminus of the Fc domain of murine IgG₁. The murine surrogate biological fusion protein RANK-Fc was not immunotoxic in the mouse influenza host resistance model since there was no effect on clearance of infectious virus. Furthermore, there was no effect on influenza-specific IgG and no effect on IL-1β, IL-6, or TNFα (Miller et al., 2007).

Genetically Engineered Mice

Advanced techniques in molecular biology have allowed sophisticated genetic engineering of mice to develop animal models of disease, to study biomarkers of disease, validate drug targets in vivo, probe the mechanism of diseases, and model a drug before it is made. For example, a genetically engineered knockout mouse could be useful if it has a target that is genetically deleted and analogous to the intended target of the pharmacological agent. In knockout mice, a single gene is heterozygously or homozygously deleted or rendered defective by genetic engineering techniques. Knockin mice contain an inserted transgene bearing a mutation of interest, flanked by sequences that are homologous to the endogenous locus. A transgenic mouse is created by the introduction of an isolated gene (the transgene) into the genome of a whole mouse embryo. The transgene is expressed along with the recipient's own genes. Expression of a transgene may be controlled either by its natural transcriptional promoters and enhancers, or by inserted exogenous regulatory elements. These genetically engineered mice have been used to study autoimmune diseases, critical cells of the immune system, and resistance to intracellular and extracellular bacteria, viruses, and immunological functions associated with the deletion of particular cell types (Arbeit and Hirose. 1999). For example, knockout mice have been used to further understand the immunotoxicity of ablated TNFα. As described previously, the lack of TNFα or its receptors results in mice more susceptible to S. pneumoniae infections (O'Brien et al., 1999; Wellmer et al., 2001).

Humanized Mice

Humanized mice, or mouse-human chimeras, are immunodeficient mice engrafted with hematopoietic cells or tissues that allow the in vivo study of human cells, tissues and organs. Further development of these models, and their eventual availability at a reduced cost, will allow targeted testing of pharmaceutical agents on the human immune system. Future generations of humanized mice will be powerful tools in pre-clinical testing and in the investigation of human biological processes (Shulstz et al., 2007). The new generation humanized mice have built on previous mouse strains, including (a) nude mice that are homozygous for a mutation in the forkhead box N1 gene that results in absence of hair and impaired thymic development of mature T-lymphocytes, (b) SCID-hu mice (severe combined immunodeficiency mice) that were engrafted with human fetal liver and thymus tissue; and (c) Hu-SRC-SCID mice that are SCID mice exposed to sublethal irradiation and injected with human hematopoietic stem cells. There are currently limitations on the use of humanized mice with modifications and improvements required before their immense promise is achieved.

Non-human Primates

Non-human primates have gained importance in vaccine testing and evaluation of immunogenicity of biologicals. More recently, immunotoxicity evaluations using screening and functional assays in non-human primates have gained popularity for the testing of biologicals, based on the close phylogenetic relationship of primates and humans. However, while non-human primates may respond similarly to humans when given biological therapeutics, there are drawbacks to their use. These include: (a) expense and lack of availability, (b) limited, often inadequate, group sizes that may preclude a statistically valid interpretation of results (c) the outbred status of the non-human primates that magnifies the variability already inherent with the small group size, (d) lack of validation of immune function assays in non-human primates, and (e) omission of the required positive immunomodulatory controls in order to confirm negative results.

DISCUSSION

Immunotoxicology as a scientific discipline began in the mid to late 1970s with the early efforts toward establishing the immune system as a target organ for xenobiotic-induced damage, developing methods to best detect immunotoxicity, and utilizing data in safety assessment (reviewed by Burleson and Dean, 1995). The overwhelming majority of immunotoxicology studies in the past 25 years of research have been performed in the mouse and rat. Greater than 95% of the principles of immunology were discovered in studies using the mouse and most appear to be applicable to human immunology. Rodent models exist for nearly every human disease. New immunological and molecular biology approaches such as PCR and RT-PCR, gene arrays, and genetic susceptibility markers using single nucleotide polymorphisms (SNPs) (Yucesoy et al., 2001a and b) are being used to better evaluate alterations to the immune system.

Many human biologicals target components of the immune system. Biologicals that are pharmacologically active in rodents can be evaluated using well-characterized rodent host resistance assays. Surrogate biologicals are often prepared in tandem with the human biological and can be used for testing in rodent host resistance assays. Biologicals not active in rodents may use genetically engineered mice for host resistance assays that mimic the effect of the human biological pharmacological agent.

The purpose of immunotoxicity testing is to obtain data that is meaningful for immunotoxicity safety assessment. Host resistance models are the ultimate predictor of adverse effects on the immune system (Germolec, 2004) and as such are important in the safety assessment of pharmaceuticals. Host resistance assays are the most relevant for both validating the usefulness of other detection methods and for extrapolating the potential of a substance, drug or chemical, to alter host susceptibility in the human population (Germolec, 2004). It is not known what percent decrease in a screening or functional assay constitutes an adverse result, i.e., increased susceptibility to infectious or neoplastic disease. This uncertainty creates a problem for an immunotoxicity safety assessment. However, dosing with a test article and a resultant decrease in clearance of an infectious agent in a host resistance model is unambiguously adverse and by definition demonstrates immunotoxicity of the test article in the host resistance assay used for the evaluation. The major objective of immunotoxicity safety testing is to determine the significance of an immunotoxic effect with respect to increased susceptibility to infectious or neoplastic disease (Gleichmann et al., 1989). Host resistance models provide the only means to directly assess the functional reserve of the immune system.

Immune function assays are an important aspect of immunotoxicity safety assessment. Measurement of immune function aids in the mechanistic understanding of immunotoxicity. Mechanistic immune function testing may be performed in one of three possible testing strategies. (1) Immune function testing may be performed as the first immunotoxicological evaluation. In this case, a defect in a particular immune cell type may direct subsequent studies to a targeted host resistance model. For example, if a 30% inhibition of respiratory burst function induced by compound XYZ is demonstrated by in an in vitro assay, what is the true biological importance of this inhibition? A targeted host resistance assay is required. Another example: histology reports indicate a drug-induced loss of splenic marginal zone morphology. The question becomes “Does this loss of splenic marginal zone region have an effect on the overall immune response to a blood-borne infection? (2) A second testing strategy is to do mechanistic immune function tests concomitant with a general host resistance assay. In this scenario, an effect on host resistance will have an immunological basis and understanding of how the pharmaceutical agent affected clearance of an infectious agent. (3) The third testing strategy involves an initial evaluation of the pharmaceutical agent in a host resistance assay. Immune function evaluations are then scheduled if an effect on clearance of an infectious agent is observed. The testing strategy choice is dependent on the drug development timeline, budget and company preference.