Immunotoxicity Testing in Non-Rodent Species

Abstract

Evaluation of the immunotoxicity potential of some pharmaceuticals, including immunomodulatory chemicals and biologics, cannot be limited to testing in rodents. Thus, immune function tests have also been applied in studies with non-human primates and more recently dogs that assess various components of the immune system. These assays include TDAR responses with various immunogens, lymphocyte phenotyping, natural-killer cell activity, delayed-type hypersensitivity, and macrophage function assays. Approaches for incorporating immune function testing in non-rodent species, results from these tests, their interpretation and limitations with respect to drug safety assessment will be reviewed.

Keywords :

• primate
• dog
• immune function
• immunotoxicity

INTRODUCTION

Increased expectations from a number of regulatory agencies, as summarized in the ICH S8 Guidance document (FDA, 2006), have resulted in increased attention to the importance of determining the potential adverse effects of pharmaceuticals on the immune system. Hazard assessment for immunotoxicology is primarily conducted in rats or mice, However, the evaluation of the immunotoxic potential of some pharmaceuticals, including immunomodulatory chemicals and biologics, cannot be limited to testing in rodents. Many protein therapeutics and even some small molecules are either not active or significantly less active in rodent species. Furthermore, the immunogenicity of human protein therapeutics can sometimes be minimized by testing in non-human primates. In addition, drug-induced immune-related findings may be observed that are limited to a non-rodent species. In these cases, the non-rodent species may be a more relevant model and thus methodologies that assess various components of the immune system have also been applied in studies with non-human primates and, more recently, dogs. This manuscript will review the current practice for assessing immunotoxicity in the non-rodent species based on personal experience and interaction with other immunotoxicologists in the pharmaceutical industry.

Non-Humans Primates and Dogs as Models for Immunotoxicity Assessment

Since many monoclonal antibodies directed against epitopes on human proteins (e.g., cytokines and cell-surface markers) cross-react with corresponding proteins of non-human primate, the development of assays to assess the competency of the immune system of non-human primates has advanced at a much more rapid rate than that for other non-rodent species, such as the dog. These antibodies have allowed for the analysis of a variety of immune cells, the receptors they express, and the cellular mediators that they produce (such as cytokines and immunoglobulins) that are involved in many immunologic reactions. Over time, antibodies have also been raised against monkey proteins, for which cross-reactivity has not yet been demonstrated. Thus, the non-human primate is often the species of choice to assess the effect of a pharmaceutical on immune function. Clearly, their is a need for more antibodies for canine cell-surface markers and cytokines, so that assays to assess immune function in the dog may be developed.

There are a variety of assays that have been used to assess the immune system of non-human primates and they include, but are not limited to, immunophenotyping, serum cytokine profiling, antibody responses to various immunogens, natural killer (NK) cell activity, delayed-type hypersensitivity, and both macrophage and neutrophil functions. A few of these assays (e.g., lymphocyte phenotyping and NK cell activity) are beginning to be developed for dogs, but are much more limited and are still in the optimization stage.

Currently there are no standardized protocols or methods, and validation is needed for many of these assays. Since the assays are being conducted with non-rodents species, the ability to optimize and validate is severely restricted due to the number of animals that are required. Thus while the assays may be verified for use within a laboratory, they are often conducted and the data analyzed with a variety of different methods between laboratories. For these reasons, it can be very difficult to compare data across laboratories and compounds. Furthermore, as non-human primates and dogs are out-bred animals, a high degree of variability is often observed that is compounded by the small sample size evaluated. This variability can sometimes be minimized to some extent by comparing responses to individual animal pre-study values (internal control) and collecting multiple samples overtime within an individual animal.

Immunophenotyping

The competency of the non-rodent species immune system can be assessed by basic parameters commonly incorporated into toxicology studies such as hematology (including differential counts) and clinical chemistry values, gross pathology finding. lymphoid organ weights (thymus and spleen) and microscopic changes in lymphoid organs (thymus, spleen, draining and distal lymph nodes, Peyer's patches within the intestine, and bone marrow). The evaluation can be further enhanced by the enumeration of immune cells of the peripheral blood or spleen by the use of flow cytometry. A number of antibodies are available for the non-human primate to measure the subpopulations of leukocytes, by evaluating cell-surface markers [e.g., T-lymphocytes (CD2 or CD3, CD4 and CD8); B-lymphocytes (CD20); NK cells (CD16); and, monocytes (CD14)] (Neubert et al., 1996). For dogs, the antibodies available are limited for lymphocytes [e.g., T-lymphocytes (CD3, CD4, and CD8) and B-lymphocytes (CD21-like or antiIg)].

A list of cell-surface markers useful in immunotoxicity studies are shown in Table 1. The use of flow cytometry to measure lymphocyte subpopulations in peripheral blood allows for relatively accurate and reproducible analysis that can be easily incorporated into a study. In addition to differences in the antibody clones that may be used, these data can be analyzed and reported in a numerous ways. For example, data may be normalized against hematology data or calibrated flourescent beads, absolute cell numbers or percentage of cells may be reported, or gates may be set on the percentage of lymphocytes, leukocytes, or all events. In addition to peripheral blood phenotyping, the population of cells in lymphoid organs and the architecture of these tissues can be evaluated using immunohistochemistry. Numerous differences in staining technique, reagents, and analysis exist as well for immunohistochemistry. For example, staining can be preformed with frozen tissue, preferably, or formalin-fixed, antibodies may be conjugated to enzymes or fluorescence.

TABLE 1 Leukocyte phenotyping in monkeys and dogs

Cell Types expressing molecule	Molecule	Dog
Pan T-cells	CD2^+	CD5+
Mature T-cells	CD3^+	CD3^+
Helper T-cells	CD3^+/CD4^+	CD3^+/CD4^+
Cytotoxic T-cells	CD3^+/CD8^+	CD3^+/CD8a^+
Pan B-cells	CD3^+/CD20^+	CD21-like or anti-Ig
Natural killer cells	CD3^−/CD16^+	—
Monocytes	CD14^+	—

Immunologic Mediators

In addition to cell-type evaluation, there are a number of soluble factors, such as serum, immunoglobulins, cytokines, complement components, anti-nuclear antibodies, and other immunological mediators that can be measured as indicators or biomarkers of immune status. If cells need to be stimulated for the mediator of interest to be produced, then monkeys can be treated in vivo with the stimulus, or monkey blood cells can be cultured ex vivo and the supernatants measured following immunologic stimulation or challenge. For example, monkeys can be challenged in vivo with lipopolysaccharide (LPS) and blood collected over time to assess for the release of cytokines, such as tumor necrosis factor (TNF) in vivo or blood collected from animals can be cultured in vitro, and then challenged with LPS. The ability of immune cells to produce or secrete a LPS-induced cytokines, such as TNF, can then be measured via standard ELISAs or multiplex-bead array assays. It is however, critical to understand the time course of a response as mediator may only be transiently present following stimulation/activation.

Antibody Responses

As with rodents, the ability of a non-human primate or dog to elicit an antibody response following antigen challenge provides information on the function of multiple components of the acquired immune system (antigen-presenting cells. T-lymphocytes, and B-lymphocytes), which must work together in order to produce antigen-specific antibodies. If any of these cell types are affected by drug treatment, then a measurable decrease in antigen-specific antibodies is often detected. A number of different T-dependent as well as T-independent immunogens have been employed in non-rodent studies using various routes of administration. Some of the more commonly used immunogens used in primates studies are keyhole limpet hemocyanin (KLH), as shown in Figure 1, bacteriophage, tetanus toxoid vaccine, or sheep erythrocytes. It has been our experience that the sheep erythrocytes administered intravenously is not an appropriate antigen in the dog, as an anaphylactic-like response, which was associated with a significant increase in complement activation, was observed following intravenous administration, compromising the health of the animal. However, KLH administered as an intramuscular injection can be an acceptable T-dependent antigen for the dog, as shown in Figure 2. Antibody responses are usually evaluated by measuring the antibody titer or concentration of specific antibody in the serum using an ELISA or similar technology. When evaluating non-rodents, multiple samples can be collected from the same animal over time. This is especially important when characterizing the kinetics of the response, primary and secondary responses, and the reversal of the response. By using the same cohort of animals, the number of animals and drug supply needed is significantly reduced. In addition, multiple antigens can be used so that the effect of drug treatment on a primary, secondary, as well as memory response can be evaluated in parallel.

FIG. 1 Effect of Abatacept on the toral KLH-specific antibody response in monkeys. Cynomolgus monkeys (n = 6 females/group) were immunized with a single 10-mg intramuscular injection of KLH (Pierce) following 10 mg/kg intravenous dose of abatacept [a selective CD80/86 costimulatory inhibitor; CTLA4Ig]. Serum was collected prior to dosing and on post-KLH-immunization Days 8, 15, 22, and 29 and evaluated for total KLH-specific antibodies, Results are expressed as endpoint titer ± SD, defined as the reciprocal of the interpolated dilution equal to 5 times the mean plate background. Abatacept substantially diminished the KLH-specific antibody response (Bigwarfe et al., 2004).

FIG. 2 KLH-specific antibody response in dogs. A single 5- or 10-mg intramuscular injection of KLH (Pacific Biomarine) was administered to ∼17-month old Beagle dogs (n = 2). Serum was collected at 5, 7, 9, 11, 14, 21, and 28 days post-KLH-immunization and evaluated for KLH-specific IgM or IgG antibodies. Results are expressed as EPT, defined as the reciprocal of the interpolated dilution equal to 5 times the mean plate background. A single 5 mg intramuscular dose of KLH was sufficient to produce a robust antibody response in dogs, with a mean peak IgM titer of 12,000 on Day 7 and a mean peak IgG titer of > 119,000 on Day 21 (Bigwarfe et al., 2004).

Natural Killer Cell Activity

NK cells are the immune systems first line of defense against tumor cells, virus-infected cells and MHC Class I mismatched cells. NK cells can recognize and kill these target cells without prior immunity. This response is assessed ex vivo without in vivo antigen exposure. Effeetor cells (peripheral blood) are mixed with labeled target cells (species-specific tumor cell line), incubated, and cytotoxicity measured. For primates, the target cells line K562, human erythroleukemia cells (Condevaux et al, 2001) is usually used, while a canine osteosareoma cell line, D-17, has been used for dogs (Rodrigue et al., 2004). There are many different variations to NK cell cytotoxicity assays, all of which have a fair degree of variability and are not always sensitive. Types of methods include the traditional chromium-release-assay (Condevaux et al., 2001) and fluorometric assays (Giavedoni et al., 2000). While the NK assay has been optimized sufficiently by some labs for use in primates, the canine assay still requires additional work.

Delayed-Type Hypersensitivity

Cellular immunity protects against intracellular bacteria, viruses, and cancer and is mediated by antigen-specific memory T-lymphocytes. These cells release inflammatory and toxic substances, which attract other white blood cells (lymphocytes, macrophages), ultimately resulting in tissue injury at the site of antigen contact. It can only be transferred from sensitized to normal animals via lymphocytes (thus cell-mediated), not humoral (IgE-mediated), and takes several days to develop. In general, to assess the delayed-type hypersensitivity response, animals that have been previously sensitized with an antigen, or cocktail of antigens, are challenged with a lower concentration of that antigen at a distal site and the subsequent induration and erythema at the injection site quantitated after 24–72 hours. In primates, clinical reactions have not been robust and the assays have shown a very high inter-animal and inter-site variability, positive controls have not been identified, and protocols have been very inconsistent.

It is important that the antigen challenge be given greater than two weeks after the last sensitization, as early reactions may be due to immune-complex deposition and subsequent complement activation, which attract PMN, and is not a T-lymphocyte-mediated response. While a clinical response (erythema and induration) should be present, it is important that the pathology of a biopsy collected 48 hours following challenge be conducted to confirm the nature of the response, which should consist of lymphocytes and macrophages and be quantified by immunohistochemistry. Primates have remained sensitized for up to at least 24 weeks post-sensitization and, thus, could potentially be reused for multiple evaluations. Dexamethasone may be a potential positive control; oral administration at 2 or 20 mg/kg/day given three days prior to and during challenge caused a dose-related inhibition of the DTH reaction. The addition of oligo-CpG to the antigen cocktail during sensitization did not enhance the challenge DTH response, although it did augment humoral immunity (Bugelski et al., 1990; Bleavins et al., 1995; Bussiere et al., 2001; Price et al., 2004).

Macrophage and Neutrophil Function

Macrophages and neutrophils serve as an important part of the first line of host defense against infectious or allergenic agents. There are several cell-based assays that can be used to evaluate drug toxicities ot the innate immune-cell defense mechanisms, including respiratory burst and phagocytic functions of macrophages/monocytes and neutrophils. Flow-based methodologies used to evaluate drug effects on respiratory burst and phagocytosis in both monocytes and neutrophils from peripheral blood of human, mice and rats, can also be applied to dogs and monkeys. Both respiratory burst and phagocytosis assays employ commercially available kits manufactured by Orpegen Pharma (PhagoBurst and PhagoTest) (Robinson and Carter, 1993; Miyagawa and Klingemann, 1997; 〈http://www.biocarta.com/TDS/10-0100.pdf〉, 2007; 〈http://www.biocarta.com/TDS/10-0200.pdf〉, 2007).

The manufacturer's protocols can be adapted for high throughput analyses on a flow cytometer and modified for individual experimental optimization. The respiratory burst assay involves the timed and temperature sensitive stimulation of the immune cells within the blood with phorbal 12-myristate 13-acetate (PMA), N-formyl-Met-Leu-Phe (fMLP), and/or opsonized Escherichia coli treatment. After a short stimulation, a fluorogenic reactive oxidant substrate, such as dihydrorhodamine 123, is added to the blood and following erythrocyte lysis, the remaining leukocytes are analyzed on the flow cytometer. The neutrophils or monocytes are gated based on forward and side scatter and the fluorescence intensity measured and compared with control (unstimulated and/or vehicle treated), stimulated and/or drug treated cells. The phagocytosis assay evaluates the timed and temperature sensitive uptake of fluorescein labeled-E coli by phagocytic leukocytes in the blood. As with the respiratory burst assay, neutrophils and monocytes are gated and the mean fluorescent intensities of both populations are compared among control (iced and/or vehicle treated) and drug treated cells. These assays can be used for both ex vivo and in vitro evaluation. However, day-to-day variability is an issue that must be addressed to evaluate studies with a small number of subjects. Classical inhibitory immunomodulators, including rottierin, wortmanin, and p38 inhibitor SB203550, have been used to determine the sensitivity and reproducibility of in vitro results.

SUMMARY

In summary, there are a number of immune function evaluations that can be applied to the non-human primate model and a few that are now being applied to dogs to assess the immune status. Some assays require antigen challenge in vivo (TDAR, DTH), however, there are several that can be conducted ex vivo (cytokine release, NK). Many of these assays are applicable to the clinic, although the high degree of variability in responses combined with low animal numbers can make the data very difficult to interpret. However, more work is still needed to optimize some primate models, such as the DTH, but due to the large number of animals that are needed to optimize and verify these assays this has been difficult to achieve. Furthermore, it is critical that more antibodies specific for canine cell-surface markers and cytokines become available so that assays to evaluate the canine immune system can continue to be developed and implemented as the canine model at times may be the most appropriate model for evaluating an effect on the immune system, but assays are very limited due to lack of canine-specific reagents. Although many of these assays have not been standardized or formally validated, they have been verified by the laboratory with positive controls and can be useful in evaluating the immune status of the animal. However, improvements in the performance, of the current assays could be gained by laboratories sharing their methods and data, and combining their expertise and resources to allow for further optimization, especially given the cost and large number of animals that are required for this work.