New Developments in the Assessment of Developmental Immunotoxicology

Abstract

Recent progress has been made in directly comparing the risk of immunotoxicity following exposure to various drugs and environmental chemicals during different stages of life. With the availability of an increased developmental immunotoxicology database, new concepts of effective immunotoxicological risk assessment have emerged. From the standpoint of risk assessment, recent results suggest that there is greater value obtained from exposure-assessment of non-adults than can be derived solely from adult-exposure-outcome data. This is hardly surprising given the fact that, for the vast majority of known immunotoxicants compared across age groups, the non-adult stages are more sensitive than adults for risk of clinically important immunomodulation. Therefore, if immunotoxicity testing is to identify risk for the more susceptible subpopulations, the adult is not the informative model. This brief review, based on the Immuntoxicology III conference presentation, describes the data supporting age-based differences in sensitivity to immunotoxicants, differences in immunotoxic outcomes, and the potential benefits of utilizing non-adult exposures and life-stage-relevant immune assessment. In essence, the issue is whether historic adult immunotoxicity testing strategies can continue to ensure adequate protection of the most vulnerable subpopulations in the face of recent developmental immunotoxicological data. The review describes the possible benefits of substituting non-adult exposures for adult exposures in future assessment protocols.

Keywords :

• developmental immunotoxicology
• risk assessment
• immune balance
• DIT testing

INTRODUCTION

Developmental immunotoxicology (DIT) is an area of research that dates back at least to the 1970s (Luster et al., 1978; Faith et al., 1979). Despite the fact that it was an active area of research during the earliest days of immunotoxicology, recognition of DIT as a necessary component of effective risk assessment has been a recent development. This has emerged with the realization that developing systems, such as the immune and neurological systems, are inherently more vulnerable to environmentally induced disruption than are their fully matured counterparts. While this concept makes perfect biological sense given the nature of in utero development, it presents a special dilemma in light of the historic approaches to safety testing. This presentation will consider the reasons why embryos, fetuses, neonates, and juveniles appear to be more susceptible to a majority of immunotoxicants than are adults, and the implications this would appear to have for future safety testing.

Differences Between Early-Life Immunotoxic Responses and Those of Adults

Certainly, the developing immune system presents a moving target for immunotoxic insult compared with a fully matured and dispersed immune system. Dynamic events such as the seeding of the thymus by pre-T-cells and T-cell “education” within the thymus occur over discrete early periods of development never to be experienced again in later life. Therefore, it should not be surprising that adult exposure assessment would provide little information concerning the capacity of a xenobiotic to disrupt immune maturation and to potentially enhance the risk of later life disease including atopy (Lutz et al., 1999; Reichrtova, et al., 1999; Snyder et al., 2000), asthma (Shaheen et al., 2005; reviewed in Armstrong et al., 2005) and autoimmunity (Noller et al., 1988; reviewed in Holladay, 1999; Blaylock et al., 2005). For this reason, modeling the developing immune system and its vulnerabilities using adults is inherently problematic.

The increased sensitivity of the early life stages can take several forms based on recent research. In some cases, the doses required to produce immunotoxicity are significantly lower when exposure occurs at an early age. With other examples, the range of immune alterations is more severe after early exposure. In some cases, the immune changes are qualitatively different across life stages. Additionally, immunotoxicity seems to persist for many chemicals following early exposure while the same chemical or drug produces only transitory effects in adults. Finally, significant gender differences can exist for the immunotoxic response to some xenobiotics even in the early embryo. Clearly, these age-related differences in sensitivity can dramatically alter the risk associated with a given exposure. The types of age-based differences in sensitivity are considered in the following sections.

Persistence

Diazepam (Valium) is a drug falling within the benzodiazepine family of compounds. It acts via specific receptors in both the central and peripheral nervous systems. The drug crosses the placenta and the immunological ramifications following exposure of the non-adult are a potential concern. The immunotoxicity of diazapam across life stages was recently reviewed by Luebke et al. (2005). Age comparisons suggest that fetal rats are sensitive to much lower doses of the drug than are adults and that the adverse immunological effects are far more persistent following early life exposures.

This is a similar finding to that obtained in age-based comparisons of the drug cyclosporin A (CYP-A) (Hussain et al., 2005). Exposure of fetal vs. adult female rats to CYP-A resulted in persistent immunotoxic effects in the offspring exposed in utero. However, the adults recovered immune function after the drug had been removed. This reinforces the finding that early life stages are more likely to show persistence of immunotoxicological effects and adult-exposure assessment may not identify this problem.

Dose Response/Severity

With life stage comparisons of exposure to dexamethsone (DEX), the range of immunotoxicity overlapped between offspring exposed in utero and exposed adults. However, the in utero exposed rats had a greater severity of symptoms and at lower doses than did the adults (Dietert et al., 2003). Additionally, histopathology of the thymus was very ineffective as a predictor of functional alterations among these animals. This latter observation in this DIT model is consistent with the recent reports for adult-exposure of mice under the National Toxicology Program (NTP) immunotoxicology assessment regime. These reports compared the relative predictive value of immune function tests with histopathology (Germolec et al., 2004a, 2004b). These authors reported that a single immune functional test (the antibody plaque-forming cell assay) had a greater predictive value (90%) for identifying immunotoxicants (within the data set employed) than did histopathology on one (60%) or multiple (80%) immune tissues.

In addition to the differences in the severity of alterations based on age of exposure, the lowest observed adverse effects levels (LOAELs) can differ across age groups as well. For example, adult vs. fetal rats appear to have approximately a magnitude difference in dose sensitivity for lead (Heo et al., 1996; Miller et al., 1998; Snyder et al., 2000; Bunn et al., 2001a; Chen et al., 2004) and greater than a magnitude difference for TCDD (Smialowicz et al., 1994; Gehrs and Smialowicz, 1999) (reviewed in Luebke et al., 2005).

Nature of the Immune Alterations

Developmental stage-specific effects have been described for some toxicants and target physiological systems. For example, critical windows of increased vulnerability appear to exist for the capacity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) to cause certain changes within the male reproductive system (Lin et al., 2002; Ohsako et al., 2002). It also appears that for some immunotoxicants, the very nature of the immune alterations can differ depending upon the age at which the exposure occurred. Hence, adult exposure assessment may not predict the type of immune changes evident following fetal or perinatal exposure.

Skewed T-helper 1 (T_H1)1/T-helper 2 (T_H2) balance is one of the hallmarks of lead-induced immunotoxicity (Heo et al., 1996; McCabe et al., 1999; Snyder et al., 2000). Lead exposure can disrupt T-helper balance if exposure occurs after the mid-point of gestation (Dietert et al., 2004). However, earlier exposure targets the macrophage in the absence of major T-cell alterations (Bunn et al., 2001b; Lee et al., 2001). Such age-based differences reflect the pitfalls of extrapolating adult exposure assessment results to earlier life stages.

Gender Differences

While early life exposure to immunotoxicants does not always produce sex-specific differences in outcome (Vorderstrasse et al., 2004), such differences can occur for some xenobiotics under certain exposure regimes. With this in mind, potential gender differences following in utero exposure should not be overlooked in immunotoxicological risk assessment. Despite the fact that exposure may occur during gestation, instances of significant postnatal immunotoxic outcomes among exposed male and female embryos have been reported for TCDD (Gehrs and Smialowicz, 1999), atrazine (Rooney et al., 2003), mercury (Silva et al., 2005), and lead (Bunn et al., 2001a). Hence, gender comparisons are likely to be important for identifying the most susceptible subpopulation.

Assessment Considerations

It is important that historic assessments so useful in adult immunotoxicity testing and for the detection of immunosuppression not be assumed to have equal utility for the analysis of early exposure immune alterations. The reason to approach our prior testing regimes with a healthy skepticism is based on the fact that the intact adult immune system with a well-established functional balance presents a very different picture and target from that of a late-gestation embryo.

In fact, the critical developmental change from the late embryo to the early neonate is to attain effective T-helper-associated immune balance to the extent afforded by the individual's genotype. Maternal-fetal based genetic incompatibilities (e.g., Human Leukocyte Antigen and other genotypic differences) in the womb present the potential for mutual T_H1-mediated tissue rejection thereby jeopardizing the pregnancy (Kruse et al., 2004; Zhu et al., 2005). To protect against this, the local maternal immune environment is skewed toward T_H2 cytokine production (Wegmann et al., 1993; Lim et al., 2000; Poole and Clayman, 2004). This suppresses maternal rejection of the fetus. Additionally, the fetus is delayed in acquiring a robust T_H1 functional capacity (thereby avoiding a graft vs. host type response), which is normally adjusted after birth (Morein et al., 2002; Upham et al., 2002; Protonotariou et al., 2003). In fact, Malamitsi-Puchner et al. (2005) recently found that human vaginal delivery promotes both soluble IL-4 receptor scavaging of IL-4 and elevated IFN-gamma production (thereby altering effective T_H1/T_H2 functional balance) when compared with delivery by Cesarean section. The authors proposed that the vaginal delivery environment may help initiate needed balance of neonatal T_H1/T_H2 functional capacity thereby reducing the risk of atopic disease.

One may then predict that T_H1 vs. T_H2 functions might not be equally susceptible to developmental immunotoxicants. T_H2 is the embryonic default capacity, and T_H1 functional capacity must be acquired or matured during the perinatal period (Leibnitz, 2005) to avoid an increased risk of T_H2-associated diseases (Holt and Sly, 2002). Hence, acquisition of T_H1 functional capacity is in greatest jeopardy during embryonic-early neonatal periods (Protonotariou et al., 2003). In fact, based on the drugs and chemicals compared to date, this appears to be the case (Holladay, 2005; Luebke et al., 2005). There are few, if any, reports of developmental immunotoxicants that selectively impair T_H2 function following in utero exposure. In contrast, many of the developmental immunotoxicants examined to date either target T_H1-dependent cell-mediated immune function (e.g., the delayed-type hypersensitivity or similar responses or graft vs. host responses) or impair both T_H1 and T_H2 capacities. Examples in which exposure to developmental immunotoxicants either impaired later-life T_H1-dependent functions or shifted T_H1/T_H2 balance toward T_H2 have been reported for alflatoxin-B1 (Dietert et al., 1985), atrazine (Rooney et al., 2003), chlordane (Barnett et al., 1985), cyclosporine A (Hussain et al., 2005), dexamethasone (Dietert et al., 2003), diethylstilbesterol (Ways et al., 1980), heptachlor (Smialowicz, 2002), hexachlorobenzene (Barnett et al., 1987), hexachlorocyclohexane (Das et al., 1990), lead (Miller et al., 1998; Chen et al., 1999; Snyder et al., 2000), organochlorine compounds (Reichrtova et al., 1999), paracetamol (Shaheen et al., 2005) and TCDD (Gehrs and Smialowicz, 1999; Walker et al., 2004).

Simply because T_H skewing may be a frequent outcome of early life exposures to toxicants, one should not assume that the most sensitive target of developmental immunotoxicants is inherently T-lymphocytes or more specifically T_H cells. Instead, dendritic and other antigen-presenting cells may be critically important in the acquisition of a more T_H1 oriented functional capacity in the newborn (Holt and Jones, 2000; Holt and Sly, 2002). In fact, Mainali et al. (2005) recently demonstrated that human cord blood-derived dendritric cells (DCs) were preferentially sensitive to DEX-induced skewing toward a T_H2 phenotype (IL-10/IL-12 ratio) when compared with adult-derived DCs. Therefore, in many cases, chemically induced T_H skewing may reflect immunotoxic effects on DCs and other antigen presenting cell populations as opposed to direct effects on T lymphocytes.

What this means is that in contrast with adult immunotoxicity assessment, developmental immunotoxicity assessment should include overt functional measures of T_H1 capacity and, hopefully, T_H1/T_H2 functional balance for effective risk assessment. This is the one measure that might be expected to change following exposure to a developmental immunotoxicant based on the current literature. For this reason, there is a greater need to include T_H-dependent functional tests in DIT assessment than were necessarily justified when adult testing protocols were originally designed and subsequently validated decades ago.

This raises the question of where future testing should be headed. Based on the disparity in sensitivity across life stages, it seems clear that the best data for effective risk assessment would result from early life exposure as opposed to adult exposure. Therefore, it is prudent to consider the options for replacing routine adult immunotoxicity safety testing with a different test paradigm. Toward this end, Luster et al. (2003), Kimmel et al. (2005), and Holsapple et al. (2005) have all recommended that exposure extending across all possible critical developmental windows should be utilized. A similar version of this exposure protocol was utilized by Luster et al. (1978), Faith et al., (1979) and, more recently, by Chapin et al. (1997). The rat would be the default test species. Additionally, immune assessment would be married to reproductive and/or developmental assessment to the extent possible (Luster et al., 2003; Holsapple et al., 2005; Kimmel et al., 2005). Such a protocol might involve exposure throughout gestation and lactation and extending until the offspring were approximately 6 weeks of age. Certainly, a broad DIT protocol would ensure that the most sensitive developmental window had been exposed to the test chemical or drug. By combining reproductive and immune parameters, it may be possible to have a more efficient assessment and one that can provide direct comparisons across the two systems (Dietert et al., 2003; Hussain et al., 2005; Piepenbrink et al., 2005).

Rather than DIT testing being an extra requirement for safety assessment, it seems likely that it might replace adult immunotoxicity testing with little impact. One area not readily predictable from DIT testing would be the impact of chemical exposure in later life on geriatric-associated decline in immune function, but current adult testing provides little insight regarding geriatric immune status either.

Prior reports have proposed potential DIT assessment screens based on available data and consensus opinion (Luster et al., 2003; Holsappple et al., 2005). Both forums recommended the measurement of T-dependent antibody production (either as a plaque-forming cell assay or in an ELISA) as well as a cell-mediated immune measurement such a DTH in a primary screen. In general, histopathology was not thought to be as useful in a primary screen as were combined functional tests (Holsapple et al., 2005). The recent experience of our laboratory with DIT evaluation (Dietert et al., 2003; Chen et al., 2004; Hussain et al., 2005; Piepenbrink et al., 2005) would confirm the merits of these recommendations. Combining T-cell-dependent antibody measurements (particularly if immunoglobulin class/subclass can be measured) with CMI-assessment (such as DTH) in such a manner that evaluates T_H functional capacity and balance is a very powerful tool for DIT assessment. However, since some early exposures have the potential to target non-lymphoid function (i.e., involving macrophages and dendritic cells) (Bunn et al., 2001b), there may be merit in including a highly sensitive and clinically important innate immune parameter such a nitric oxide (NO) production by macrophages and possibly cytokine production by DCs.

The recent DIT forums have not included host resistance to disease in suggested primary or even secondary screens (Luster et al., 2003; Holsapple et al., 2005). While combinations of host resistance assays can provide useful information as to xenobiotically-induced gaps in host defense, the assays are very costly in that they usually require different sets of animals for each disease challenge. Additionally, not all disease models have equally informative and validated endpoints for measurement. Certainly, the exclusion of host resistance in initial assessment appears to be a prudent decision given the requirements of employing multiple disease challenges. However, this places an even greater requirement to use functional tests in combination for DIT assessment. For example, there are situations where early exposure to lead induce immunotoxicity that is not evident among resting circulating immune cells. However, it becomes easy to identify lead-exposed animals when the immune system is stressed via an infection or immune challenge (Bunn et al., 2001a, 2001b; Lee et al., 2002). Therefore, requiring the immune system to mount functional responses as part of assessment is helpful and needed in the absence of host resistance data.

SUMMARY

In summary, an increasing number of examples of chemicals evaluated for immunotoxicity across life stages all suggest the merits of future DIT testing. Protocols have been recommended that should address differential age-based windows of vulnerability. However, testing regimes should consider the unique transition between the late embryo and the neonate and the potential need for new combinations of functional assays to be employed for effective evaluation of DIT risk. To this end, it is suggested that assessment include an evaluation of T_H-dependent functional balance and possibly accessory cell (macrophages and dendritic cells) function.

The research of the author described in the review was supported by funding from the American Chemistry Council.