Introduction

Halogenated aromatic hydrocarbons (HAHs) are ubiquituous in the environment. The most toxic and best-studied of these compounds is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Numerous studies on the toxic effects of TCDD have been done, most of them on rodents. TCDD is tumor promoting, teratogenic, and embryotoxic (1). One of the earliest and most sensitive markers of TCDD toxicity in experimental animals is impairment of the immune system, which is evident at doses that do not lead to overt signs of general toxicity. TCDD leads to atrophy of lymphoid organs, such as the thymus, spleen, and lymph nodes (2). Moreover, TCDD was shown to suppress cellular and humoral immune functions in experimental animals (2). Using a variety of in vivo exposure schemes or in vitro assays, TCDD was found to impair cytotoxic T-lymphocytes and natural killer cell functions, or inhibit antibody production by B-cells, for example (2). The capacity to respond to mitogenic stimuli, such as phytohemagglutinin (PHA) and lipopolysaccharide (LPS), is also affected by TCDD (3,4).

Susceptibility to the toxicity of TCDD is genetically determined by the aryl hydrocarbon receptor (AhR) locus. This gene codes for a cytosolic, TCDD-binding protein that is activated to a DNA-binding state upon ligand engagement (5) and induces the expression of a gamut of genes, including CYP1A1, genes of fatty acid metabolism, cytokines, and/or growth and differentiation factors (6). The binding affinity of AhR to TCDD and thus susceptibility varies between different animal species and also interindividually in outbred populations. Neither the overall toxicity of TCDD to humans nor TCDD-induced effects on the human immune system is known.

We examined 11 workers with defined TCDD body burdens who had been inadvertently exposed to TCDD for several years. We evaluated possible deviations in peripheral blood lymphocyte subsets using multi-parameter immunofluorescence of relevant markers (7) and examined in vitro effects of TCDD on immunocompetence of T- and B-cells.

Materials and Methods

Eleven workers, 45-63 years of age, participated in this study. They had been exposed to high doses of TCDD and other polychlorinated dibenzo-p-dioxins between 1966 and 1976 during production and maintenance operations at a chemical factory producing 2,4,5-trichlorophenol. Six other exposed workers declined to participate. The level of TCDD in blood lipids was determined in 1989 and 1992 by the ERGO-Forschungsgesellschaft, Hamburg, Germany, or Bioscientia, Moers, Germany, respectively, according to standard mass spectrometry (8). Ten age-matched, healthy, males with no known TCDD exposure history, working in the same company in office jobs, volunteered as controls. Informed consent was obtained from all subjects.

Peripheral blood (50 ml) was drawn by venipuncture into a sterile tube, containing 100 U of heparin/ml blood. Peripheral blood mononuclear cells (PBMC) were isolated by Ficoll-Hypaque (Pharmacia Fine Chemicals, Uppsala, Sweden) density centrifugation.

For single or dual fluorescence analysis, aliquots of 1 x 10⁶ peripheral blood lymphocytes were spun down and incubated for 10 min at 6-8°C with fluoresceinated antibodies as recommended by the manufacturer. Cells were washed two times after staining, and data from 10,000 cells were collected immediately in list-mode on a FACScan flow cytometer, using appropriate compensation settings (Becton-Dickinson, Mountain View, California). Forward and rectangular light scatter gates were used.

The following fluorochromed antibodies were used: mouse anti-CD3^FITC (clone SK7), mouse anti-CD4^FITC (clone SK3), mouse anti-CD8^PE/CD8^FITC (clone SK1), mouse anti-CD19^FITC (clone 4G7), all from Becton-Dickinson; mouse anti-CD45RO^PE (clone UCH-L1) and anti-CD45RA^PE (clone F8-11-3), both from Serotec (Oxford, UK); mouse anti-CD56^PE (clone B 159.5; Pharmingen, San Diego, California); and anti-CD57 (Leu7; Becton-Dickinson).

We cultured 5 X 10⁴ cells/well in 96-well round-bottom plates containing 150 l RPMI 1640 culture medium, supplemented with 5% heat-inactivated pooled human AB0 serum, l-glutamine (2 mM), sodium pyruvate (1 mM), nonessential amino acids, 50 U/ml penicillin, and 50 U/ml streptomycin.

Cells were cultured in triplicate either with medium alone or in the presence of 9 g/ml PHA (Sigma, St. Louis, Missouri) or 6 g/ml pokeweed mitogen (PWM; Sigma) for 3 days at 37°C in a water-saturated 5% CO₂ atmosphere. On day 4 of culture, cells were pulsed with 1 Ci of ³H-thymidine and then harvested 20 hr later to determine incorporated radioactivity (ß counter LS 6000 IC, Beckman Instruments, München, Germany).

The human lymphocyte antigen (HLA) phenotype of peripheral blood mononuclear cells was determined by standard microlymphotoxicity assays using A,B,C plates (One-lambda, Canoga Park, California) and HLA-class II plates (Biotest, Dreieich, Germany). For mixed lymphocyte cultures, washed cells were resuspended in serum-free CG-medium (Vitronex, Vilshofen, Switzerland), supplemented with 60 g/ml gentamcycin. We seeded 5 X 10⁴ responder cells in triplicate into 96-well round-bottom microtiter plates. For "one-way" mixed lymphocyte cultures, the same number of irradiated (30 Gy) stimulator cells were added. Stimulator cells were from a pool of cells from 20 allogeneic, HLA-unrelated individuals. For "third-party" suppressor assays, the lymphocytes from dioxin-exposed or control persons were irradiated (30 Gy) and added to the pooled allogeneic cells (in this case unirradiated).

Interleukin-2 (IL-2)-inducible proliferation of T-cells was measured after adding 30 U/ml recombinant IL-2 per 5 X 10⁴ cells. Recombinant IL-2 was a gift of P. Loeliger (Sandoz, Switzerland).

Results

The concentration of TCDD in the blood of 11 workers exposed between 2 and 11 years before 1976 (with one exception, see Table 1) was determined in 1989 or 1992, i.e., 13-15 years after the last exposure. Table 1 summarizes the exposure parameters and the health status of the TCDD-exposed workers determined at a thorough general medical examination in 1992. The TCDD values in blood fat differed up to 20-fold between individuals, yet even the lowest burdens were well above the average level of the German population, which is about 4 pg/g blood fat (9). At the time of the study, five persons still suffered from chloracne, of which one had chronic gastritis and one hyperthyreosis. Two subjects displayed a disturbance of fatty acid metabolism. The others appeared healthy.

The frequencies of various lymphocyte subsets were determined by flow cytometry. The percentages of B-cells (CD19), T-cells (CD3), and subsets thereof were determined, e.g., T-helper cells (CD4), cytotoxic T-cells (CD8), virgin helper/inducer T-cells (CD4CD45RA) as well as primed helper/inducer (CD4CD45RO). Moreover, natural killer cells (CD56 and CD57) and HLA-DR expression were measured. Some of the subset markers were previously shown to be sensitive parameters of TCDD-exposure in marmosets, a new world primate species (7). Table 2 shows the result of the analysis for each of the 11 TCDD-exposed individuals and for the controls. No difference between controls and the mean values of the TCDD-exposed group was evident for any of the lymphocytes subsets analyzed. Note that the percentage of CD56-expressing cells (i.e., natural killer cells) is comparatively lower than reported in the literature for younger individuals, possibly a phenomenon of age (27).

Figure 1. Proliferation capacity of (A) peripheral blood T-cells in the presence of phytohemagglutinin and (B) T-cells and B-cells in the presence of pokeweed mitogen in individuals exposed to dioxins about 20 years ago. We cultured 1 X 10⁵ Ficoll-isolated peripheral blood mononuclear cells in the presence of mitogen for 4 days, and pulsed with ³H-thymidine 20 hr before harvesting. Shown are the means and SDs (bars) of triplicate cultures (TCDD exposed and controls). The order of symbols for dioxin-exposed persons are arranged from 1 to 7 and 9 to 11 as they appear in Tables 1 and 2; the value for person 8 was not determined.

Peripheral blood lymphocytes were tested for their capacity to respond to mitogens PHA and PWM. As shown in Figure 1, lymphocytes from TCDD-exposed and control persons responded equally well to mitogen stimulation, with some individual variation. There was no statistically significant difference between the response of the two groups (p = 0.5641). Moreover, when we corrected the data for the interindividual differences in T-cell frequencies in the blood (see Table 2), it becomes even more clear that TCDD exposure does not affect the mitogen-induced proliferative capacity (p = 0.7823; data for PHA stimulation not shown). No correlation existed between individual TCDD levels in the blood nor the age of the persons and the respective proliferative capacity of their lymphocytes (not shown).

We tested the capacity of T-cells ("responder cells") from dioxin-exposed persons to specifically react against irradiated, HLA-different, allogeneic lymphocytes ("stimulator cells") in a mixed-lymphocyte culture. In another experiment, we added irradiated responder cells as a third party to a pool of unirradiated peripheral blood mononuclear cells from 20 different donors. Whereas the former assay measures the response to allo-major histocompatibility complex, the latter is used to detect suppressive factors/cells in the responder cell population, which would inhibit the respective response of the unirradiated cells (10). As shown in Figure 2, the responder cells of the TCDD-exposed persons proliferated less in response to irradiated stimulator cells (p<0.05 by Student's t-test). Moreover, the third-party mixed lymphocyte reaction against unirradiated stimulator cells revealed a small amount of suppressive activity in the responder cells of dioxin-exposed individuals, resulting in a decreased overall proliferation of T-cells (Fig. 2,). This significant suppression (p <0.01 by Student's t-test) is indicative of a reduced T-helper cell response (11). However, the actual number of T-helper cells was unaffected by TCDD (see Table 2).

Figure 2. Proliferation of peripheral blood mononuclear cells (PBMCs) of dioxin-exposed persons and controls. Tested were 1) the spontaneous response, 2) the response toward irradiated stimulator PBMCs of unrelated HLA-haplotypes (MLR), 3) the third-party response of irradiated PBMCs of TCDD-exposed persons toward the mixed-lymphocyte reaction of the pool of unirradiated stimulator cells (MLR*), and 4) the proliferative capacity in response to interleukin-2 stimulation (IL-2). Cultures were done in triplicate. Each dot represents the mean of the count per minute values of the PBMCs of one person. For p-values from Student´s t-test, see text. One dioxin-exposed person (value in parentheses) displayed a very high response in all assays, probably indicating preactivated T-cells (e.g., due to an undetected, ongoing infection). This value was disregarded for statistical calculations (see text).

The capacity of a pool of T-cells to proliferate upon IL-2 stimulation is a parameter of normal T-cell function and correlates with the presence of preactivated T-cells in the pool, which would result in a higher overall proliferative response. Peripheral blood mononuclear cells of TCDD-exposed individuals and control persons were co-cultured with a low dose of IL-2 for 4 days. The cells of TCDD-exposed persons revealed a reduced capacity to proliferate with IL-2 (Fig. 2; p<0.05). The values for one exposed individual were not included in the data due to an excessively high proliferation rate. Those particular values reflected an extreme preactivation of the T-cells, which is typical for the beginning of an acute, but undetected, infection. No atopic status is known for that person. However, it must be noted that the statistically significant differences between the TCDD-exposed and control persons disappear if this one value is included.

Discussion

Ever since the major accident at Seveso, Italy, the danger of dioxin has been recognized. The acute toxicity dose of TCDD is among the lowest for any known chemical substance. Fortunately, no people died at Seveso, and most scientists believe humans to be at the lower end of the susceptibility scale. However, TCDD continues to be inadvertently released into the environment, and little is known on the long-term effects of low doses. In Germany, the body burden of the general population is about 4 ppt TCDD in blood fat, with an estimated daily uptake, mostly by food, of 26 pg (9).

The immune system seems to be the most sensitive target of TCDD action in experimental animals. Humoral as well as cellular components of immune responses are suppressed by TCDD (2,12,13). The interaction of the human immune system with TCDD has remained controversial. Studies with an evolutionarily closely related species, the primate Callithrix jacchus (marmosets), revealed sensitive parameters for TCDD-induced alterations in peripheral blood lymphocyte subsets. Using combinations of surface markers, changes in the frequency of memory (CD4⁺, CDw29⁺), suppressor-inducer (CD4⁺, CD45RA⁺) T-cells and B-cells (CD20⁺) were found (7). In vitro studies on human blood leukocytes after accidental or occupational exposure to TCDD were ambiguous with respect to the immunological alterations induced (14). Some authors reported suppressed responses (15), whereas others reported enhanced mitogen-induced lymphoproliferative responses (16) and increased percentages in suppressor/cytotoxic T-cells and the absolute number of natural killer cells in the peripheral blood (17,18). Others found no effect on mitogen stimulation (19) and no deviations in peripheral blood subsets (20,21). With the exception of one of these studies (21), the actual body burden had not been determined. It is interesting to note that also in experimental animals the immunosuppressive effects of TCDD are commonly detected by functional tests (i.e., following antigen-specific stimulation), rather than unspecifically, by mitogen stimulation, for example. Thus, it was important to choose an appropriately sensitive test system.

We tested possible immunosuppression in TCDD-exposed workers using three parameters of lymphocyte competence. We found no difference compared to the age-matched control persons for the distribution of lymphocyte subsets in the blood. This result is in accordance with the more elaborate study of Neubert and co-workers (7), in which they applied the experience gained in Callithrix jacchus on TCDD-sensitive lymphocyte subsets to humans, but failed to detect any decrease in leukocyte subsets associated with elevated dioxin body burdens. In the present study, we demonstrated a small but significant difference in the allo-response of T-cells and in their proliferative response to IL-2. Moreover, the lymphocytes of TCDD-exposed persons displayed a suppressive activity, which inhibited an ongoing allo-response of HLA-unrelated lymphocytes. This is probably due to an increased proliferation response of T-cells counteracting the CD4⁺ T-helper-1 cells, representing the primary responder type in mixed lymphocyte cultures.

As reported previously (22), the changes in immunocompetence we observed in vitro did not correlate with obvious diseases related to severe immunodeficiency such as certain cancers and infections. The workers were generally healthy and, with one exception, had no history of increased susceptibility to infections. The functional reserve of the immune system is enormous, so that impaired immune responsiveness need not have pathological consequences (23). Indeed, the term "immunotoxic" has not been properly defined (24). Only a large, well-controlled epidemiological study might reveal the actual health effects of subtle changes in immunocompetence.

The reduced immunocompetence in vitro observed here fails to correlate with a reduced number of lymphocyte subsets. Thus, TCDD-induced immunosuppression is more likely mediated by a reduced functionality of individual cells rather than by a reduction in cell numbers circulating in the blood.

Thymus involution, a reduction of thymus weight and cellularity, is a hallmark of TCDD exposure. In mice, TCDD skews the distribution of lymphocyte subsets (25,26). However, although a link between thymus events and peripheral immunosuppresssion is often implicitly assumed, nothing is known about the balance between thymocyte/T-cell generation and migration to the periphery under TCDD treatment. The thymus normally begins to atrophy at about the time of sexual maturity in mice and humans (27). T-cell numbers and other leukocytes in the periphery remain unaffected in TCDD-exposed versus nonexposed persons (28), thus an additional atrophic effect on the thymus by TCDD might not be detectable in people exposed to TCDD as adults over about 30 years of age.

We found no correlation between TCDD levels in blood and performance of peripheral blood mononuclear cells in the assays. This is not surprising, since the development of chloracne, the major effect of dioxin exposure on humans, in individuals affected by the Seveso accident did not correlate to the severity of TCDD exposure. It is now generally accepted that the Ah receptor mediates dioxin toxicity, and the reason for variability in individual responses is likely due to genetic differences in Ah receptor alleles in different individuals.

We have presented here data where the actual body burden of the individuals analyzed were known. We know of only two other such studies. In accordance with our data, no phenotypic difference between lymphocyte subsets in the exposed and unexposed groups was found, with the exception of CD8 cells, which were slightly increased (21,22). However, the present study is the first to demonstrate reduced immunocompetence of lymphocytes from TCDD-exposed humans.