Modeling and Predicting Selected Immunological Effects of a Chemical Stressor (3,4-Dichloropropionanilide) Using the Area under the Corticosterone Concentration versus Time Curve

Stephen B. Pruett1, Ruping Fan, Qiang Zheng, L. Peyton Myers and Pamela Hébert

Department of Cellular Biology and Anatomy, Louisiana State University Health Sciences Center, Shreveport, Louisiana 71130

Received May 8, 2000; accepted July 27, 2000


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Many chemicals and drugs can induce a neuroendocrine stress response that can be immunosuppressive. Mathematical models have been developed that allow prediction of the immunological impact of such stress responses in mice on the basis of exposure to the important stress-related mediator corticosterone. The area under the corticosterone concentration vs. time curve (AUC) has been used as an indicator of cumulative corticosterone exposure in these modeling studies. In the present study, an immunotoxicant known to induce a stress response, 3,4-dichloropropionanilide (propanil), was evaluated to determine if corticosterone AUC values are related to suppression of immunological parameters in mice treated with this chemical. Linear relationships between corticosterone AUC values and suppression of the following parameters were noted in B6C3F1 female mice: thymus cellularity and thymus subpopulation percentages, splenic subpopulation percentages, natural killer cell activity, MHC class II protein expression, and IgG1 and IgG2a antibody responses to antigen. Linear models derived in previous studies using mice treated with exogenous corticosterone or with restraint stress effectively predicted the immunological effects of 3,4-dichloropropionanilide on the basis of corticosterone AUC values. The models derived using immobilization stress were more effective (r2 for observed vs. predicted = 0.90) than the models derived using mice treated with exogenous corticosterone (r2 for observed vs. predicted = 0.65). This was expected, because most stressors induce a variety of immunomodulatory mediators, not just corticosterone. These findings have implications for risk assessment in immunotoxicology.

Key Words: stress; corticosterone; 3,4-dichloropropionanilide; predictive models.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Exposure of humans or rodents to a wide variety of chemicals and drugs induces a neuroendocrine stress response that can be immunosuppressive (Pruett et al., 1993Go; Pruett et al., 1999Go). In toxicological assessments, chemicals that induce stress responses can be detected in some cases by the observation of adrenal hypertrophy. However, there are no quantitative data that could be used to indicate the degree to which such a stress response may suppress any particular immunological parameter. Toxicological assessments in rodent models often include doses of the test compound that are near the maximum tolerated dose, and these high doses can produce a neuroendocrine stress response that does not occur at lower doses (Brown et al., 1988Go; Clement, 1985Go; Kunimatsu et al., 1996Go). Such nonspecific stress responses may lead to erroneous identification of the test compound as an immunotoxicant. Associations between stress responses and immunosuppression are important because of their implications for immunotoxicity of the test compound in humans. If the immunosuppression noted in animals occurs at only high doses and is caused by the stress response to the compound, this would suggest that the compound is likely to be immunosuppressive in humans only if a similar stress response occurs in persons exposed occupationally or environmentally to the compound.

In an effort to understand the quantitative relationships between stress responses and suppression of selected immunological parameters, this laboratory has obtained data from mice exposed to various dosages of exogenous corticosterone or to immobilization (restraint) stress for various periods of time. Previous work demonstrated that corticosterone accounts for some or all of the suppression of several immunological parameters in mice treated with the potent chemical stressor ethanol (Collier et al., 1998Go; Collier et al., 1999; Han and Pruett, 1995Go; Han et al., 1993Go; Weiss et al., 1996Go; Wu and Pruett, 1997Go). Therefore, corticosterone was selected as the neuroendocrine mediator to be used in these initial modeling efforts. The area under the corticosterone concentration vs. time curve was determined for mice treated with exogenous corticosterone or restraint. Linear models were then developed from these data, and these models indicate a clear relationship between cumulative corticosterone exposure and suppression of several immunological parameters.

The purpose of the present study was to examine a stress-inducing chemical, 3,4-dichloropropionanilide (propanil), to determine if the corticosterone AUC values induced by different dosages of that chemical could be used to predict the amount of suppression of several immunological parameters. Linear models were obtained that allow such predictions. In addition, the results for propanil were compared to results obtained in previous studies for exogenous corticosterone and restraint stress to determine if the immunological effects of propanil more closely matched the effects of exogenous corticosterone or restraint. Our hypothesis was that propanil would induce the whole range of neuroendocrine mediators induced by most stressors (including restraint), and that its effects would therefore more closely resemble the effects of restraint than the effects of exogenous corticosterone.

Propanil was selected as the chemical stressor for these studies because it is a widely used herbicide, and propanil-induced thymic atrophy seems to be mediated by corticosterone. Thymic atrophy occurred in normal mice but not in adrenalectomized mice that were treated with propanil (Cuff et al., 1996Go). Propanil suppresses a number of immunological parameters in mice, and T-dependent antibody responses and thymus cellularity seem to be among the more sensitive parameters (Barnett and Gandy, 1989Go). Some direct effects of propanil have been identified using in vitro techniques, but it has not been determined if these effects are responsible for the alterations in immunological parameters noted in vivo (Xie et al., 1997Go; Zhao et al., 1998Go). Thus, it remains possible that suppression of a number of immunological parameters in vivo may be mediated by the neuroendocrine stress response to propanil.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Mice.
Female C57BL/6xC3H F1 (B6C3F1) mice were obtained from the National Cancer Institute's animal program. These mice were selected because of their extensive use in studies sponsored by the National Toxicology Program. Mice were allowed to recover from shipping stress for at least 2 weeks prior to use in experiments, and they were 8–12 weeks of age at the initiation of experiments. They were maintained on a 12-h light–dark cycle (on 7:00 A.M. and off 7:00 P.M.) and were given food (Purina Lab Chow) and water ad libitum. Sentinel mice housed in the same room as used in this study were negative for antibodies to a panel of adventitious agents common to mice, including mouse hepatitis virus and Sendai virus. Animal care and use was conducted in accord with the National Institutes of Health Guide and the regulations of Lousiana State University Health Sciences Center in a facility that is accredited by the American Association for Accreditation of Laboratory Animal Care.

Administration of propanil.
Propanil was obtained from Chem Service (West Chester, PA). It was administered intraperitoneally at dosages of 50, 75, or 100 mg/kg body weight in food-grade corn oil. In some experiments a dosage of 150 mg/kg was also used. Preliminary experiments demonstrated that the corticosterone response to vehicle administration subsided within 4 h. This yields a negligible corticosterone AUC value, which only serves to add an additional data point near the 0 value for corticosterone AUC to any models developed (Pruett et al., 1999Go). Therefore, naive control groups, not vehicle control groups, were used for these modeling studies.

Determination of corticosterone AUC values for propanil-treated mice.
Seven groups of mice (five mice per group) were treated with propanil at each of the four dosages used in this study (50, 75, 100, or 150 mg/kg body weight) at 9:00 P.M. One group at each dosage was bled by decapitation at 1, 2, 4, 6, 8, 12, and 18 h after propanil administration. Naive control groups were designated for each time point, and these mice remained undisturbed in a separate room from the treated mice until they were bled at the appropriate time point to match each treatment group. Mice were bled in an adjacent room, so neither naive nor treated mice were exposed to odors or noise associated with the bleeding process. Serum was separated from each blood sample and the corticosterone was analyzed using a radioimmunoassay kit (DPC, Los Angeles, CA) (Pruett et al., 1999Go). The data were graphed using DeltaGraph software, scanned to produce a digital image, and the area between the naive control values and the treated values was determined, as described previously, using NIH Image software (Pruett et al., 1999Go).

Measurement of immunological parameters.
Immunological parameters were evaluated exactly as described in our previous studies at the time after administration of the stressor previously shown to produce maximum suppression (Han and Pruett, 1995Go; Pruett et al., 1999Go; Pruett et al., 2000Go; Weiss et al., 1996Go; Wu et al., 1994Go). In all experiments, the group size was five mice per group. Cell suspensions were prepared by pressing the spleen or thymus between the frosted ends of glass microscope slides, and spleen and thymus cellularity were determined using an electronic cell counter (Coulter model Z1, Coulter Electronics, Hialeah, FL). A previous time course study indicated that the greatest suppression of cell number in the spleen and thymus occurred 24 h after the initial administration of corticosterone (Pruett et al., 2000Go). Therefore, this time point was used for the present study both for assessing cell number and examining cellular subpopulations by flow cytometry.

Labeling for flow cytometry was conducted using fluorescent-labeled antibodies specific for CD4 and CD8 from Pharmingen (San Diego, CA) and B220 from Gibco/BRL (Grand Island, NY). Cells were labeled in V-bottom microtiter plates at a concentration of 106 cells per well. An isotype control was used for each antibody. Gates were set using the isotype controls, and less than 2% of the cells labeled with the isotype control were outside the gate defined as negative in any experiment.

Splenic natural killer (NK) cell activity was measured 12 h after initiation of propanil treatment (the same time used previously in experiments with exogenous corticosterone or restraint stress) (Wu et al., 1994Go) using a standard 4-h 51Cr release assay with YAC-1 target cells. Effector-to-target cell ratios of 100:1, 50:1, and 25:1 were used. Values were expressed as lytic units per 107 splenocytes (Bryant et al., 1992Go) to provide a single value for modeling. Expression of MHC class II protein on splenocytes was evaluated using samples of the same splenocyte preparations used for the NK cell lytic assay. Cells were labeled as described above using anti-MHC class II protein (I-Ap,b,k,j,q,r,s) (Pharmingen, San Diego, CA) (Weiss et al., 1996Go).

The IgG1 and IgG2a response to keyhole limpet hemocyanin (KLH, Pierce Chemical Co.) was measured by immunizing mice with KLH (100 µg/mouse) 12 h after treatment with propanil (the same as the optimum time previously identified for treatment with exogenous corticosterone) (manuscript submitted). Mice were then bled (under methyoxyflurane anesthesia) 2 weeks later (the time at which near maximum antibody levels occur) (Kruszewska et al., 1995Go). Serum was diluted 1/25, 1/50, and 1/100, and IgG1 and IgG2a antibodies specific for KLH were measured by ELISA. Wells in 96-well microtiter plates were coated with 100 µl KLH at 10 µg/ml in phosphate-buffered saline (PBS). The antigen solution was removed, and a blocking solution (1% bovine serum albumin in PBS) was added for 2 h at room temperature. Serum (100 µl diluted in PBS with 0.05% Tween 20, referred to hereafter as wash buffer) was added to the wells, and incubated 1.5 h at room temperature; the wells were washed three times with wash buffer using an automated plate washer. Secondary antibody (anti-mouse IgG1 or IgG2a-alkaline phosphatase conjugate, Caltag, Burlingame, CA) diluted 1/1600 in wash buffer was added to each well, and the wells were washed three times after a 1.5-h incubation period at room temperature. One hundred microliters of PNPP alkaline phosphatase substrate solution (Sigma Chemical Co, St. Louis, MO) was added to each well. Absorbance at 405 nm was determined using a plate reader (UV-3550, BioRad, Hercules, CA). Because models developed using all three serum dilutions were very similar, the 1/25 dilution only was used to obtain a single value for each mouse for modeling purposes. Serum from nonimmunized mice was diluted to the same extent as the test serum and included as a control in each experiment. Absorbance values for these controls were always less than 0.05.

Statistical and modeling methods.
Linear regression analysis was performed using the Prism software package (GraphPad, San Diego, CA). Replicate values for immunological parameters within each group were analyzed as separate values, not means. The program was also used to calculate and graphically illustrate the 95% confidence intervals of each line. The runs test (Motulsky, 1994Go) indicated no significant nonlinearity in any of the immunological parameters vs. corticosterone AUC plots. Statistical comparison of pairs of lines was done as described (Zar, 1984Go) and as implemented by Prism software. This approach determines if the slope or the elevation (or Y-intercept) differ significantly between any pair of lines. A variety of nonlinear models were also examined, and in one case (spleen cell number) a nonlinear model yielded a larger correlation coefficient and a plot that more closely represented the apparent threshold effect in this data set. Therefore, the nonlinear model is shown for this parameter.

Linear models of the effects of exogenous corticosterone and restraint stress on the same immunological parameters examined in the present study were developed in previous studies in this lab (Pruett et al., 1999Go; Pruett et al., 2000Go). These models were used to predict the effects of propanil (100 mg/kg). The regression equations were used to calculate the value for the immunological parameter that corresponds with the corticosterone AUC measured in mice treated with propanil at 100 mg/kg (4383 ng/ml •h). The 95% confidence intervals for these calculated values were determined using a table of calculated confidence intervals over the entire range of the regression line, which is calculated by the Prism software package. In some cases, the regression equations and confidence intervals were confirmed using StatView software (SPSS, Chicago, IL), and the results were the same in all cases examined.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Determination of Corticosterone AUC Values for Propanil
The peak corticosterone concentration induced by propanil was similar for all dosages used in this study (~600 ng/ml). This is less than the peak levels induced by exogenous corticosterone (~1100 ng/ml) or restraint (~800 ng/ml) (Pruett et al., 1999Go). The duration of increased corticosterone concentrations in propanil-treated mice was dose dependent (Fig. 1Go). Therefore, the AUC values were also dose dependent. In addition, the AUC values fall within a similar range as noted for restraint for periods of 1–8 h or for administration of exogenous corticosterone at dosages of 9–18 mg/kg (Pruett et al., 1999Go). Because some of the immunological parameters evaluated in this study were assessed 12 h after dosing with propanil (NK cell function and MHC class II expression), AUC values were determined for 12 h. A second set of AUC values was determined for 18 h, at which time corticosterone concentrations were at or near basal levels in propanil-treated mice (Fig. 1Go). Previous studies indicate that handling and injection of vehicle induce a small increase in corticosterone levels (AUC = 303 ng/ml • h) (Pruett et al., 1999Go), which are not sufficient to affect most immunological parameters or even to contribute effectively to linear models. Thus, only naive control groups were used in this study.



View larger version (42K):
[in this window]
[in a new window]
 
FIG. 1. Determination of the area under the corticosterone concentration vs. time curve (AUC). Mice were treated with propanil at the indicated dosages (closed circles), and a group was bled at each of the time points shown. A naive control group was included at each time point (closed squares) to determine basal corticosterone levels. Corticosterone in serum samples was measured by radioimmunoassay, and the values shown are means ± SE (n = five mice per group). Corticosterone AUC values were determined using NIH Image software. The value for 12 h was obtained by drawing a vertical line between the control and treated groups at the 12-h time point. The dotted line in the graph of the 50 mg/kg dosage indicates that the 18-h time point was extrapolated on the basis of the values for the other dosages.

 
Linear Models of Corticosterone AUC versus Thymus and Spleen Cell Number and Subpopulation Percentages in Propanil-Treated Mice
Results shown in Figure 2Go demonstrate that propanil caused a dose-responsive decrease in thymocyte number. There was no significant difference between the slope or the elevation of the regression line for propanil and the line obtained in a previous study for restraint (Pruett et al., 2000Go). However, the elevation was significantly different for the regression line obtained with propanil and the line obtained in a previous study using exogenous corticosterone (Pruett et al., 2000Go). Cell loss was less in mice treated with propanil or restraint than in mice treated with exogenous corticosterone, at equivalent corticosterone AUC values. The percentages of the four major thymocyte subpopulations followed this same general pattern, in which changes induced by propanil were more similar to those induced by restraint than to those induced by exogenous corticosterone. Only in the case of CD4CD8 cells was the line obtained using data from propanil-treated mice significantly different from the line obtained using data from restrained mice. The data were particularly striking in the case of CD4+CD8+ cells, CD4+CD8 cells, and CD4CD8+ cells, in that the effect of corticosterone was very different from the effects of propanil and restraint. As noted previously (Pruett et al., 2000Go), exogenous corticosterone preferentially depleted CD4+CD8+ cells and thus decreased the percentage of these cells and increased the percentage of the other subpopulations. However, restraint decreased all subpopulations nearly equally, resulting in little change in the percentages of each subpopulation. Similarly, propanil did not have substantial effects of the percentages of thymic subpopulations, although it did decrease the total number of thymocytes.



View larger version (36K):
[in this window]
[in a new window]
 
FIG. 2. Linear models of the effects of propanil on thymus cell number and subpopulation percentages. The values shown for propanil are means ± SE (n = five mice per group). The formula for the regression line is shown, and the 95% confidence intervals are indicated by dashed lines. The AUC values used here for each dosage of propanil were obtained as noted in Figure 1Go. The values for cell number in the upper panel and for the percentage of total thymocytes in each of the four major subpopulations were all normalized to percent of the naive control group to facilitate comparison between experiments. The actual values for the naive control group were as follows: cell number, 8.5 x 107 ± 7.8 x 106 cells per thymus; CD4+ CD8+ cells, 83.5 ± 0.7%; CD4+ CD8 cells, 11.54 ± 0.5%; CD4 CD8+ cells, 2.6 ± 0.2%; CD4 CD8 cells, 2.3 ± 0.2%. Data and regression lines obtained in previous studies (Pruett et al., 2000Go) from mice treated with exogenous corticosterone and restraint are shown for comparison.

 
As shown in Figure 3Go, a nonlinear model was more appropriate than a linear model for describing the effects of propanil on the number of nucleated cells in the spleen. Visual identification of threshold or other nonlinear relationships is typically as effective as mathematical methods (Luster et al., 1993Go), and a threshold effect is suggested by visual inspection of the data shown in the upper panel of Figure 3Go. However, the percentage of cellular subpopulations in the spleen was effectively described by linear models. As shown in Figure 3Go, propanil had little effect on the percentage of B220+ cells (mature B lymphocytes) in the spleen. The slope was not significantly different from 0, indicating no significant relationship between the corticosterone AUC and the percentage of B lymphocytes in the spleen in propanil-treated mice. Similarly, propanil did not affect the percentage of CD4+ (Th) lymphocytes (Fig. 3Go) or the percentage of CD8+ (Tc) lymphocytes (Fig. 3Go) in the spleen. This indicates propanil decreased all cell populations similarly, leading to little or no change in the percentage of major subpopulations. Similar effects were noted in a previous study for exogenous corticosterone and restraint (manuscript submitted), and these data are shown for comparison (Fig. 3Go).



View larger version (27K):
[in this window]
[in a new window]
 
FIG. 3. Modeling the effects of propanil on total cell number and percentages of selected subpopulations in the spleen. The values shown for propanil are means ± SE (n = five mice per group). The AUC values used here for each dosage of propanil were obtained as noted in Figure 1Go. A third-order polynomial equation fit the data for spleen cell number better than a linear model. The formula for this polynomial equation and the formulas for the linear regression equations for the other parameters are shown. The 95% confidence intervals are indicated by dashed lines. The AUC values used here for each dosage of propanil were obtained as noted in Figure 1Go. The values for cell number in the upper panel and for the percentage of total splenocytes in each of three major subpopulations were all normalized to percent of the naive control group to facilitate comparison between experiments. The actual values for naive control group were as follows: cell number, 7.3 x 107 ± 3.3 x 106 cells per spleen; B cells, 51 ± 0.4%; CD4+ CD8 cells, 21.5 ± 0.8%; CD4 CD8+ cells, 12.1 ± 0.3%. Data and regression lines obtained in a previous study (Pruett et al., 2000Go) from mice treated with exogenous corticosterone and restraint are shown for comparison.

 
Linear Models of Corticosterone AUC versus Expression of MHC Class II Protein on Splenocytes in Mice Treated with Propanil
As shown in Figure 4Go, propanil dose-responsively decreased the expression of MHC class II molecules on splenocytes. The data are shown as the natural logarithm of the fraction of cells that express the highest levels of MHC class II proteins (expressed as percent of the naive control value) (Pruett et al., 1999Go). For example, a value of 50% indicates that only half as many cells with the highest levels of MHC II molecules are found in treated animals as compared to naive control animals. As a point of reference, 50% corresponds to a natural logarithm value of 3.9. Expressing the data this way provides a better indication of changes in MHC II than reporting the total number of MHC II–positive cells, because most cells do not completely lose MHC II expression in response to stress, but express lower levels (Weiss et al., 1996Go). The total percent of MHC II–positive cells in this study (53.6 ± 0.8 for the naive control group) corresponded well to reported percentages of B lymphocytes in the spleen (Holsapple et al., 1988Go). Because almost all B cells in the spleen are MHC II positive (Melchers and Rolink, 1999Go), this value seems appropriate. The decrease in MHC class II expression was very similar in mice treated with propanil as reported previously for mice treated with exogenous corticosterone or restraint (at equivalent AUC values) (Pruett et al., 1999Go). The linear models obtained using these three agents were compared statistically. There was a significant difference in the elevation of the lines for the restraint and propanil models, but the exogenous corticosterone and propanil models were not significantly different. However, visual inspection indicates that all three treatments yielded very similar linear models (Fig. 4Go).



View larger version (33K):
[in this window]
[in a new window]
 
FIG. 4. Linear models of the effects of propanil on the expression of MHC class II proteins on splenocytes and examination of the effect of exogenous corticosterone on expression of MHC class II proteins on peripheral blood mononuclear cells. The values shown for propanil are means ± SE (n = five mice per group). The formula for the regression line is shown, and the 95% confidence interval is indicated by dashed lines. The AUC values used here for each dosage of propanil were obtained as noted in Figure 1Go. The values were normalized to percent of the naive control group to facilitate comparison between experiments, and the percentage of splenocytes that expressed the hightest level of MHC II protein was 20.5 ± 1.3. Data and regression lines obtained in previous studies (Pruett et al., 1999Go) from mice treated with exogenous corticosterone and restraint are shown for comparison. The lower panels illustrate flow cytometry of peripheral blood mononuclear cells to determine if exogenous corticosterone (one dose at 18 mg/kg) decreases expression of MHC II protein on these cells as well as splenocytes. Values in the bar graph represent mean ± SE for groups of five mice, and a statistically significant difference between control and treatment groups is indicated by * (p < 0.05).

 
The effect of exogenous corticosterone (18 mg/kg, one dose) on MHC class II expression on peripheral blood lymphocytes was also examined. This was not done to develop a mathematical model, but to establish that this parameter is affected by corticosterone and would therefore be suitable for evaluation and modeling in future studies in humans. Results shown in Figure 4Go also demonstrate that corticosterone significantly decreases MHC II protein expression on peripheral blood lymphocytes in mice. Exogenous corticosterone (one dose at 18 mg/kg) decreased the percentage of cells expressing the highest level of MHC class II protein to 34% of the naive control value on peripheral blood mononuclear cells (Fig. 4Go, lower right panel), as compared to 36% on splenocytes (Fig. 4Go, upper panel).

Linear Models of Corticosterone AUC versus NK Cell Activity in Mice Treated with Propanil
Data shown in Figure 5Go indicate that propanil dose-responsively decreases splenic NK cell activity. For modeling purposes, the highest dosage of propanil is not shown. It induced only a slight decrease in NK cell lytic function (26 ± 4 LU/107 splenocytes compared to 30 ± 3 for the control group) that was not consistent with the dose-responsive decreases produced by the lower dosages. This may represent NK cell activation associated with increased spleen weight and cellularity reported after high dosages of propanil (Barnett and Gandy, 1989Go), or it may represent direct activation of NK cells by propanil at high dosages, which overcomes the suppression induced by stress-related mediators. However, the effect of propanil on NK cell function at dosages of 0, 50, 75, and 100 mg/kg yielded a linear model that was used to describe suppression of NK activity within this dosage range. Linear models obtained in a previous study (Pruett et al., 1999Go) using mice treated with exogenous corticosterone or restraint were consistent with data from mice treated with propanil at 50, 75, and 100 mg/kg. The line obtained using data from propanil-treated mice was not significantly different from the line obtained using data from restrained mice, but it was different from the line obtained from corticosterone treated mice (the elevations of these lines differ significantly; p = 0.0015). The elevation of the propanil line was lower than the exogenous corticosterone line, suggesting greater suppression of NK cell function by propanil than by an equivalent exposure to corticosterone alone.



View larger version (28K):
[in this window]
[in a new window]
 
FIG. 5. Linear models of the effects of propanil on splenic NK cell activity. The values shown for propanil are means ± SE (n = 5 mice per group). The formula for the regression line is shown, and the 95% confidence interval is indicated by dashed lines. The AUC values used here for each dosage of propanil were obtained as noted in Figure 1Go. The values were normalized to percent of the naive control group to facilitate comparison between experiments, and the actual lytic unit value for the naive control group was 29.8 ± 3.3. Data and regression lines obtained in previous studies (Pruett et al., 1999Go) from mice treated with exogenous corticosterone and restraint are shown for comparison.

 
Linear Models of Corticosterone AUC versus IgG1 and IgG2a Antibody Responses to KLH in Mice Treated with Propanil
Propanil dose-responsively decreased the IgG1 and IgG2a responses to KLH, as shown in Figure 6Go. Data obtained in a previous study (manuscript submitted) using mice treated with exogenous corticosterone or restraint are shown for comparison. The IgG1 response to KLH was affected similarly by propanil and restraint, at equivalent corticosterone AUC values. However, exogenous corticosterone had a smaller effect on the IgG1 response to KLH than either propanil (the elevation was significantly different, p = 0.011) or restraint (the elevation was significantly different; p = 0.001). The IgG2a response to KLH was suppressed more by propanil than either restraint or exogenous corticosterone. However, the difference in elevation was greater for propanil and exogenous corticosterone (p = 0.00033) than for propanil and restraint (p = 0.012).



View larger version (24K):
[in this window]
[in a new window]
 
FIG. 6. Linear models of the effects of propanil on the IgG1 and IgG2a antibody responses to KLH. The values shown for propanil are means ± SE (n = five mice per group). The formulas for the regression lines are shown, and the 95% confidence intervals are indicated by dashed lines. The AUC values used here for each dosage of propanil were obtained as noted in Figure 1Go. The values were normalized to percent of the naive control group to facilitate comparison between experiments. The actual absorbance values for the naive control groups were IgG1, 0.466 ± 0.006, and IgG2a, 1.162 ± 0.057. In this experiment, naive control means that the mice were not treated with corticosterone, restraint, or propanil, but all groups were immunized. Data and regression lines obtained in a previous study (manuscript submitted) from mice treated with exogenous corticosterone and restraint are shown for comparison.

 
Use of Linear Models from Mice Treated with Exogenous Corticosterone or Restraint to Predict Immunological Effects in Mice Treated with Propanil
The linear regression equations derived in previous studies (Pruett et al., 1999Go; Pruett et al., 2000Go) or from results that have been submitted for review (antibody responses to KLH) were used to predict the effect of propanil at 100 mg/kg on several immunological parameters (Table 1Go). As indicated in Figure 1Go, a corticosterone AUC value of 4383 ng/ml • h was used for parameters measured after 12 h, and a value of 3725 ng/ml • h was used for parameters measured at 12 h (NK cell activity and MHC class II expression). In most cases, the observed values were within or near the 95% confidence intervals for the predicted values. A graphical representation of observed and predicted values is shown in Figure 7Go. There was a strong correlation between observed values and values predicted using the restraint models (r2 = 0.90). The relationship between the observed values and the values predicted using the exogenous corticosterone was not as strong (r2 = 0.65). The difference was particularly evident in the case of immunological parameters for which exogenous corticosterone and restraint yielded significantly different models (the IgG1 and IgG2a responses to KLH, thymus cell number, and the percentages of all four major cellular subpopulations in the thymus). For these immunological parameters, the models derived from restrained mice provided more accurate predictions of the effects of propanil than did the models derived from mice treated with exogenous corticosterone.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Prediction of the Effects of Propanil on Selected Immunological Parameters on the Basis of the Corticosterone AUC Value Measured in Propanil-Treated Mice
 


View larger version (23K):
[in this window]
[in a new window]
 
FIG. 7. Correlation between observed effects of propanil on immunological parameters and effects predicted using models derived from mice treated with exogenous corticosterone or restraint stress. The data shown in Table 1Go were subjected to regression analysis to determine the correlation coefficients.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The results reported here demonstrate the feasibility of using measurements of a single neuroendocrine mediator (corticosterone) to predict a broad range of immunological effects induced a by chemical stressor. Increased concentrations of cortisol or other glucocorticoids are associated with suppression in humans of many of the same immunological parameters as evaluated here in mice (Asadullah et al., 1995Go; Blazar et al., 1986Go; Cupps et al., 1984Go; Hanson et al., 1999Go). Cortisol AUC values were not determined in these human studies, but cortisol AUC would not be difficult to measure in human subjects. In addition, data obtained in the present study indicate that one of the immunological parameters measured (MHC class II expression) is affected similarly by exogenous corticosterone in lymphocytes from the spleen and from the blood (Fig. 4Go). Another study noted similar results for other immunological parameters in rodents (White et al., 1991Go). Thus, it seems feasible to develop linear models describing the effects of cortisol on a wide range of immunological parameters using peripheral blood mononuclear cells from human subjects, and to obtain the same type of data for mice. This would allow a conversion factor to be determined that would permit immunosuppression in humans to be predicted on the basis of corticosterone AUC values in mice. Of course, it would also be possible to predict immunological consequences on the basis of cortisol measurements in persons acutely exposed to a chemical or drug. Completion of a mouse-human risk assessment parallelogram (Selgrade et al., 1995Go) for cortisol/corticosterone and selected immunological parameters may have broader application than the completion of a parallelogram relating in vitro immunological effects in humans and mice and in vivo effects in mice for one chemical at a time (Selgrade et al., 1995Go). Numerous drugs and chemicals can act as stressors (summarized in Pruett et al., 1999), and mathematical models relating the cortisol/corticosterone AUC values in mice and humans could be used to extrapolate from mice to humans for a wide variety of compounds. The major limitation of this approach would be that some chemicals clearly act as direct immunotoxicants or act through indirect mechanisms unrelated to the neuroendocrine stress axis. However, models of the type described here should allow at least the minimum expected immunological effects (the portion due to the stress response alone) to be determined.

Although glucocorticoid-induced changes in lymphocyte trafficking could be involved in the changes in cell types both in the spleen and in the blood, this seems unlikely. Glucocorticoid-induced changes in lymphocyte trafficking typically occur very quickly, and values return to normal within 4 h (Dhabhar et al., 1995Go). In addition, the finding that MHC class II protein expression is decreased similarly at 12 h in the spleen and in peripheral blood (Fig. 4Go) is more consistent with decreased MHC II expression than with altered trafficking. Altered trafficking would be a more reasonable explanation if an increase in MHC II expression was noted in the blood, suggesting that the cells lost from the spleen were mobilized to the blood.

Only one of the immunological parameters (the IgG2a response to KLH) evaluated in the present study was suppressed by propanil to a significantly greater extent than predicted from corticosterone AUC values using either restraint or exogenous corticosterone models. This suggests that most of the effects of propanil noted here could have been mediated by the neuroendocrine stress response to this compound. The additional suppression of the IgG2a response to KLH may be a consequence of the ability of propanil to directly suppress the production of some cytokines (Zhao et al., 1999Go). In addition, it should be noted that propanil exerts immunosuppressive effects that were not evaluated in the present study, and it remains possible that some of these effects are mediated directly by propanil or indirectly by mechanisms that do not involve the neuroendocrine stress response. The failure of the highest dosage of propanil to suppress NK cell activity in a manner consistent with the suppression expected from corticosterone exposure suggests that some dosages of some immunotoxicants may counteract the effects of stressors, leading to nonlinear (biphasic) relationships. Additional data are needed to determine how common such occurrences will be, but such departures from linearity can be easily identified by visual inspection or by mathematical methods (Luster et al., 1993Go). In any case, the effectiveness of the restraint models in predicting important immunological effects of a chemical immunotoxicant suggests that further development of this approach is warranted.

The basis for the nonlinear effect of propanil on spleen cell number is not clear. A number of immunotoxic chemicals can induce hematopoiesis, and it is possible that increased hematopoiesis in the spleen overcomes the effect of corticosterone at low propanil dosages. In any case, the departure from a linear model is informative. It demonstrates that the effects of corticosterone alone or in combination with the other neuroendocrine mediators induced by restraint are not the only factors involved in the effects of propanil on spleen cellularity.

It is not surprising that models developed using data from mice subjected to restraint stress provide better predictions of the effect of propanil on a number of immunological parameters than models developed using data from mice treated with exogenous corticosterone. We had hypothesized that this would be the case, because a wide variety of stressors induce similar (though not identical) patterns of increased concentrations of several neuroendocrine mediators (Pacák et al., 1998Go). In contrast, mice treated with exogenous corticosterone presumably only exhibit changes in these neuroendocrine mediators for a few minutes as a result of the brief period of handling and restraint involved in dosing. Our previous studies and studies from other labs demonstrate that some immunological parameters can be affected by corticosterone and other neuroendocrine mediators (Ader et al., 1990Go; Wu and Pruett, 1997Go), and that interactions between mediators can occur. For example, the presence of restraint stress-induced mediators apparently modifies the effect of corticosterone on the thymus so that all subpopulations are depleted equally, without preferential loss of CD4+CD8+ cells noted in mice treated with exogenous corticosterone (Pruett et al., 1999Go). Fortunately, these mechanistic considerations do not preclude the use of modeling on the basis of corticosterone AUC alone. Data presented here demonstrate that corticosterone AUC measured in restrained mice effectively predicts suppression of several important immunological parameters. Presumably, the other neuroendocrine mediators that contribute to these effects exhibit covariance with corticosterone AUC such that corticosterone AUC acts as a suitable surrogate for these mediators for modeling purposes. Considering that the data for modeling the effects of restraint and data for modeling the effects of propanil were obtained during a 3-year time period and that different personnel were involved, the results seem quite robust.


    ACKNOWLEDGMENTS
 
This work was supported by NIEHS grant ES09158, and S.B.P. is supported by a Research Career Development award from NIAAA.


    NOTES
 
1 To whom correspondence should be addressed at Department of Cellular Biology and Anatomy, Louisiana State University Health Sciences Center, 1501 Kings Highway, Shreveport, Louisiana 71130. Fax: (318) 675-5889. E-mail: spruet{at}lsumc.edu. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Ader, R., Felten, D., and Cohen, N. (1990). Interactions between the brain and the immune system. Ann. Rev. Pharmacol. Toxicol. 30, 561–602.[ISI][Medline]

Asadullah, K., Woiciechowsky, C., Docke, W. D., Egerer, K., Kox, W. J., Vogel, S., Sterry, W., and Volk, H. D. (1995). Very low monocytic HLA-DR expression indicates high risk of infection-immunomonitoring for patients after neurosurgery and patients during high dose steroid therapy. Eur. J. Emerg. Med. 2, 184–190.[Medline]

Barnett, J. B., and Gandy, J. (1989). Effect of acute propanil exposure on the immune response of C57Bl/6 mice. Fundam. Appl. Toxicol. 12, 757–764.[ISI][Medline]

Blazar, B. A., Rodrick, M. L., O`Mahony, J. B., Wood, J. J., Bessey, P. Q., Wilmore, D. W., and Mannick, J. A. (1986). Suppression of natural killer-cell function in humans following thermal and traumatic injury. J. Clin. Immunol. 6, 26–36.[ISI][Medline]

Brown, L. D., Wilson, D. E., and Yarbrough, J. D. (1988). Alterations in the hepatic glucocorticoid response to mirex treatment. Toxicol. Appl. Pharmacol. 92, 203–213.[ISI][Medline]

Bryant, J., Day, R., Whiteside, T. L., and Herberman, R. B. (1992). Calculation of lytic units for the expression of cell-mediated cytotoxicity. J. Immunol. Methods 146, 91–103.[ISI][Medline]

Clement, J. G. (1985). Hormonal consequences of organophosphate poisoning. Fundam. Appl. Toxicol. 5, S61–S67.[ISI][Medline]

Collier, S. D., Wu, W.-J., and Pruett, S. B. (1998). Endogenous glucocorticoids induced by a chemical stressor (ethanol) cause apoptosis in the spleen in B6C3F1 female mice. Toxicol. Appl. Pharmocol. 148, 176–182.[ISI][Medline]

Collier, S. D., Wu, W.-J., and Pruett, S. B. (2000). Ethanol suppresses NK cell activation by polyinosinic-polycytidilic acid (poly I:C) in female B6C3F1 mice: Role of endogenous corticosterone. Alcohol. Clin. Exp. Res. 24, 291–299.[ISI][Medline]

Cuff, C., Zhao, W., Nukui, T., Schafer, R., and Barnett, J. B. (1996). 3,4-dichloropropionanilide-induced atrophy of the thymus: mechanisms of toxicity and recovery. Fundam. Appl. Toxicol. 33, 83–90.[ISI][Medline]

Cupps, T. R., Edgar, L. C., Thomas, C. A., and Fauci, A. S. (1984). Multiple mechanisms of B cell immunoregulation in man after administration of in vivo corticosteroids. J. Immunol. 132, 170–175.[Abstract/Free Full Text]

Dhabhar, F. S., Miller, A. H., McEwen, B. S., and Spencer, R. L. (1995). Effects of stress on immune cell distribution. Dynamics and hormonal mechanisms. J. Immunol., 154, 5511–5527.[Abstract/Free Full Text]

Han, Y. C., Lin, T. L., and Pruett, S. B. (1993). Thymic atrophy caused by ethanol in a mouse model for binge drinking: involvement of endogenous glucocorticoids. Toxicol. Appl. Pharmacol. 123, 16–25.[ISI][Medline]

Han, Y.-C., and Pruett, S. B. (1995). Mechanisms of ethanol-induced suppression of a primary antibody response in a mouse model for binge drinking. J. Pharmacol. Exp. Ther. 275, 950–957.[Abstract]

Hanson, J. A., Sohaib, S. A., Newell-Price, J., Islam, N., Monson, J. P., Trainer, P. J., Grossman, A., Besser, G. M., and Reznek, R. H. (1999). Computed tomography appearance of the thymus and anterior mediastinum in active Cushing`s syndrome. J. Clin. Endocrinol. Metab. 84, 602–605.[Abstract/Free Full Text]

Holsapple, M. P., White, K. L., Jr., McCay, J. A., Bradley, S. G., and Munson, A. E. (1988). An immunotoxicological evaluation of 4,4'-thiobis-(6-t-butyl-m-cresol) in female B6C3F1 mice. 2. Humoral and cell-mediated immunity, macrophage function, and host resistance. Fundam. Appl. Toxicol. 10, 701–716.[ISI][Medline]

Kruszewska, B., Felten, S. Y., and Moynihan, J. A. (1995). Alterations in cytokine and antibody production following chemical sympathectomy in two strains of mice. J. Immunol. 155, 4613–4620.[Abstract]

Kunimatsu, T., Kamita, Y., Isobe, N., and Kawasaki, H. (1996). Immunotoxicological insignificance of fenitrothion in mice and rats. Fundam. Appl. Toxicol. 33, 246–253.[ISI][Medline]

Luster, M. I., Portier, C., Pait, D. G., Rosenthal, G. J., Germolec, D. R., Corsini, E., Blaylock, B. L., Pollock, P., Kouchi, Y., Craig, W., White, K. L., Munson, A. E., and Comment, C. E. (1993). Risk assessment in immunotoxicology. II. Relationships between immune and host resistance tests. Fundam. Appl. Toxicol. 21, 71–82.[ISI][Medline]

Melchers, F., and Rolink, A. (1999). B-lymphocyte development and biology. In Fundamental Immunology (Eds.), pp. 183–224. Lippincott-Raven Publishers, Philadelphia.

Motulsky, H. (1994). Prism Software User's Manual. GraphPad, Inc.: San Diego:

Pacák, K., Baffi, J. S., Kvetnansky, R., Goldstein, D. S., and Palkovits, M. (1998). Stressor-specific activation of catecholaminergic systems: implications for stress-related hypothalamic-pituitary-adrenocortical responses. Adv. Pharmacol. 42, 561–564.[Medline]

Pruett, S. B., Collier, S., Wu, W.-J., and Fan, R. (1999). Quantitative relationships between the supression of selected immunological parameters and the area under the corticosterone concentration vs. time curve in B6C3F1 mice subjected to exogenous corticosterone or to restraint stress. Toxicol. Sci. 49, 272–280.[Abstract]

Pruett, S. B., Ensley, D. K., and Crittenden, P. (1993). The role of chemical-induced stress responses in immunosuppression: A review of quantitative associations and cause-effect relationships between chemical-induced stress responses and immunosuppression. J. Toxicol. Environ. Health 39, 163–192.[ISI][Medline]

Pruett, S. B., Fan, R., Myers, L. P., Wu, W.-J., and Collier, S. D. (2000). Quantitative analysis of the neuroendocrine-immune axis: linear modeling of the effects of exogenous corticosterone and restraint stress on lymphocyte subpopulations in the spleen and thymus in female B6C3F1 mice. Brain, Behavior, and Immunity (in press).

Selgrade, M., Cooper, K. D., Devlin, R. B., van Loveren, H., Biagini, R. E., and Luster, M. I. (1995). Immunotoxicity—bridging the gap between animal research and human health effects. Fundam. Appl. Toxicol. 24, 13–21.[ISI][Medline]

Weiss, P. A., Collier, S. D., and Pruett, S. B. (1996). Role of glucocorticoids in ethanol-induced decreases in expression of MHC class II molecules on B cells and selective decreases in spleen cell number. Toxicol. Appl. Pharmacol. 139, 153–162.[ISI][Medline]

White, K. L., Murray, M. J., Butterworth, L. M., McCay, J. A., Wang, J., Musgrove, D. L., Brown, R. B., Stern, M. L., and Munson, A. E. (1991). Comparison of immunocompetence of peripheral blood and spleen cells in cyclophosphamide (CTX) treated Fischer 344 rats. Toxicologist 11, 203.

Wu, W.-J., and Pruett, S. B. (1997). Involvement of catecholamines and glucocorticoids in ethanol-induced suppression of splenic natural killer cell activity in a mouse model for binge drinking. Alcohol. Clin. Exp. Res. 21, 1030–1036.[ISI][Medline]

Wu, W.-J., Wolcott, R. M., and Pruett, S. B. (1994). Ethanol decreases the number and activity of splenic natural killer cells in a mouse model for binge drinking. J. Pharmacol. Exp. Ther. 271, 722–729.[Abstract]

Xie, Y. C., Schafer, R., and Barnett, J. B. (1997). Inhibitory effect of 3,4-dichloro-propionaniline on cytokine production by macrophages is associated with LPS-mediated signal transduction. J Leukoc. Biol. 61, 745–752.[Abstract]

Zar, J. H. (1984). Simple Linear Regression. In Biostatistical Analysis (Ed.), pp. 261–291. Prentice-Hall, Inc., Englewood Cliffs, NJ.

Zhao, W., Schafer, R., and Barnett, J. B. (1998). Cytokine production by C57BL/6 mouse spleen cells is selectively reduced by exposure to propanil. J. Toxicol. Environ. Health 55, 107–120.[ISI]

Zhao, W., Schafer, R., and Barnett, J. B. (1999). Propanil affects transcriptional and posttranscriptional regulation of IL-2 expression in activated El-4 cells. Toxicol. Appl. Pharmacol. 154, 153–159.[ISI][Medline]