Modeling and Predicting Stress-Induced Immunosuppression in Mice Using Blood Parameters

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

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

1 To whom correspondence should be addressed. Fax: (318) 675-5889. E-mail: spruet{at}lsuhsc.edu.

Received August 11, 2004; accepted October 19, 2004


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Previous studies have shown that the area under the corticosterone concentration vs. time curve (AUC) can be used to model and predict the effects of restraint stress and chemical stressors on a variety of immunological parameters in the mouse spleen and thymus. In order to complete a risk assessment parallelogram, similar data are needed with blood as the source of immune system cells, because this is the only tissue routinely available from human subjects. Therefore, studies were conducted using treatments for which the corticosterone AUC values are already known: exogenous corticosterone, restraint, propanil, atrazine, and ethanol. Immunological parameters were measured using peripheral blood from mice treated with a series of dosages of each of these agents. Flow cytometry was used to quantify MHC II, B220, CD4, and CD8 cells. Leukocyte and differential counts were done. Spleen cell number and NK cell activity were evaluated to confirm similarity to previous studies. Immune parameter data from mouse blood indicate that MHC II expression has consistent quantitative relationships to corticosterone AUC values, similar to but less consistent than those observed in the spleen. Other immune parameters tended to have greater variability in the blood than in the spleen. The pattern observed in the spleen in which the chemical stressors generally produced very similar effects as noted for restraint stress (at the same corticosterone AUC values) was not observed for blood leukocytes. Nevertheless, MHC class II expression seems to provide a reasonably consistent indication of stress exposure in blood and spleen.

Key Words: modeling; stress; biomarker; ethanol; atrazine; propanil.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Guidance documents have been released by the U.S. Environmental Protection Agency and Food and Drug Administration for immunotoxicity safety testing of chemicals and drugs in rodents (Anonymous, 1998Go, 1999Go). In assessing the safety of chemicals or drugs, it is recommended that some animals receive a relatively high dosage of the agent being tested. Immunotoxicity can then be quantitatively assessed on the basis of changes in histological characteristics, blood parameters, or a few immune parameters that have previously been demonstrated to be good indicators of immunotoxicity (Luster et al., 1987Go, 1992Go, 1993Go). However, the guidance documents do not include specific recommendations for distinguishing genuine immunotoxicity from immunotoxicity that is secondary to a generalized stress response. Because many drugs and chemicals administered at high dosages induce immunosuppressive stress responses (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go), methods are needed to allow identification of stress-induced immunosuppression, preferably without the assessment of multiple additional parameters or the use of more rodents. If consistent quantitative patterns of change are seen in immune parameters as stressor dosages increase, then these patterns could be used to suggest the presence of a stress effect.

It has been established that stress can affect immune function through the activation of the hypothalamic pituitary adrenal axis resulting in the production of a number of neuroendocrine mediators (Riley, 1981Go; Zwilling et al., 1993Go). Some of these mediators, such as corticosterone (in rodents) or cortisol (in humans), have been shown to be immunosuppressive in both rodents and humans (Dhabhar et al., 1994Go). Plasma corticosterone levels have been used for many years as an indicator of stress in mice. The effect that a stress response has on immunological parameters can be quantified by relating immunosuppression to the quantity and duration of the stress response represented by the area under the corticosterone concentration vs. time curve (AUC) (Pruett et al., 1999Go). Linear regression analysis can then be used to indicate correlations between AUC and immunosuppression of each immunological parameter examined. Quantitatively consistent results showing similar effects on several parameters by both chemical and physical stressors, at comparable corticosterone AUC values have been demonstrated in a series of studies using spleen and thymus (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). In the present study, similar procedures were used to determine if an immunological biomarker for stress can be identified using blood samples instead of spleen or thymus. Finding such a biomarker would improve the accuracy of risk assessment in humans because blood is the only tissue routinely available for such comparative studies.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Animal care. Female B6C3F1 mice, obtained through the National Cancer Institute's Animal Program, were housed and treated according to NIH and LSUHSC-S guidelines in a facility accredited by the American Association for Accreditation of Laboratory Animal care. Mice were allowed to acclimate to the facility and recover from shipping stress for at least two weeks. Mice were given food (Purina Lab Chow) and water ad libitum and were maintained on a 12-h light/dark cycle. They were used in experiments at 8–12 weeks of age.

General dosing. Five cages of five mice were used for each chemical or restraint stressor. Each cage comprised one of five dosage groups. One of these cages of mice was used as a control group that was either naive (untreated) or treated with the vehicle for the chemical that was administered. All doses were administered at 2130–2200 h to allow for corticosterone circadian rhythms to be similar with those used to calculate AUC values (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go).

Restraint. The treated groups were placed in 50 ml conical tubes for 2, 4, 6, or 8 h. These conical tubes were ventilated by a longitudinal slit used to carefully pull each mouse into the tube. After placing each mouse in its tube, the tubes were placed back in the cage for the duration of the restraint period. The mice in the groups receiving the two highest dosages of restraint (6 and 8 h) were allowed access to food and water for 10 min at the 4-h time point.

Corticosterone dosing. A corticosterone (Sigma, St. Louis, MO) suspension was made in a vehicle of phosphate buffered saline (Sigma, St. Louis, MO) containing 2% ß-cyclodextrin (Sigma, St. Louis, MO). Corticosterone doses were administered via sc injections at concentrations of 9 mg/kg, one dose at 18 mg/kg, two doses at 18 mg/kg (2 h apart), or three doses at 18 mg/kg (2 h apart).

Propanil dosing. A propanil (Chem Service, West Chester, PA) suspension was made by mixing the chemical in corn oil (Mazola). Propanil was given by ip injection at dosages of 50, 75, 100, or 150 mg/kg (in 0.2 ml).

Atrazine dosing. Atrazine (Chem Service, West Chester, PA) was mixed in corn oil (Mazola) and administered by ip injection at dosages of 75, 150, 225, or 300 mg/kg (in 0.2 ml).

Ethanol dosing. Ethanol (AAPER Alcohol Chemical Co., Shelbyville, KY) used for dosing was diluted in sterile tissue culture-grade water (Sigma, St. Louis, MO) to 32% by volume. Ethanol was given by po gavage at dosages of 4, 5, 6, or 7 g/kg.

Blood cell harvesting. Mice were placed under halothane anesthesia. Their blood was collected in heparin coated tubes (Becton Dickinson and Company, Franklin Lakes, NJ) by bleeding from the retrorbital plexus 12 h after dosing.

Flow cytometric analysis. Blood from each mouse (0.15 ml) was labeled with anti-MHC II FITC (BD Pharmingen, San Diego, CA) and anti-CD45R/B220 (BD Pharmingen, San Diego, CA) by adding 5 µl of each antibody diluted 1/10 in FACS buffer (phosphate buffered saline with 0.1% bovine serum albumin and 0.1% sodium azide pH 7.4). Another blood sample from each mouse was labeled with 5 µl anti-CD4 PE (BD Pharmingen, San Diego, CA) and 5 µl anti-CD8 Cychrome (BD Pharmingen, San Diego, CA) diluted 1/10 in FACS buffer. These samples were allowed to incubate at 4°C in the dark for 30 min. After incubation, RBCs were lysed by adding 8 ml of ammonium chloride buffer (4.13 g NH4Cl, 0.5 g NaHCO3, 0.03 g EDTA per 500 ml water, pH 7.0) warmed to 37°C. The samples were allowed to incubate at 37°C for 10 min. They were then washed with FACS buffer. Following one wash the samples were resuspended in 1% paraformaldehyde (in PBS) and incubated in the dark for 10 min at room temperature. The samples were washed twice using FACS buffer, resuspended, and stored at 4°C in FACS buffer until analyzed by flow cytometry (FAC Calibor BD, Franklin Lakes, NJ) no more than five days after fixation.

Spleen cell counts. Each spleen was removed and placed in 3 ml of RPMI (Invitrogen, Carlsbad, CA), which was kept on ice. Frosted slides were then used to press the spleens yielding single cell suspensions in RPMI. The cells were centrifuged at 300 x g for 7 min, resuspended in 3 ml of RPMI 1640, and 20 µl of each cell suspension was added to 10 ml of isoton (Beckman Coulter, Miami, FL) in a counting vial. Three drops of Manual Lyse reagent (J&S Medial Associates Inc., Framingham, MA) were added, and the samples were then counted using a Coulter Z1 counter (Coulter Corporation, Miami, FL).

White blood cell counts. White blood cells were counted by placing 20 µl of whole blood in 10 ml of isoton (Beckman Coulter, Miami, FL). Manual Lyse Reagent (J&S Medial Associates Inc., Framingham, MA) was then added to lyse erythrocytes. The samples were counted using a Coulter Z1 counter (Coulter Corporation, Miami, FL).

NK assay. Spleen cells were diluted to 1.0 x 107 cells/ml and were plated in a 96 well v-bottom plate using triplicate samples for at least three effector/target ratios. YAC-1 target cells were labeled with 51Cr (ICN Biomedicals, Irvine, CA), then diluted to 1 x 105 cells/ml and plated in a 96 well plate, and incubated for 4 h at 37°C. A gamma counter (Perkin-Elmer, Wellesley, MA) was then used to measure release of 51CR into culture supernatants, as an indication of NK cell activity. The assay and calculation of lytic units were carried out as described previously (Pruett et al., 1999Go).

Differential counts. Manual differential counts were made using blood smears for each animal. These slide were then stained using a Diff-Quik three step staining kit (Dade Behring Inc., Newark, DE). Cells were then counted by observation under the microscope. At least 100 cells were counted for each sample.

Statistical analysis. Statistical analysis was carried out by using Microsoft Excel to first normalize data to control values to facilitate comparison of results from different experiments. Data were then transferred to Prism Graph Pad 4.0 (San Diego, CA) where linear regression models were generated by comparing the change in each immune parameter to the area under the corticosterone AUC value corresponding to the dose that caused the change. These AUC values were obtained in our previously published studies (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go), which indicate that the values are quite reproducible (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). Each regression generated was analyzed using the "runs test" to determine if there was a significant nonlinear component in the regression models. None was detected for any of the data shown. The linear regression models were then compared by evaluating overlap in their 83.7% confidence intervals. These confidence intervals were calculated using StatView software (v4.5 for Macintosh). Overlap of the 83.7% confidence intervals for linear regression models indicate that the values are not significantly different at the p = 0.05 level: lack of overlap indicates significance at the 0.05 level (Barr, 1969Go; Nelson, 1989Go).


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Effects of Stress on Blood Lymphocyte and Neutrophil Populations
Differential cell counts shown in Figures 1 and 2 demonstrate that increases in plasma corticosterone are associated with a decrease in the number of lymphocytes and an increase in the number of neutrophils present in the blood. These results show that restraint had a smaller effect on the lymphocyte to neutrophil ratio than did atrazine, propanil, exogenous corticosterone, or ethanol. Direct comparisons of the regression lines using Prism 4.0 software indicated that the lines for restraint were significantly different from the corresponding lines for all other treatments. However, there was also considerable variability among these other treatments. For example, ethanol induced a much greater increase in neutrophils than did exogenous corticosterone at similar AUC levels (Fig. 2).



View larger version (36K):
[in this window]
[in a new window]
 
FIG. 1. Effects of stress on lymphocyte populations in the blood. Changes in lymphocyte populations in peripherial blood samples were determined through differential counts. Each of the graphs represents the percentage of lymphocytes normalized to express the average percentage of lymphocytes in the control group for each stressor as 100. The actual values of percent lymphocytes for control groups in each experiment were: exogenous corticosterone, 88 ± 1.63, n = 4; restraint, 90.4 ± 1.54, n = 5; EtOH, 88.6 ± 1.96, n = 5; atrazine, 85 ± 1.64, n = 5; propanil, 78.6 ± 2.56, n = 5. The table shows the relationships between the slopes and intercepts for the linear regressions generated for atrazine (ATZ), propanil (Prop), ethanol (EtOH), exogenous corticosterone (Cort), and restraint (Rest).

 


View larger version (37K):
[in this window]
[in a new window]
 
FIG. 2. Effects of stress on neutrophil populations in the blood. The number of neutrophils in peripherial blood samples increased as animals were treated with increasing dosage or duration of each stressor. The data points on these graphs represent the percentage of neutrophils observed in differential counts normalized to the average number of neutrophils seen in the control groups. The actual raw percentage values for naive control groups from each experiment follow: exogenous corticosterone, 11.75 ± 1.65, n = 4; restraint, 8.8 ± 1.62, n = 5; EtOH, 8.8 ± 2.2, n = 5; atrazine, 11.6 ± 1.66, n = 5; propanil, 18.8 ± 2.48, n = 5. The table shows the relationships between the slopes and intercepts for each of the linear regressions generated for atrazine (ATZ), propanil (Prop), ethanol (EtOH), exogenous corticosterone (Cort), and restraint (Rest).

 
Since the percentage of lymphocytes and neutrophils had been examined separately, it was important to assess the effect stress had on the leukocyte population as a whole. It has been reported that the total number of white blood cells in animals decreases when they are subjected to stress (Li et al., 2001Go). This decrease has been attributed to increases in adrenal hormones including plasma corticosterone levels (Cunnick et al., 1990Go). Figure 3 shows the relationship between white blood cell number and increasing dosages of each stressor. The slope indicates decreasing WBC counts with increasing dosage of ethanol, atrazine, and corticosterone. However, there were increases in WBC counts with increasing dosages of the stressors restraint and propanil. The slopes for ethanol and corticosterone were similar indicating a similar decrease in WBC count at similar corticosterone AUC values.



View larger version (35K):
[in this window]
[in a new window]
 
FIG. 3. Effect of stress on white blood cell number in the blood. The number of white blood cells tended to decrease with an increase in stressor dosage. All of the stressors followed this trend except propanil. Data were first normalized by dividing the number of white blood cells/ml in each mouse by the average number of white blood cells/ml in the control groups. The absolute WBC count for the naive control group from each experiment follows: exogenous corticosterone, 5.95 x 106 ± 3.26 x 105, n = 5; restraint, 6.86 x 106 ± 1.22 x 106, n = 5; EtOH, 9.22 x 106 ± 1.34 x 106, n = 5; atrazine, 1.18 x 107 ± 1.08 x 106, n = 5; propanil, 1.16 x 107 ± 1.39 x 107, n = 5. Linear regressions were then generated for each chemical using the percentage of white blood cells relative to AUC values. The table shows the relationships between the slopes and intercepts for each of the linear regressions generated for atrazine (ATZ), propanil (Prop), ethanol (EtOH), exogenous corticosterone (Cort), and restraint (Rest).

 
Effects of Stress on CD4+ and CD8+ T Cells in the Blood
Increases in plasma corticosterone levels have been shown to effect T-cell mediated immunity (DePasquale-Jardieu and Fraker, 1980Go). In these blood studies both helper T-cell (CD4+) and cytotoxic T-cell (CD8+) populations decreased in association with an increase in corticosterone AUC values for all treatments. In Figures 4 and 5 the responses of both CD4+ and CD8+ cells to propanil, atrazine, and ethanol are more similar to exogenous corticosterone than to restraint stress. Restraint had very little effect. These results are quite different from the effects of these agents on CD4+ and CD8+ T cells in the spleen and thymus studies (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). Only atrazine had a substantial effect in the spleen, and the percentage of both CD4+ and CD8+ cells increased substantially, concomitant with a decrease in the percentage of B cells (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). In the present study all agents except restraint caused substantial decreases in CD4+ and CD8+ T cells in the blood. The decreases were generally similar for similar corticosterone AUC values, but the slope for ethanol was significantly greater than for any other agent, indicating greater effects at comparable corticosterone AUC values and suggesting the possibility of direct effects on these cells or effects caused by neuroendocrine mediators other than corticosterone.



View larger version (35K):
[in this window]
[in a new window]
 
FIG. 4. Effects of stress on CD4+ T cells in the blood. Each of the stressors decreased the expression of CD 4+ cells in the blood of mice. As the dosage or duration of the stressor increased, corresponding to an increase in AUC, the concentration of CD 4+ cells decreased. These graphs represent data that were normalized to the average expression of CD 4+ cells in control animals. The average concentration of CD 4+ cells seen in naive control groups for each experiment were as follows: exogenous corticosterone, 26.912% ± 1.6394, n = 5; restraint, 22.616% ± 1.1216, n = 5; EtOH, 27.32% ± 1.6011, n = 5; atrazine, 29.172% ± 1.628, n = 5; propanil, 32.68% ± 2.2440, n = 5. The table shows the relationships between the slopes and intercepts for each of the linear regressions generated for atrazine (ATZ), propanil (Prop), ethanol (EtOH), exogenous corticosterone (Cort), and restraint (Rest).

 


View larger version (34K):
[in this window]
[in a new window]
 
FIG. 5. Effects of stress on CD8+ T cells in the blood. The percentage of Cytotoxic T cells tended to decrease with an increase in stressor dosage or duration. These graphs represent the concentration of CD 8+ cells normalized to the average concentration of CD 8+ cells in control animals. The average concentration of CD 8+ cells in the naive control groups for each experiment were as follows: exogenous corticosterone, 14.490% ± 0.8924, n = 5; restraint, 11.008% ± 0.5607, n = 5; EtOH, 12.722% ± 0.8245, n = 5; atrazine, 15.558% ± 0.6635, n = 5; propanil, 15.704% ± 1.0836, n = 5. The table shows the relationships between the slopes and intercepts for each of the linear regressions generated for atrazine (ATZ), propanil (Prop), ethanol (EtOH), exogenous corticosterone (Cort), and restraint (Rest).

 
Effects of Stress on B-Cell Expression of the MHC II Phenotype in the Blood
Expression of major histocompatibility complex class II proteins shown in Figure 6 decreases in association with an increase in corticosterone AUC values for all treatments. MHC II expression is shown as percent MHC II positive cells divided by the percentage of B cells (B220+) multiplied by 100. This function was used to insure that decreases in B-cell numbers, the most abundant MHC II expressing cell type in blood or spleen, were not directly causing the observed decreases in MHC II. Unlike the T-cell population, expression of MHC II proteins shows reasonably consistent quantitative relationships for all stressors indicating that their expression might be a good indicator of stress-induced immunological changes. Before this parameter can be implemented in a predictive model, it was important to examine the MHC II response to cyclophosphamide (Anonymous, 1998Go). Cyclophosphamide is widely used as a positive control in immunotoxicity studies, and it seemed an ideal agent to confirm that the MHC II/B cell parameter is not suppressed indiscriminately by all immunotoxicants but that it could be relatively specific for chemical stressors.



View larger version (34K):
[in this window]
[in a new window]
 
FIG. 6. Effects of stress on B cell expression of the MHC II phenotype in the blood. The average number of cells expressing MHC II in the blood decreased in the most consistent quantitative manner with increasing duration or dosage of stressor, which corresponds to an increase in AUC value. Data were normalized to the average concentration of MHC II expressing B cells seen in control animals. The average concentration of MHC II expressing cells seen in naive control groups were as follows for each experiment: exogenous corticosterone, 78.932% ± 1.4862, n = 5; restraint, 83.551% ± 1.405, n = 5; EtOH, 77.722% ± 1.915, n = 5; atrazine, 88.126% ± 0.6684, n = 5; propanil, 86.049% ± 0.3478, n = 5. These values were determined by adding both the MHC II+ and MHC II – B cells to determine the total percentage of B cells. The raw total B cell percentages were then normalized to naive control groups. The normalized MHC II+ B cells were then divided by the normalized total B cell values and multiplied by 100. This procedure was carried out to ensure that loss of MHC II was not caused by loss of B cells. The table shows the relationships between the slopes and intercepts for each of the linear regressions generated for atrazine (ATZ), propanil (Prop), ethanol (EtOH), exogenous corticosterone (Cort), and restraint (Rest). Since this data showed the most consistent quantitative relationships it was implemented in a predictive model.

 
Effects of Cyclophosphamide Compared to Stressors on Plasma Corticosterone Levels and MHC II Expression in the Blood
Figure 7 presents plasma corticosterone levels in mice 2 h after treatment with 200 mg/kg of cyclophosphamide. This figure shows that plasma corticosterone levels do significantly increase with this large dosage of CyP. However, as compared with corticosterone levels in mice dosed with high dosages of chemical stressors used in this study, the increase in corticosterone after CyP exposure is quite small (91.58 ± 25.97 ng/ml). The effect that 150 mg/kg and 200 mg/kg of cyclophosphamide had on MHC II expression was determined to be insignificant (p > 0.05). These results are shown in Figure 8. These results indicate that MHC II expression is not suppressed by all immunotoxicants.



View larger version (24K):
[in this window]
[in a new window]
 
FIG. 7. Effects of cyclophosphamide compared to stressors on plasma corticosterone levels. Cyclophosphamide did significantly increase serum corticosterone levels in mice dosed with 200 mg/kg 2 h after treatment (naïve, 22.22 ± 4.752; cyclophosphamide, 91.58 ± 25.53). It is important, however, to realize that this increase, although significant (p < 0.05), was low relative to equivalent dosages of the stressors used in this study at a 2 h time point.

 


View larger version (37K):
[in this window]
[in a new window]
 
FIG. 8. Effects of cyclophosphamide compared to stressors on B cell expression of MHC II phenotype in the blood. Unlike exogenous corticosterone, restraint, propanil, atrazine, and ethanol, cyclophosphamide had an insignificant (p > 0.05) effect on MHC II expression in the blood 12 h after exposure (naïve, 79.612 ± 1.280; cyclophosphamide 150 mg/kg, 73.585 ± 3.3023; cyclophosphamide 200 mg/kg, 71.325 ± 3.275). Data were analyzed by normalizing the concentration of MHC II cells seen in treated animals by the average concentration of MHC II cells in control animals.

 
MHC II Expression in the Blood as a Predictor of Suppression of Other Immune Parameters and MHC II Expression
To examine MHC II and its viability as a predictor of stress induced immunosuppression, regression lines for the MHC II/B cell parameter in the blood generated from restraint and exogenous corticosterone treated mice were used to develop a predictive model. Numbers predicted by using these two regression equations were compared to actual data points collected throughout the course of the experiments. The ability of the corticosterone and restraint regression equations to predict experimental values was tested by first using the area under the corticosterone curve value at which 50% suppression of MHC II occurred in the restraint regression equation (AUC value of 5471.77 ng/ml•h) and the corticosterone regression equation (AUC value of 3841.87 ng/ml•h). Experimentally observed values for each immune parameter were determined by calculating them from the linear regression equations for propanil, atrazine, and ethanol. The observed values for each immune parameter were determined at AUC values of 5471.77 or 3841.87 ng/ml•h. If the 83.7% confidence interval of the observed values for a parameter overlapped with the 83.7% confidence interval of the value predicted by the restraint or corticosterone regression equation for a parameter at the aforementioned AUC values, this indicated no significant difference (p > 0.05), and the model was deemed predictive for that particular parameter (Barr, 1969Go; Nelson, 1989Go).

It is commonly assumed that if the 95% confidence intervals for points on different lines do not overlap, this means that they are significantly different at the p < 0.05 level. In reality, comparing 95% confidence intervals in this way is more stringent than the p < 0.05 level. Independent investigators have determined by mathematical analysis that the critical confidence interval that can be used to precisely identify differences where p < 0.05 is the 83.7% confidence interval (Barr, 1969Go; Nelson, 1989Go). Thus, if the 83.7% confidence intervals of points on two lines do not overlap, then these points are significantly different at the p < 0.05 level.

The predictive ability of both the exogenous corticosterone and restraint models using the AUC value at which suppression of MHC II/B cell is 50% is shown in Table 1. The results in Table 1 demonstrate that only one of the parameters, WBC, can be accurately predicted for all three chemical stressors using the corticosterone AUC values with 50% MHC II/B-cell expression as a reference point. Implementing the corticosterone regression in this predictive model proved to be more accurate in predicting the lymphocyte and neutrophil subpopulations reactions to stress than did the restraint based model. The slopes of the regression lines generated for the MHC II/B-cell parameter are reasonably similar, and the actual effect of stressors on MHC II/B cell was predicted accurately for all of the stressors when either the corticosterone or restraint regression equation was used. This is consistent with the results shown in Figure 6, which indicate generally similar but not identical effects of all treatments on MHCII/B cell expression in the blood.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Predictive Ability of the Corticosterone and Restraint Models Using Area under the Corticosterone vs. Time Curve Values Where 50% Suppression of MHC II Was Seen in the Corticosterone (AUC = 3841.87) and Restraint (AUC = 5471.77) Models

 
Although generating AUC values may help in creating predictive models, the process is labor intensive and would be impractical for testing large numbers of drugs or chemicals. As a result, another way of using these models is needed that would circumvent the process of generating AUC values for all dosages of a drug or chemical to be tested. One such approach uses only the dosage of a compound that causes a 50% suppression of MHC II/B-cell expression. The regressions for exogenous corticosterone and restraint (vs. dosage) are then employed to give predictions of the suppression or augmentation of other immune parameters at that particular dosage, using the points on these lines at which 50% suppression of MHC II/B cells occurs for comparison. Thus, the dosage at which 50% suppression of MHC II/B cells occurred with each treatment was determined, and the values for the other parameters were either determined from the regression line for each treatment (experimentally observed values) or predicted from the regression line for corticosterone or restraint-treated animals at the dosage that yielded 50% suppression of MHC II/B cell. This approach is described in greater detail in a previous publication (Pruett et al., 2003Go). In Table 2 this method is tested by comparing different values predicted by the exogenous corticosterone and restraint models to the experimentally observed values for each chemical stressor. If the 83.7% confidence intervals, represented in parenthesis, of the predicted and experimental values overlapped then the model was deemed predictive for that particular data point (p > 0.05) (Barr, 1969Go; Nelson, 1989Go).


View this table:
[in this window]
[in a new window]
 
TABLE 2 Predictive Ability of the Corticosterone and Restraint Models Using the Dosage Where 50% Suppression of MHC II Was Seen

 
The results shown in Table 2 demonstrate that this approach provides an accurate estimate of only one parameter in the blood, WBC number. The restraint and exogenous corticosterone models accurately predict the effect that all the chemicals tested had on WBC number. The corticosterone model predicts the change in the neutrophil and lymphocyte subpopulations after stress induced immunosuppression for two of the three chemicals. The confidence intervals for the third chemicals effect, propanil, on the subpopulations is within 10–20% of overlapping those predicted for the treatment.

Quality Control
To examine consistency in the amount of stress induced by the treatment used in this study as compared to previous studies, NK cell activity in the spleen was examined using a chromium 51-release assay. This allowed for a comparison of responses seen in these studies to those seen in previous studies of chemical stressor effects on immune parameters in the spleen (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). Linear regressions for NK cell activity generated in previous experimentation were compared to those generated in this study. Comparison of NK cell activity linear regressions generated in the present study and in previous studies indicated that there was no significant difference (p > 0.05) between the slopes of the linear regressions generated for each individual stressor. No significant difference in the slopes of the linear regressions indicates that the effects of these stressors are sufficiently consistent to suggest that prediction of stressor effects on immune parameters is feasible.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The system investigated here would provide a tool that would give some insight into whether or not observed immunosuppression was caused by stress or direct immunotoxicity. Normally, determining the role of stress in chemical-induced immunosuppression requires sets of animals dedicated to specialized experiments that are not typically included in safety assessment (Weiss et al., 1996Go; Wu and Pruett, 1997Go). The goal of the series of studies, which includes the present one, was to identify patterns of change in particular immune parameters that are predictably associated with stress and which can be easily measured in the same animals used for safety testing. In both the present and past reports one parameter examined, MHC II, had a consistent quantitative relationship to corticosterone AUC values (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). Increased plasma corticosterone levels have also been associated with decreases in MHC II expression in studies conducted using peritoneal macrophages and splenic B cells (Weiss et al., 1996Go; Zwilling et al., 1990Go). These findings further suggest that corticosterone plays a major role in the regulation of B cells and their immunological capabilities. In addition, there is now substantial evidence that decreased MHC II expression following trauma or surgery is remarkably predictive of risk of infection in humans (Lekkou et al., 2004Go). Although the ratio of lymphocytes to neutrophils in humans is the inverse of what is seen in mice the MHC II/B220 parameter could still be applicable to a human model since there are still plenty of B cells present for such a calculation. As a result, MHC II expression could be considered a possible biomarker for stress-induced immunosuppression. Such a biomarker would be helpful in risk assessment analysis by indicating that immunosuppression is stress related.

The expression of a suggested biomarker for stress-induced immunosuppression would also need to be unaffected by immunotoxicants that do not cause dramatic increases in plasma corticosterone levels. A review of literature indicates that there are no reports of MHC class II suppression by immunotoxicants that are not stressors (indicated by a Medline search using the terms immunotoxic and MHC class II). This was further investigated in this study by using a well-known broad-spectrum immunotoxicant, cyclophosphamide, which had no effect on MHC II expression. Even though MHC II expression seems to correlate well with stress-induced immunosuppression, the present study did reveal limitations to predictive models based on the blood parameters.

Previous studies showed that chemical stressors affected immune parameters in spleen and thymus more like restraint stress than exogenous corticosterone (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). Differences seen in these studies indicate that the same endpoints in blood are affected differently by the stressors than in the spleen. As compared to restraint stress, exogenous corticosterone and chemical stressors caused more dramatic decreases and increases in the percentages of lymphocytes and neutrophils, respectively. It is conjectured that increases in neutrophil number may act to counter balance some of the immunosuppressive effects of corticosterone (Dhabhar, 2000Go). Increases in neutrophil number have been associated with an increase in the amount of glucocorticoid in circulation (Miller et al., 1994Go). It has also been suggested that an increase in glucocorticoid, such as corticosterone, increases both the longevity and rate of production of neutrophils (Fauci and Dale, 1974Go; Friedman et al., 1995Go; Mishler, 1977Go). Since chemical stressors and exogenous corticosterone seem to affect the leukocyte population to a greater extent and plasma corticosterone levels are similar in restrained mice, it can be suggested that restraint stress elicits either the production of other stress mediators or different amounts of these mediators, which may counteract the effects of corticosterone (Ader and Cohen, 1993Go).

Exogenous corticosterone and chemical stressors also cause a more dramatic change in CD4+ cells and CD 8+ cells as compared to restraint stress. Decreases in both CD4+ and CD8+ T-cell subpopulations with increasing plasma glucocorticoid levels did correspond to previously observed decreases in thymus weight and cellularity (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go). However, the relationship between these decreases and the corticosterone AUC values were more variable among the three chemicals than in previous studies in which these cells were evaluated in the thymus. The data presented here suggest that the blood parameters are more sensitive to exogenous corticosterone and chemical stressors than to restraint stress. The chemicals themselves may also have some direct effect on the cell types examined in the blood.

Testing the suitability of MHC II expression on blood leukocytes as a predictive parameter was accomplished by implementing the MHC II/B220 parameter in a predictive model. Comparison of the predictive ability of the blood models showed that using the particular dosage of the stressor at which 50% suppression of MHC II/B220 was observed was more accurate in predicting experimental values for each parameter measured than the corticosterone AUC based model, but it was not quite as effective as models based on decreased MHC II expression in the spleen (Pruett et al., 1999Go, 2000aGo,bGo, 2003Go).

No data yet collected completely explains the differences in predictive value of spleen and thymus parameters as compared to the same parameters in blood. However, results from other studies suggest some of the factors that may be involved. Blood seems to show a greater sensitivity to the differences in stressors. One explanation of difference in the sensitivity of blood, as compared to spleen, may be the extremely dynamic nature of change in blood leukocyte populations in response to stress. Previous studies have shown that both catecholamines and glucocorticoids are involved in these changes (Dhabhar, 2002Go; Shephard, 2003Go). The relative amounts of norepinephrine and glucocorticoids may vary following treatment with different chemicals (Pacák et al., 1998Go), causing substantial changes in leukocyte trafficking (Richter et al., 1996Go) which may not be reflected in the spleen or thymus.

Although it is now clear that stressors can significantly decrease resistance to infection (Cohen et al., 1998Go, 1999Go; Kiecolt-Glaser et al., 1996Go; Vedhara et al., 1999Go; Zhang et al., 1998Go), quantitative estimations are not yet possible. To obtain such quantitative predictions of the effect of stressors on host resistance to infection in humans one could employ the parallelogram approach (Loveren et al., 1998Go). The parallelogram approach is based upon the idea that if the effects of a chemical on immune functions and host resistance are known for an animal model and its effects are known for the same immune functions in humans, then using these three corners of the parallelogram it is possible to extrapolate the fourth, host resistance to infection in humans. However, it is not usually possible to obtain immune function data from humans exposed to chemicals under controlled conditions. If the chemical is immunosuppressive primarily because it induces a stress response, it should be possible to model at least some of the immunotoxic effects of the chemical in humans by administering the major immunosuppressive stress hormone, cortisol, to attain stress inducible cortisol levels (Blazar et al., 1986Go; Davis et al., 1991Go; Tonnesen et al., 1987Go; Vedhara et al., 1999Go). This would allow extrapolation of an endpoint that would otherwise be unattainable without a predictive model, the effect of chemical-induced stress on resistance to infection.

Although the differences between blood, spleen, and thymus are distinct in their reactions to stress, they still share one parameter that is effected similarly, MHC II. Using MHC II expression coupled with observing for a pattern of change in other blood parameters affected by stress (e.g., >the ratio of neutrophils to lymphocytes and the number of WBC) could indicate a stress effect of a chemical. Predicting the stressor's effect on other parameters would be less reliable than hoped (Tables 1 and 2). Nevertheless, identifying MHC II as a reliable biomarker for stress induced immunosuppression in mouse blood has brought the first corner of the parallelogram to completion.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Ader, R., and Cohen, N. (1993). Psychoneuroimmunology: Conditioning and stress. Annu. Rev. Psychol. 44, 53–85.[CrossRef][ISI][Medline]

Anonymous (1998). Immunotoxicity. EPA: Heath Effects Test Guidelines, 1–11.

Anonymous (1999). Immunotoxicity Testing Guidance. Guidance for Industry and FDA Reviewers, 1–16.

Barr, D. R. (1969). Using confidence intervals to test hypotheses. J. Quality Technol. 1, 256–258.

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]

Cohen, S., Doyle, W. J., and Skoner, D. P. (1999). Psychological stress, cytokine production, and severity of upper respiratory illness. Psychosom. Med. 61, 175–180.[Abstract/Free Full Text]

Cohen, S., Frank, E., Doyle, W. J., Skoner, D. P., Rabin, B. S., and Gwaltney, J. (1998). Types of stressors that increase susceptibility to the common cold in healthy adults. Health Psychol. 17, 214–223.[CrossRef][ISI][Medline]

Cunnick, J. E., Lysle, D. T., Kucinski, B. J., and Rabin, B. S. (1990). Evidence that shock-induced immune suppression is mediated by adrenal hormones and peripheral beta-adrenergic receptors. Pharmacol. Biochem. Behav. 36, 645–651.[CrossRef][ISI][Medline]

Davis, J. M., Albert, J. D., Tracy, K. J., Calvano, S. E., Lowry, S. F., Shires, G. T., and Yurt, R. W. (1991). Increased neutrophil mobilization and decreased chemotaxis during cortisol and epinephrine infusions. J. Trauma 31, 725–732.[ISI][Medline]

DePasquale-Jardieu, P., and Fraker, P. J. (1980). Further characterization of the role of corticosterone in the loss of humoral immunity in zinc-deficient A/J mice as determined by adrenalectomy. J. Immunol. 124, 2650–2655.[Free Full Text]

Dhabhar, F. S. (2000). Acute stress enhances while chronic stress suppresses skin immunity. The role of stress hormones and leukocyte trafficking. Ann. N.Y. Acad. Sci. 917, 876–893.[Abstract/Free Full Text]

Dhabhar, F. S. (2002). Stress-induced augmentation of immune function—the role of stress hormones, leukocyte trafficking, and cytokines. Brain Behav. Immun. 16, 785–798.[CrossRef][ISI][Medline]

Dhabhar, F. S., Miller, A. H., Stein, M., McEwen, B. S., and Spencer, R. L. (1994). Diurnal and acute stress-induced changes in distribution of peripheral blood leukocyte subpopulations. Brain Behav. Immun. 8, 66–79.[CrossRef][ISI][Medline]

Fauci, A. S., and Dale, D. C. (1974). The effect of in vivo hydrocortisone on subpopulations of human lymphocytes. J. Clin. Invest. 53, 240–246.[ISI][Medline]

Friedman, H., Klein, T. W., and Friedman, A. L. (1995). Psychoneuroimmunology, Stress, and Infection, pp. 173–194. CRC Press, Boca Raton, FL.

Kiecolt-Glaser, J. K., Glaser, R., Gravenstein, S., Malarkey, W. B., and Sheridan, J. (1996). Chronic stress alters the immune response to influenza virus vaccine in older adults. Proc. Natl. Acad. Sci. U.S.A. 93, 3043–3047.[Abstract/Free Full Text]

Lekkou, A., Karakantza, M., Mouzaki, A., Kalfarentzos, F., and Gogos, C. A. (2004). Cytokine production and monocyte HLA-DR expression as predictors of outcome for patients with community-acquired severe infections. Clin. Diagn. Lab. Immunol. 11, 161–167.[Abstract/Free Full Text]

Li, Y. F., Yuan, L., Xu, Y. K., Yang, M., Zhao, Y. M., and Luo, Z. P. (2001). Antistress effect of oligosaccharides extracted from Morinda officinalis in mice and rats. Acta Pharmacol. Sin 22, 1084–1088.[ISI][Medline]

Loveren, H. V., De Jong, W. H., Vandebriel, R. J., Vos, J. G., and Garssen, J. (1998). Risk assesment and immunotoxicology. Toxicol. Lett. 102–103, 261–265.[CrossRef]

Luster, M. I., Germolec, D. R., Burleson, G. R., Jameson, C. W., Ackermann, M. F., Lamm, K. R., and Hayes, H. T. (1987). Selective immunosuppression in mice of natural killer cell activity by ochratoxin A. Cancer Res. 47, 2259–2263.[Abstract]

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.[CrossRef][ISI][Medline]

Luster, M. I., Portier, C., Pait, D. G., White, K. L., Gennings, C., Munson, A. E., and Rosenthal, G. J. (1992). Risk assessment in immunotoxicology. I. Sensitivity and predictability of immune tests. Fundam. Appl. Toxicol. 18, 200–210.[ISI][Medline]

Miller, A. H., Spencer, R. L., Hassett, J., Kim, C., Rhee, R., Ciurea, D., Dhabhar, F., McEwen, B., and Stein, M. (1994). Effects of selective type I and II adrenal steroid agonists on immune cell distribution. Endocrinology 135, 1934–1944.[Abstract]

Mishler, J. M. T. (1977). Glucocorticoid effects on neutrophils. Lancet 2, 95.[CrossRef]

Nelson, L. S. (1989). Evaluating overlapping confidence intervals. J. Quality Technol. 21, 140.[ISI]

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 suppression 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., Fan, R., Myers, L. P., Wu, W.-J., and Collier, S. D. (2000a). 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 Behav. Immun. 14, 270–287.[CrossRef][ISI][Medline]

Pruett, S. B., Fan, R., Zheng, Q., Myers, L. P., and Hebert, P. (2000b). Modeling and predicting selected immunological effects of a chemical stressor (3,4-dichloropropionanilide) using the area under the corticosterone concentration versus time curve. Toxicol. Sci. 58, 77–87.[Abstract/Free Full Text]

Pruett, S. B., Fan, R., Zheng, Q., Myers, L. P., and Hebert, P. (2003). Modeling and predicting immunological effects of chemical stressors: Characterization of a quantitative biomarker for immunological changes caused by atrazine and ethanol. Toxicol. Sci. 75, 343–354.[Abstract/Free Full Text]

Richter, S. D., Schurmeyer, T. H., Schedlowski, M., Hadicke, A., Tewes, U., Schmidt, R. E., and Wagner, T. O. F. (1996). Time kinetics of the endocrine response to acute psychological stress. J. Clin. Endocrinol. Metab. 81, 1956–1960.[Abstract]

Riley, V. (1981). Psychoneuroendocrine influences on immunocompetence and neoplasia. Science 212, 1100–1109.[ISI][Medline]

Shephard, R. J. (2003). Adhesion molecules, catecholamines and leucocyte redistribution during and following exercise. Sports Med. 33, 261–284.[ISI][Medline]

Tonnesen, E., Christensen, N. J., and Brinklov, M. M. (1987). Natural killer cell activity during cortisol and adrenaline infusion in healthy volunteers. Eur. J. Clin. Invest. 17, 497–503.[ISI][Medline]

Vedhara, K., Cox, N. K., Wilcock, G. K., Perks, P., Hunt, M., Anderson, S., Lightman, S. L., and Shanks, N. M. (1999). Chronic stress in elderly caregivers of dementia patients and antibody response to influenza vaccination [see comments]. Lancet 353, 627–631.[CrossRef][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.[CrossRef][ISI][Medline]

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]

Zhang, D., Kishihara, K., Wang, B., Mizobe, K., Kubo, C., and Nomoto, K. (1998). Restraint stress-induced immunosuppression by inhibiting leukocyte migration and Th1 cytokine expression during the intraperitoneal infection of Listeria monocytogenes. J. Neuroimmunol. 92, 139–151.[CrossRef][ISI][Medline]

Zwilling, B. S., Brown, D., Christner, R., Faris, M., Hilburger, M., McPeek, M., Van Epps, C., and Hartlaub, B. A. (1990). Differential effect of restraint stress on MHC class II expression by murine peritoneal macrophages. Brain Behav. Immun. 4, 330–338.[ISI][Medline]

Zwilling, B. S., Brown, D., Feng, N., Sheridan, J., and Pearl, D. (1993). The effect of adrenalectomy on the restraint stressed induced suppression of MHC class II expression by murine peritoneal macrophages. Brain Behav. Immun. 7, 29–35.[CrossRef][ISI][Medline]





This Article
Abstract
FREE Full Text (PDF)
All Versions of this Article:
83/1/101    most recent
kfi014v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Schwab, C. L.
Articles by Pruett, S. B.
PubMed
PubMed Citation
Articles by Schwab, C. L.
Articles by Pruett, S. B.