The JI Acurri Cytometers
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Keil, D.
Right arrow Articles by Pruett, S. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Keil, D.
Right arrow Articles by Pruett, S. B.
The Journal of Immunology, 2001, 167: 4543-4552.
Copyright © 2001 by The American Association of Immunologists

Quantifying the Relationship Between Multiple Immunological Parameters and Host Resistance: Probing the Limits of Reductionism1

Deborah Keil2,*, Robert W. Luebke{ddagger} and Stephen B. Pruett3,4,{dagger}

* {dagger} Department of Biological Sciences, Mississippi State University, Mississippi State, MS 39762; and {ddagger} Immunotoxicology Branch, National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Although reductionist experimental designs are excellent for identifying cells, molecules, or functions involved in resistance to particular microbes or cancer cells, they do not provide an integrated, quantitative view of immune function. In the present study, mice were treated with either dexamethasone (DEX) or cyclosporin A (CyA), and immune function and host resistance were evaluated. Multivariate statistical methods were used to describe the relative importance of a broad range of immunological parameters for host resistance in mice treated with various dosages of DEX. Multiple regression and logistic regression analysis indicated that changes in 24 immunological parameters explained a substantial portion of the changes in resistance to B16F10 tumor cells or streptococcus group B. However, at least 40% of the change in host resistance remained unexplained. DEX at all dosages substantially suppressed numerous relevant immunological parameters, but significantly decreased resistance to Listeria monocytogenes only at the highest dosage. In contrast, CyA substantially decreased resistance to L. monocytogenes at dosages that caused relatively minor suppression of just a few immunological parameters (unfortunately, CyA data and host resistance data for L. monocytogenes were not suitable for multivariate analysis). These results illustrate that mathematical models can be used to explain changes in host resistance on the basis of changes in immune parameters, and that moderate changes in relevant immunological parameters may not produce the types of changes in host resistance expected on the basis of results from reductionist experimental designs.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
In many respects, the immune system resembles an ecosystem. It is remarkably complex and has many redundant or overlapping components capable of achieving the same ultimate function (host resistance to microbes and cancer). However, statistical modeling approaches that have proved successful in predicting changes in ecosystems and other complex natural systems (1, 2, 3) have not been applied to the immune system. It is clear that drastic suppression of one or more immune functions decreases host resistance to infection (4) or cancer (5). However, it is not presently possible, even in animal models (6), to predict the cumulative effects of several smaller changes in immune system parameters on host resistance.

Reductionist experimental designs evaluate one immunological parameter at a time to determine whether it plays any role in resistance to a particular pathogen, but most studies do not include dose-response experiments that would establish quantitative relationships between the immune parameter and host resistance. Furthermore, studies in which all relevant immune parameters (or even a representative subset) are evaluated simultaneously to determine their quantitative contributions to host resistance have been rare. Consequently, the portion of host resistance to infectious diseases or cancer that is contributed by each immune mechanism is not generally known. In addition, the roles of interactions between immune mechanisms, redundancy of immune functions, and compensation by one function when another is diminished have not been systematically examined. This represents a fundamental gap in our understanding of the immune system, and it has important practical implications. For example, regulatory agencies (e.g., Environmental Protection Agency and Food and Drug Administration) evaluate environmental chemicals and drugs for potential adverse effects on immunocompetence, typically by using a panel of functional assays (7). Such agents seldom affect a single immunological parameter (8), and there is no effective way at present to predict the impact on host resistance of small to moderate changes in multiple immunological parameters. These considerations are also relevant in estimating the effects of stress, malnutrition, and other environmental influences on resistance to infection or cancer. Although changes in host resistance can be measured experimentally in animals, this cannot be done in humans. Therefore, models that can predict changes in host resistance on the basis of changes in immunological parameters are potentially important.

Multivariate statistical methods have been used successfully in fields such as ecology and sociology to predict changes in a single dependent variable using multiple explanatory variables (9). These methods seem well suited for analysis of the complex relationships between immune parameters and host resistance. In a recent feasibility study, we demonstrated that the sequential use of two multivariate methods (factor analysis and multiple regression) can effectively model relationships between immune function end points and host resistance in mice treated with various dosages of the immunosuppressant, dexamethasone (DEX)5 (10). In the present study, factor analysis followed by multiple regression or logistic regression was used to quantitatively evaluate the contributions of immune system parameters to host resistance to B16F10 tumor cells and to streptococcus group B. Immune parameters for this study were selected on the basis of three criteria: small coefficients of variation (10), holistic end points (i.e., an assay of a relevant final function is preferable to an assay of any of the molecular or cellular components required for that function), and assessment of all major cell types and/or functions known to mediate resistance to the pathogens and tumor cells selected for use in this study.

Four different dosages of the prototypical immunosuppressant DEX were used to suppress numerous immune system parameters in the present study. In addition, similar experiments were performed with cyclosporin A (CyA), another prototypical immunosuppressant with a narrower range of effects than DEX. Although multivariate methods proved to be inappropriate for analysis of the CyA data, conventional statistical methods and comparison with the DEX data set permitted some interesting conclusions regarding immune function parameters and host resistance.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Mice

Specific pathogen-free B6C3F1 female mice were obtained through the animal program of the National Cancer Institute. They were allowed to recover from shipping stress for at least 1 wk before use in experiments at the age of 8–10 wk. They were housed in an American Association for Accreditation of Laboratory Animal Care accredited animal facility, and animal care and experimentation were performed in accordance with the National Institutes of Health Guide and the policies of the Institutional Animal Care and Use Committee at Mississippi State University. Mice were maintained on a 12-h light/dark cycle with continual access to food (Purina Lab Chow; Purina, St. Louis, MO) and water.

Immunosuppressive treatments and experimental design

Mice were treated with either DEX-21-phosphate (Sigma, St. Louis, MO) or CyA (kindly provided by Novartis Pharmaceuticals, East Hanover, NJ) by s.c. injection. A control group in each experiment was given an equal volume of the vehicle for each agent: Dulbecco’s PBS for DEX and olive oil (as suggested by Novartis Pharmaceuticals) for CyA. The dosages used in this study were determined using data from preliminary experiments, and the maximum dosage of CyA was based on the occurrence of nephrotoxicity in mice at dosages greater than 100 mg/kg/day (11).

Most of the immunological parameters assessed in this study were evaluated using the same set of mice. However, parameters that required immunization or immune responses (CTL function, Ab response, and host resistance assays) were evaluated using separate sets of mice. Each immunological and host resistance parameter was evaluated in two identical experiments. In each experiment, there was a vehicle group and four groups treated with different dosages of DEX or CyA every day for 16 days, except in mice used to evaluate Tc activity (see below). There were 8 mice per group in each experiment, for a total of 80 mice (8 per group x 5 groups x 2 independent experiments). The only exception was the determination of the effect of DEX on resistance to Listeria monocytogenes. This was performed in a single experiment in which the group size was 20.

The decision to examine immunological parameters 24 h after the last dose of DEX was based on evaluating the status of the immune system just before the next daily dose, demonstrating the minimal effects of DEX on these immunological parameters. The t1/2 of DEX in rats is 5.5 h (12), and drug t1/2 in mice is typically about one-half that in rats (13). Thus, it is likely that DEX was mostly cleared within 24 h in our study. In at least some cases, it is likely that suppression was greater at earlier times after DEX administration than when measured at 24 h. Parameters were measured at the same time in the CyA experiments to facilitate comparison of the data from animals treated with CyA and DEX.

Immunological and host resistance assessments

Hematology parameters. Mice were bled from the retroorbital plexus (under methoxyflurane anesthesia) into EDTA tubes. White blood cell counts were determined following lysis of erythrocytes with manual hemoglobin and lysing reagent (Baxter Scientific, McGaw Park, IL) using an electronic cell counter (model Zf; Coulter, Hialeah, FL). RBC counts were determined similarly, using samples without the lysing reagent. Differential counts were determined using blood smears stained with Wright’s stain.

Spleen and thymus weight and cellularity. Spleen and thymus cell suspensions were prepared by pressing the organs between the frosted ends of sterile glass microscope slides and suspending the cells in 3 ml of RPMI 1640 culture medium. Cells were enumerated using an electronic cell counter (model Zf; Coulter Instruments). These parameters were assessed 24 h after the last (16th) dose of CyA or DEX.

Macrophage number and function. Resident peritoneal macrophages were obtained by peritoneal lavage, as described previously (10). The ability of these cells to phagocytose opsonized L. monocytogenes (strain 19303) and to produce nitrite in response to IFN-{gamma} (1000 U/ml; Genzyme, Cambridge, MA), bacterial LPS from Escherichia coli 0111:B4 (10 µg/ml; Sigma), or both was determined in ex vivo assays, as described previously (10). The results are expressed as number of bacteria phagocytosed per macrophage, and nitrite production is expressed as nitrite (µM) in the supernatants of 24-h cultures with 2 x 105 peritoneal cells in 100 µl of medium. These assessments were begun 24 h after the last (16th) dose of CyA or DEX.

Hemolytic complement assay. Blood was obtained from the retroorbital plexus under methoxyflurane anesthesia and allowed to clot. Serum was removed and stored frozen until used in the complement assay. Hemolytic complement was quantified by measuring the ability of serially diluted mouse serum samples to lyse rabbit erythrocytes coated with mouse anti-rabbit erythrocyte Abs, essentially as described (14). Lysis was detected as a decrease in light scattering, which was measured with a microplate reader (Bio-Rad UV3550, Richmond, CA). Results are expressed as the reciprocal of the dilution that produced 50% maximum lysis, using control serum at 1/20 to obtain the 100% value.

Tc function. Tc were induced by i.p. administration of 107 P815 allogeneic tumor cells (freshly passaged in syngeneic DBA/2 mice), as described (15). Tc activity in splenocyte preparations was determined using a standard 4-h 51Cr release assay using labeled P815 target cells and E:T ratios of 100:1, 30:1, 10:1, and 3:1. LU per 107 splenocytes were calculated as described (16). This value was then converted to LU per spleen for each mouse on the basis of the number of spleen cells obtained from that mouse. Mice were immunized on the third day of dosing with DEX or CyA, and spleens were removed for analysis on day 10 after immunization (the optimum time for the Tc response (17). This was 1 day after the last dose of DEX or CyA.

NK cell lytic function. Splenic NK cell activity was measured using a standard 4-h 51Cr release assay with labeled YAC-1 tumor cells as targets (18). Results are expressed as LU per spleen, which were calculated as described (16). These assays were begun 24 h after the last (16th) dose of CyA or DEX.

Flow cytometric analysis of cellular subpopulations in the spleen, thymus, and peritoneal cavity. Cells were obtained as described above and suspended in PBS with 0.1% BSA and 0.1% sodium azide at 107 cells/ml. Cell suspension (100 µl) was placed in the wells of a 96-well U-bottom microtiter plate and labeled for 30 min at 4°C with one of the following pairs of Abs: anti-CD4 PE and anti-CD8 FITC (Life Technologies, Grand Island, NY); anti-B220 PE (Life Technologies) and anti-MHC class II FITC (BD PharMingen, San Diego, CA). After labeling, erythrocytes were lysed using ammonium chloride lysis buffer (NH4Cl, 4.13 g; NaHCO3, 0.5 g; EDTA, 0.03 g in 500 ml of water, pH 7), and the cells were washed and fixed with 1% paraformaldehyde in PBS, as in our previous studies (10, 18). Because of the number of samples, isotype controls were not routinely included in these studies. However, the results obtained in this study were comparable with results we have obtained in several other studies in which isotype controls were included (19, 20). These parameters were evaluated 24 h after the final (16th) dose of DEX or CyA. Values are expressed as number of cells of each subpopulation per organ (spleen, thymus, or peritoneal cavity), which was determined by multiplying the percentage of each subpopulation determined by flow cytometry by the cell number for each organ in each mouse.

Ab response to SRBC. Mice were immunized by i.v. administration of 5 x 108 SRBC, and IgG Abs to SRBC were measured by ELISA, as described previously (21). Immunization was performed on day 2 of administration of DEX or CyA, and mice were bled 24 h after the final (16th) dose of DEX or CyA. This time frame was selected because it was consistent with the 16-day dosing protocol used in most other experiments in this study and because a preliminary experiment indicated that the peak IgG response to SRBC occurs on day 16 (22).

Host resistance assays; resistance to streptococcus group B. B6C3F1 mice have little innate resistance to these bacteria (the LD50 is ~5 bacteria/mouse), but resistance increases substantially if the mice are immunized with heat-killed streptococcus group B bacteria (23). Resistance in this model is mediated by Abs, complement, and phagocytic cells (23). The methodology for this experiment has been described previously (23). Briefly, mice were immunized with heat-killed streptococcus group B (1 or 2 x 106/mouse) on days 2 and 9 of the 16 daily doses of DEX or CyA. Mice were challenged with a lethal dose of live streptococcus group B 1 day after the last dose of CyA or DEX, and mortality was noted. A nonimmunized control group was included in each experiment to demonstrate that the challenge dose was lethal to most mice that were not immunized.

Resistance to L. monocytogenes. Resistance to L. monocytogenes involves early clearance by neutrophils, followed by the activation of T cells; the latter is required to activate macrophages that finally eliminate the remainder of the bacteria (24, 25, 26). Mice received DEX or CyA 2 days before challenge with L. monocytogenes (strain 19303; provided by A. E. Munson, National Institute for Occupational Safety and Health, Morgantown, WV), and daily for 14 consecutive days. The mice were challenged by i.v. injection with 7.5 x 103 CFU (approximately an LD10, as determined in our laboratory) or 3.1 x 103 CFU (a nonlethal dose) of viable L. monocytogenes. Mice were observed for mortality for 14 days.

Resistance to B16F10 melanoma cells. The B16F10 melanoma line forms tumor nodules almost exclusively in the lungs following i.v. administration, and it is widely used to assess host resistance to nonimmunogenic or minimally immunogenic tumors in mice. NK cells and possibly Tc are involved in resistance to this tumor (7, 27). Mice were challenged by i.v. administration of 1 x 105 B16F10 tumor cells on day 2 of DEX or CyA administration, and lungs were removed for enumeration of nodules (using a dissecting microscope) 1 day after the last (16th) dose.

Statistical analysis and modeling

To determine whether immunological or host resistance parameters were significantly affected by DEX or CyA, continuous data were evaluated by ANOVA, followed by Dunnett’s post hoc test to determine whether values for treatment groups were significantly different from values for control groups. Discrete data (survival in host resistance assays) were evaluated using Fisher’s exact test to determine whether survival of any group treated with DEX or CyA was significantly different from its matching control group.

Multivariate statistical analysis was performed as described in our previous study (10). Briefly, principal component analysis with promax rotation was performed on 24 of the immunological parameters evaluated in this study to place these parameters into groups (referred to as factors) on the basis of the slope of the dose-response lines. These parameters were selected from the total of 33 examined in this study (32 shown in Table IGo, and the Ab response to SRBC shown in Fig. 1Go), because they met the criterion determined in our previous study (10): a coefficient of variation under 50% in all groups. For each group of parameters, this analysis produces a factor score, which is a mathematical expression representing all variables in that factor. These factor scores were then used as the explanatory (independent) variables in multiple regression analysis (for the continuous data obtained with the B16F10 model) or logistic regression analysis (for the discrete data obtained in the streptococcus group B host resistance model). In both cases, host resistance was the dependent variable. In the case of the Ab response to SRBC and the lytic function of complement, the cumulative (titer-based) measures of activity did not meet the coefficient of variation requirement, but absorbance values for most dilutions did. Therefore, these absorbance values were used in the regression models. Our previous study demonstrated that evaluation of CyA-induced effects by multivariate analysis was not appropriate, so only data from DEX-treated animals were subjected to multivariate analysis.


View this table:
[in this window]
[in a new window]
  		Table I. Effects of daily administration of DEX for 16 days on selected immunological parameters1

 


View larger version (30K):
[in this window]
[in a new window]
  		FIGURE 1. Effects of DEX and CyA on the IgG Ab response to SRBC. Mice were immunized with SRBC on the third day of daily administration of DEX or CyA, and blood samples were obtained for analysis 24 h after the last dose. Abs were quantified by ELISA. Serum from nonimmunized mice was included as a control, and at a 1/20 dilution it produced an absorbance value of 0.14 for the CyA experiment and <0.05 for the DEX experiments. The values shown are means ± SE (n = 16 mice per group), and all dosages of DEX significantly suppressed the absorbance values at all Ab dilutions (p < 0.05 by Dunnett’s test). However, CyA at all dosages except 100 mg/kg/day significantly increased absorbance at all serum dilutions except 1/20. Only the 100 mg/kg/day dosage significantly decreased absorbance at all serum dilutions.

 
Statistical diagnostics provided by SAS Institute (Cary, NC) indicated no significant departures from the assumptions of the methods used in this study. Tests for problems that often occur in these analyses were also performed (e.g., multicollinearity), and no problems were noted. Outliers were evaluated, but they were retained in the data set for three reasons: 1) there were few outliers, 2) no obvious experimental problems could be identified to justify their deletion, and 3) analysis of a data set that excluded the outliers did not produce substantially different models than those with outliers.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
DEX significantly suppresses a wide range of immunological parameters, and enhances a few parameters

The results shown in Table IGo indicate that most of the immunological parameters evaluated were profoundly suppressed by administration of DEX for 16 days. Most of these parameters were also evaluated after 3 or 10 days of dosing, and the effects on most parameters increased with time, but were similar after 10 or 16 days of dosing (data not shown). Parameters involving lymphocyte number and lytic function (NK or Tc) were significantly suppressed even at the lowest dosage. The total number of peritoneal cells (mostly macrophages) was decreased less than the percentage of blood lymphocytes, and the number and percentage of neutrophils in the blood increased substantially at dosages of 3 mg/kg/day and greater. Flow cytometry demonstrated that all major subpopulations of thymocytes (defined by the CD4 and CD8 surface markers) decreased significantly at almost all dosages of DEX. However, the loss of CD4+CD8+ cells was proportionately greater than the loss of the other subpopulations. All dosages of DEX caused significant loss of CD4+ T cells, CD8+ T cells, CD4+CD8+ T cells, and CD4-CD8- cells (mostly B cells) in the spleen. Flow cytometry of splenocytes labeled with Abs specific for B220 and MHC class II demonstrated significant loss of all four subpopulations defined by these markers at all dosages of DEX. The decrease in number of these subpopulations in the peritoneal cavity was significant only for B220- MHC II- cells, which is consistent with the small decreases in the total number of resident peritoneal cells. Production of IgG Abs to SRBC was significantly diminished by all dosages of DEX used in this study, and the higher dosages almost eliminated this Ab response (Fig. 1Go). Peritoneal macrophages from mice treated with DEX exhibited significant increases in production of nitrite when stimulated in vitro with LPS or IFN-{gamma} and LPS. This was unexpected, because incubation of macrophages with DEX in vitro causes a decrease in nitrite production (28). It is likely that the 24-h period between the last dose of DEX and the evaluation of nitrite production allowed sufficient time for recovery of this function. Thus, a picture emerges in which DEX markedly suppresses almost all parameters related to lymphocyte number and function, suppresses macrophage number and phagocytic capability to a lesser extent, and increases the number of neutrophils and the production of nitrite by macrophages.

DEX decreases host resistance, but to a lesser extent than several relevant immunological parameters

DEX decreases host resistance to bacteria and cancer cells (Fig. 2Go), but significant effects required a minimum DEX dosage of 3 mg/kg/day for streptococcus group B (at an immunizing dose of 1 x 106 heat-killed bacteria), 10 mg/kg for B16F10 tumor cells, and 30 mg/kg for L. monocytogenes (at 3100 CFU/mouse). This contrasts dramatically with the effects of DEX on the immunological parameters examined in this study (Table IGo and Fig. 1Go). Eighteen parameters were significantly suppressed at a DEX dosage of 0.3 mg/kg, and four additional parameters were suppressed at 10 mg/kg. Parameters important in resistance to L. monocytogenes, streptococcus group B, and B16F10 tumor cells (Tc generation, Ab production, and NK cell activity) were substantially suppressed at 0.3 mg/kg/day of DEX and profoundly suppressed at 3 mg/kg/day.



View larger version (33K):
[in this window]
[in a new window]
  		FIGURE 2. Effects of DEX on host resistance. Values shown in the upper two graphs are percentage of mortality values for groups of 20 (Listeria) or 16 (streptococcus) mice. Groups for which survival was significantly different from the vehicle control group (0 mg/kg DEX) are indicated by ** (p < 0.01 by Fisher’s exact test). In the streptococcus group B experiment, all mice were immunized with the indicated dosage of heat-killed streptococci, except for the nonimmunized control groups (N. I.), which were included to verify that the dose of bacteria used was lethal to most nonimmunized mice. Values shown in the bottom graph are means ± SE for the number of tumor nodules on both lungs of mice (n = 16 mice/group). Values significantly different from the vehicle control are indicated by ** (p < 0.01 by ANOVA, followed by Dunnett’s post hoc test). The data for streptococcus and B16F10 were obtained in two independent experiments (n = 8 mice/group in each experiment). The data for Listeria were obtained in a single experiment.

 
CyA suppresses a few immunological parameters and enhances an approximately equal number

CTL generation, the number of CD4+CD8- and CD4-CD8+ cells in the thymus, thymus weight (but not total cell number), and the number of CD4+CD8- and CD4+CD8+ cells in the spleen decreased in a dose-responsive manner (Table IIGo). The IgG response to SRBC was significantly suppressed only at the highest dosage of CyA and was slightly enhanced at lower dosages (Fig. 1Go). Only the loss of mature (single-positive) thymocytes and the decrease in CTL activity were as great as the suppression of many of the parameters by DEX. Among the parameters significantly increased by CyA were spleen weight and cell number (only at the highest dosage level), NK cell activity, nitrite production by macrophages, blood neutrophil number (only at the highest dosage level), and the number of non-T cells (CD4-CD8-) and MHC II+ cells (B cells and macrophages) in the spleen (Table IIGo).


View this table:
[in this window]
[in a new window]
  		Table II. Effects of daily administration of CyA for 16 days on selected immunological parameters1

 
CyA suppresses host resistance to L. monocytogenes, but not to B16F10 tumor cells

At dosages of 25 mg/kg/day or greater, mortality in mice challenged with L. monocytogenes (7.5 x 105 CFU/mouse) increased from 25% (in the control group) to 100% (Fig. 3Go). In contrast, few of the immunological parameters important in resistance to L. monocytogenes were decreased at 25 mg/kg/day, and some were enhanced (e.g., NK cell activity, and nitrite production by macrophages). Unfortunately, it was not possible to use multivariate methods to analyze these data. A previous validation study using a subset of the data reported in this work revealed that CyA did not suppress a sufficient number of parameters to a sufficient degree to permit the use of multivariate methods on these data obtained from multiple sets of mice (10).



View larger version (32K):
[in this window]
[in a new window]
  		FIGURE 3. Effects of CyA on host resistance. Values for Listeria challenge are percentage of mortality. Results were obtained in two separate experiments (n = 8 mice per group in each experiment). Values significantly different from the vehicle control group (0 mg/kg CyA) are indicated by *** (p < 0.001 by Fisher’s exact test). Values shown for B16F10 are means ± SE for the number of tumor nodules on both lungs of mice (n = 16 mice/group). No values were significantly different from the vehicle control (0 mg/kg CyA) (as indicated by ANOVA, followed by Dunnett’s post hoc test). These results were obtained in two independent experiments (n = 8 mice/group in each experiment).

 
Resistance to B16F10 tumor cells was not affected by CyA at any dosage (Fig. 3Go), and this is consistent with the lack of suppressed NK cell lytic function, a major mechanism of resistance to B16F10 cells (27).

Multivariate analysis

In our previous study, it was demonstrated that multivariate statistical methods could be used for some data sets (i.e., data from DEX-treated mice), even though more than one experiment was required to obtain the data (10). This has typically been considered a violation of one of the assumptions underlying these methods, but we demonstrated that if the coefficient of variation for each group is less than 50% and there are a sufficient number of parameters that are strongly dose responsively affected, valid results may be obtained (10). Of the 33 immune parameters evaluated in the present study (Table IGo and Fig. 1Go), 24 met the coefficient of variation requirements, and factor analysis was used to assign these parameters to groups with similar dose-response relationships (Tables IIIGo and IVGo). The factor scores for each of these groups were then used as explanatory variables in multiple regression, with the number of B16F10 tumor nodules in the lungs as the dependent variable. Unfortunately, similar analysis was not possible for L. monocytogenes, because there was not a significant linear relationship between DEX dosage and resistance to this organism.


View this table:
[in this window]
[in a new window]
  		Table III. Summary of principal component analysis and multiple regression results using immunological parameters measured after 16 days of DEX to predict resistance to B16F10 tumor cells

 

View this table:
[in this window]
[in a new window]
  		Table IV. Multiple regression analysis1

 
As shown in Tables IIIGo and VGo, two of these factors correlated significantly with changes in resistance to B16F10 cells. The adjusted r2 value for the multiple regression model indicates that ~57% of the variance in the number of B16F10 nodules is explained by the immunological parameters in all factors. The factor that contributes most to this relationship contains the variables nitrite production by macrophages, number of white blood cells, and number of neutrophils in blood (factor 4). However, it should be noted that these parameters were increased by DEX, so the relationship identified by multiple regression analysis reflects an inverse correlation between these parameters and resistance to B16F10 cells. Factor 1 includes variables that would be expected to influence resistance to B16F10 cells (NK cell activity and CTL generation), and these variables are suppressed by DEX. Thus, it is not surprising that this factor also correlates significantly with host resistance. Although a substantial portion of the variance in resistance to B16F10 cells is explained by the immunological parameters examined in this study (as indicated by the R2 value for the multiple regression), >40% of the variance remains unexplained.


View this table:
[in this window]
[in a new window]
  		Table V. Summary of principle component analysis followed by logistic regression analysis of the role of selected immunological parameters in resistance to Streptococcus group B in mice treated with different dosages of DEX for 16 days

 
The results for principal component analysis and logistic regression for mice treated with DEX and challenged with streptococcus group B are shown in Tables VGo and VIGo. Factor 3 (lymphocyte number in the blood; B220+ MHC II+ and B220- MHC II- cell number in the peritoneal cavity) significantly contributed to correct prediction of the outcome (survival or death) in logistic regression. Although the r2 value (0.197) may seem to indicate that the model is not effective, it should be noted that the goal of logistic regression is not explaining variance (as in multiple regression), but correctly predicting the classification of statistical units (e.g., alive or dead) (29). Thus, the ability of the logistic regression model to correctly classify 65.71% of cases in the present study is an indication that changes in the parameters measured predict a substantial portion of the death or survival in the streptococcus group B model.


View this table:
[in this window]
[in a new window]
  		Table VI. Logistic regression analysis1

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The results of this study demonstrate that evaluating the suppression of several immunological parameters concurrently provides insights into changes in host resistance that would not be revealed by examining one parameter at a time. For example, reductionist experimental approaches have demonstrated that depletion of NK cells (30), CD8+ T cells (31), or CD4+ T cells (25) substantially diminishes resistance to L. monocytogenes. Because the number of CD4+ and CD8+ T cells and the function of NK cells are decreased by DEX (Table IGo), substantial decreases in host resistance were anticipated. This would also have been consistent with reports that DEX suppresses the production of a variety of cytokines, including IFN-{gamma} (32, 33, 34), which is critical in resistance to L. monocytogenes (30). However, it is also important to note that DEX caused a 4-fold increase in the concentration of neutrophils in the blood (Table IGo), and neutrophils are important in the early stages of resistance to L. monocytogenes (24, 35). It seems likely that this increase, possibly in conjunction with increased production of NO 24 h after dosing with DEX (Table IGo), was sufficient to compensate for the diminution of other host resistance mechanisms. However, in the group that received the highest dosage of DEX, suppression of other parameters was apparently too great to be compensated by the increased number of neutrophils, leading to decreased resistance to L. monocytogenes.

The effects of DEX and CyA on individual immunological parameters observed in this study are generally consistent with those reported in previous studies. DEX suppressed virtually all lymphocyte-related parameters and increased the percentage and number of circulating neutrophils (36, 37, 38, 39, 40, 41, 42) (Fig. 1Go and Table IGo). Suppression of the Tc response and decreases in the number of CD4+CD8- T cells in the thymus and spleen are hallmarks of CyA (43, 44), and these were among the most prominent effects in the present study (Table IIGo). Enhancement of macrophage nitrite production by DEX may seem surprising in view of reported suppression in vitro (45), but we noted that the persistence of this suppression was dependent on the concentration of DEX to which the macrophages had been exposed and on the time after exposure at which the macrophages were stimulated (28). Another group has reported similar results (46). Although some investigators have reported suppression of NK cell activity by CyA (47), others have noted no effect or enhancement (48, 49), as observed in the present study (Table IIGo). Suppression of NO production has been reported following in vitro CyA exposure (50), but a recent study strongly suggests that nephrotoxicity produced by CyA may be mediated by enhanced production of inducible NO in vivo (51). Thus, the increased nitrite production noted in this study was not surprising.

It is well known that the efficacy of immunization or the numbers of infectious particles encountered by the host often determine whether disease follows infection. Thus, different challenge doses of L. monocytogenes and different immunizing doses in the streptococcus group B model were used to determine whether immunosuppression is more evident when challenge doses are greater as has been proposed (6), or when immunizing doses are less (for streptococcus group B). The results demonstrate that resistance to L. monocytogenes was less sensitive to suppression when a higher challenge dose was used, whereas resistance to streptococcus group B was suppressed more when a smaller immunizing dose of heat-killed bacteria was used. The latter situation is analogous to the use of a higher challenge dose of live bacteria in a model that does not require immunization. Thus, the present results with streptococcus group B seem consistent with the more extensive study of Luster et al. (6), indicating increased sensitivity to suppressed host resistance with increasing challenge doses of microorganisms or cancer cells. However, our results with L. monocytogenes suggest that this may not be the case for all pathogens or all experimental systems.

The concept of immune system reserve capacity suggests that total inducible effector function exceeds that which is required to overcome infection; thus, some degree of suppressed functional capacity can be tolerated without loss of resistance to infection. Our experimental approach demonstrated examples of this concept. Resistance to B16F10 melanoma cells and streptococcus group B changed in ways that are generally consistent with changes in the immunological parameters known to be important in resistance to these agents. For example, NK cell activity, which is a major mechanism of defense against B16F10 tumor cells (27, 52), was decreased in DEX-treated mice. However, there was a disparity in the dosage required to suppress NK cell activity and host resistance. NK cell activity was suppressed by ~50% at 0.3 mg/kg/day of DEX and by ~85% at 3 mg/kg/day, and Tc activity was suppressed to an even greater extent. Yet, the number of B16F10 nodules was not significantly increased compared with the control value at either of these dosages. Using a reductionist experimental design, we demonstrated previously that depletion of NK cells by administration of a mAb decreases by 10-fold or more the dose of tumor cells required to produce a similar number of nodules in the lungs as observed in control mice (27). On the basis of this result alone, one would conclude that NK cells are the major defense mechanism against B16F10 tumors. Nevertheless, the results shown in this study demonstrate that substantial suppression of NK cell activity (at lower doses of DEX) does not necessarily impair resistance to B16F10 cells (Table IGo and Fig. 2Go). One interpretation of this finding is that there is considerable reserve capacity in the immune system. Thus, NK cells may be essential for resistance to B16F10 tumor cells, but the number (or the amount of lytic function) required may be much less than present in a normal animal. Alternatively, there may be an additional, unidentified mechanism of resistance that is enhanced or not affected by DEX, and thus partially compensates for the loss of NK cell function.

The situation is similar for streptococcus group B. One of the most important aspects of resistance to this organism is the ability to produce opsonizing and complement-fixing Abs (23, 53, 54, 55). DEX at dosages of 0.1 mg/kg/day and greater significantly suppresses the IgG Ab response to a model T-dependent Ag (SRBC). However, a significant decrease in resistance to streptococcus group B was only noted at 3 and 30 mg/kg/day. Neutrophils are also important in resistance to these bacteria (56), and it is possible that the increased number of neutrophils induced by DEX partially compensated for the diminished Ab response. It should be noted that the increase in neutrophils was not significant at 0.3 mg/kg/day, a dosage that significantly suppressed the Ab response (Table IGo and Fig. 1Go). Therefore, it is possible that the amount of Ab induced by our immunization protocol is more than adequate to control infection, and that the Ab response can be suppressed significantly without adversely affecting host resistance to this pathogen. This may be similar to the situation with many human vaccines. Most vaccines produce an increase in Ab titer greater than the titer required for protective immunity (57, 58). Immunosuppressed persons whose increase in titer was less than average would still be protected from infection, but only if their titer was greater than the minimal protective value (57).

Perhaps the most interesting aspect of the results reported in this work is the remarkable suppression of resistance to L. monocytogenes by CyA at dosages that decrease only one of the major mechanisms of resistance to these organisms. Previous investigators have reported that CyA decreases resistance to L. monocytogenes, but those studies did not include simultaneous assessment of a wide range of immunological parameters (59, 60). Our results are particularly surprising in view of the minimal effects of DEX on resistance to this organism, even though DEX substantially suppressed a number of relevant mechanisms of resistance (Table IGo and Fig. 1Go). Perhaps the key difference in the effects of DEX and CyA is the absence of an increase in neutrophil number in mice treated with CyA. In the absence of this increase, the suppression of Tc activity, which is important in clearance of Listeria (61), may have been sufficient to produce the observed suppression of host resistance. Interestingly, significant decreases in resistance to Listeria and significant decreases in Tc activity began at the same dosage. Thus, immunological reserve capacity cannot be assumed to occur in all cases; it seems to be dependent on the pathogen and the immunosuppressive agent.

Multiple regression analysis indicated that changes in the parameters measured in this study account for 57% of the change in resistance to B16F10 cells (the adjusted r2 value in Table IVGo). Compared with typical results in other studies involving complex systems, a substantial portion of the variance in the dependent variable is explained, suggesting that most of the decrease in resistance to B16F10 melanoma cells in DEX-treated mice is mediated by changes in the immunological parameters measured in this study. Similar results were obtained when logistic regression analysis was used to evaluate the effects of DEX on resistance to streptococcus group B. The model correctly classified the outcome (survival or death) for 65% of the mice. Again this suggests that a substantial portion of the changes in resistance to streptococcus group B is mediated by changes in the parameters measured in this study. Thus, multivariate analysis seems preferable to using individual variables in single regression analyses. This latter approach yields very low r2 values for most immune function parameters (6), and none of these simple linear models explained a greater amount of the variance in host resistance or predicted more outcomes correctly than noted in this work for multivariate methods. However, it is interesting that the only factor in our streptococcus group B model that contributed significantly to the model contained three variables (number of blood lymphocytes and number of two cellular subpopulations in the peritoneal cavity) that one would not have expected to be the most important in resistance to streptococcus group B (Tables VGo and VIGo). Because the bacteria were administered i.p., it is possible that lymphocytes at that location are particularly important and that lymphocytes in the blood prevent hematogenous spread of the bacteria. However, it is also possible that the importance of this factor simply reflects a similar dose-response relationship as host resistance, and this does not necessarily signify a cause-effect relationship.

This observation illustrates one of the limitations of multivariate regression methods. These methods use both positive and negative linear correlations in multidimensional space to explain the dependent variable (host resistance). The models are not "aware" of evidence indicating that increases in a variable in one factor (e.g., nitrite production in Table IGo) would tend to increase host resistance, whereas decreases in others (e.g., NK cell activity in Table IGo) would tend to decrease host resistance. Nonregression-based multivariate methods such as neural network analysis may ultimately prove to be more effective (62). To date, the use of such methods in evaluating the immune system has been limited to developing theoretical models that will predict the behavior of particular immune parameters (e.g., Ab responses) (63). Neural networks can "learn" the likely impact on the dependent variable of increases or decreases in various parameters in the context of increases or decreases in other parameters. However, a large and consistent data set will be required to achieve adequate training. The development of methodologies such as gene chip arrays and proteomics techniques, which allow assessment of changes in large numbers of mRNAs and proteins, may very well generate the large data sets needed for effective neural network analysis, and may allow simultaneous determination of many parameters for each individual mouse that will permit more effective multivariate regression analysis.

The results presented in this work illustrate the importance of a holistic approach in understanding the relative quantitative importance of various immunological parameters in host resistance to particular microbes or cancer cells. In addition, these results suggest the existence of immunological reserve capacity, which indicates that moderate suppression of immunological parameters may not affect host resistance to some pathogens or tumor cells.


   Footnotes
 
1 This work was supported by a Cooperative Agreement from the U.S. Environmental Protection Agency (CR819682-01-0). This report has been reviewed by the Environmental Protection Agency’s Office of Research and Development and approved for publication. Approval does not signify that the contents necessarily reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. Back

2 Current address: Department of Medical Laboratory Sciences, Medical University of South Carolina, Charleston, SC. Back

3 Current address: Department of Cellular Biology and Anatomy, Louisiana State University Health Sciences Center, Shreveport, LA 71130. Back

4 Address correspondence and reprint requests to Dr. Stephen B. Pruett at the current address: Department of Cellular Biology and Anatomy, Louisiana State University Health Sciences Center, 1501 Kings Highway, Shreveport, LA 71130. E-mail address: spruet{at}LSUHSC.edu Back

5 Abbreviations used in this paper: DEX, dexamethasone; CyA, cyclosporin A; Tc, cytotoxic T cell. Back

Received for publication May 29, 2001. Accepted for publication August 16, 2001.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Robinson, T., D. Rogers, B. Williams. 1997. Mapping tsetse habitat suitability in the common fly belt of southern Africa using multivariate analysis of climate and remotely sensed vegetation data. Med. Vet. Entomol. 11:235.[Medline]
  2. Struck, B. D., R. Pelzer, P. Ostapczuk, H. Emons, C. Mohl. 1997. Statistical evaluation of ecosystem properties influencing the uptake of As, Cd, Co, Cu, Hg, Mn, Ni, Pb and Zn in seaweed (Eucus vesiculosus) and common mussel (Mytilus edulis). Sci. Total Environ. 207:29.[Medline]
  3. Yao, H., Z. He, M. J. Wilson, C. D. Campbell. 2000. Microbial biomass and community structure in a sequence of soils with increasing fertility and changing land use. Microb. Ecol. 40:223.[Medline]
  4. Cohen, O., D. Weissman, A. S. Fauci. 1999. The immunopathogenesis of HIV infection. W. E. Paul, ed. Fundamental Immunology 1455. Lippincott-Raven Publishers, Philadelphia.
  5. Penn, I.. 2000. Post-transplant malignancy: the role of immunosuppression. Drug Saf. 23:101.[Medline]
  6. Luster, M. I., C. Portier, D. G. Pait, G. J. Rosenthal, D. R. Germolec, E. Corsini, B. L. Blaylock, P. Pollock, Y. Kouchi, W. Craig, et al 1993. Risk assessment in immunotoxicology. II. Relationships between immune and host resistance tests. Fundam. Appl. Toxicol. 21:71.[Medline]
  7. Luster, M. I., A. E. Munson, P. T. Thomas, M. P. Holsapple, J. D. Fenters, K. L. White, L. D. Lauer, D. R. Germolec, G. J. Rosenthal, J. H. Dean. 1988. Methods evaluation: development of a testing battery to assess chemical-induced immunotoxicity: National Toxicology Program’s guidelines for immunotoxicity evaluation in mice. Fundam. Appl. Toxicol. 10:2.[Medline]
  8. Luster, M. I., C. Portier, D. G. Pait, K. L. White, C. Gennings, A. E. Munson, G. J. Rosenthal. 1992. Risk assessment in immunotoxicology. I. sensitivity and predictability of immune tests. Fundam. Appl. Toxicol. 18:200.[Medline]
  9. Hair, J. F., R. E. Anderson, R. L. Tatham, W. C. Black. 1992. Multivariate Data Analysis Macmillan, New York.
  10. Keil, D., R. W. Luebke, M. Ensley, P. D. Gerard, S. B. Pruett. 1999. Evaluation of multivariate statistical methods for analysis and modeling of immunotoxicology data. Toxicol. Sci. 51:245.[Abstract/Free Full Text]
  11. Boland, J., K. Atkinson, K. Britton, P. Darveniza, S. Johnson, J. Biggs. 1984. Tissue distribution and toxicity of cyclosporin A in the mouse. Pathology 16:117.[Medline]
  12. Varma, D. R., S. Mulay. 1980. Anti-inflammatory and ulcerogenic effects and pharmacokinetics of dexamethasone in protein-deficient rats. J. Pharmacol. Exp. Ther. 214:197.[Abstract/Free Full Text]
  13. Mordenti, J.. 1986. Man versus beast: pharmacokinetic scaling in mammals. J. Pharm. Sci. 75:1028.[Medline]
  14. White, K. L., A. C. Anderson. 1984. Measurement of modulation of mouse complement levels in vivo, utilizing a microtiter hemolytic assay. Agents Actions 15:562.[Medline]
  15. Murray, M. J., L. D. Lauer, M. I. Luster, R. W. Luebke, D. O. Adams, J. H. Dean. 1985. Correlation of murine susceptibility to tumor, parasite, and bacterial challenge with altered cell-mediated immunity following systemic exposure to the tumor promoter phorbol myristate acetate. Int. J. Immunopharmacol. 7:491.[Medline]
  16. Bryant, J., R. Day, T. L. Whiteside, R. B. Herberman. 1992. Calculation of lytic units for the expression of cell-mediated cytotoxicity. J. Immunol. Methods 146:91.[Medline]
  17. Prell, R. A., N. I. Kerkvliet. 1997. Involvement of altered B7 expression in dioxin immunotoxicity: B7 transfection restores the CTL but not the alloantibody response to the P815 mastocytoma. J. Immunol. 158:2695.[Abstract]
  18. Wu, W.-J., R. M. Wolcott, S. B. Pruett. 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.[Abstract/Free Full Text]
  19. Pruett, S. B., R. Fan, Q. Zheng, L. P. Myers, P. Hebért. 2000. Modeling and predicting selected immunological effects of a chemical stressor (3,4 dichloropropionanalide) using the area under the corticosterone concentration vs. time curve. Toxicol. Sci. 58:77.[Abstract/Free Full Text]
  20. Pruett, S. B., S. Collier, W.-J. Wu, R. Fan. 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.[Abstract/Free Full Text]
  21. Crittenden, P. L., R. Carr, S. B. Pruett. 1998. Immunotoxicological assessment of methyl parathion in female B6C3F1 mice. J. Toxicol. Environ. Health 54:1.
  22. Thomas, P. T., R. Fugmann, C. Aranyi, P. Barbera, R. Gibbons, J. Fenters. 1985. Effect of dimethylnitrosamine on host resistance and immunity. Toxicol. Appl. Pharmacol. 77:219.[Medline]
  23. Barnes, D. B., J. M. Hardin, S. B. Pruett. 1992. Acute infection of mice with highly virulent group B streptococci as a host resistance model for immunotoxicity assessment. Arch. Toxicol. 66:423.[Medline]
  24. Czuprynski, C. J., J. F. Brown, N. Maroushek, R. D. Wagner, H. Steinburg. 1994. Administration of anti-granulocyte mAb RB6-8C5 impairs the resistance of mice to Listeria monocytogenes infection. J. Immunol. 152:1836.[Abstract]
  25. Czuprynski, C. J., J. F. Brown, K. M. Young, A. J. Cooley. 1989. Administration of purified anti-L3T4 monoclonal antibody impairs the resistance of mice to Listeria monocytogenes infection. Infect. Immun. 57:100.[Abstract/Free Full Text]
  26. Higginbotham, J. L., T. L. Lin, S. B. Pruett. 1992. Effect of macrophage activation on killing of Listeria monocytogenes: roles of reactive oxygen or nitrogen intermediates, rate of phagocytosis, and retention of bacteria in endosomes. Clin. Exp. Immunol. 88:492.[Medline]
  27. Wu, W.-J., S. B. Pruett. 1999. Ethanol decreases host resistance to pulmonary metastases in a mouse model: role of natural killer cells and the ethanol-induced stress response. Int. J. Cancer 82:886.[Medline]
  28. Keil, D. E., R. W. Luebke, S. B. Pruett. 1995. Differences in the effects of dexamethasone on macrophage nitrite production: dependence on exposure regimen (in vivo or in vitro) and activation stimuli. Int. J. Immunopharmacol. 17:157.[Medline]
  29. Inglefinger, J. A., F. Mosterller, L. A. Thibodeau, J. H. Ware. 1994. Biostatistics in Clinical Medicine McGraw-Hill, New York.
  30. Dunn, P. L., R. J. North. 1991. Early {gamma} interferon production by natural killer cells is important in defense against murine listeriosis. Infect. Immun. 59:2892.[Abstract/Free Full Text]
  31. Sasaki, T., M. Mieno, H. Udono, K. Yamaguchi, T. Usui, K. Hara, H. Shiku, E. Nakayama. 1990. Roles of CD4+ and CD8+ cells, and the effect of administration of recombinant murine interferon {gamma} in listerial infection. J. Exp. Med. 171:1141.[Abstract/Free Full Text]
  32. Kelso, A., A. Munck. 1984. Glucocorticoid inhibition of lymphokine secretion by alloreactive T lymphocyte clones. J. Immunol. 133:784.[Abstract]
  33. Franchimont, D., J. Galon, M. Gadina, R. Visconti, Y. Zhou, M. Aringer, D. M. Frucht, G. P. Chrousos, J. J. O’Shea. 2000. Inhibition of Th1 immune response by glucocorticoids: dexamethasone selectively inhibits IL-12-induced Stat4 phosphorylation in T lymphocytes. J. Immunol. 164:1768.[Abstract/Free Full Text]
  34. Tsutsui, N., T. Kamiyama. 1998. Suppression of in vitro IFN-{gamma} production by spleen cells of Plasmodium chabaudi-infected C57BL/10 mice exposed to dexamethasone at a low dose. Int. J. Immunopharmacol. 20:141.[Medline]
  35. Dalrymple, S. A., L. A. Lucian, R. Slattery, T. McNeil, D. M. Aud, S. Fuchino, F. Lee, R. Murray. 1995. Interleukin-6-deficient mice are highly susceptible to Listeria monocytogenes infection: correlation with inefficient neutrophilia. Infect. Immun. 63:2262.[Abstract]
  36. Komoriya, A., B. Z. Packard, M. J. Brown, M. L. Wu, P. A. Henkart. 2000. Assessment of caspase activities in intact apoptotic thymocytes using cell-permeable fluorogenic caspase substrates. J. Exp. Med. 191:1819.[Abstract/Free Full Text]
  37. Screpanti, I., S. Morrone, D. Meco, A. Santoni, A. Gulino, R. Paolini, A. Crisanti, B. J. Mathieson, L. Frati. 1989. Steroid sensitivity of thymocyte subpopulations during intrathymic differentiation: effects of 17 {beta}-estradiol and dexamethasone on subsets expressing T cell antigen receptor or IL-2 receptor. J. Immunol. 142:3378.[Abstract]
  38. Stevenson, J. R., J. A. Kreiling, R. Taylor. 1989. Effects of corticosterone on responses of murine splenic B and T cells to phytohemagglutinin, concanavalin A, and lipopolysaccharide. Immunol. Invest. 18:951.[Medline]
  39. Dhabhar, F. S., A. H. Miller, B. S. McEwen, R. L. Spencer. 1996. Stress-induced changes in blood leukocyte distribution: role of adrenal steroid hormones. J. Immunol. 157:1638.[Abstract]
  40. Garvy, B. A., P. J. Fraker. 1991. Suppression of the antigenic response of murine bone marrow B cells by physiological concentrations of glucocorticoids. Immunology 74:519.[Medline]
  41. Todo, T., S. D. Rabkin, A. Chahlavi, R. L. Martuza. 1999. Corticosteroid administration does not affect viral oncolytic activity, but inhibits antitumor immunity in replication-competent herpes simplex virus tumor therapy. Hum. Gene Ther. 10:2869.[Medline]
  42. Cox, W. I., N. J. Holbrook, R. J. Grasso, S. Specter, H. Friedman. 1982. Suppression of the natural killer cell activity of murine spleen cell cultures by dexamethasone. Proc. Soc. Exp. Biol. Med. 171:146.[Medline]
  43. Takeuchi, Y., S. Habu, K. Okumura, G. Suzuki. 1989. Cyclosporin A and anti-Ia antibody cause a maturation defect of CD4+8- cells in organ-cultured fetal thymus. Immunology 66:362.[Medline]
  44. Koprak, S., A. Sirotina, H. Ok, F. J. Dumont. 1996. Depletion of the mature CD4+8- thymocyte subset by FK506 analogs correlates with their immunosuppressive and calcineurin inhibitory activities. Transplantation 61:926.[Medline]
  45. Di Rosa, M., M. Radomski, R. Carnuccio, S. Moncada. 1990. Glucocorticoids inhibit the induction of nitric oxide synthase in macrophages. Biochem. Biophys. Res. Commun. 172:1246.[Medline]
  46. Fantuzzi, G., G. Galli, M. Zinetti, M. Fratelli, P. Ghezzi. 1995. The up-regulating effect of dexamethasone on tumor necrosis factor production is mediated by a nitric oxide-producing cytochrome P450. Cell. Immunol. 160:305.[Medline]
  47. Attallah, A. M., A. Urritia-Shaw, T. J. Yeatman, T. Folks. 1981. Effect of the new immunosuppressive agent, cyclosporin-A, on natural killer cell and antibody-dependent cell-mediated cytotoxic activity. Int. Arch. Allergy Appl. Immunol. 65:465.[Medline]
  48. Selgrade, M. K., M. J. Daniels, J. H. Dean. 1992. Interpreting immunotoxicity testing in terms of enhanced risk of disease: correlation between suppression of natural killer cell activity in mice and susceptibility to cytomegalovirus. J. Toxicol. Environ. Health 37:123.[Medline]
  49. Kosugi, A., G. M. Shearer. 1991. Effect of cyclosporin A on lymphopoiesis. III. Augmentation of the generation of natural killer cells in bone marrow transplanted mice treated with cyclosporin A. J. Immunol. 146:1416.[Abstract]
  50. Kaibori, M., K. Sakitani, M. Oda, Y. Kamiyama, Y. Masu, M. Nishizawa, S. Ito, T. Okumura. 1999. Immunosuppressant FK506 inhibits inducible nitric oxide synthase gene expression at a step of NF-{kappa}B activation in rat hepatocytes. J. Hepatol. 30:1138.[Medline]
  51. Vos, I. H., J. A. Joles, M. Schurink, G. Weckbecker, T. Stojanovic, T. J. Rabelink, H. J. Grone. 2000. Inhibition of inducible nitric oxide synthase improves graft function and reduces tubulointerstitial injury in renal allograft rejection. Eur. J. Pharmacol. 391:31.[Medline]
  52. Wiltrout, R. H., R. B. Herberman, S.-R. Zhang, M. A. Chirigos, J. R. Ortaldo, Jr K. M. Green, J. E. Talmadge. 1985. Role of organ-associated NK cells in decreased formation of experimental metastases in lung and liver. J. Immunol. 134:4267.[Abstract]
  53. Baker, C. J., D. L. Kasper. 1985. Group B streptococcal vaccines. Rev. Infect. Dis. 7:458.[Medline]
  54. Stanton, B. F., R. S. Baltimore, D. G. Shedd. 1981. Effects of route and time of administration of antiserum on protection of mice from lethal infection due to group B streptococcus type III. Infect. Immun. 31:391.[Abstract/Free Full Text]
  55. Wennerstrom, D. E., R. W. Schutt. 1978. Adult mice as a model for early onset group B streptococcal disease. Infect. Immun. 19:741.[Abstract/Free Full Text]
  56. Edwards, M. S., M. R. Wessels, C. J. Baker. 1993. Capsular polysaccharide regulates neutrophil complement receptor interactions with type III group B streptococci. Infect. Immun. 61:2866.[Abstract/Free Full Text]
  57. Vedhara, K., N. K. Cox, G. K. Wilcock, P. Perks, M. Hunt, S. Anderson, S. L. Lightman, N. M. Shanks. 1999. Chronic stress in elderly carers of dementia patients and antibody response to influenza vaccination. Lancet 353:627.[Medline]
  58. Newell, K. W., D. R. Leblanc, G. Edsall, L. Levine, H. Christensen, M. H. Montouri, N. Ramirez. 1971. The serological assessment of a tetanus toxoid field trial. Bull. W. H. O. 45:773.[Medline]
  59. Strauss, R., B. Heymer, H. Hof. 1985. Effects of cyclosporin A on experimental infection with Listeria monocytogenes. Clin. Exp. Immunol. 62:491.[Medline]
  60. Hugin, A. W., A. Cerny, M. Wrann, H. Hengartner, R. M. Zinkernagel. 1986. Effect of cyclosporin A on immunity to Listeria monocytogenes. Infect. Immun. 52:12.[Abstract/Free Full Text]
  61. Baldridge, J. R., R. A. Barry, D. J. Hinrichs. 1990. Expression of systemic protection and delayed-type hypersensitivity to Listeria monocytogenes is mediated by different T-cell subsets. Infect. Immun. 58:654.[Abstract/Free Full Text]
  62. Cerrito, P. B., S. C. Koenig, D. J. VanHimbergen, S. F. Jaber, D. L. Ewert, P. A. Spence. 1999. Neural network pattern recognition analysis of graft flow characteristics improves intra-operative anastomotic error detection in minimally invasive CABG. Eur. J. Cardiothorac. Surg. 16:88.[Abstract/Free Full Text]
  63. Faro, J., S. Velasco. 1993. Studies on a recent class of network models of the immune system. J. Theor. Biol. 164:271.[Medline]



This article has been cited by other articles:


Home page
 Infect. Immun.Home page
W. G. Jaoko, E. Michael, D. W. Meyrowitsch, B. B. A. Estambale, M. N. Malecela, and P. E. Simonsen
Immunoepidemiology of Wuchereria bancrofti Infection: Parasite Transmission Intensity, Filaria-Specific Antibodies, and Host Immunity in Two East African Communities
Infect. Immun., December 1, 2007; 75(12): 5651 - 5662.
[Abstract] [Full Text] [PDF]

Home page
 J. Appl. Physiol.Home page
M. Gleeson
Immune function in sport and exercise
J Appl Physiol, August 1, 2007; 103(2): 693 - 699.
[Abstract] [Full Text] [PDF]

Home page
 Poult. Sci.Home page
B. Ask, E. H. van der Waaij, E. J. Glass, and S. C. Bishop
Modeling Immunocompetence Development and Immunoresponsiveness to Challenge in Chicks
Poult. Sci., July 1, 2007; 86(7): 1336 - 1350.
[Abstract] [Full Text] [PDF]