Main

Glucocorticoids (GC), the final mediators of hypothalamic-pituitary-adrenal (HPA) activation, are important regulators of various physiological systems, including the immune system, and play a major role in the adaptation of organisms to stressful situations. Previous studies in humans and animals have shown that circulating GCs are beneficial during the adaptive process in the short run, but during long-term or repeated exposure to stressors, the effects of GCs on immune function are detrimental (1,2).

Furthermore, it is well known that there are crucial interactive loops between GCs and cytokines. Proinflammatory cytokines, produced by activated immune cells, are potent activators of the HPA axis. GCs in turn suppress cytokine production and, by this mechanism, are able to terminate immune processes to protect the organism from an overactive immune system (3,4). An increasing number of studies suggest that GCs may cause alterations in cytokine production, which favor humoral immune responses while suppressing cellular immunity (5,6). Although this model of immune deviation could be an adaptive mechanism to prevent the immune response from causing tissue damage, maladaptive responses to stress-induced immune alterations may contribute to increased disease susceptibility (7).

In addition to peripheral GC levels, the GC sensitivity of different target cells from organisms exposed to stressors should also be considered when evaluating adaptive processes, potential imbalances, and increased health risks (3,4). Several studies have supported the hypothesis that social stressors affect the steroid sensitivity of immune cells in animals and humans. As shown in mice, repeated social disruption stress may cause reduced GC sensitivity in splenocytes (810). In addition, the corticosteroid sensitivity of peripheral blood lymphocytes was decreased in chronically stressed caregivers of patients with dementia (11).

However, acute modulation of GC sensitivity in response to short-term psychosocial stress has only been investigated in a small number of studies. In students, it has been demonstrated that stress associated with academic examinations provokes an activation of the HPA axis with increased levels of cortisol followed by a transient decrease in the GC sensitivity of leukocytes ex vivo (12). Similarly, the laboratory Trier Social Stress Test induced changes in GC sensitivity for proinflammatory cytokine production by lipopolysaccharide (LPS)-stimulated whole blood cultures of healthy humans (4,13). Moreover, there is also evidence for age-related changes in GC sensitivity of different target tissues in human and animal models, with greater sensitivity observed in younger individuals (14,15).

Psychosocial stress in early life, such as social deprivation and maternal separation, has been shown to induce robust alterations in the physiological mechanisms of adaptation. In humans, early life stress is viewed as a major risk factor for the development of mental disorders (15,16) and immune-related diseases in later life (17). Although the importance of psychosocial factors for offspring development is well established, limited information exists about how the immune and stress systems interact to cope with psychosocial stressors in early postnatal life.

Recent findings from our group indicate that social isolation of piglets reliably activates the release of stress hormones and causes changes in the proportions of blood lymphocytes, reflecting also the negative emotions experienced by this treatment (18,19). In contrast to rodents, the pig HPA axis is well developed and functional at birth (20). Therefore, this social isolation model in pigs may be useful for studying the effects of psychosocial stress on the responsiveness of the immune-neuroendocrine system early in life. Based on our previous research, we hypothesized that social stress in neonatal pigs alters circulating levels of cytokines and also affects the GC sensitivity of peripheral immune cells. To test this hypothesis, we examined the effects of an activated HPA axis in piglets exposed to a single social isolation for 4 h at different age categories on cytokines. Plasma concentrations of stress hormones (ACTH and cortisol) and proinflammatory cytokines (TNF-α and IL-6) were analyzed. Furthermore, the reactivity of cells of the immune system and the sensitivity of these cells to GC inhibition were assessed by measuring the in vitro proliferation of peripheral blood mononuclear cells (PBMCs) in response to the T-cell mitogen ConA and the B-cell mitogen LPS under a range of increasing cortisol concentrations.

METHODS

Animals and experimental design.

The procedures involving animal treatments were approved by the Committee on Animal Care and Use of the Agricultural Department of Mecklenburg-Vorpommern, Germany. Piglets were taken from eight German Landrace litters that were born and raised in the experimental pig unit of our institute. After birth, the litter size was standardized to 10 piglets. During the suckling period, sows and their piglets were housed in a loose farrowing pen (6 m2) at a room temperature of 28 ± 1°C. At 7, 21, or 35 d of age, two piglets from each litter were randomly allocated to an isolation treatment or to a nonisolated control group. The allocation of male and female piglets within both groups was approximately equivalent. The piglets were isolated once from their mother and siblings in a separate test room located in the same experimental station for 4 h in the morning (0700–1100 h). Each piglet was placed alone into an opaque plastic box (68 × 50 × 65 cm) with sawdust on the floor, adequate air passage, and temperature control (28 ± 1°C). The control piglets remained undisturbed in the farrowing pen during this time. Blood samples were taken while piglets were in a supine position by anterior vena cava puncture (the whole procedure lasted 30 s) from both isolated (n = 8 per age group) and nonisolated control piglets (n = 8 per age group) both before, for basal levels of stress hormones, and immediately after the social isolation. One aliquot of EDTA blood samples was centrifuged at 2000 × g for 15 min at 4°C to separate plasma, which was stored at −20°C until analysis of ACTH, cortisol, TNF-α, and IL-6. Another aliquot of heparinized blood samples was stored on ice until processing for immunological measurements.

Hormone analyses.

The analyses of ACTH concentrations were performed in duplicate on 200 μL plasma using a commercial 125I-RIA kit (DSL, Inc., Sinsheim, Germany) according to the instructions of the manufacturer. The lowest level of ACTH that could be detected by this assay was 3.5 pg/mL, and intra- and interassay coefficients of variation were 6.9 and 9.6%, respectively. Plasma cortisol concentrations were analyzed in duplicates using a commercially available 125I-RIA kit (DSL, Inc.) according to the manufacturer's instructions. The test sensitivity was 8.1 nmol/L, and intra- and interassay coefficients of variation were 8.2 and 9.8%, respectively. The ratio of ACTH/cortisol (A/C) was calculated. Both assays were validated for porcine plasma.

Measurement of mitogenic response and cortisol sensitivity.

The mitogens ConA (5 μg/mL) and LPS (10 μg/mL) were used in a lymphocyte proliferation/viability assay as previously described (19). Briefly, PBMCs were isolated from heparinized blood by density gradient centrifugation, and the cell concentration was adjusted to 5 × 106 cells/mL complete RPMI 1640 medium. To test the sensitivity of cells to inhibition by GCs, aliquots from each cell suspension were treated with increasing concentrations of cortisol (0, 0.05, 0.1, 0.5, and 1 μM) diluted in a buffer of 0.2% ethanol in complete medium. All reagents were obtained from Sigma Chemical Co. (St. Louis, MO). Cell suspensions were added in triplicate to flat-bottom 96-well plates at a volume of 200 μL/well, and plates were incubated for 72 h in a 5% CO2-humidified incubator at 37°C. Cell proliferation/viability was evaluated using 3-[4,5-dimethyldiazol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT) assay (Roche Diagnostics, Mannheim, Germany). The OD was measured by a microplate reader (Dynatech, Denkendorf, Germany) using a test wavelength of 550 nm and a reference wavelength of 690 nm. The results were expressed as mitogen-stimulated proliferation (OD in the presence of mitogen subtracted by OD in the absence of mitogen) and as cortisol resistance index, which was calculated as the OD for a culture treated with cortisol divided by the OD for cells from the same group not treated with cortisol × 100 as described by Stark et al. (8).

Cytokine assays.

TNF-α and IL-6 concentrations were analyzed in plasma samples using commercially available pig ELISA kits (DRG Instruments GmbH, Marburg, Germany) according to the manufacturer's instructions. Samples were analyzed in duplicate at a 1:2 dilution. The sensitivity of the TNF-α assay was 6 pg/mL and intra- and interassay coefficients of variation were 6.2 and 8.2%, respectively. The limit of detection of the IL-6 assay was 45 pg/mL. Intra- and interassay coefficients of variation were 4.5 and 9.8%, respectively.

Statistical analysis.

Statistical analyses were performed using the SAS System for Windows, release 9.2 (21). Data were evaluated by ANOVA using the MIXED procedure. The model included the fixed classification variables social isolation (isolation and control), age (d 7, 21, and 35), sex (male and female), the interaction age × isolation, and the random sow effect. The ANOVA model for the cortisol resistance index contained the fixed effects social isolation (isolation and control), age (d 7, 21, and 35), and mitogen (ConA and LPS) and the repeated factor cortisol concentration with measures at four levels (0.05, 0.1, 0.5, and 1 μM) using an unstructured residual covariance matrix and the random sow effect, and all interactions between the fixed effects. In addition, least-squares means (LS-means) and SE were computed for each effect in the models. Significance of differences between LS-means was tested by the Tukey-Kramer procedure (p < 0.05).

RESULTS

Plasma hormones and cytokines.

Before the social isolation on d 7, 21, and 35, there were no significant differences in plasma ACTH and cortisol concentrations observed between control piglets and those that would be isolated (p > 0.89, data not shown).

ACTH (Fig. 1A), cortisol (Fig. 1B), and TNF-α (Fig. 2A) concentrations were affected by social isolation (p < 0.01), whereas IL-6 concentrations were not influenced by isolation (p = 0.23; Fig. 2B). Pair-wise comparisons of the LS-means of isolated and control piglets indicated a significantly higher ACTH concentration on d 35 (p < 0.05) and significantly higher cortisol concentrations on d 7 (p < 0.05) and 21 (p < 0.05) in isolated animals (Fig. 1). As shown in Fig. 2A, the TNF-α concentration in plasma from isolated piglets was significantly lower compared with controls on d 7 (p < 0.05).

Figure 1
figure 1

Plasma ACTH (A) and cortisol (B) concentrations in isolated (▪; n = 8 at each age) and control piglets (; n = 8 at each age) after social isolation of 4 h on d 7, 21, and 35. Data are expressed as LS-means + SE. Significant differences between isolated and control piglets are indicated by a dagger (†p < 0.05).

Figure 2
figure 2

Plasma TNF-α (A) and IL-6 (B) concentrations in isolated (▪; n = 8 at each age) and control piglets (; n = 8 at each age) after social isolation of 4 h on d 7, 21, and 35. Data are expressed as LS-means + SE. Significant differences between isolated and control piglets are indicated by a dagger (†p < 0.05).

Mitogenic response and cortisol sensitivity.

There was no effect of social isolation on the proliferation of PBMCs in response to ConA (p = 0.455). In contrast, the response to LPS was influenced by isolation (p < 0.001). As shown in Table 1, LPS-stimulated proliferation was significantly higher in isolated piglets compared with controls on d 21 (p < 0.05) and 35 (p < 0.01). In addition, the factor age had an effect on mitogen-induced PBMC proliferation in response to both ConA and LPS (p < 0.0001). Sensitivity of PBMCs to increasing cortisol concentrations, expressed as resistance indices (%), was assessed in cultures stimulated with ConA (Fig. 3) or LPS (Fig. 4). Repeated-measures ANOVA indicated that the sensitivity of PBMCs from piglets to cortisol was affected by social isolation (p < 0.001), concentration of cortisol (p < 0.001), type of mitogen (p < 0.05), and age of the piglets (p < 0.001). Furthermore, significant interactions of social isolation × age (p < 0.01), cortisol concentration × mitogen (p < 0.001), and mitogen × age (p < 0.01) were found. ConA-stimulated PBMCs from isolated piglets showed significantly less inhibition by all cortisol concentrations than cells from control piglets on d 7 (Fig. 3A), 21 (Fig. 3B), and 35 (Fig. 3C; in all cases, p < 0.001). LPS-induced stimulation of PBMCs from isolated piglets on d 7 of age was also significantly less inhibited by cortisol at all concentrations (p < 0.001; Fig. 4A), whereas on d 21, PBMCs from isolated piglets seemed to be more resistant to cortisol only at 0.5 μM (p < 0.01) and 1 μM (p < 0.01; Fig. 4B). However, there was no difference in resistance of LPS-stimulated cultures to cortisol between isolated and control piglets on d 35 (p > 0.56; Fig. 4C). Furthermore, a dose-dependent decrease of cortisol resistance was found in ConA-stimulated cultures from both isolated and control piglets at 7, 21, and 35 d of age (p < 0.01; Fig. 3A–C) and, in LPS-stimulated PBMCs, only from control piglets on d 7 (p < 0.02, Fig. 4A). Pair-wise comparisons of resistance indices between ConA- and LPS-stimulated PBMCs from isolated piglets indicated that ConA-induced cells seemed to be more resistant to cortisol than LPS-stimulated cells at 0.05 μM cortisol concentration on d 21 (92.4 ± 1.2 versus 85.2 ± 1.5%, p = 0.004, Figs. 3B and 4B) and 35 (95.8 ± 1.2 versus 88.8 ± 1.5%, p = 0.008, Figs. 3C and 4C). However, in control piglets, ConA-stimulated PBMCs showed a lower cortisol resistance than LPS-stimulated cultures only at the highest cortisol concentration on d 35 (62.9 ± 1.8 versus 80.3 ± 1.9%, p < 0.001, Figs. 3C and 4C). Cortisol resistance did not differ among the three age categories, in ConA-stimulated cultures from isolated and control piglets (p > 0.92) or in LPS-stimulated cultures from isolated piglets (p > 0.81). In contrast, resistance to cortisol in LPS-stimulated cultures from control piglets was significantly lower on d 7 (Fig. 4A) compared with d 35 (Fig. 4C) at 0.1, 0.5, and 1 μM cortisol (70.9 ± 2.2 versus 82.5 ± 2.2%, p < 0.02; 67.2 ± 1.9 versus 77.7 ± 1.9%, p < 0.01; 62.8 ± 1.9 versus 80.3 ± 1.9%, p < 0.001, respectively) and on d 21 (Fig. 4B) compared with d 35 (Fig. 4C) at 0.5 and 1 μM cortisol (68.5 ± 1.9 versus 77.7 ± 1.9%, p < 0.05; 68.4 ± 1.9 versus 80.3 ± 1.9%, p < 0.01, respectively). The factor sex had no significant effect on any of the traits investigated (p > 0.29; data not shown).

Table 1 ConA- and LPS-stimulated proliferation of PBMCs in isolated and control piglets after social isolation on d 7, 21, and 35
Figure 3
figure 3

Cortisol resistance index (%) of ConA-stimulated PBMCs cultured with increasing cortisol concentrations (0.05, 0.1, 0.5, and 1 μM) for isolated (▪; n = 8 at each age) and control piglets (; n = 8 at each age) after social isolation of 4 h on d 7 (A), 21 (B), and 35 (C). The index was calculated as the OD for a culture treated with cortisol divided by the OD for cells from the same group not treated with cortisol × 100. Data are expressed as LS-means + SE. Significant differences between isolated and control piglets are indicated by asterisks (**p < 0.001).

Figure 4
figure 4

Cortisol resistance index (%) of LPS-stimulated PBMCs cultured with increasing cortisol concentrations (0.05, 0.1, 0.5, and 1 μM) for isolated (▪; n = 8 at each age) and control piglets (; n = 8 at each age) after social isolation of 4 h on d 7 (A), 21 (B), and 35 (C). The index was calculated as the OD for a culture treated with cortisol divided by the OD for cells from the same group not treated with cortisol × 100. Data are expressed as LS-means + SE. Significant differences between isolated and control piglets are indicated by asterisks (*p < 0.01 and **p < 0.001).

DISCUSSION

There are numerous challenges within a domestic piglet's environment such as handling by humans and abrupt weaning that involve psychosocial factors, which may play critical role in adaptive responses of the neuroendocrine and immune systems. In this study, a single exposure to social isolation for 4 h was perceived as a stressful condition by piglets at each of the three age categories. This was indicated by increased plasma ACTH and cortisol levels, which confirm that social isolation is a robust paradigm of psychosocial stress in pigs (22,23).

Furthermore, a single social isolation stress in our experiments resulted in lower plasma TNF-α levels but had no significant effect on IL-6 levels. Recent findings in animal and human research suggest that short-term psychological stress may alter the circulating levels of inflammatory cytokines (24,25). TNF-α is an important proinflammatory cytokine that regulates inflammatory responses to infection and stress and activates the HPA axis. GCs in turn act as a negative feedback regulator and suppress further release of proinflammatory cytokines, thereby protecting against overstimulation of the immune system (3,4). Therefore, we suppose that the increased cortisol levels in socially isolated piglets are responsible for the lower TNF-α levels of these animals.

Several reports have shown that repeated or chronic activation of the HPA axis can cause increased systemic GC levels and may lead to altered activation and diminished GC responsiveness of immune cells resulting in excessive inflammatory responses and increased health risks (8,9,11). In our study, the consequences of short-term psychosocial stress on cell proliferation and GC sensitivity of T lymphocytes were investigated by stimulating PBMCs with ConA (26), and LPS was used for stimulation of B cells and monocytes (27). We found that social isolation had no effect on lymphocyte proliferation in response to ConA, whereas in vitro LPS-induced cell proliferation was increased in isolated piglets. Our results support previous findings in which mice exposed to social disruption demonstrated enhanced proliferation of LPS-stimulated splenocytes (8,10,28). In addition, our data suggest that short-term social stress promotes humoral immune responses. In this study, the mechanism underlying the different effects of isolation stress on modulation of ConA- and LPS-stimulated proliferation responses were not examined. However, animal studies suggest that those effects could be mediated by altered glucocorticoid receptor (GR) expression and/or function (29). For example, ligand-induced GR down-regulation is a significant mechanism to inhibit GC signaling in various tissues and cell types but not in T lymphocytes (30).

An in vitro cortisol sensitivity assay showed that ConA- and LPS-stimulated PBMCs from socially isolated piglets were less sensitive to the inhibitory effects of cortisol, even in the presence of high concentrations. To our knowledge, this is the first study to demonstrate that a short-term psychosocial stress may induce cortisol resistance in PBMC cultures from neonates. Similarly, GC resistance of blood immune cells has been reported after a single physical exercise (3) or mental stress in humans (4,12,13). However, most previous work using socially defeated mice has shown decreased GC sensitivity of spleen cells, which was more pronounced in animals with higher number of injuries because of fighting (8,9). In particular, a study from Merlot et al. (10) reported that wounded and nonwounded mice presented similar responses to social stress. In general, development of GC resistance after repeated or chronic stress is viewed as maladaptive if the resistant individual has a predisposition to an autoimmune disease or is exposed to an infectious challenge but it may be also adaptive for healing wounds and protection against bacterial contamination (8,28,31). The effects of GC resistance may be different depending on the nature and timing of the stressor and the individual state of the organism (32). Therefore, the rapid development of cortisol resistance observed in the present isolation model could be adaptive for the organism to preserve cell function and prepare the immune system for potential unpredictable danger. However, identification of the mechanisms mediating such a rapid modulation of GC sensitivity and their possible consequences remains to be investigated.

Although in our study, LPS-stimulated cells from isolated piglets showed higher cell proliferation compared with controls on d 21 and 35, we found that these cells were more sensitive to inhibition by minimal cortisol concentrations than ConA-stimulated cell cultures. Similarly, in academically stressed students, there was a positive correlation between the degree of lymphocyte activation and cortisol inhibition in vitro (12). It has been suggested that GC sensitivity of lymphocytes varies with the state of immunological activation of the cells and can be modulated by the GR number or expression level in the cells (33) and by the hormone binding affinity of the GR (4).

LPS-stimulated PBMCs of younger control piglets were much more sensitive to inhibition of the proliferative response by cortisol than cells of piglets at 35 d of age. This finding confirms an age-dependent effect, because it was documented that neonatal lymphocytes are more sensitive to GC inhibition than are those from older pigs (15). In human infants and children, a decreasing GC sensitivity with advancing age has also been shown (14). In both human and pigs, the immune system is not fully developed at birth (34,35). Therefore, age-related changes in GC sensitivity may be related to a still functionally immature immune system. It is known that sudden separation of suckling piglets from the mother at an early age impairs antibody-mediated immunity (36), which may be partially explained by an increased sensitivity of neonatal lymphocytes to GC inhibition (15). Hence, with respect to the present results, an increasing weaning age could be associated with enhanced protection of piglets against frequently occurring bacterial infections of the gastrointestinal and respiratory tracts provoked by stressful experiences after weaning.

Recently, we have demonstrated that repeated social isolation of neonatal pigs may cause long-term changes in neuroendocrine and immune regulation and can modulate coping mechanisms against a later risk of infection (2,23). A single exposure to social stress in rodents has been found to induce short-term effects on immune responses (37) and long-term consequences on behavior, HPA responsiveness, and immune functions (17,38). Therefore, future studies are needed to investigate whether a single social isolation of piglets during the early neonatal period can also modify sensitivity to stress and infection later in life.

In summary, this study shows that a single exposure to isolation stress in piglets causes activation of the HPA axis and suppression of circulating TNF- α. Furthermore, isolation induces a state of cortisol resistance in blood immune cells, which may be an adaptive advantage to maintain cellular immune responses in the short term. These results extend previous findings about acute modulation of GC sensitivity in immune tissue after psychological stress. This may have implications for the assessment of mental experiences in young animals and humans.