Activation of the ovine hypothalamic-pituitary-adrenal axis and febrile response by interleukin-6: A comparative study with bacterial lipopolysaccharide endotoxin

Abstract

Sheep were subjected to immune challenge with either recombinant human interleukin-6 (rhIL-6; 2.0 μg/kg; n = 5), Escherichia coli lipopolysaccharide (LPS) endotoxin (400 ng/kg; n = 7), or saline (n = 6) to determine if IL-6 activates the febrile and hypothalamic-pituitary-adrenal axis (HPAA) responses in sheep, and to compare these responses with those associated with endotoxemia. Blood was collected over time to measure plasma adrenocorticotropic hormone (ACTH) and serum cortisol concentrations as indicators of HPAA activity. Unlike LPS, rhIL-6 was not pyrogenic in sheep at this challenge dose. In contrast, rhIL-6 elicited ACTH and cortisol responses that peaked earlier than those induced by LPS. These results suggest that this dose of IL-6, alone, is not sufficient to elicit the febrile response in sheep, however, it is a potent activator of the ovine HPAA response.

Introduction

Bi-directional communication occurs between the neuroendocrine and immune systems during gram-negative bacterial infections as a means to maintain or restore physiological homeostasis (1). Toll-like receptor 4 (TLR-4) complexes expressed on the surface of host sentinel cells initially mediate pathogen recognition by ligating to lipopolysaccharide (LPS) endotoxin derived from the bacterial membrane. This ligation initiates the induction of numerous gene products, including the cytokines TNF-α, IL-1, and IL-6. These pro-inflammatory cytokines contribute to the activation of thermoregulatory neurons within the hypothalamus (2), and to the activation of the hypothalamic-pituitary-adrenal axis (HPAA), which leads to secretion of corticotrophin-releasing factor (CRF) and arginine vasopressin (AVP) from the hypothalamus. Both CRF and AVP subsequently initiate the synthesis and release of adrenocorticotrophic hormone (ACTH) from the anterior pituitary into the circulatory system, which stimulates the adrenal cortex to secrete glucocorticoids such as cortisol into the circulatory system (1). Glucocorticoids have a wide range of immunomodulatory properties, one of which includes controlling the potentially damaging host inflammatory response that was mounted against the pathogen (3).

Interleukin-6 is a pleiotropic cytokine with reported bioactivities that include activation of the febrile response in rats during endotoxemia (4), and activation of the HPAA response in mice (5), and primates (6). A number of studies, several of which were carried out in our laboratory, have reported increased circulating IL-6 concentrations in sheep experiencing endotoxemia (7–11). However, we have been unable to correlate increased IL-6 concentration with the febrile and HPAA response in sheep, raising question about its involvement in eliciting these responses in this species (9–11).

Therefore, the purpose of this study was to determine if IL-6 activates the febrile and HPAA responses in sheep, and to compare these responses with those associated with endotoxemia. Sheep were systemically challenged with either recombinant human IL-6 (rhIL-6), Escherichia coli LPS, or saline, and the febrile and HPAA responsiveness were assessed over time; previous studies have demonstrated that rhIL6 is bioactive in guinea pigs (12), rabbits (13), and ruminants (14).

Materials and methods

Experimental animals

Six-month-old, Rideau-Arcott ewe lambs were subjected to bolus IV challenge with either LPS from E. coli (serotype 0111:B4, n = 7, 400 ng/kg; Sigma Chemical, St. Louis, Missouri, USA), rhIL-6 (n = 5, 2.0 μg/kg, R&D Systems, Minneapolis, Minnesota, USA), or saline as a control (n = 6). The LPS dose was determined from a previous dose-response study (10), whereas the dose of rhIL-6 was estimated from the literature (12). The challenge studies were carried out over a period of 6 d to facilitate sample collection. Body temperatures were monitored during the challenges by measuring rectal temperature with a standard digital thermometer. All animals were held at the Ontario Ministry of Agriculture and Food and Rural Affairs Ponsonby Sheep Research Station, Ontario, and housed in individual pens with access to food and water ad libitum during the challenge. The University of Guelph Animal Care Committee approved all procedures involving these animals.

Blood collection

Jugular blood was collected into either heparinized, or silicon gel and clot activator Vacutainer tubes (Becton Dickinson and Company, Oakville, Ontario) 0, 0.25, 0.5, 1, 2, and 3 h post immune challenge to obtain plasma and serum, respectively. Blood was centrifuged at 1000 × g for 15 min at room temperature to obtain plasma and serum. For serum collection, the blood was allowed to clot for approximately 45 min at room temperature prior to centrifugation. Plasma and serum were aliquoted into microcentrifuge tubes and stored at −80°C.

ACTH and cortisol response to immune challenge

Plasma ACTH concentrations were determined using a commercially available chemiluminescence enzyme-linked immunosorbent assay (ELISA) kit (Calbiotech, Spring Valley, California, USA) and a Victor 3 plate reader (Perkin-Elmer, Wellesley, Massachusetts, USA). The ACTH response in plasma samples from all subject animals was assessed for the time periods 0, 0.25, 0.5, 1, 2, and 3 h post-immune challenge. Samples were analyzed in triplicate with an average intra-assay CV of 3.8% for all plates.

Serum cortisol concentrations were measured using a commercially available luminescence immunoassay kit (IBL Hamburg, Minneapolis, Minnesota, USA) and a Victor 3 plate reader. The cortisol response was measured 0, 1, 2, and 3 h post-immune challenge. Samples were analyzed in triplicate with an average intra-assay coefficient of variation (CV) of 8.5% for all plates.

Statistical analysis

Data were analyzed as a complete block design using SAS (SAS 2002; SAS Institute, Cary, North Carolina, USA), and considered significant when P < 0.05. Residual plots were examined to assess variance homogeneity, and natural log transformations were utilized when required. Statistical analysis was carried out using the PROC MIXED procedure with repeated measurements over time, and the mixed procedure, incorporating the best fitting covariance structure, included in the model (15). Time trends across the 3 h study were compared among treatment groups (LPS, rhIL-6, saline) from differences of linear and quadratic orthogonal polynomial contrasts across time among treatment groups (16).

Results

Sheep responded significantly to the LPS challenge with an increase in body temperature (Figure 1). Significant LPS*saline (P < 0.01) and LPS*rhIL-6 (P < 0.01) linear contrasts were observed between treatments. Significant rhIL-6*saline quadratic contrasts (P = 0.02) were also observed; however, they are not deemed to be biologically significant given that the difference between these treatments was largely determined by different 0 h temperature measurements.

Figure 1.

Ovine febrile response to systemic challenge with recombinant human IL-6 (rhIL-6; 2.0 μg/kg; n = 5), Escherichia coli lipopolysaccharide (LPS) endotoxin (400 ng/kg; n = 7), or saline (n = 6). Results are presented as the mean concentration ± standard deviation (s).

Both LPS and IL-6 challenges induced serum cortisol in sheep (Figure 2). Significant linear contrasts between LPS*saline (P < 0.01) and LPS*rhIL-6 (P < 0.01), and quadratic contrasts between LPS*saline (P < 0.01) and rhIL-6*saline (P < 0.01) were measured (Figure 2A). The cortisol response to LPS peaked within 2 h and remained elevated for the 3-h duration of the study. In contrast, the peak cortisol response to rhIL-6 challenge occurred around 1 h and returned to basal levels within 2 h post-challenge.

Figure 2.

Ovine cortisol response (A) and ACTH response (B) to systemic challenge with recombinant human IL-6 (rhIL-6; 2.0 μg/kg; n = 5), Escherichia coli lipopolysaccharide (LPS) endotoxin (400 ng/kg; n = 7), or saline (n = 6). Results are presented as the mean concentration ± standard deviation (s).

Plasma ACTH was induced during LPS and rhIL-6 challenge (Figure 2B). Significant LPS*saline (P > 0.01) and LPS*rhIL-6 (P < 0.02) linear contrasts, and rhIL-6*saline (P < 0.01) quadratic contrasts were observed between the treatments. The ACTH response to LPS peaked within 2 h and remained elevated for the duration of the study, whereas the peak ACTH response to rhIL-6 challenge occurred around 30 min and returned to basal levels within 2 h post-challenge.

Discussion

Circulating concentrations of the pro-inflammatory cytokine IL-6 increase as part of the host response to bacterial infection. This cytokine is critical for regulating the innate and acquired immune systems (17) and is, in part, responsible for inducing the febrile and HPAA response associated with endotoxemia (4,5).

Circulating concentrations of ovine IL-6 have also been reported to increase during endotoxemia (7–11). A study by Kabaroff et al (10), for example, showed that systemic challenge with LPS (400 or 600 ng/kg) caused a significant and sustained increase in serum IL-6 concentration 3–7 h post-challenge; however, serum IL-6 concentrations did not correlate with either serum cortisol concentration, or the febrile response. Another study, conducted by You et al (11), also demonstrated that LPS-induced serum IL-6 concentrations did not correlate with cortisol concentrations in sheep that had been selected based on their phenotypic extreme cortisol response to LPS (400 ng/kg) (11). Kabaroff et al later carried out a LPS challenge study (400 ng/kg) using pregnant and lactating sheep and showed that ovine IL-6 was non-responsive to LPS during early-to-mid-pregnancy, yet febrile and cortisol responses were observed. This was in contrast to late pregnancy, where the IL-6 response to LPS occurred, and correlated with the cortisol but not febrile response to LPS (9). Lastly, a study carried out by McClure et al (8) demonstrated that during late pregnancy, LPS (300 ng/kg) elicited a febrile and IL-6 response that is attenuated in comparison to non-pregnant ewes. Given the variable results of these studies, some of which were likely attributed to hormonal changes associated with pregnancy, it is difficult to assess the significant contribution of IL-6 in the induction of the ovine febrile and HPAA responses during endotoxemia.

The purpose of this study, therefore, was to determine whether or not IL-6, alone, induces the febrile and HPAA response in sheep. The results presented herein suggest that IL-6 is not pyrogenic but elicits transient activation of the HPAA, as determined by increased blood ACTH and cortisol concentrations. The dose of rhIL-6 used in this study was equivalent to 33 ng/mL of plasma [estimated on the basis of 60 mL/kg body weight (BW) for a 35 kg sheep], which is physiologically relevant but higher than peak IL-6 concentrations of 5 ng/mL reported for sheep challenged with 300 ng/kg LPS (8). Since rhIL-6 binding affinity to the IL-6 receptor was not assessed in this study, it is likely that the bioactive concentration was lower than this, and that higher doses of rhIL-6 may be required to induce a febrile response in sheep.

A number of studies have demonstrated that IL-6 is weakly pyrogenic in various species, depending on the route of administration and dose. One study reported that rhIL-6, but not heat-treated rhIL-6, was pyrogenic when administered to rabbits by intracerebroventricular (ICV) injection; the heat treatment demonstrated that rhIL-6 was bioactive as opposed to antigenic (13). An earlier rat study also demonstrated that rhIL-6 was pyrogenic when administered by ICV injection, but not when it was administered by either IV, or IP injection (19). In Guinea pigs, rhIL-6 was not pyrogenic when administered IV at a dose of 1 μg/kg, but caused a dose-independent febrile response between 5–20 μg/kg, and a dose-dependent febrile response at 50 and 100 μg/kg (12). Lastly, low dose infusion of rhIL-6 (140 pg/mL) administered to humans did not elicit significant changes in body temperature despite increases in plasma cortisol concentration (19). We are unaware of any ruminant studies aside from the present study that have evaluated the pyrogenic activity of rhIL-6.

A number of studies have demonstrated that IL-6 is critical signal for HPAA activation. Studies using IL-6 and IL-1 deficient mice, and neutralizing IL-6 antiserum in normal C57BL/6 mice for example, demonstrated that IL-6 is a circulating afferent signal to the HPAA during inflammation (5). Earlier studies by Perlstein et al (20) demonstrated that rh-IL-6 alone, and synergistically with rhIL-1 induced circulating ACTH in mice. Lastly, a human study demonstrated that infusion with rhIL-6 induces HPAA activation (19). These studies combined with the present study suggest that IL-6 plays a significant role in the activation of the HPAA response across various species during systemic inflammation.

Temporal differences in HPAA response were observed between the rhIL-6 and LPS challenges in the present study, and this may, in part, explain why IL-6 concentrations were not previously associated with the ovine febrile and HPAA response to LPS in other ovine studies (9–11). These differences may be attributed to potential interaction with other cytokines, such as IL-1, IL-1ra, and IL-10, and the timing of their secretion. Turnbull et al (5) for example, demonstrated that activation of the HPAA by IL-6 was dependent on prior activation of an IL-1 type I receptor. Additionally, Perlstein et al (20) demonstrated that HPAA activation was dependent on synergistic IL-1 and IL-6 signaling. Lastly, Steensberg et al (19) showed that rhIL-6 induced the secretion of circulating IL-1ra and IL-10; both of these anti-inflammatory cytokines are likely to decrease circulating concentrations of IL-1, which may shorten the duration of HPAA activation.

In summary, we have shown that 2 μg/kg of rhIL-6 is not pyrogenic in sheep, but transiently activated the HPAA when compared with LPS. This suggests that IL-6, alone, is not sufficient to activate the febrile response in sheep at the dose used in this study; however, it is an important afferent signal to the HPAA in sheep during systemic inflammation.