Temporal relationships between plasma cortisol, corticosteroid-binding globulin (CBG), and the free cortisol index (FCI) in pigs in response to adrenal stimulation or suppression

Abstract

The objective of this study was to document changes in plasma concentrations of total cortisol, porcine corticosteroid-binding globulin (pCBG), and the free cortisol index (FCI) in pigs over a 6-h period in response to adrenal stimulation or suppression. Twenty-four 8-week old pigs allotted in equal numbers were administered ACTH, dexamethasone or saline, and blood samples were collected every 15 min via an indwelling jugular catheter for 1 h prior to and 5 h following treatment. Total plasma cortisol increased in ACTH-treated pigs and decreased in dexamethasone-treated pigs within 0.25 and 0.5 h, respectively. In contrast, pCBG concentration was altered in an inverse fashion subsequent to the changes exhibited in total cortisol. FCI reflected the changes observed in total cortisol. These results further document the negative relationship that exists between circulating concentrations of plasma cortisol and pCBG, and illustrate that this association exists under conditions of acute stress in the pig.

• ACTH
• CBG
• cortisol
• dexamethasone
• FCI
• pig

Introduction

The active form of cortisol in the circulation is that which is unbound and that which is loosely bound to albumin, thus biologically available to the cell (Siiteri et al. 1982). The majority of circulating cortisol in humans (Siiteri et al. 1982) and swine (Kattesh et al. 1990) is bound to its specific carrier glycoprotein, corticosteroid-binding globulin (CBG), which both transports and modulates cortisol availability in the circulation (Siiteri et al. 1982). We have reported finding significant changes in circulating CBG concentrations, examined over days or weeks, in pregnant sows (Kattesh et al. 1980), pseudopregnant gilts following cortisol administration (Behrens et al. 1993), and pigs following exposure to heat and social stress (Heo et al. 2005). We recently reported that CBG in young pigs follows a diurnal pattern over a 24 h period (Adcock et al. 2006). Acute changes in circulating CBG in relation to cortisol change have not been examined in swine.

Interpretation of plasma total cortisol concentrations, in situations where CBG changes significantly, does not address the free cortisol fraction and thus may not provide a true representation of its biological impact (le Roux et al. 2003). The free cortisol index (FCI), a ratio of the concentration of plasma cortisol to CBG, has been shown to correlate well with plasma free cortisol in both humans (le Roux et al. 2003) and swine (Adcock et al. 2006). The objective of this study was to document the changes in plasma total cortisol, pCBG and FCI in pigs sampled frequently over a 6-h period in response to adrenocortical stimulation or suppression.

Materials and methods

Animals and diets

All animal procedures were reviewed and approved by the University of Tennessee Animal Care and Use Committee, which were in accordance with the applicable portions of the Animal Welfare Act and the “Guide for the Care and Use of Agricultural Animals in Agricultural Research and Teaching” by FASS. Twenty-four 8-week old female pigs (Premier x QMax 100) selected from eight litters were assigned to two replicate groups (four litters/group). Replicate groups of 12 pigs, separated by 1 week of age, were used due to the frequency of blood sampling and space limitations. One pig from each litter was randomly selected to receive one of the three treatments. Pigs were weaned at 25 days of age and housed in elevated pens (2.4 × 2.4 m) with slotted floors. Animals were given ad libitum access to a commercial diet (3.315 kcal/kg of ME, 18% CP, 1.15% lysine; as fed basis) and water. Artificial lights were provided for 13 h, starting at 06:30 h, and a red light source was activated during the dark period to aid in blood sampling.

Treatments, plasma collection and analyses

Three days prior to initiation of the experiment, pigs from group one (15.7 ± 2.4 kg) and group two (22.2 ± 2.8 kg) were placed in individual 0.61 × 1.22 m pens. On the morning of the experiment, pigs were initially restrained while halothane inhalant (5%; Halocarbon Laboratories; River Edge, NJ, USA) was administered with oxygen (3 l/min) via a modified face mask attached to a closed circuit anesthesia machine, and non-surgically fitted with an indwelling jugular vein catheter (Carroll et al. 1998). The pigs were returned to their individual pens and assigned to one of three treatments: ACTH (11.1 μg/kg BW, 111 μg/ml in 0.9% NaCl, Sigma A6303), dexamethasone (16 μg/kg BW, 100 μg/ml in 0.9% NaCl and 5% ethanol, Sigma D1756) or saline (2.5 ml 0.9% NaCl). All treatments were administered via the indwelling catheter.

Blood sampling began at 14:30 h, a minimum of 4 h after the catheter was inserted. Piglets are reported to recover in less than 4 h from this catheterization procedure and display no clinical signs of pain or discomfort (Carroll et al. 1998). Blood samples were collected over a 6-h period every 15 min, beginning at − 1 h. Immediately prior to collection of each sample, 2 ml of blood was drawn from the catheter and discarded. A blood sample (5 ml) was collected in a heparinized syringe (Li-Heparin LH/4.5 ml, Sarstedt Monovette, Newton, NC, USA), and the catheter was flushed with 2.5 ml of 0.9% NaCl followed by 0.5 ml of heparinized saline (20 IU/ml in 0.9% NaCl). Samples were immediately centrifuged at 1520g for 15 min, the plasma was removed and stored at − 20°C.

Plasma total cortisol concentration was determined by radioimmunoassay as reported previously (Adcock et al. 2006). Cross-reactivity as listed by the manufacturer was less than 0.10% for dexamethasone. Intra- and inter-assay CV was 3.6 and 13.5% for low (110 nmol/l), 4.4 and 15.7% for medium (325 nmol/l), and 3.3 and 12.2% for high (772 nmol/l) cortisol standards. The concentration of porcine corticosteroid-binding globulin (pCBG) was measured by a direct ELISA (Roberts et al. 2003). Intra- and inter-assay CV of a pooled pig plasma sample was 18 and 20%, respectively. The FCI was calculated using the ratio of plasma total cortisol to pCBG concentration.

Statistical analysis

Data were analyzed using the MIXED procedure of SAS (SAS Inc., v9.0 Cary, NC, USA) for a randomized block design. Litter and sow parity were tested and found to be insignificant, so were removed from the model to increase power. The fixed effects of treatment (ACTH, dexamethasone and saline), time, and treatment × time interaction were used to analyze for differences in plasma cortisol, pCBG and FCI. Random effect of pig within treatment was used to test treatment differences, and repeated measures over time were accounted for in the model. Samples taken prior to treatment administration were averaged to provide a single baseline value for each pig. Data were represented as least squares means with standard errors, and significant differences were separated using Fisher's least significant difference test. Pearson correlation coefficients were used to determine relationships between cortisol and pCBG within individual treatments.

Results

One pig assigned to the saline treatment in the first replication died due to an adverse reaction to the anesthesia. During the second replication, the catheter of a saline treated pig lost patency 1 h after sampling began and was removed from the study. Preliminary analysis indicated there was no effect due to replicate for any of the variables measured; hence, data from both replicates were combined.

Mean plasma total cortisol concentrations were different (p < 0.0001) due to treatment, time and the interaction of treatment × time (Figure 1). Prior to treatment administration, cortisol concentrations were similar (p>0.14) between the saline (77.3 ± 11.0 nmol/l), ACTH (77.6 ± 9.5 nmol/l), and dexamethasone (93.5 ± 9.5 nmol/l) treated animals. Plasma total cortisol concentration in saline-treated pigs decreased (p < 0.001) from baseline values beginning at 0.5 h (52.7 ± 12.5 nmol/l) and remained low for the remainder of the experiment. Pigs administered ACTH exhibited a greater (p < 0.001) cortisol concentration compared to baseline by 0.25 h (213.8 ± 10.9 nmol/l), which continued until 2.75 h (101.67 ± 10.9 nmol/l). At 3 h following ACTH injection, plasma total cortisol values returned to baseline and were lower (p < 0.001) than baseline by 3.75 h (45.7 ± 10.9 nmol/l) where they persisted for the remainder of the experiment. In the dexamethasone treated group, plasma cortisol values were lower (p < 0.001) than baseline at 0.5 h (54.5 ± 10.9 nmol/l) and remained suppressed for the duration of the experiment.

Figure 1 Changes in plasma cortisol and pCBG concentrations, and FCI in pigs following intravenous administration of adrenocorticotrophin (ACTH; 11.1 μg/kg BW), saline (Saline; 0.9% NaCl), or dexamethasone (Dex; 16 μg/kg BW). Treatment was administered immediately following the 0 h sample. Each point represents the mean value ( ± SEM, n = 7–8 per group). See text for statistical analysis.

Mean plasma pCBG concentrations were similar (p>0.10) among treatment groups prior to administration of saline (9.9 ± 1.6 mg/l), ACTH (13.0 ± 1.3 mg/l) or dexamethasone (9.3 ± 1.3 mg/l) (Figure 1). Overall, neither treatment (p = 0.67) nor time (p = 0.10) differences were observed for pCBG, however, there was a strong interaction (p < 0.0001) indicating different treatment responses across time. This interaction was caused by the overall pattern across time of increasing pCBG concentration for dexamethasone treatment, no changes for saline treated pigs, and a decrease, then increase for ACTH treated pigs. The pCBG concentrations for saline-treated pigs remained similar to baseline values throughout the course of the experiment. For ACTH-treated pigs, pCBG concentrations were lower (p < 0.001) than the mean baseline value from 1.75 to 3.25 h post-treatment, and were similar to the baseline value for the remainder of the experiment. The concentration of pCBG in dexamethasone treated pigs increased (p < 0.05) within 1 h post-injection (13.6 ± 1.5 mg/l) compared to baseline values and returned to baseline by 4.75 h. At the end of the experiment (5 h), pCBG concentration for ACTH treated pigs was greater (p < 0.05) than that of the dexamethasone or saline treatment groups.

The calculated FCI revealed (p < 0.0001) differences for the main effects of treatment and time and the interaction between treatment and time (Figure 1). Dexamethasone treated pigs had a higher (p < 0.05) pre-treatment FCI baseline value (10.9 ± 1.1 nmol/mg) compared to the saline and ACTH treatment groups (9.5 and 6.7 ± 1.1 nmol/mg, respectively). FCI decreased (p < 0.001) in saline-treated pigs at 0.5 h (6.0 ± 1.8 nmol/mg) and remained low for the remainder of the experiment. Following administration of ACTH, FCI increased (p < 0.001) at 0.25 h to 20.8 ± 1.9 nmol/mg and remained elevated until 2.75 h. FCI returned to baseline by 3 h post-ACTH injection and was lower (p < 0.001) than the baseline value from 3.75 to 5 h. In pigs administered dexamethasone, FCI increased (p < 0.001) to 14.5 ± 1.6 nmol/mg within 15 min following treatment administration. However, by 0.5 h the FCI decreased (p < 0.001) to 5.5 ± 1.6 nmol/mg and remained lower (p < 0.001) than baseline, and the saline or ACTH treatment group values at similar sampling times, for the remainder of the experiment.

No relationship was found between plasma total cortisol and pCBG concentrations for any of the groups sampled prior to treatment administration. Weak inverse relationships (p < 0.05) were observed between plasma cortisol and pCBG for the ACTH (r = − 0.19), saline (r = − 0.36) and dexamethasone (r = − 0.19) treated pigs for all samples collected following treatment administration. The negative correlations are evident from inspection of Figure 1. The ACTH treatment data exhibit elevated total cortisol concentration for 2 h followed by an equivalent decline, whereas pCBG concentration followed the opposite pattern, producing a negative correlation. The dexamethasone and saline-treated groups exhibited declining total cortisol concentrations while pCBG increased, again producing a negative correlation.

Discussion

The dose of ACTH administered (11.1 μg/kg body weight) was sufficient to generate a 3–4-fold increase in plasma cortisol concentration in our pigs peaking within 1 h, similar to that reported previously in swine (Haussmann et al. 2000). This rapid elevation of plasma cortisol in pigs following ACTH injection, and its return to pre-injection concentrations within 3 h, was similar to that reported in our previous study (Adcock et al. 2006). The observed decrease in total cortisol concentration over the sampling period following the administration of saline, and eventually for the ACTH-treated animals as well, may be a reflection of the normal diurnal changes in circulating cortisol for swine, where cortisol concentrations decrease over the afternoon hours during the time we sampled in the present study (Adcock et al. 2006). The extended suppression of plasma cortisol concentration beginning within an hour following dexamethasone administration was expected. Dexamethasone has been shown to have an inhibitory effect on the hypothalamic–pituitary–adrenal (HPA) axis via glucocorticoid receptors within the anterior pituitary gland (Cole et al. 2000). However, the cortisol spike immediately following the administration of dexamethasone was unexpected. This transient increase in cortisol might be a consequence of the alcohol used to dissolve the dexamethasone, since plasma cortisol concentration has been reported to increase within 30 min of an intravenous infusion of ethanol in men (Jenkins and Connolly 1968).

We have shown previously that circulating pCBG is reduced in swine following chronic exposure to environmental or management related stress (Kattesh et al. 1980, Heo et al. 2005) or administration of hydrocortisone acetate (Behrens et al. 1993). However, due to infrequent daily or weekly blood sampling in these earlier studies, a temporal relationship of the acute pCBG response to elevated cortisol concentrations could not be defined. In the present study, the weak inverse relationship exhibited between total cortisol and pCBG for the saline-treated group may represent the normal diurnal pattern of change within the circulation for these two variables. Daytime CBG variation related to surges in total plasma cortisol has been implicated as having a biological role for daytime cortisol variation (Lewis et al. 2006). Likewise, a brief but significant reduction in plasma pCBG in the ACTH-treated pigs was seen 1.75 h following the increase in circulating cortisol. This confirms our earlier observation of a delayed effect of elevated cortisol on circulating pCBG concentrations in young pigs sampled at 30 min intervals for 4 h following ACTH administration (Adcock et al. 2006). Rats subjected to inescapable tail shock exhibit an immediate elevation in corticosterone followed by a reduction in CBG approximately 6 h later (Fleshner et al. 1995). Similarly, CBG capacity has been shown to decline within 30–60 min of capture in five species of birds (Breuner et al. 2006). It has not been demonstrated directly whether stress-related changes in plasma CBG concentration result from changes in production, metabolic degradation and/or transfer to target tissues (Kanitz et al. 2006). A reduction in CBG in response to elevated total cortisol concentrations would effectively increase the clearance rate and thus lower total circulating cortisol concentrations (Seralini et al. 1990). Alternatively, the decrease in CBG could serve to increase the concentration of cortisol reaching tissues during acute stress (Breuner et al. 2006).

As a glucocorticoid agonist, dexamethasone has less than 0.1% binding affinity to CBG compared to cortisol and therefore, should not interfere with cortisol binding to CBG (De Kloet et al. 1975). As a result of frequent sampling, we were able to document a significant increase in pCBG occurring 30 min after the immediate suppression of plasma cortisol in the dexamethasone-treated pigs. The increase in pCBG is most likely a consequence of the effect of dexamethasone in reducing circulating cortisol levels. In that CBG contributes to determining plasma cortisol concentrations relative to its production (Bright and Darmaun 1995), a decrease in circulating cortisol would prompt an increase in CBG concentrations to maintain physiological concentrations of the glucocorticoid. Indeed, adrenalectomy has been found to enhance, and glucocorticoid administration to inhibit, the rate of CBG production and secretion from isolated liver in rats (Feldman et al. 1979). Dexamethasone treatment has been reported to increase plasma CBG concentrations within the ovine fetus (Berdusco et al. 1995) and in the liver and kidney of the immature mouse (Zhao et al. 1997).

The significance of the observed plasma cortisol–CBG relationship reported here and in previous studies may reside in the calculation of the FCI. The FCI increased following the administration of ACTH as a result of an increase in total cortisol and a decrease in pCBG concentrations. In the dexamethasone treated group, the FCI decreased as a result of a suppression of total cortisol and an increase in pCBG. Arguably, the changes in pCBG concentration over time appear trivial, especially for the ACTH and dexamethasone treated pigs, and hence measuring plasma total cortisol concentration alone seems to almost completely predict FCI. However, since approximately 60% of the total cortisol in the circulation is bound to pCBG in the pig (Kattesh et al. 1990), a change in pCBG could have a significant effect on the amount of biologically available cortisol. As an example, our earlier study in heat-stressed pigs found that the FCI did not change because there was a simultaneous decrease in plasma total cortisol and pCBG concentrations (Heo et al. 2005). The FCI may provide a better illustration of temporal changes in the adrenal response in animals than by measuring total plasma cortisol alone.

Acknowledgements

This study is published with the approval of the Dean of the Tennessee Agricultural Experiment Station and was supported in part by State and Hatch Funds allocated to the station.

Mention of trade names or proprietary products does not constitute a guarantee or warranty of the product by the USDA and does not imply its approval to the exclusion of other products that may also be suitable.