Original Paper Neuroendocrinology 1992;55:97-104
William S. Middleton Memorial Veterans Hospital; Department of Medicine, University of Wisconsin; Department of Comparative Biosciences, University of Wisconsin, Madison, Wise. USA
Pulsatile ACTH and Cortisol in Goats: Effects of Insulin-Induced Hypoglycemia and Dexamethasone
Key Words
Abstract
ACTH, pulsatile Ultradian rhythm Hypoglycemia, goat
Insulin-induced hypoglycemia is a metabolic stress that stimulates secretion of adrenocorti cotropic hormone (ACTH) and cortisol in a number of animal species. Dexamethasone is a potent synthetic glucocorticoid that suppresses the secretion of ACTH and cortisol. Both ACTH and cortisol exhibit complex secretory patterns demonstrating ultradian and circa dian rhythms. This work investigated the pattern of ACTH and cortisol response to hypogly cemia in goats and the effect of dexamethasone on this response. Five goats were pretreated with dexamethasone (0.1 mg/kg) and 5 with saline. Blood samples were taken every 2 min for 60 min before and 60 min after administration of insulin (2.5 lU/kg, i.v.). Immunoreactive ACTH and cortisol were measured in all samples and glucose in selected samples. Data sets were analyzed for significant pulses with the Cluster Analysis program. Complete data sets were compared as well as those for each 30-min interval. Plasma glucose was lower than preinsulin levels at 10 min, declined rapidly between 10 and 30 min, and remained low 30-60 min after insulin injection in both treatment groups. Controls showed a rapid rise in ACTH and cortisol beginning 30 ± 10 min postinsulin. The increase in mean plasma hor mone levels during hypoglycemia was predominantly due to an increase in amplitude of secretory pulses for ACTH and cortisol compared with the 30 min before insulin. Dexa methasone significantly lowered mean ACTH and cortisol levels and prevented alteration in plasma ACTH and cortisol secretion during hypoglycemia but did not totally ablate pul satile activity of either hormone. The amplitude of ACTH and cortisol pulses was signifi cantly decreased by dexamethasone treatment. The frequency of cortisol but not ACTH pulses was also significantly decresed. The highest cross-correlation between plasma ACTH and cortisol occurred at a lag of 0 min in control goats. Cross-correlation was lower and no consistent lag was seen in dexamethasone-treated goats. In control goats, during the fall in plasma glucose, before the rapid rise in plasma ACTH and cortisol, secretion appeared to be relatively quiescent compared to the prior 30 min. Specifically, a slight reduction oc curred in frequency, amplitude, and area of ACTH pulses, in amplitude and area of cortisol pulses, and in cortisol levels. While this unexpected observation may have been an artifact of the sampling protocol, it bears further investigation.
Received: March I, 1991 Accepted after revision: June 25, 1991
Molly Carnes, MD Geriatric Section VA Medical Center 2500 Overlook Terrace Madison. WI 53705 (USA)
Downloaded by: King's College London 137.73.144.138 - 10/9/2018 12:29:43 PM
Molly Carnesab Mark Brownfield1 Stephanie J. Lentb Kalen Nicholsc Linda Schuler1'
Materials and Methods Animals Ten female goats (Nubian, LaMancha, or French Alpine Crosses) (weight 38-66 kg) were placed in grouped pens the even ing before the study. Animals were fasted for 18 h before blood sampling. Water was available ad libitum. Six hours before blood sampling began, the animals were given local anesthesia and in dwelling catheters were placed in the external jugular vein, taped in place, and filled with heparinized saline (5 U/ml). Study Design One hour before blood sampling, 5 goats were injected intra muscularly with dexamethasone (0.1 mg/kg) and 5 with an equal volume of saline (0.9% NaCl). Bolus blood samples (I ml) were taken every 2 min for 60 min before and 60 min after administration of insulin (2.5 IU/kg, i.v.). Samples were collected into tubes con taining EDTA and placed immediately on ice. Plasma was sepa rated by cold centrifugation within I h of collection and stored at -70 °C until assayed. Blood samples were obtained between 15.00 and 17.00 h. Assays ACTH. ACTH immunoreactivity was measured in singlet 100-pl aliquots of unextracted plasma with antiserum IgG-ACTH-1 (IgG Corporation, Nashville, Tenn.) as previously described [8, 9]. This antiserum does not crossreact significantly with a-MSH, (3-MSH, [3-lipotropin, ACTH(11-24), ACTH(1-16)-NH2, or (3-endorphin. It cross-reacts 100% with ACTH(l-39). Human synthetic ACTH(139) (Sigma Chemical, St. Louis, Mo.) was the reference standard and human synthetic l25I-labeled ACTH(l-24) (Incstar Corpora tion, Stillwater, Minn.) was used as the radioactive tracer. The sensitivity of the assay was 3.0 pg/ml (B/Bo = 95%). The intraassay
98
coefficient of variation calculated from pooled plasma replicates (n = 20) was 6-10% for mean ACTH concentrations of 20-67 pg/ ml. All samples were run in one assay. Cortisol. Plasma cortisol was measured in 20-pl singlets by ra dioimmunoassay with the Coat-A-Count® RIA kit obtained from Diagnostic Products Corporation (Los Angeles, Calif.) The sensi tivity of the assay was 2.0 ng/ml (B/Bo = 95%). The intraassay co efficient of variation calculated on standard replicates (n = 20) was 58% at 2.0 ng/ml, 3% at 22.5 ng/ml, and 4% at 44.0 ng/ml. All samples were run in the same assay. Glucose. Plasma glucose was assayed 30, 15, and 0 min before and 10, 20, 30 and 60 min after insulin injection in duplicate sam ples with a glucose oxidase enzymatic determination kit from Sigma Diagnostics (St. Louis, Mo.). Analysis. Significant peaks in the time series were identified by the Cluster Analysis program developed by Veldhuis and Johnson [39]. This program searches for significant increases and decreases in a data series by evaluating the behavior of groups of points to test for nadirs and peaks. Peaks are defined as areas bounded by signifi cant increases and decreases. A constant CV was used in the pro gram based on assay replicates plus an arbitrary addition of 2% for potential error at the time of sample collection. Peak and nadir sizes were set at 2; 3 x 3 was used for the t statistic. In order for the pro gram to read the entire data series, two low and two high values which were ignored in calculation of means and pulse parameters were entered into the program at the end of each data set. Using the definition of Veldhuis et al. [37] for a false-positive peak, with these parameters, the false-positive rates on 231 plasma replicates at a mean ACTH level of 30 pg/ml was 1.7%. The data sets were divided into 30-min time periods for analysis as follows: Tl =60-32 min before insulin = adjustment to experi mental conditions; T2 = 30-2 min before insulin = unstimulated; T3 = 2-30 min after insulin = rapid fall in plasma glucose; T4 = 32-60 min after insulin = sustained hypoglycemia. Individual animal means were calculated for the complete time series as well as each 30-min segment for plasma ACTH and cortisol levels, number of pulses identified, pulse amplitude, and pulse area. Due to the heterogeneity of variance, nonparametric statistics were used for between and within group comparisons. Friedman’s test was used to compare hormone levels and characteristics of identi fied peaks at 30-min intervals within animals. The multiple compar isons procedure suggested for use with the Friedman test [21] was used in post hoc pairwise comparisons. If a pulse was split between time periods, it was considered to fall in the time period containing the identified peak, in the analysis of pulse parameters, if no signifi cant pulse was identified, the amplitude and area of the largest pulse not identified as significant by the Cluster Analysis program was used for calculation. The Mann-Whitney test was used for twosample comparisons between treatment groups. Cross-correlation analysis was performed between plasma ACTH and cortisol con centrations at time lags o f-34 to + 34 min. Autoregressive model ing was used to remove any spurious cross-correlations introduced by autocorrelations within each ACTH and cortisol time series. All analyses were two-tailed and a p value less than 0.05 was considered significant.
C arnes/B row nfield/Lent/N ichols/Schuler
Pulsatile ACTH and Cortisol
Downloaded by: King's College London 137.73.144.138 - 10/9/2018 12:29:43 PM
Insulin-induced hypoglycemia is a metabolic stress that stimulates secretion of adrenocorticotropic hormone (ACTH) and cortisol in humans and a number of other animal species [4, 16, 22, 25, 27, 31, 32, 34, 38, 41]. Dexamethasone is a potent synthetic glucocorticoid that suppresses the secretion of ACTH and cortisol [28]. Both ACTH and cortisol are secreted in episodic bursts that exhibit high- and low-frequency components [7-11, 18, 23]. Sampling intervals greater than 2 min are unable to resolve the higher-frequency secretory bursts and may re sult in lost information or a distorted signal [6, 23]. The physiological significance of these secretory bursts and their contribution to the integration and plasticity of a complex stress-responsive system have not been studied. The intent of this work was to observe in detail the change in the pattern of ACTH and cortisol secretion during a particular stress, hypoglycemia, and to observe the effect of dexamethasone on this pattern. Through these initial observations we hope to better understand the physiolog ical significance of ACTH and cortisol secretory bursts.
Results In both dexamethasone and saline treatment groups, plasma glucose was lower than preinsulin levels at 10 min, had its most rapid decline from 10 to 30 min, and continued to fall 30-60 min after insulin injection in most animals (fig. 1, 2). Glucose levels significantly decreased over time in both treatment groups (p
William S. Middleton Memorial Veterans Hospital; Department of Medicine, University of Wisconsin; Department of Comparative Biosciences, University of Wisconsin, Madison, Wise. USA
Pulsatile ACTH and Cortisol in Goats: Effects of Insulin-Induced Hypoglycemia and Dexamethasone
Key Words
Abstract
ACTH, pulsatile Ultradian rhythm Hypoglycemia, goat
Insulin-induced hypoglycemia is a metabolic stress that stimulates secretion of adrenocorti cotropic hormone (ACTH) and cortisol in a number of animal species. Dexamethasone is a potent synthetic glucocorticoid that suppresses the secretion of ACTH and cortisol. Both ACTH and cortisol exhibit complex secretory patterns demonstrating ultradian and circa dian rhythms. This work investigated the pattern of ACTH and cortisol response to hypogly cemia in goats and the effect of dexamethasone on this response. Five goats were pretreated with dexamethasone (0.1 mg/kg) and 5 with saline. Blood samples were taken every 2 min for 60 min before and 60 min after administration of insulin (2.5 lU/kg, i.v.). Immunoreactive ACTH and cortisol were measured in all samples and glucose in selected samples. Data sets were analyzed for significant pulses with the Cluster Analysis program. Complete data sets were compared as well as those for each 30-min interval. Plasma glucose was lower than preinsulin levels at 10 min, declined rapidly between 10 and 30 min, and remained low 30-60 min after insulin injection in both treatment groups. Controls showed a rapid rise in ACTH and cortisol beginning 30 ± 10 min postinsulin. The increase in mean plasma hor mone levels during hypoglycemia was predominantly due to an increase in amplitude of secretory pulses for ACTH and cortisol compared with the 30 min before insulin. Dexa methasone significantly lowered mean ACTH and cortisol levels and prevented alteration in plasma ACTH and cortisol secretion during hypoglycemia but did not totally ablate pul satile activity of either hormone. The amplitude of ACTH and cortisol pulses was signifi cantly decreased by dexamethasone treatment. The frequency of cortisol but not ACTH pulses was also significantly decresed. The highest cross-correlation between plasma ACTH and cortisol occurred at a lag of 0 min in control goats. Cross-correlation was lower and no consistent lag was seen in dexamethasone-treated goats. In control goats, during the fall in plasma glucose, before the rapid rise in plasma ACTH and cortisol, secretion appeared to be relatively quiescent compared to the prior 30 min. Specifically, a slight reduction oc curred in frequency, amplitude, and area of ACTH pulses, in amplitude and area of cortisol pulses, and in cortisol levels. While this unexpected observation may have been an artifact of the sampling protocol, it bears further investigation.
Received: March I, 1991 Accepted after revision: June 25, 1991
Molly Carnes, MD Geriatric Section VA Medical Center 2500 Overlook Terrace Madison. WI 53705 (USA)
Downloaded by: King's College London 137.73.144.138 - 10/9/2018 12:29:43 PM
Molly Carnesab Mark Brownfield1 Stephanie J. Lentb Kalen Nicholsc Linda Schuler1'
Materials and Methods Animals Ten female goats (Nubian, LaMancha, or French Alpine Crosses) (weight 38-66 kg) were placed in grouped pens the even ing before the study. Animals were fasted for 18 h before blood sampling. Water was available ad libitum. Six hours before blood sampling began, the animals were given local anesthesia and in dwelling catheters were placed in the external jugular vein, taped in place, and filled with heparinized saline (5 U/ml). Study Design One hour before blood sampling, 5 goats were injected intra muscularly with dexamethasone (0.1 mg/kg) and 5 with an equal volume of saline (0.9% NaCl). Bolus blood samples (I ml) were taken every 2 min for 60 min before and 60 min after administration of insulin (2.5 IU/kg, i.v.). Samples were collected into tubes con taining EDTA and placed immediately on ice. Plasma was sepa rated by cold centrifugation within I h of collection and stored at -70 °C until assayed. Blood samples were obtained between 15.00 and 17.00 h. Assays ACTH. ACTH immunoreactivity was measured in singlet 100-pl aliquots of unextracted plasma with antiserum IgG-ACTH-1 (IgG Corporation, Nashville, Tenn.) as previously described [8, 9]. This antiserum does not crossreact significantly with a-MSH, (3-MSH, [3-lipotropin, ACTH(11-24), ACTH(1-16)-NH2, or (3-endorphin. It cross-reacts 100% with ACTH(l-39). Human synthetic ACTH(139) (Sigma Chemical, St. Louis, Mo.) was the reference standard and human synthetic l25I-labeled ACTH(l-24) (Incstar Corpora tion, Stillwater, Minn.) was used as the radioactive tracer. The sensitivity of the assay was 3.0 pg/ml (B/Bo = 95%). The intraassay
98
coefficient of variation calculated from pooled plasma replicates (n = 20) was 6-10% for mean ACTH concentrations of 20-67 pg/ ml. All samples were run in one assay. Cortisol. Plasma cortisol was measured in 20-pl singlets by ra dioimmunoassay with the Coat-A-Count® RIA kit obtained from Diagnostic Products Corporation (Los Angeles, Calif.) The sensi tivity of the assay was 2.0 ng/ml (B/Bo = 95%). The intraassay co efficient of variation calculated on standard replicates (n = 20) was 58% at 2.0 ng/ml, 3% at 22.5 ng/ml, and 4% at 44.0 ng/ml. All samples were run in the same assay. Glucose. Plasma glucose was assayed 30, 15, and 0 min before and 10, 20, 30 and 60 min after insulin injection in duplicate sam ples with a glucose oxidase enzymatic determination kit from Sigma Diagnostics (St. Louis, Mo.). Analysis. Significant peaks in the time series were identified by the Cluster Analysis program developed by Veldhuis and Johnson [39]. This program searches for significant increases and decreases in a data series by evaluating the behavior of groups of points to test for nadirs and peaks. Peaks are defined as areas bounded by signifi cant increases and decreases. A constant CV was used in the pro gram based on assay replicates plus an arbitrary addition of 2% for potential error at the time of sample collection. Peak and nadir sizes were set at 2; 3 x 3 was used for the t statistic. In order for the pro gram to read the entire data series, two low and two high values which were ignored in calculation of means and pulse parameters were entered into the program at the end of each data set. Using the definition of Veldhuis et al. [37] for a false-positive peak, with these parameters, the false-positive rates on 231 plasma replicates at a mean ACTH level of 30 pg/ml was 1.7%. The data sets were divided into 30-min time periods for analysis as follows: Tl =60-32 min before insulin = adjustment to experi mental conditions; T2 = 30-2 min before insulin = unstimulated; T3 = 2-30 min after insulin = rapid fall in plasma glucose; T4 = 32-60 min after insulin = sustained hypoglycemia. Individual animal means were calculated for the complete time series as well as each 30-min segment for plasma ACTH and cortisol levels, number of pulses identified, pulse amplitude, and pulse area. Due to the heterogeneity of variance, nonparametric statistics were used for between and within group comparisons. Friedman’s test was used to compare hormone levels and characteristics of identi fied peaks at 30-min intervals within animals. The multiple compar isons procedure suggested for use with the Friedman test [21] was used in post hoc pairwise comparisons. If a pulse was split between time periods, it was considered to fall in the time period containing the identified peak, in the analysis of pulse parameters, if no signifi cant pulse was identified, the amplitude and area of the largest pulse not identified as significant by the Cluster Analysis program was used for calculation. The Mann-Whitney test was used for twosample comparisons between treatment groups. Cross-correlation analysis was performed between plasma ACTH and cortisol con centrations at time lags o f-34 to + 34 min. Autoregressive model ing was used to remove any spurious cross-correlations introduced by autocorrelations within each ACTH and cortisol time series. All analyses were two-tailed and a p value less than 0.05 was considered significant.
C arnes/B row nfield/Lent/N ichols/Schuler
Pulsatile ACTH and Cortisol
Downloaded by: King's College London 137.73.144.138 - 10/9/2018 12:29:43 PM
Insulin-induced hypoglycemia is a metabolic stress that stimulates secretion of adrenocorticotropic hormone (ACTH) and cortisol in humans and a number of other animal species [4, 16, 22, 25, 27, 31, 32, 34, 38, 41]. Dexamethasone is a potent synthetic glucocorticoid that suppresses the secretion of ACTH and cortisol [28]. Both ACTH and cortisol are secreted in episodic bursts that exhibit high- and low-frequency components [7-11, 18, 23]. Sampling intervals greater than 2 min are unable to resolve the higher-frequency secretory bursts and may re sult in lost information or a distorted signal [6, 23]. The physiological significance of these secretory bursts and their contribution to the integration and plasticity of a complex stress-responsive system have not been studied. The intent of this work was to observe in detail the change in the pattern of ACTH and cortisol secretion during a particular stress, hypoglycemia, and to observe the effect of dexamethasone on this pattern. Through these initial observations we hope to better understand the physiolog ical significance of ACTH and cortisol secretory bursts.
Results In both dexamethasone and saline treatment groups, plasma glucose was lower than preinsulin levels at 10 min, had its most rapid decline from 10 to 30 min, and continued to fall 30-60 min after insulin injection in most animals (fig. 1, 2). Glucose levels significantly decreased over time in both treatment groups (p
