Insulin Alone Increases Hypothalamo-Pituitary-Adrenal Activity, and Diabetes Lowers Peak Stress Responses
http://www.100md.com
内分泌学杂志 2005年第3期
Departments of Physiology (O.C., K.I., E.A., M.V., S.G.M.), Obstetrics and Gynecology (S.G.M.), and Medicine (M.V., S.G.M.), University of Toronto, Toronto Ontario, Canada M5S 1A8; and Department of Kinesiology and Health Science, York University (E.P., M.C.R.), Toronto, Ontario, Canada M3J 1P3
Address all correspondence and requests for reprints to: Dr. Stephen G. Matthews, Medical Sciences Building, Room 3240, 1 King’s College Circle, University of Toronto, Toronto, Ontario, Canada M5S 1A8. E-mail: stephen.matthews@utoronto.ca.
Abstract
Diabetes is associated with increased basal hypothalamo-pituitary-adrenal (HPA) activity and impaired stress responsiveness. Previously, we demonstrated that the HPA response to hypoglycemia is significantly impaired in diabetic rats. In this study our goals were to 1) differentiate between the effects of hyperinsulinemia and those of hypoglycemia per se, and 2) establish whether diabetes lowers peak stress responses. Normal and streptozotocin-diabetic rats were subjected to hyperinsulinemic-euglycemic glucose clamps to evaluate central and peripheral responses. These were compared with peak ACTH and corticosterone responses to restraint and hypoglycemia. Hyperinsulinemia increased CRH and vasopressin mRNA, and plasma ACTH and corticosterone in normal and diabetic rats. In normal animals, insulin-induced activation of ACTH and corticosterone was lower than the responses during either restraint or hypoglycemia. In contrast, ACTH and corticosterone activation in diabetic rats was similar with all three stressors. Pituitary-adrenal axis activation in diabetic animals was also much lower compared with that in normal controls. The response to hyperinsulinemia (euglycemia) was associated with increases in glucocorticoid receptor mRNA in the anterior pituitary and paraventricular nucleus. Hippocampal mineralocorticoid receptor mRNA expression was increased in normal, but not in diabetic, animals. We speculate that the ability to appropriately match the HPA response to the potency of a stressor is related to the ability to alter hippocampal mineralocorticoid receptor expression. In diabetes, this ability is impaired; hence, maximal HPA activation is greatly diminished. This is a novel observation that may have important implications in the treatment of impaired counterregulatory mechanisms in human diabetes.
Introduction
DIABETES IS A metabolic disorder that is associated with dysregulation of a number of systems within the body, including cardiovascular disorders and altered immune and central nervous system functions. One of these central nervous system dysfunctions includes activation of the hypothalamo-pituitary-adrenocortical (HPA) axis in uncontrolled or poorly controlled diabetes (1, 2, 3, 4, 5). We have previously demonstrated that basal HPA function is up-regulated in streptozotocin (STZ)-diabetic rats and that this is associated with increases in hypothalamic CRH mRNA and plasma ACTH and corticosterone levels (3). Interestingly, elevated basal HPA function in diabetes is also associated with increased hippocampal mineralocorticoid receptor (MR) mRNA expression, a major source of inhibitory input to the HPA axis. This suggests that increased central drive at or above the level of the hypothalamus is responsible for HPA hyperactivation in diabetes. More importantly, our laboratory also demonstrated that activation of the HPA axis in response to various forms of stress and, more specifically, in response to 20 min of restraint and to insulin induced-hypoglycemia was greatly diminished in diabetic animals (6, 7). Defects in stress responsiveness in diabetes appear to be the result of a combination of decreased tissue sensitivity to secretagogues as well as an impaired capacity to stimulate the axis that may stem from an inability to decrease hippocampal MR expression and thus to relieve inhibitory tone to the axis. In the hypoglycemia studies it was difficult to distinguish between effects that were the result of hyperinsulinemia and those that were attributable to hypoglycemia alone. We demonstrate in this study that insulin itself can have profound effects on the HPA axis, both centrally and peripherally, that are independent of hypoglycemia. More interestingly, by comparing the peak pituitary-adrenal responses to various activators of HPA function (hyperinsulinemia, restraint, and hypoglycemia), we show that defects in the HPA axis of diabetic animals result in a system that is incapable of responding appropriately to a stress challenge because its peak response plateaus at a significantly lower level than that in normal animals.
The central interactions between insulin and the HPA axis have primarily been examined with regard to energy homeostasis. In general, it is believed that glucocorticoids establish a deleterious environment wherein obesity and insulin resistance can develop. Centrally, insulin primarily acts to suppress food intake and prevent the onset of obesity (8). In a study that examined the central role of insulin in STZ-diabetic rats, it was shown that insulin can also increase the expression of the anorexigenic neuropeptide, CRH, in the hypothalamus (9). More specific interactions of insulin with the HPA axis in the absence of hypoglycemia have not been investigated.
Limited studies have examined the effect of insulin on pituitary-adrenal function in human subjects. Fruehwald-Schultes et al. (10) demonstrated that iv infusion of high doses of insulin to human subjects in the presence of euglycemia increases plasma ACTH and cortisol concentrations. Conversely, when a lower dose of insulin was used, pituitary-adrenal activity remained unchanged (11). These studies suggest that the stimulatory actions of insulin on HPA function may only be effective when insulin concentrations are high.
As mentioned above, our own work has shown that insulin administration has profound effects on the HPA axis of diabetic animals in regulating both basal (3) and activated function (6). Thus, the aims of this study were 1) to determine which aspects of the HPA response to hypoglycemia are the result of hyperinsulinemia and which are due to hypoglycemia per se, and 2) to ascertain whether the capacity to adjust the magnitude of the HPA response to various stressors is altered in diabetes.
Materials and Methods
Experimental animals and design
Male Sprague Dawley rats (Charles River, Québec, Canada), initially weighing 325–375 g, were individually housed in opaque cages in temperature (22 C)- and humidity-controlled rooms. The animals were fed rat chow (Ralston-Purina Co., St. Louis, MO) and water ad libitum and were acclimatized to a 12-h light cycle (lights on between 0700–1900 h) for a period of 1 wk before experimental manipulation. All experiments were approved by the animal care committee of University of Toronto and were conducted in accordance with regulations set by the Canadian Council for Animal Care.
Euglycemic clamp studies
Two groups of rats were used: normal controls (n = 6) and untreated STZ-induced diabetic (n = 6) rats. On d 0, catheters were placed into the left carotid artery and right jugular vein, as described previously (6). Diabetes was induced at the end of surgery via a single injection of STZ (65 mg/kg; Sigma-Aldrich Corp., St. Louis, MO) dissolved in sterile saline, through the penile vein. Control animals received a saline injection under similar conditions. Animals treated with STZ were given 10% sucrose water for the first 24 h after STZ injection to prevent hypoglycemia (12). This is a model of moderate diabetes characterized by fasting hyperglycemia, normal fasting plasma insulin, and impaired plasma insulin responses after marked hyperglycemia (534 ± 2.4 mg/dl) (13). Blood glucose was monitored twice daily with a glucometer (Glucometer Elite 3903, Bayer, Inc., Etobicoke, Canada) in all animals to ensure that fasting normoglycemia was maintained in control animals and that adequate hyperglycemia (>15 mM) was achieved in the uncontrolled diabetic group. The contents of the catheters were aspirated and reprimed with a 60% polyvinylpyrolidone solution (wt/vol; 1000 U/ml heparin) on d 2, 4, and 6 to maintain catheter patency and to acclimatize the rats to being handled.
On d 7, the rats were fasted for 24 h before the start of the experiment. On d 8, the experiments were carried out in unrestrained, conscious animals. The catheters were extended outside the cage to minimize investigator interaction and were connected to infusion pumps. The animals were left undisturbed for 2.5 h before basal hormone samples were collected (1030 and 1100 h), just before the start of the insulin infusion (1100 h). The rats then underwent a hyperinsulinemic-euglycemic glucose clamp. A constant insulin (50 mU/kg·min regular porcine insulin; Eli Lilly & Co., Indianapolis, IN) and variable dextrose (45%; Abbott Laboratories Ltd., Montréal, Canada) infusion through the jugular vein catheter were used to maintain plasma glucose levels at 6.7 ± 0.6 mM for 130 min. Large doses of insulin were used because even in moderately diabetic animals these doses were necessary to induce hypoglycemia (6, 14, 15). The concentration of glucose was determined every 5 min from a sample of plasma (10 μl) using a glucose analyzer II (Beckman Coulter, Palo Alto, CA). Arterial blood samples were obtained from the carotid catheter at regular time intervals throughout the glucose clamp. For ACTH, glucagon, and insulin measurements, blood was collected in chilled tubes containing EDTA (Sangon Ltd. Canada, Scarborough, Canada) and Trasylol (Bayer, Inc.). Blood samples for catecholamine determinations were collected in chilled tubes containing 1.5 mg reduced glutathione (Roche, Mannheim, Germany) and 5 μl EGTA (Sigma-Aldrich Corp.). Serum was collected for corticosterone measurements. Plasma was aliquoted into storage tubes and stored at –20 C (or –80 C for catecholamine determination). After removal of plasma, red blood cells were resuspended in heparinized saline (10 U/ml) and reinfused after each blood sampling to prevent volume depletion and anemia. Hematocrit, determined at the beginning and the end of the experiment, was maintained above 35%. At the end of the experiment, the rats were euthanized by decapitation. Trunk blood samples were collected, and brains and pituitary glands were removed and stored at –80 C.
In situ hybridization
The method of in situ hybridization has been described in detail previously (16). Briefly, coronal cryosections (12 μm) were obtained through selected hypothalamic (bregma –2.00 mm) and hippocampal (bregma –3.80 mm) regions according to the stereotaxic coordinates of Paxinos and Watson (17). The sections were then thaw-mounted onto poly-L-lysine (Sigma-Aldrich Corp.)-coated slides, fixed for 5 min in 4% phosphate-buffered paraformaldehyde, rinsed in PBS (2 min), dehydrated in an ethanol series (70% and 95%), and stored in 95% ethanol at 4 C until use.
The 45-mer antisense CRH (bases 536–580) (18), AVP (bases 588–632) (18), proopiomelanocortin (POMC; bases 572–616) (18), MR (bases 2942–2986) (19), and glucocorticoid receptor (GR; bases 1321–1365) (20) oligonucleotide probes were synthesized by Dalton Chemical Laboratories, Inc. (Toronto, Canada). The probes were labeled using terminal deoxynucleotidyl transferase (Pharmacia Biotech, Baie d’Urfé, Canada) and [35S]deoxy-ATP (1300 Ci/mmol; NEN Life Science Products, DuPont Canada, Mississauga, Canada) to a specific activity of 1.0 x 109 cpm/μg. Labeled probe in hybridization buffer (180 μl) was applied to each slide at a concentration of 1.0 x 106 cpm/μl. Slides were incubated overnight in a moist chamber at 42.5 C. After washing in 1x standard saline citrate (1x SSC; 20 min at room temperature) and 1x SSC (35 min at 55 C), the slides were rinsed twice with 1x SSC and once with 0.1x SSC at room temperature, then dehydrated in 70% and 95% ethanol (1 min each), air-dried, and exposed to autoradiographic film (Biomax, Eastman Kodak Co., Rochester, NY). The films were developed using standard procedures (exposure times: CRH, 21 d; AVP, 2 d; POMC, 2 h; MR, 14 d; GR, 28 d).
Because tissues obtained from the current study were collected after a euglycemic clamp, we used slides containing select hypothalamic, hippocampal, and pituitary tissue sections collected from a previous basal study (3) and simultaneously ran new in situ hybridizations using these slides and those collected from the current study to determine the molecular HPA responses to hyperinsulinemia-euglycemia. The animals from which the basal tissues were collected received identical treatment regimens as described above, except they were euthanized between 1000–1100 h on d 8. By doing this, we were able to compare basal neuropeptide and corticosteroid receptor mRNA expression with that after the euglycemic clamp to ascertain the effects of hyperinsulinemic-euglycemia.
Plasma hormone and catecholamine determination
Plasma insulin was measured using a modified version of the insulin RIA by Herbert et al. (21). Plasma ACTH (Diasorin, Inc., Stillwater, MN), corticosterone (ICN Pharmaceuticals, Inc., Orangeburg, NY), and glucagon (Diagnostic Products Corp., Los Angeles, CA) concentrations were determined using commercially available RIA kits.
Plasma epinephrine and norepinephrine concentrations were determined using the simultaneous single isotope derivative radioenzymatic assay technique described previously (14, 22).
Data analysis
For in situ hybridizations, brain sections were processed simultaneously for each probe to allow direct comparison between treatment groups. Six to eight sections were analyzed from each animal for each probe based on visual inspection of the desired region. The sections were exposed together with 14C-labeled standards (American Radiochemical, St. Louis, MO) to ensure analysis in the linear region of the autoradiographic film. The relative OD (ROD) of the signal on autoradiographic film was quantified, after subtraction of background values, using a computerized image analysis system (Imaging Research, St. Catherines, Canada). Hormone data are presented as the mean ± SEM, and in situ hybridization data are expressed as the ROD (mean ± SEM). Statistical analysis was performed by one- or two-way ANOVA for independent or repeated measures, as appropriate, using Statistica 6.0 statistical software (StatSoft, Tulsa, OK), with P < 0.05 set as the criterion for statistical significance.
Results
Body weight and plasma hormone and catecholamine concentrations
Although body weights did not differ significantly between treatment groups before surgery, 8 d after the induction of diabetes, STZ-diabetic rats exhibited a significant decline in body weight (P < 0.03; normal, 348.0 ± 13.2; STZ, 298.6 ± 14.1 g). Hematocrit levels were maintained above 35% during the clamp (normal, 42.4 ± 2.3 STZ, 40.6 ± 2.3%) to prevent hemodynamic changes resulting from volume depletion and anemia.
Basal fasting plasma glucose concentrations were increased (P < 0.01) by approximately 4-fold in diabetic rats compared with normal animals (normal, 5.4 ± 0.3; STZ, 21.1 ± 1.0 mM). Plasma glucose levels during the euglycemic clamp are shown in Fig. 1. The break in Fig. 1 indicates the period during which we ensured that plasma glucose levels were matched between the two treatment groups before blood sampling was resumed. Although basal plasma glucose levels were higher in diabetic animals, they were maintained at similar levels to normal controls during the euglycemic clamping period from 80–210 min.
FIG. 1. Plasma glucose concentrations achieved in normal (solid line) and diabetic (dotted line) rats during the hyperinsulinemic-euglycemic glucose clamp. Results are presented as the mean ± SEM.
Basal plasma ACTH levels (normal, 64.9 ± 4.5; STZ, 53.6 ± 4.4 pg/ml) were not significantly different between normal and diabetic rats. However, basal plasma corticosterone concentrations were significantly (P < 0.05) elevated in diabetic compared with control animals (normal, 59.2 ± 11.3; STZ, 140.9 ± 42.0 ng/ml). During the hyperinsulinemic-euglycemic clamp, plasma ACTH and corticosterone concentrations rose significantly (P < 0.05) in both normal and diabetic rats and peak levels were similar in the two groups (Fig. 2). In comparison with other activators of HPA function, such as restraint (7) and insulin-induced hypoglycemia (6), we noted that the magnitude of the pituitary-adrenal response to hyperinsulinemia-euglycemia was significantly less in nondiabetic animals (Fig. 3). In contrast, the pituitary-adrenal responses of diabetic animals to all three of these stressors were similar in magnitude (Fig. 3). These effects appear to be specific to the infusion of insulin and are not the result of repeated sampling, because the same sampling protocol using saline infusion instead of insulin did not significantly affect corticosterone levels at any of the sampling points (0, 30, 60, 80, 100, 150, and 210 min: normal, 47.9 ± 9.6, 43.1 ± 4.5, 65.9 ± 1.8, 68.5 ± 6.1, 71.2 ± 9.6, 70.0 ± 5.8, and 68.6 ± 5.4, respectively; diabetic, 289.8 ± 30.9, 282.9 ± 29.5, 327.7 ± 15.4, 342.6 ± 16.3, 370.2 ± 12.1, 354.9 ± 16.2, and 353.2 ± 13.5 ng/ml, respectively).
FIG. 2. Changes in plasma ACTH (top panel) and corticosterone (bottom panel) concentrations from basal levels in normal (solid line; n = 6) and diabetic (dotted line; n = 6) rats during the hyperinsulinemic-euglycemic glucose clamp. Results are expressed as the mean ± SEM. No significant differences were observed between the responses of normal and diabetic animals to hyperinsulinemia-euglycemia.
FIG. 3. Comparison of peak ACTH (top panel) and corticosterone (bottom panel) responses of normal and diabetic rats to three different stressors: a mild activator of HPA function, hyperinsulinemia-euglycemia (; n = 6 for normal rats; n = 6 for diabetic rats), and the more potent activators, insulin-induced hypoglycemia (; n = 6 for normal rats; n = 5 for diabetic rats) (6 ) and restraint stress (; n = 6 for normal rats; n = 6 for diabetic rats) (7 ). *, P < 0.01; , P < 0.05 (vs. normal, hyperinsulinemia-euglycemia). Normal animals exhibit graded responses to stimuli of varying intensities whereas the response in diabetic animals seem to plateau at the same low level for all stimuli.
Basal plasma glucagon concentrations were significantly higher in diabetic animals compared with controls (P < 0.01; Table 1). During the clamp, plasma glucagon levels were suppressed compared with baseline levels in both normal and diabetic animals (P < 0.05; Table 1).
TABLE 1. Plasma glucagon concentrations (picograms per milliliter) during the hyperinsulinemic-euglycemic glucose clamp
No significant differences were observed in plasma catecholamine concentrations between the normal and diabetic groups either at baseline or during the clamp (Table 2). However, infusion of high doses of insulin while maintaining euglycemia did increase (P < 0.02) plasma epinephrine levels in both normal and diabetic animals by the end of the glucose clamp (Table 2).
TABLE 2. Plasma catecholamine concentrations (picograms per milliliter) under basal conditions (0 min) and during the hyperinsulinemic-euglycemic glucose clamp (30–150 min)
Fasting plasma insulin levels were not significantly different between normal and diabetic animals (Table 3). Insulin levels achieved during the euglycemic clamping period (Table 3) were also similar between the two treatment groups and to those obtained previously by our laboratory during hypoglycemic glucose clamps in identically treated groups of animals (6).
TABLE 3. Plasma insulin concentrations (microunits per milliliter) under basal conditions (0 min) and during the hyperinsulinemic-euglycemic glucose clamp (150 and 210 min)
Hypothalamic and pituitary neuropeptide expression
CRH mRNA was expressed throughout the hypothalamic paraventricular nucleus (PVN), with the highest abundance in the medial parvocellular region (Fig. 4). Diabetes resulted in significantly (P < 0.05) elevated basal CRH mRNA levels compared with those in nondiabetic controls. After 2 h of hyperinsulinemia-euglycemia, hypothalamic CRH mRNA levels rose significantly (P < 0.01) in both control and diabetic animals. However, CRH mRNA levels increased almost 4-fold in normal controls, whereas levels rose by less than 2-fold in diabetic animals. CRH mRNA levels during hyperinsulinemia-euglycemia were not significantly (P = 0.36) different between normal and diabetic animals.
FIG. 4. Computerized images and densitometric analysis after in situ hybridization of CRH mRNA in normal and diabetic rats under basal () conditions and after a hyperinsulinemic-euglycemic clamp (). Results are expressed as the mean ± SEM ROD. *, P < 0.01 vs. basal. , P < 0.05 vs. normal, basal. Basal CRH mRNA is elevated in diabetic animals and hyperinsulinemia-euglycemia increases hypothalamic CRH mRNA levels in both normal and diabetic animals.
Arginine vasopressin mRNA was expressed at high levels in magnocellular subfields of the PVN and supraoptic nucleus (SON) of all groups. No significant differences in basal AVP mRNA levels were observed between normal and diabetic rats (Table 4). After 2 h of hyperinsulinemia-euglycemia, AVP mRNA expression in the PVN and SON rose significantly (P < 0.02) in both treatment groups compared with their respective baseline levels.
TABLE 4. Densitometric analysis after in situ hybridization of arginine vasopressin (AVP) mRNA in the PVN and SON and of POMC mRNA in the anterior pituitary of normal and diabetic rats under basal conditions (Basal) and after a hyperinsulinemic-euglycemic clamp (EUG)
Expression of POMC mRNA was measured in the anterior pituitary. No significant differences in POMC mRNA expression were observed between the two groups under basal conditions (Table 4). Two hours of hyperinsulinemia-euglycemia resulted in significantly (P < 0.001) increased POMC mRNA content in both groups compared with baseline levels.
Corticosteroid receptor expression
MR mRNA expression was localized to limbic structures within the rat brain, particularly the CA1/2, CA3, and CA4 fields of the hippocampus and the dentate gyrus (Fig. 5A). Basal MR mRNA levels were significantly (P < 0.01) elevated throughout the hippocampus in uncontrolled diabetic animals. After 2 h of hyperinsulinemia-euglycemia, MR mRNA expression increased in normal control animals (P < 0.01), but not in diabetic animals.
FIG. 5. Computerized images and densitometric analysis after in situ hybridization of hippocampal MR mRNA (A) and GR mRNA (B) in the PVN and anterior pituitary (AP) of normal and diabetic rats under basal conditions and after hyperinsulinemic-euglycemic clamp. A, Representative results of hippocampal MR mRNA are shown for the CA3 region only under basal () and hyperinsulinemic-euglycemic () conditions and are expressed as the mean ± SEM ROD. *, P < 0.01 vs. normal, basal. Basal MR mRNA levels were increased in diabetic animals, and hyperinsulinemia-euglycemia increased MR mRNA in normal, but not diabetic, animals. B, Representative results of GR mRNA in both the PVN and AP for normal () and diabetic () animals under basal conditions and for normal (dark gray bars) and diabetic (light gray bars) animals after hyperinsulinemia-euglycemia. ?, P < 0.01 vs. normal, basal. #, P < 0.01 vs. diabetic, basal. Hyperinsulinemia-euglycemia increases GR mRNA levels in both the PVN and AP of normal animals, but only in the AP of diabetic animals.
GR mRNA expression was detected in the limbic system (CA1, CA2, CA3, and CA4 fields of the hippocampus and the dentate gyrus), PVN, and anterior pituitary. No significant changes in GR mRNA expression were observed between normal and diabetic rats under basal conditions. However, after 2 h of hyperinsulinemia-euglycemia, GR mRNA expression increased significantly (P < 0.01) in the PVN of normal animals and in the anterior pituitary of both normal and diabetic animals (Fig. 5B). There was no effect of hyperinsulinemia-euglycemia on GR mRNA levels in the hippocampus or dentate gyrus (Table 5).
TABLE 5. Densitometric analysis after in situ hybridization of GR mRNA expression in the limbic system [hippocampal CA1/2, CA3, and CA4 fields and dentate gyrus (DG)] of normal and diabetic rats under basal (Basal) conditions and after hyperinsulinemic-euglycemia (EUG)
Discussion
We previously demonstrated that the pituitary-adrenal response to hypoglycemia (6) was significantly blunted in uncontrolled STZ-diabetic rats despite elevated basal HPA function. Because insulin itself can stimulate the HPA axis, it was difficult to establish which aspects of the HPA response were attributable to the rise in plasma insulin levels and which were attributable to hypoglycemia in that earlier study. In the current study we showed that a rise in plasma insulin appears to moderately stimulate the axis of both normal and diabetic animals. These data, when examined in conjunction with results from our previous studies (6, 7), suggest that the ability to generate an appropriate stress response is compromised in uncontrolled STZ-diabetes, and this impairment may be centrally mediated.
Consistent with our previous findings, basal plasma corticosterone levels were significantly increased in diabetic rats. Although we previously reported increased basal plasma ACTH concentrations in uncontrolled STZ-diabetic rats (3, 6, 7), this was not observed in the present study. Such discrepancies in ACTH levels have been reported in studies from other laboratories as well (23). The pulsatile nature of ACTH release can increase signal noise, which can mask subtle differences in corticotroph activity. Although the experiments were performed at the same time of day to control for this variable, without taking more frequent samples we cannot be certain that we were not sampling during one of the peaks or troughs. Thus, it is not uncommon to find activated HPA function in the absence of altered plasma ACTH concentrations. Hence, it is generally accepted that a more reliable indicator of increased HPA function is the plasma corticosterone concentration. We previously reported that chronic treatment of diabetic animals with physiological doses of insulin lowered basal plasma ACTH and corticosterone concentrations to normal. In contrast, the current study showed that acutely administering supraphysiological doses of insulin mildly activated the axis. In human studies performed by Fruehwald-Schultes et al. (11), it was noted that the stimulatory effects of insulin on the HPA axis occurred only when high doses of insulin were administered. No such effects were observed when normal physiological doses of insulin were administered. Similarly, our data suggest that when insulin is administered at physiological doses, either the restoration of plasma insulin levels or the restoration of blood glucose concentrations was sufficient to normalize pituitary-adrenal function in diabetic animals. However, when much larger doses of insulin were used, as in the current study, the axis was stimulated.
Plasma ACTH and corticosterone concentrations increased moderately in both normal controls and diabetic rats during the hyperinsulinemic-euglycemic clamp. This observation is consistent with reports in human studies (10, 11). In the present study, activation of the pituitary-adrenal axis during hyperinsulinemia-euglycemia was slightly delayed in diabetic animals. This may be due to a decrease in pituitary and adrenal sensitivities (7). Peak ACTH and corticosterone responses to stimuli of varying intensities (hyperinsulinemia being a mild activator of HPA function, and restraint and hypoglycemia being more potent activators) were also compared, which highlight the magnitude of the hormonal responses to these different challenges. Normal animals responded to more potent stimuli with much greater HPA activation, whereas diabetic animals responded to all three stimuli with a similar degree of activation. Although diabetic rats responded similarly to normal animals to the milder stimulus of hyperinsulinemic-euglycemia, they appeared to lack the ability to generate greater responses to more potent stimuli. In this regard, the system is impaired in the diabetic group, and we suggest that perhaps the pituitary-adrenal response of diabetic animals has already reached a maximum following a mild challenge, such that when the system is presented with a more potent stress challenge, these animals are unable to drive the HPA system further. The adrenocortical response to stress appeared reduced in uncontrolled diabetic animals despite the fact that basal plasma corticosterone levels were higher in this group. In another study, in which we normalized the hyperglycemic condition in diabetic rats with phloridzin, we observed much more robust pituitary-adrenal responses to hypoglycemia even when basal plasma corticosterone concentrations remained elevated (24). Thus, the potential to generate a much greater response remains intact in diabetic animals despite elevated basal pituitary-adrenal activity. The impaired ability of diabetic animals to appropriately grade the HPA response to the intensity of a stressor is a novel observation that may have important implications in how patients with diabetes adapt to different stressors, especially to the metabolic stress of hypoglycemia.
Basal plasma glucagon concentrations were significantly elevated in diabetic animals compared with controls. As expected, during the euglycemic clamping period, glucagon levels gradually decreased and were significantly different from basal concentrations by the end of the clamp. The gradual decline in glucagon levels can be attributed to the suppressive effects of exogenous insulin on glucagon secretion from the pancreatic -cells (25, 26).
No differences in plasma catecholamine levels were observed between normal and diabetic animals either at baseline or during the euglycemic clamp. The variability in the responses of diabetic animals may be attributed to the fact that some of these animals reached their peak levels at an earlier time point than others. The gradual rise in plasma epinephrine levels during the clamp may be attributed to either the increase in CRH (27) or the effects of hyperinsulinemia in stimulating catecholamine secretion (28, 29). Brown et al. (27) showed that CRH injected into the brains of rats produced increases in plasma epinephrine. Kern et al. (28) and Rashid et al. (29) demonstrated that during periods of hyperinsulinemia, plasma epinephrine concentrations could rise, even in the presence of euglycemia. In the study by Rashid et al. (29), it was shown in dogs that during stress (an intracerebroventricular infusion of carbachol), when normoglycemia was maintained using a hyperinsulinemic-euglycemic glucose clamp protocol, epinephrine and norepinephrine levels increased with increasing doses of insulin, suggesting that insulin may play a role in stimulating catecholamine release. From our studies it is difficult to determine which of the two factors, CRH or insulin, predominates in stimulating epinephrine release or whether it is due to a combination of both hormones.
The expression of CRH mRNA in the hypothalamic PVN increased in both normal and diabetic animals after exposure to hyperinsulinemia, although the increase in diabetic animals was smaller (<2-fold) than that in normal animals (4-fold). Increases in hypothalamic CRH mRNA after the central administration of insulin have been reported previously (30). Although the mechanism by which insulin acts to increase CRH expression is not known, it is thought to be indirect and involve other hypothalamic areas (30). Physiologically, a large increase in plasma insulin levels may increase the synthesis and presumably the secretion of ACTH secretagogues to prepare the body for an impending fall in blood glucose. More importantly, the release of glucocorticoids may be important in increasing the body’s sensitivity to other counterregulatory hormones (31).
To our knowledge, little information is available concerning the effects of high insulin levels on hypothalamic AVP expression. High insulin concentrations increased AVP mRNA expression in both the PVN and SON of normal and diabetic rats despite plasma glucose levels being maintained at euglycemia. Although insulin receptors have been identified in a number of hypothalamic regions, the mechanism of interaction between insulin and AVP has yet to be determined.
POMC mRNA expression increased in both nondiabetic and diabetic rats after 2 h of hyperinsulinemia-euglycemia. Kim et al. (32) showed that insulin can stimulate pituitary POMC mRNA expression in STZ-diabetic animals. We have now shown that insulin can do this in the absence of hypoglycemia. Our data suggest that a large rise in plasma insulin increases all components of the HPA axis, and this mechanism appears to be centrally mediated at the hypothalamic level. This is consistent with findings that insulin increases POMC mRNA expression in corticotrophs (32). However, this effect may, again, be dose-dependent. At this point, the mechanisms underlying the actions of insulin on the HPA axis at these different doses are still unclear, and additional studies are required.
Interestingly, hippocampal MR mRNA levels rose significantly after 2 h of hyperinsulinemia-euglycemia in normal rats, but not in diabetic rats. Changes in hippocampal MR mRNA have been shown to occur as early as 30 min after the onset of a stress challenge (33). We previously suggested that a decrease in hippocampal MR mRNA expression may be required for full activation of the HPA axis in response to insulin-induced hypoglycemia (6). The increase in hippocampal MR mRNA expression in nondiabetic rats after exposure to hyperinsulinemia-euglycemia may be a mechanism to prevent additional activation of the system, because the rise in plasma insulin concentrations was not associated with a decline in plasma glucose levels. This would correlate well with what was observed for plasma ACTH and corticosterone levels. Compared with the robust pituitary-adrenal response observed during a hyperinsulinemic-hypoglycemic glucose clamp, plasma ACTH and corticosterone levels only increased modestly in the presence of euglycemia in normal animals. Conversely, the fact that hippocampal MR mRNA levels did not change in diabetic rats after 2 h of hyperinsulinemia-euglycemia may partially explain why plasma ACTH and corticosterone concentrations reached similar levels in response to different stress challenges. The absence of changes in hippocampal MR expression in diabetic animals may preclude appropriate activation of the axis during periods of stress. The data clearly demonstrate that diabetic animals exhibit impaired adaptive abilities of their central MR system in response to stress. This may play a very important role in determining appropriate HPA responsiveness to a given stress challenge.
In our studies we primarily noted changes in hippocampal MR expression during various stress challenges. Although hippocampal MRs have a similar affinity for both aldosterone and corticosterone, there is a 100- to 1000-fold greater concentration of corticosterone over aldosterone in the circulation. As such, hippocampal MRs are believed to be more responsive to changes in glucocorticoid, rather than mineralocorticoid, concentrations. Moreover, the specificity of renal MRs for aldosterone is conveyed by the expression of 11?-hydroxysteroid dehydrogenase-2, which acts to inactivate circulating glucocorticoids (34). This enzyme is not expressed in the hippocampus. Instead, another isoform, 11?-hydroxysteroid dehydrogenase-1, acts to locally amplify the glucocorticoid signal by converting the inactive glucocorticoid to its active counterpart (34). Although studies have shown that plasma aldosterone levels do increase during hypoglycemia (35, 36), to our knowledge, the effects of this increase on hippocampal MR function have not been investigated. Although this would be an interesting question to explore, we could not incorporate this parameter into our sampling protocol, because the additional blood sampling can potentially affect other hemodynamic parameters that may ultimately influence HPA output. Hence, aldosterone was not examined in our studies. With respect to electrolytes, we previously determined plasma osmolarity both under baseline conditions (0 min) and during hypoglycemia (80 and 210 min), and they revealed no differences (normal: 0 min, 289.7 ± 3.7; 80 min, 279.5 ± 4.3; 210 min, 300.2 ± 11.5; diabetics: 0 min, 298.5 ± 5.5; 80 min, 2280.8 ± 16.7; 210 min, 276.7 ± 6.4 mosmol) (our unpublished observations). In this regard, we do not believe that changes in osmolarity during stress are the cause of these changes in hippocampal MR mRNA levels.
In response to hyperinsulinemia-euglycemia, GR mRNA expression increased in the PVN and anterior pituitary of both normal and diabetic rats. Changes in GR mRNA in these areas after insulin treatment have been previously reported (3). It appears that insulin itself is capable of regulating GR mRNA expression, at least at the level of the PVN and anterior pituitary gland. We are currently unaware of any studies that have examined the mechanism by which this regulation occurs. The increase in GR mRNA expression after hyperinsulinemia-euglycemia may also contribute, at least in part, to averting additional activation of the HPA axis. Our previous study demonstrated that 4 d of insulin treatment caused an increase in GR mRNA expression in the anterior pituitary of diabetic rats and a corresponding decrease in ACTH and corticosterone. In the current study, infusion of pharmacological doses of insulin to both normal and diabetic animals revealed that insulin, given over even a short period of time, can cause an increase in GR mRNA expression in the anterior pituitary. We speculate that the increase in GR mRNA in these specific regions may aid in curtailing HPA activation, because the perceived rise in plasma insulin was not associated with a decline in plasma glucose levels. In the case of diabetic animals receiving insulin treatment, we previously suggested that the increase in pituitary GR mRNA may contribute to an increase in GR protein expression and, thus, normalization of pituitary-adrenal activity in diabetic animals. In the current study a similar argument can be made. With hyperinsulinemia-hypoglycemia, no differences in GR mRNA expression were observed, whereas hyperinsulinemia, in the absence of hypoglycemia, caused changes in GR mRNA expression. The data suggest that corticosteroid receptors as a whole may adjust their expression level in accordance with the type of stress challenge to modulate the output of the HPA axis. Because both the PVN and anterior pituitary have been shown to express insulin receptor substrate-1 (37) and insulin receptors (38, 39, 40), it is plausible that changes in insulin levels may play a role in modulating neuroendocrine function in these brain regions.
Pharmacological doses of insulin have been shown to exert vasodilatory effects, which can potentially affect local blood flow. Although in vitro studies by Zhao et al. (41) and Hasdai et al. (42) suggest that both insulin and IGF have vasodilatory effects on vascular smooth muscle, the in vivo data in both rats (43) and humans (44) suggest that the effects of insulin on cardiac output are minimal, and in the case of diabetes, that changes in vascular tone may be the result of hyperglycemia more than hypoinsulinemia. Nuutila et al. (45) showed that the insulin-induced increase in blood flow occurs much later than the effect of insulin on glucose uptake. More recently, Gudbjornsdottir et al. (46) measured the permeable surface area of insulin and glucose to assess the importance of the vasodilatory effects of insulin and noted that the permeable surface area for glucose was markedly increased by oral glucose, whereas additional vasodilation exerted by high insulin concentrations may not be physiologically relevant for capillary delivery of either glucose or insulin in muscle. In this respect, although blood flow was not measured in our studies, we do not believe that hyperinsulinemia would have altered blood flow significantly during our acute study to amend the interpretation of our results.
In conclusion, we have shown that insulin at high doses can stimulate the HPA axis both centrally and peripherally, effects that are independent of hypoglycemia. Indeed, although insulin activates pituitary-adrenal function to a similar degree in both normal and diabetic rats, the molecular mechanisms that underlie regulation of this response are different in the two groups. More importantly, in normal rats, although hyperinsulinemia per se stimulated the HPA axis, the response was much greater when hyperinsulinemia was associated with hypoglycemia or when the rats were exposed to restraint stress. In diabetic rats, the HPA responses to hyperinsulinemia, hyperinsulinemia, hypoglycemia, and restraint were equal, but were greatly diminished compared with those in normal rats. Based on these results, we propose that the peak response to stress is diminished by diabetes. Moreover, it appears that the ability to grade the HPA response to various forms of stress, which is observed in normal rats, is lost in diabetes and this may be related to defects in the corticosteroid receptor system. This latter finding may have important implications in the way patients with diabetes adapt to stressful stimuli, especially to the metabolic challenge of hypoglycemia.
Acknowledgments
We thank Elena Burdett and Debra Bilinski for their excellent technical assistance with this study.
References
Cameron OG, Kronfol Z, Grenden JF, Carroll BJ 1984 Hypothalamic-pituitary-adrenocortical activity in patients with diabetes mellitus. Arch Gen Psychiatry 41:1090–1095
Cameron OG, Thomas B, Tiongco D, Hariharan M, Greden JF 1987 Hypercortisolism in diabetes mellitus. Diabetes Care 10:662–664
Chan O, Chan S, Inouye K, Vranic M, Matthews SG 2001 Molecular regulation of the hypothalamo-pituitary-adrenal axis in streptozotocin-induced diabetes: effects of insulin treatment. Endocrinology 142:4872–4879
Roy M, Collier B, Roy A 1990 Hypothalamic-pituitary-adrenal axis dysregulation among diabetic out-patients. Psychiatry Res 31:31–37
Shapiro ET, Polonsky KS, Copinschi G, Bosson D, Tillil H, Blackman J, Lewis G, Van Cauter E 1991 Nocturnal elevation of glucose levels during fasting in noninsulin-dependent diabetes. J Clin Endocrinol Metab 72:444–454
Chan O, Chan S, Inouye K, Shum K, Matthews SG, Vranic M 2002 Diabetes impairs hypothalamo-pituitary-adrenal (HPA) responses to hypoglycemia, and insulin treatment normalizes HPA but not epinephrine responses. Diabetes 51:1681–1689
Chan O, Inouye K, Vranic M, Matthews SG 2002 Hyperactivation of the hypothalamo-pituitary-adrenocortical axis in streptozotocin-diabetes is associated with reduced stress responsiveness and decreased pituitary and adrenal sensitivity. Endocrinology 143:1761–1768
Baskin DG, Figlewicz LD, Seeley RJ, Woods SC, Porte Jr D, Schwartz MW 1999 Insulin and leptin: dual adiposity signals to the brain for the regulation of food intake and body weight. Brain Res 848:114–123
Havel PJ, Hahn TM, Sindelar DK, Baskin DG, Dallman MF, Weigle DS, Schwartz MW 2000 Effects of streptozotocin-induced diabetes and insulin treatment on the hypothalamic melanocortin system and muscle uncoupling protein 3 expression in rats. Diabetes 49:244–252
Fruehwald-Schultes B, Kern W, Bong W, Wellhoener P, Kerner W, Born J, Fehm HL, Peters A 1999 Supraphysiological hyperinsulinemia acutely increases hypothalamic-pituitary-adrenal secretory activity in humans. J Clin Endocrinol Metab 84:3041–3046
Fruehwald-Schultes B, Kern W, Born J, Fehm HL, Peters A 2001 Hyperinsulinemia causes activation of the hypothalamus-pituitary-adrenal axis in humans. Int J Obes Relat Metab Disord 25(Suppl 1):S38–S40
Shi ZQ, Rastogi KS, Lekas M, Efendic S, Drucker DJ, Vranic M 1996 Glucagon response to hypoglycemia is improved by insulin-independent restoration of normoglycemia in diabetic rats. Endocrinology 137:3193–3199
Dimitrakoudis D, Ramlal T, Rastogi S, Vranic M, Klip A 1992 Glycaemia regulates the glucose transporter number in the plasma membrane of rat skeletal muscle. Biochem J 284:341–348
Inouye K, Shum K, Chan O, Mathoo J, Matthews SG, Vranic M 2002 Effects of recurrent hyperinsulinemia with and without hypoglycemia on counterregulation in diabetic rats. Am J Physiol 282:E1369–E1379
Simpson IA, Vannucci SJ 2002 Glucose transport into brain: effects of hypoglycemia. Diabetes Nutr Metab 15:281–284
Matthews SG, Challis JR 1995 Regulation of CRH and AVP mRNA in the developing ovine hypothalamus: effects of stress and glucocorticoids. Am J Physiol 268:E1096–E1107
Paxinos G, Watson C 1997 The rat brain in stereotaxic co-ordinates. 3rd ed. San Diego: Academic Press
Lightman SL, Young III WS 1988 Corticotrophin-releasing factor, vasopressin and pro-opiomelanocortin mRNA responses to stress and opiates in the rat. J Physiol 403:511–523
Matthews SG 1998 Dynamic changes in glucocorticoid and mineralocorticoid receptor mRNA in the developing guinea pig brain. Brain Res Dev Brain Res 107:123–132
Miesfeld R, Rusconi S, Godowski PJ, Maler BA, Okret S, Wikstrom AC, Gustafsson JA, Yamamoto KR 1986 Genetic complementation of a glucocorticoid receptor deficiency by expression of cloned receptor cDNA. Cell 46:389–399
Herbert V, Lau KS, Gottlieb CW, Bleicher SJ 1965 Coated charcoal immunoassay of insulin. J Clin Endocrinol Metab 25:1375–1384
Peuler JD, Johnson GA 1977 Simultaneous single isotope radioenzymatic assay of plasma norepinephrine, epinephrine and dopamine. Life Sci 21:625–636
Scribner KA, Walker CD, Cascio CS, Dallman MF 1991 Chronic streptozotocin diabetes in rats facilitates the acute stress response without altering pituitary or adrenal responsiveness to secretagogues. Endocrinology 129:99–108
Chan O, Inouye K, Akirav EM, Park E, Riddell MC, Vranic M, Matthews SG 2002 Hyperglycemia impairs the hypothalamo-pituitary-adrenal (HPA) response to insulin-induced hypoglycemia in streptozotocin-diabetic rats. Stress 5(Suppl):27
Rastogi KS, Cooper RL, Shi ZQ, Vranic M 1997 Quantitative measurement of islet glucagon response to hypoglycemia by confocal fluorescence imaging in diabetic rats: effects of phlorizin treatment. Endocrine 7:367–375
Taborsky Jr GJ, Ahren B, Havel PJ 1998 Autonomic mediation of glucagon secretion during hypoglycemia: implications for impaired -cell responses in type 1 diabetes. Diabetes 47:995–1005
Brown MR, Fisher LA, Spiess J, Rivier C, Rivier J, Vale W 1982 Corticotropin-releasing factor: actions on the sympathetic nervous system and metabolism. Endocrinology 111:928–931
Kern W, Fittje A, Fohr W, Kerner W, Born J, Fehm HL 2000 Increase in systolic blood pressure and catecholamine level during hyperinsulinemia in a placebo-controlled euglycemic clamp in healthy subjects. Exp Clin Endocrinol Diabetes 108:498–505
Rashid S, Shi ZQ, Niwa M, Mathoo JM, Vandelangeryt ML, Bilinski D, Lewis GF, Vranic M 2000 ?-Blockade, but not normoglycemia or hyperinsulinemia, markedly diminishes stress-induced hyperglycemia in diabetic dogs. Diabetes 49:253–262
Sipols AJ, Baskin DG, Schwartz MW 1995 Effect of intracerebroventricular insulin infusion on diabetic hyperphagia and hypothalamic neuropeptide gene expression. Diabetes 44:147–151
Dirlewanger M, Schneiter PH, Paquot N, Jequier E, Rey V, Tappy L 2000 Effects of glucocorticoids on hepatic sensitivity to insulin and glucagon in man. Clin Nutr 19:29–34
Kim EM, Grace MK, Welch CC, Billington CJ, Levine AS 1999 STZ-induced diabetes decreases and insulin normalizes POMC mRNA in arcuate nucleus and pituitary in rats. Am J Physiol 276:R1320–R1326
Fujikawa T, Soya H, Fukuoka H, Alam KS, Yoshizato H, McEwen BS, Nakashima K 2000 A biphasic regulation of receptor mRNA expressions for growth hormone, glucocorticoid and mineralocorticoid in the rat dentate gyrus during acute stress. Brain Res 874:186–193
De Kloet ER, Vreugdenhil E, Oitzl MS, Joels M 1998 Brain corticosteroid receptor balance in health and disease. Endocr Rev 19:269–301
Hochberg Z, Dickerman Z, Kaufman H, Laron Z 1980 Evaluation of the renin-aldosterone system during hypo- and hyperglycemia in children and adolescents. Horm Res 12:16–21
Ramirez G, Ganguly A, Bruegemeyer C, Gomez-Sanchez C 1989 Aldosterone response to insulin-induced hypoglycemia in hemodialysis patients. Am J Nephrol 9:215–221
Baskin DG, Sipols AJ, Schwartz MW, White MF 1993 Immunocytochemical detection of insulin receptor substrate-1 (IRS-1) in rat brain: colocalization with phosphotyrosine. Regul Pept 48:257–266
Yamashita S, Melmed S 1986 Effects of insulin on rat anterior pituitary cells. Inhibition of growth hormone secretion and mRNA levels. Diabetes 35:440–447
Unger JW, Lange W 1997 Insulin receptors in the pituitary gland: morphological evidence for influence on opioid peptide-synthesizing cells. Cell Tissue Res 288:471–483
Unger JW, Betz M 1998 Insulin receptors and signal transduction proteins in the hypothalamo-hypophyseal system: a review on morphological findings and functional implications. Histol Histopathol 13:1215–1224
Zhao G, Sutliff RL, Weber CS, Wang J, Lorenz J, Paul RJ, Fagin JA 2001 Smooth muscle-targeted overexpression of insulin-like growth factor I results in enhanced vascular contractility. Endocrinology 142:623–632
Hasdai D, Rizza RA, Holmes Jr DR, Richardson DM, Cohen P, Lerman A 1998 Insulin and insulin-like growth factor-I cause coronary vasorelaxation in vitro. Hypertension 32:228–234
Hilzenrat N, Yaari A, Maislos M, Sikuler E 2001 Splanchnic and systemic hemodynamic effects of sustained euglycemic hyperinsulinemia in rats. Scand J Clin Lab Invest 61:383–388
Williams SB, Goldfine AB, Timimi FK, Ting HH, Roddy MA, Simonson DC, Creager MA 1998 Acute hyperglycemia attenuates endothelium-dependent vasodilation in humans in vivo. Circulation 97:1695–1701
Nuutila P, Raitakari M, Laine H, Kirvela O, Takala T, Utriainen T, Makimattila S, Pitkanen OP, Ruotsalainen U, Iida H, Knuuti J, Yki-Jarvinen H 1996 Role of blood flow in regulating insulin-stimulated glucose uptake in humans. Studies using bradykinin, [15O]water, and [18F]fluoro-deoxy-glucose and positron emission tomography. J Clin Invest 97:1741–1747
Gudbjornsdottir S, Sjostrand M, Strindberg L, Wahren J, Lonnroth P 2003 Direct measurements of the permeability surface area for insulin and glucose in human skeletal muscle. J Clin Endocrinol Metab 88:4559–4564(Owen Chan, Karen Inouye, )
Address all correspondence and requests for reprints to: Dr. Stephen G. Matthews, Medical Sciences Building, Room 3240, 1 King’s College Circle, University of Toronto, Toronto, Ontario, Canada M5S 1A8. E-mail: stephen.matthews@utoronto.ca.
Abstract
Diabetes is associated with increased basal hypothalamo-pituitary-adrenal (HPA) activity and impaired stress responsiveness. Previously, we demonstrated that the HPA response to hypoglycemia is significantly impaired in diabetic rats. In this study our goals were to 1) differentiate between the effects of hyperinsulinemia and those of hypoglycemia per se, and 2) establish whether diabetes lowers peak stress responses. Normal and streptozotocin-diabetic rats were subjected to hyperinsulinemic-euglycemic glucose clamps to evaluate central and peripheral responses. These were compared with peak ACTH and corticosterone responses to restraint and hypoglycemia. Hyperinsulinemia increased CRH and vasopressin mRNA, and plasma ACTH and corticosterone in normal and diabetic rats. In normal animals, insulin-induced activation of ACTH and corticosterone was lower than the responses during either restraint or hypoglycemia. In contrast, ACTH and corticosterone activation in diabetic rats was similar with all three stressors. Pituitary-adrenal axis activation in diabetic animals was also much lower compared with that in normal controls. The response to hyperinsulinemia (euglycemia) was associated with increases in glucocorticoid receptor mRNA in the anterior pituitary and paraventricular nucleus. Hippocampal mineralocorticoid receptor mRNA expression was increased in normal, but not in diabetic, animals. We speculate that the ability to appropriately match the HPA response to the potency of a stressor is related to the ability to alter hippocampal mineralocorticoid receptor expression. In diabetes, this ability is impaired; hence, maximal HPA activation is greatly diminished. This is a novel observation that may have important implications in the treatment of impaired counterregulatory mechanisms in human diabetes.
Introduction
DIABETES IS A metabolic disorder that is associated with dysregulation of a number of systems within the body, including cardiovascular disorders and altered immune and central nervous system functions. One of these central nervous system dysfunctions includes activation of the hypothalamo-pituitary-adrenocortical (HPA) axis in uncontrolled or poorly controlled diabetes (1, 2, 3, 4, 5). We have previously demonstrated that basal HPA function is up-regulated in streptozotocin (STZ)-diabetic rats and that this is associated with increases in hypothalamic CRH mRNA and plasma ACTH and corticosterone levels (3). Interestingly, elevated basal HPA function in diabetes is also associated with increased hippocampal mineralocorticoid receptor (MR) mRNA expression, a major source of inhibitory input to the HPA axis. This suggests that increased central drive at or above the level of the hypothalamus is responsible for HPA hyperactivation in diabetes. More importantly, our laboratory also demonstrated that activation of the HPA axis in response to various forms of stress and, more specifically, in response to 20 min of restraint and to insulin induced-hypoglycemia was greatly diminished in diabetic animals (6, 7). Defects in stress responsiveness in diabetes appear to be the result of a combination of decreased tissue sensitivity to secretagogues as well as an impaired capacity to stimulate the axis that may stem from an inability to decrease hippocampal MR expression and thus to relieve inhibitory tone to the axis. In the hypoglycemia studies it was difficult to distinguish between effects that were the result of hyperinsulinemia and those that were attributable to hypoglycemia alone. We demonstrate in this study that insulin itself can have profound effects on the HPA axis, both centrally and peripherally, that are independent of hypoglycemia. More interestingly, by comparing the peak pituitary-adrenal responses to various activators of HPA function (hyperinsulinemia, restraint, and hypoglycemia), we show that defects in the HPA axis of diabetic animals result in a system that is incapable of responding appropriately to a stress challenge because its peak response plateaus at a significantly lower level than that in normal animals.
The central interactions between insulin and the HPA axis have primarily been examined with regard to energy homeostasis. In general, it is believed that glucocorticoids establish a deleterious environment wherein obesity and insulin resistance can develop. Centrally, insulin primarily acts to suppress food intake and prevent the onset of obesity (8). In a study that examined the central role of insulin in STZ-diabetic rats, it was shown that insulin can also increase the expression of the anorexigenic neuropeptide, CRH, in the hypothalamus (9). More specific interactions of insulin with the HPA axis in the absence of hypoglycemia have not been investigated.
Limited studies have examined the effect of insulin on pituitary-adrenal function in human subjects. Fruehwald-Schultes et al. (10) demonstrated that iv infusion of high doses of insulin to human subjects in the presence of euglycemia increases plasma ACTH and cortisol concentrations. Conversely, when a lower dose of insulin was used, pituitary-adrenal activity remained unchanged (11). These studies suggest that the stimulatory actions of insulin on HPA function may only be effective when insulin concentrations are high.
As mentioned above, our own work has shown that insulin administration has profound effects on the HPA axis of diabetic animals in regulating both basal (3) and activated function (6). Thus, the aims of this study were 1) to determine which aspects of the HPA response to hypoglycemia are the result of hyperinsulinemia and which are due to hypoglycemia per se, and 2) to ascertain whether the capacity to adjust the magnitude of the HPA response to various stressors is altered in diabetes.
Materials and Methods
Experimental animals and design
Male Sprague Dawley rats (Charles River, Québec, Canada), initially weighing 325–375 g, were individually housed in opaque cages in temperature (22 C)- and humidity-controlled rooms. The animals were fed rat chow (Ralston-Purina Co., St. Louis, MO) and water ad libitum and were acclimatized to a 12-h light cycle (lights on between 0700–1900 h) for a period of 1 wk before experimental manipulation. All experiments were approved by the animal care committee of University of Toronto and were conducted in accordance with regulations set by the Canadian Council for Animal Care.
Euglycemic clamp studies
Two groups of rats were used: normal controls (n = 6) and untreated STZ-induced diabetic (n = 6) rats. On d 0, catheters were placed into the left carotid artery and right jugular vein, as described previously (6). Diabetes was induced at the end of surgery via a single injection of STZ (65 mg/kg; Sigma-Aldrich Corp., St. Louis, MO) dissolved in sterile saline, through the penile vein. Control animals received a saline injection under similar conditions. Animals treated with STZ were given 10% sucrose water for the first 24 h after STZ injection to prevent hypoglycemia (12). This is a model of moderate diabetes characterized by fasting hyperglycemia, normal fasting plasma insulin, and impaired plasma insulin responses after marked hyperglycemia (534 ± 2.4 mg/dl) (13). Blood glucose was monitored twice daily with a glucometer (Glucometer Elite 3903, Bayer, Inc., Etobicoke, Canada) in all animals to ensure that fasting normoglycemia was maintained in control animals and that adequate hyperglycemia (>15 mM) was achieved in the uncontrolled diabetic group. The contents of the catheters were aspirated and reprimed with a 60% polyvinylpyrolidone solution (wt/vol; 1000 U/ml heparin) on d 2, 4, and 6 to maintain catheter patency and to acclimatize the rats to being handled.
On d 7, the rats were fasted for 24 h before the start of the experiment. On d 8, the experiments were carried out in unrestrained, conscious animals. The catheters were extended outside the cage to minimize investigator interaction and were connected to infusion pumps. The animals were left undisturbed for 2.5 h before basal hormone samples were collected (1030 and 1100 h), just before the start of the insulin infusion (1100 h). The rats then underwent a hyperinsulinemic-euglycemic glucose clamp. A constant insulin (50 mU/kg·min regular porcine insulin; Eli Lilly & Co., Indianapolis, IN) and variable dextrose (45%; Abbott Laboratories Ltd., Montréal, Canada) infusion through the jugular vein catheter were used to maintain plasma glucose levels at 6.7 ± 0.6 mM for 130 min. Large doses of insulin were used because even in moderately diabetic animals these doses were necessary to induce hypoglycemia (6, 14, 15). The concentration of glucose was determined every 5 min from a sample of plasma (10 μl) using a glucose analyzer II (Beckman Coulter, Palo Alto, CA). Arterial blood samples were obtained from the carotid catheter at regular time intervals throughout the glucose clamp. For ACTH, glucagon, and insulin measurements, blood was collected in chilled tubes containing EDTA (Sangon Ltd. Canada, Scarborough, Canada) and Trasylol (Bayer, Inc.). Blood samples for catecholamine determinations were collected in chilled tubes containing 1.5 mg reduced glutathione (Roche, Mannheim, Germany) and 5 μl EGTA (Sigma-Aldrich Corp.). Serum was collected for corticosterone measurements. Plasma was aliquoted into storage tubes and stored at –20 C (or –80 C for catecholamine determination). After removal of plasma, red blood cells were resuspended in heparinized saline (10 U/ml) and reinfused after each blood sampling to prevent volume depletion and anemia. Hematocrit, determined at the beginning and the end of the experiment, was maintained above 35%. At the end of the experiment, the rats were euthanized by decapitation. Trunk blood samples were collected, and brains and pituitary glands were removed and stored at –80 C.
In situ hybridization
The method of in situ hybridization has been described in detail previously (16). Briefly, coronal cryosections (12 μm) were obtained through selected hypothalamic (bregma –2.00 mm) and hippocampal (bregma –3.80 mm) regions according to the stereotaxic coordinates of Paxinos and Watson (17). The sections were then thaw-mounted onto poly-L-lysine (Sigma-Aldrich Corp.)-coated slides, fixed for 5 min in 4% phosphate-buffered paraformaldehyde, rinsed in PBS (2 min), dehydrated in an ethanol series (70% and 95%), and stored in 95% ethanol at 4 C until use.
The 45-mer antisense CRH (bases 536–580) (18), AVP (bases 588–632) (18), proopiomelanocortin (POMC; bases 572–616) (18), MR (bases 2942–2986) (19), and glucocorticoid receptor (GR; bases 1321–1365) (20) oligonucleotide probes were synthesized by Dalton Chemical Laboratories, Inc. (Toronto, Canada). The probes were labeled using terminal deoxynucleotidyl transferase (Pharmacia Biotech, Baie d’Urfé, Canada) and [35S]deoxy-ATP (1300 Ci/mmol; NEN Life Science Products, DuPont Canada, Mississauga, Canada) to a specific activity of 1.0 x 109 cpm/μg. Labeled probe in hybridization buffer (180 μl) was applied to each slide at a concentration of 1.0 x 106 cpm/μl. Slides were incubated overnight in a moist chamber at 42.5 C. After washing in 1x standard saline citrate (1x SSC; 20 min at room temperature) and 1x SSC (35 min at 55 C), the slides were rinsed twice with 1x SSC and once with 0.1x SSC at room temperature, then dehydrated in 70% and 95% ethanol (1 min each), air-dried, and exposed to autoradiographic film (Biomax, Eastman Kodak Co., Rochester, NY). The films were developed using standard procedures (exposure times: CRH, 21 d; AVP, 2 d; POMC, 2 h; MR, 14 d; GR, 28 d).
Because tissues obtained from the current study were collected after a euglycemic clamp, we used slides containing select hypothalamic, hippocampal, and pituitary tissue sections collected from a previous basal study (3) and simultaneously ran new in situ hybridizations using these slides and those collected from the current study to determine the molecular HPA responses to hyperinsulinemia-euglycemia. The animals from which the basal tissues were collected received identical treatment regimens as described above, except they were euthanized between 1000–1100 h on d 8. By doing this, we were able to compare basal neuropeptide and corticosteroid receptor mRNA expression with that after the euglycemic clamp to ascertain the effects of hyperinsulinemic-euglycemia.
Plasma hormone and catecholamine determination
Plasma insulin was measured using a modified version of the insulin RIA by Herbert et al. (21). Plasma ACTH (Diasorin, Inc., Stillwater, MN), corticosterone (ICN Pharmaceuticals, Inc., Orangeburg, NY), and glucagon (Diagnostic Products Corp., Los Angeles, CA) concentrations were determined using commercially available RIA kits.
Plasma epinephrine and norepinephrine concentrations were determined using the simultaneous single isotope derivative radioenzymatic assay technique described previously (14, 22).
Data analysis
For in situ hybridizations, brain sections were processed simultaneously for each probe to allow direct comparison between treatment groups. Six to eight sections were analyzed from each animal for each probe based on visual inspection of the desired region. The sections were exposed together with 14C-labeled standards (American Radiochemical, St. Louis, MO) to ensure analysis in the linear region of the autoradiographic film. The relative OD (ROD) of the signal on autoradiographic film was quantified, after subtraction of background values, using a computerized image analysis system (Imaging Research, St. Catherines, Canada). Hormone data are presented as the mean ± SEM, and in situ hybridization data are expressed as the ROD (mean ± SEM). Statistical analysis was performed by one- or two-way ANOVA for independent or repeated measures, as appropriate, using Statistica 6.0 statistical software (StatSoft, Tulsa, OK), with P < 0.05 set as the criterion for statistical significance.
Results
Body weight and plasma hormone and catecholamine concentrations
Although body weights did not differ significantly between treatment groups before surgery, 8 d after the induction of diabetes, STZ-diabetic rats exhibited a significant decline in body weight (P < 0.03; normal, 348.0 ± 13.2; STZ, 298.6 ± 14.1 g). Hematocrit levels were maintained above 35% during the clamp (normal, 42.4 ± 2.3 STZ, 40.6 ± 2.3%) to prevent hemodynamic changes resulting from volume depletion and anemia.
Basal fasting plasma glucose concentrations were increased (P < 0.01) by approximately 4-fold in diabetic rats compared with normal animals (normal, 5.4 ± 0.3; STZ, 21.1 ± 1.0 mM). Plasma glucose levels during the euglycemic clamp are shown in Fig. 1. The break in Fig. 1 indicates the period during which we ensured that plasma glucose levels were matched between the two treatment groups before blood sampling was resumed. Although basal plasma glucose levels were higher in diabetic animals, they were maintained at similar levels to normal controls during the euglycemic clamping period from 80–210 min.
FIG. 1. Plasma glucose concentrations achieved in normal (solid line) and diabetic (dotted line) rats during the hyperinsulinemic-euglycemic glucose clamp. Results are presented as the mean ± SEM.
Basal plasma ACTH levels (normal, 64.9 ± 4.5; STZ, 53.6 ± 4.4 pg/ml) were not significantly different between normal and diabetic rats. However, basal plasma corticosterone concentrations were significantly (P < 0.05) elevated in diabetic compared with control animals (normal, 59.2 ± 11.3; STZ, 140.9 ± 42.0 ng/ml). During the hyperinsulinemic-euglycemic clamp, plasma ACTH and corticosterone concentrations rose significantly (P < 0.05) in both normal and diabetic rats and peak levels were similar in the two groups (Fig. 2). In comparison with other activators of HPA function, such as restraint (7) and insulin-induced hypoglycemia (6), we noted that the magnitude of the pituitary-adrenal response to hyperinsulinemia-euglycemia was significantly less in nondiabetic animals (Fig. 3). In contrast, the pituitary-adrenal responses of diabetic animals to all three of these stressors were similar in magnitude (Fig. 3). These effects appear to be specific to the infusion of insulin and are not the result of repeated sampling, because the same sampling protocol using saline infusion instead of insulin did not significantly affect corticosterone levels at any of the sampling points (0, 30, 60, 80, 100, 150, and 210 min: normal, 47.9 ± 9.6, 43.1 ± 4.5, 65.9 ± 1.8, 68.5 ± 6.1, 71.2 ± 9.6, 70.0 ± 5.8, and 68.6 ± 5.4, respectively; diabetic, 289.8 ± 30.9, 282.9 ± 29.5, 327.7 ± 15.4, 342.6 ± 16.3, 370.2 ± 12.1, 354.9 ± 16.2, and 353.2 ± 13.5 ng/ml, respectively).
FIG. 2. Changes in plasma ACTH (top panel) and corticosterone (bottom panel) concentrations from basal levels in normal (solid line; n = 6) and diabetic (dotted line; n = 6) rats during the hyperinsulinemic-euglycemic glucose clamp. Results are expressed as the mean ± SEM. No significant differences were observed between the responses of normal and diabetic animals to hyperinsulinemia-euglycemia.
FIG. 3. Comparison of peak ACTH (top panel) and corticosterone (bottom panel) responses of normal and diabetic rats to three different stressors: a mild activator of HPA function, hyperinsulinemia-euglycemia (; n = 6 for normal rats; n = 6 for diabetic rats), and the more potent activators, insulin-induced hypoglycemia (; n = 6 for normal rats; n = 5 for diabetic rats) (6 ) and restraint stress (; n = 6 for normal rats; n = 6 for diabetic rats) (7 ). *, P < 0.01; , P < 0.05 (vs. normal, hyperinsulinemia-euglycemia). Normal animals exhibit graded responses to stimuli of varying intensities whereas the response in diabetic animals seem to plateau at the same low level for all stimuli.
Basal plasma glucagon concentrations were significantly higher in diabetic animals compared with controls (P < 0.01; Table 1). During the clamp, plasma glucagon levels were suppressed compared with baseline levels in both normal and diabetic animals (P < 0.05; Table 1).
TABLE 1. Plasma glucagon concentrations (picograms per milliliter) during the hyperinsulinemic-euglycemic glucose clamp
No significant differences were observed in plasma catecholamine concentrations between the normal and diabetic groups either at baseline or during the clamp (Table 2). However, infusion of high doses of insulin while maintaining euglycemia did increase (P < 0.02) plasma epinephrine levels in both normal and diabetic animals by the end of the glucose clamp (Table 2).
TABLE 2. Plasma catecholamine concentrations (picograms per milliliter) under basal conditions (0 min) and during the hyperinsulinemic-euglycemic glucose clamp (30–150 min)
Fasting plasma insulin levels were not significantly different between normal and diabetic animals (Table 3). Insulin levels achieved during the euglycemic clamping period (Table 3) were also similar between the two treatment groups and to those obtained previously by our laboratory during hypoglycemic glucose clamps in identically treated groups of animals (6).
TABLE 3. Plasma insulin concentrations (microunits per milliliter) under basal conditions (0 min) and during the hyperinsulinemic-euglycemic glucose clamp (150 and 210 min)
Hypothalamic and pituitary neuropeptide expression
CRH mRNA was expressed throughout the hypothalamic paraventricular nucleus (PVN), with the highest abundance in the medial parvocellular region (Fig. 4). Diabetes resulted in significantly (P < 0.05) elevated basal CRH mRNA levels compared with those in nondiabetic controls. After 2 h of hyperinsulinemia-euglycemia, hypothalamic CRH mRNA levels rose significantly (P < 0.01) in both control and diabetic animals. However, CRH mRNA levels increased almost 4-fold in normal controls, whereas levels rose by less than 2-fold in diabetic animals. CRH mRNA levels during hyperinsulinemia-euglycemia were not significantly (P = 0.36) different between normal and diabetic animals.
FIG. 4. Computerized images and densitometric analysis after in situ hybridization of CRH mRNA in normal and diabetic rats under basal () conditions and after a hyperinsulinemic-euglycemic clamp (). Results are expressed as the mean ± SEM ROD. *, P < 0.01 vs. basal. , P < 0.05 vs. normal, basal. Basal CRH mRNA is elevated in diabetic animals and hyperinsulinemia-euglycemia increases hypothalamic CRH mRNA levels in both normal and diabetic animals.
Arginine vasopressin mRNA was expressed at high levels in magnocellular subfields of the PVN and supraoptic nucleus (SON) of all groups. No significant differences in basal AVP mRNA levels were observed between normal and diabetic rats (Table 4). After 2 h of hyperinsulinemia-euglycemia, AVP mRNA expression in the PVN and SON rose significantly (P < 0.02) in both treatment groups compared with their respective baseline levels.
TABLE 4. Densitometric analysis after in situ hybridization of arginine vasopressin (AVP) mRNA in the PVN and SON and of POMC mRNA in the anterior pituitary of normal and diabetic rats under basal conditions (Basal) and after a hyperinsulinemic-euglycemic clamp (EUG)
Expression of POMC mRNA was measured in the anterior pituitary. No significant differences in POMC mRNA expression were observed between the two groups under basal conditions (Table 4). Two hours of hyperinsulinemia-euglycemia resulted in significantly (P < 0.001) increased POMC mRNA content in both groups compared with baseline levels.
Corticosteroid receptor expression
MR mRNA expression was localized to limbic structures within the rat brain, particularly the CA1/2, CA3, and CA4 fields of the hippocampus and the dentate gyrus (Fig. 5A). Basal MR mRNA levels were significantly (P < 0.01) elevated throughout the hippocampus in uncontrolled diabetic animals. After 2 h of hyperinsulinemia-euglycemia, MR mRNA expression increased in normal control animals (P < 0.01), but not in diabetic animals.
FIG. 5. Computerized images and densitometric analysis after in situ hybridization of hippocampal MR mRNA (A) and GR mRNA (B) in the PVN and anterior pituitary (AP) of normal and diabetic rats under basal conditions and after hyperinsulinemic-euglycemic clamp. A, Representative results of hippocampal MR mRNA are shown for the CA3 region only under basal () and hyperinsulinemic-euglycemic () conditions and are expressed as the mean ± SEM ROD. *, P < 0.01 vs. normal, basal. Basal MR mRNA levels were increased in diabetic animals, and hyperinsulinemia-euglycemia increased MR mRNA in normal, but not diabetic, animals. B, Representative results of GR mRNA in both the PVN and AP for normal () and diabetic () animals under basal conditions and for normal (dark gray bars) and diabetic (light gray bars) animals after hyperinsulinemia-euglycemia. ?, P < 0.01 vs. normal, basal. #, P < 0.01 vs. diabetic, basal. Hyperinsulinemia-euglycemia increases GR mRNA levels in both the PVN and AP of normal animals, but only in the AP of diabetic animals.
GR mRNA expression was detected in the limbic system (CA1, CA2, CA3, and CA4 fields of the hippocampus and the dentate gyrus), PVN, and anterior pituitary. No significant changes in GR mRNA expression were observed between normal and diabetic rats under basal conditions. However, after 2 h of hyperinsulinemia-euglycemia, GR mRNA expression increased significantly (P < 0.01) in the PVN of normal animals and in the anterior pituitary of both normal and diabetic animals (Fig. 5B). There was no effect of hyperinsulinemia-euglycemia on GR mRNA levels in the hippocampus or dentate gyrus (Table 5).
TABLE 5. Densitometric analysis after in situ hybridization of GR mRNA expression in the limbic system [hippocampal CA1/2, CA3, and CA4 fields and dentate gyrus (DG)] of normal and diabetic rats under basal (Basal) conditions and after hyperinsulinemic-euglycemia (EUG)
Discussion
We previously demonstrated that the pituitary-adrenal response to hypoglycemia (6) was significantly blunted in uncontrolled STZ-diabetic rats despite elevated basal HPA function. Because insulin itself can stimulate the HPA axis, it was difficult to establish which aspects of the HPA response were attributable to the rise in plasma insulin levels and which were attributable to hypoglycemia in that earlier study. In the current study we showed that a rise in plasma insulin appears to moderately stimulate the axis of both normal and diabetic animals. These data, when examined in conjunction with results from our previous studies (6, 7), suggest that the ability to generate an appropriate stress response is compromised in uncontrolled STZ-diabetes, and this impairment may be centrally mediated.
Consistent with our previous findings, basal plasma corticosterone levels were significantly increased in diabetic rats. Although we previously reported increased basal plasma ACTH concentrations in uncontrolled STZ-diabetic rats (3, 6, 7), this was not observed in the present study. Such discrepancies in ACTH levels have been reported in studies from other laboratories as well (23). The pulsatile nature of ACTH release can increase signal noise, which can mask subtle differences in corticotroph activity. Although the experiments were performed at the same time of day to control for this variable, without taking more frequent samples we cannot be certain that we were not sampling during one of the peaks or troughs. Thus, it is not uncommon to find activated HPA function in the absence of altered plasma ACTH concentrations. Hence, it is generally accepted that a more reliable indicator of increased HPA function is the plasma corticosterone concentration. We previously reported that chronic treatment of diabetic animals with physiological doses of insulin lowered basal plasma ACTH and corticosterone concentrations to normal. In contrast, the current study showed that acutely administering supraphysiological doses of insulin mildly activated the axis. In human studies performed by Fruehwald-Schultes et al. (11), it was noted that the stimulatory effects of insulin on the HPA axis occurred only when high doses of insulin were administered. No such effects were observed when normal physiological doses of insulin were administered. Similarly, our data suggest that when insulin is administered at physiological doses, either the restoration of plasma insulin levels or the restoration of blood glucose concentrations was sufficient to normalize pituitary-adrenal function in diabetic animals. However, when much larger doses of insulin were used, as in the current study, the axis was stimulated.
Plasma ACTH and corticosterone concentrations increased moderately in both normal controls and diabetic rats during the hyperinsulinemic-euglycemic clamp. This observation is consistent with reports in human studies (10, 11). In the present study, activation of the pituitary-adrenal axis during hyperinsulinemia-euglycemia was slightly delayed in diabetic animals. This may be due to a decrease in pituitary and adrenal sensitivities (7). Peak ACTH and corticosterone responses to stimuli of varying intensities (hyperinsulinemia being a mild activator of HPA function, and restraint and hypoglycemia being more potent activators) were also compared, which highlight the magnitude of the hormonal responses to these different challenges. Normal animals responded to more potent stimuli with much greater HPA activation, whereas diabetic animals responded to all three stimuli with a similar degree of activation. Although diabetic rats responded similarly to normal animals to the milder stimulus of hyperinsulinemic-euglycemia, they appeared to lack the ability to generate greater responses to more potent stimuli. In this regard, the system is impaired in the diabetic group, and we suggest that perhaps the pituitary-adrenal response of diabetic animals has already reached a maximum following a mild challenge, such that when the system is presented with a more potent stress challenge, these animals are unable to drive the HPA system further. The adrenocortical response to stress appeared reduced in uncontrolled diabetic animals despite the fact that basal plasma corticosterone levels were higher in this group. In another study, in which we normalized the hyperglycemic condition in diabetic rats with phloridzin, we observed much more robust pituitary-adrenal responses to hypoglycemia even when basal plasma corticosterone concentrations remained elevated (24). Thus, the potential to generate a much greater response remains intact in diabetic animals despite elevated basal pituitary-adrenal activity. The impaired ability of diabetic animals to appropriately grade the HPA response to the intensity of a stressor is a novel observation that may have important implications in how patients with diabetes adapt to different stressors, especially to the metabolic stress of hypoglycemia.
Basal plasma glucagon concentrations were significantly elevated in diabetic animals compared with controls. As expected, during the euglycemic clamping period, glucagon levels gradually decreased and were significantly different from basal concentrations by the end of the clamp. The gradual decline in glucagon levels can be attributed to the suppressive effects of exogenous insulin on glucagon secretion from the pancreatic -cells (25, 26).
No differences in plasma catecholamine levels were observed between normal and diabetic animals either at baseline or during the euglycemic clamp. The variability in the responses of diabetic animals may be attributed to the fact that some of these animals reached their peak levels at an earlier time point than others. The gradual rise in plasma epinephrine levels during the clamp may be attributed to either the increase in CRH (27) or the effects of hyperinsulinemia in stimulating catecholamine secretion (28, 29). Brown et al. (27) showed that CRH injected into the brains of rats produced increases in plasma epinephrine. Kern et al. (28) and Rashid et al. (29) demonstrated that during periods of hyperinsulinemia, plasma epinephrine concentrations could rise, even in the presence of euglycemia. In the study by Rashid et al. (29), it was shown in dogs that during stress (an intracerebroventricular infusion of carbachol), when normoglycemia was maintained using a hyperinsulinemic-euglycemic glucose clamp protocol, epinephrine and norepinephrine levels increased with increasing doses of insulin, suggesting that insulin may play a role in stimulating catecholamine release. From our studies it is difficult to determine which of the two factors, CRH or insulin, predominates in stimulating epinephrine release or whether it is due to a combination of both hormones.
The expression of CRH mRNA in the hypothalamic PVN increased in both normal and diabetic animals after exposure to hyperinsulinemia, although the increase in diabetic animals was smaller (<2-fold) than that in normal animals (4-fold). Increases in hypothalamic CRH mRNA after the central administration of insulin have been reported previously (30). Although the mechanism by which insulin acts to increase CRH expression is not known, it is thought to be indirect and involve other hypothalamic areas (30). Physiologically, a large increase in plasma insulin levels may increase the synthesis and presumably the secretion of ACTH secretagogues to prepare the body for an impending fall in blood glucose. More importantly, the release of glucocorticoids may be important in increasing the body’s sensitivity to other counterregulatory hormones (31).
To our knowledge, little information is available concerning the effects of high insulin levels on hypothalamic AVP expression. High insulin concentrations increased AVP mRNA expression in both the PVN and SON of normal and diabetic rats despite plasma glucose levels being maintained at euglycemia. Although insulin receptors have been identified in a number of hypothalamic regions, the mechanism of interaction between insulin and AVP has yet to be determined.
POMC mRNA expression increased in both nondiabetic and diabetic rats after 2 h of hyperinsulinemia-euglycemia. Kim et al. (32) showed that insulin can stimulate pituitary POMC mRNA expression in STZ-diabetic animals. We have now shown that insulin can do this in the absence of hypoglycemia. Our data suggest that a large rise in plasma insulin increases all components of the HPA axis, and this mechanism appears to be centrally mediated at the hypothalamic level. This is consistent with findings that insulin increases POMC mRNA expression in corticotrophs (32). However, this effect may, again, be dose-dependent. At this point, the mechanisms underlying the actions of insulin on the HPA axis at these different doses are still unclear, and additional studies are required.
Interestingly, hippocampal MR mRNA levels rose significantly after 2 h of hyperinsulinemia-euglycemia in normal rats, but not in diabetic rats. Changes in hippocampal MR mRNA have been shown to occur as early as 30 min after the onset of a stress challenge (33). We previously suggested that a decrease in hippocampal MR mRNA expression may be required for full activation of the HPA axis in response to insulin-induced hypoglycemia (6). The increase in hippocampal MR mRNA expression in nondiabetic rats after exposure to hyperinsulinemia-euglycemia may be a mechanism to prevent additional activation of the system, because the rise in plasma insulin concentrations was not associated with a decline in plasma glucose levels. This would correlate well with what was observed for plasma ACTH and corticosterone levels. Compared with the robust pituitary-adrenal response observed during a hyperinsulinemic-hypoglycemic glucose clamp, plasma ACTH and corticosterone levels only increased modestly in the presence of euglycemia in normal animals. Conversely, the fact that hippocampal MR mRNA levels did not change in diabetic rats after 2 h of hyperinsulinemia-euglycemia may partially explain why plasma ACTH and corticosterone concentrations reached similar levels in response to different stress challenges. The absence of changes in hippocampal MR expression in diabetic animals may preclude appropriate activation of the axis during periods of stress. The data clearly demonstrate that diabetic animals exhibit impaired adaptive abilities of their central MR system in response to stress. This may play a very important role in determining appropriate HPA responsiveness to a given stress challenge.
In our studies we primarily noted changes in hippocampal MR expression during various stress challenges. Although hippocampal MRs have a similar affinity for both aldosterone and corticosterone, there is a 100- to 1000-fold greater concentration of corticosterone over aldosterone in the circulation. As such, hippocampal MRs are believed to be more responsive to changes in glucocorticoid, rather than mineralocorticoid, concentrations. Moreover, the specificity of renal MRs for aldosterone is conveyed by the expression of 11?-hydroxysteroid dehydrogenase-2, which acts to inactivate circulating glucocorticoids (34). This enzyme is not expressed in the hippocampus. Instead, another isoform, 11?-hydroxysteroid dehydrogenase-1, acts to locally amplify the glucocorticoid signal by converting the inactive glucocorticoid to its active counterpart (34). Although studies have shown that plasma aldosterone levels do increase during hypoglycemia (35, 36), to our knowledge, the effects of this increase on hippocampal MR function have not been investigated. Although this would be an interesting question to explore, we could not incorporate this parameter into our sampling protocol, because the additional blood sampling can potentially affect other hemodynamic parameters that may ultimately influence HPA output. Hence, aldosterone was not examined in our studies. With respect to electrolytes, we previously determined plasma osmolarity both under baseline conditions (0 min) and during hypoglycemia (80 and 210 min), and they revealed no differences (normal: 0 min, 289.7 ± 3.7; 80 min, 279.5 ± 4.3; 210 min, 300.2 ± 11.5; diabetics: 0 min, 298.5 ± 5.5; 80 min, 2280.8 ± 16.7; 210 min, 276.7 ± 6.4 mosmol) (our unpublished observations). In this regard, we do not believe that changes in osmolarity during stress are the cause of these changes in hippocampal MR mRNA levels.
In response to hyperinsulinemia-euglycemia, GR mRNA expression increased in the PVN and anterior pituitary of both normal and diabetic rats. Changes in GR mRNA in these areas after insulin treatment have been previously reported (3). It appears that insulin itself is capable of regulating GR mRNA expression, at least at the level of the PVN and anterior pituitary gland. We are currently unaware of any studies that have examined the mechanism by which this regulation occurs. The increase in GR mRNA expression after hyperinsulinemia-euglycemia may also contribute, at least in part, to averting additional activation of the HPA axis. Our previous study demonstrated that 4 d of insulin treatment caused an increase in GR mRNA expression in the anterior pituitary of diabetic rats and a corresponding decrease in ACTH and corticosterone. In the current study, infusion of pharmacological doses of insulin to both normal and diabetic animals revealed that insulin, given over even a short period of time, can cause an increase in GR mRNA expression in the anterior pituitary. We speculate that the increase in GR mRNA in these specific regions may aid in curtailing HPA activation, because the perceived rise in plasma insulin was not associated with a decline in plasma glucose levels. In the case of diabetic animals receiving insulin treatment, we previously suggested that the increase in pituitary GR mRNA may contribute to an increase in GR protein expression and, thus, normalization of pituitary-adrenal activity in diabetic animals. In the current study a similar argument can be made. With hyperinsulinemia-hypoglycemia, no differences in GR mRNA expression were observed, whereas hyperinsulinemia, in the absence of hypoglycemia, caused changes in GR mRNA expression. The data suggest that corticosteroid receptors as a whole may adjust their expression level in accordance with the type of stress challenge to modulate the output of the HPA axis. Because both the PVN and anterior pituitary have been shown to express insulin receptor substrate-1 (37) and insulin receptors (38, 39, 40), it is plausible that changes in insulin levels may play a role in modulating neuroendocrine function in these brain regions.
Pharmacological doses of insulin have been shown to exert vasodilatory effects, which can potentially affect local blood flow. Although in vitro studies by Zhao et al. (41) and Hasdai et al. (42) suggest that both insulin and IGF have vasodilatory effects on vascular smooth muscle, the in vivo data in both rats (43) and humans (44) suggest that the effects of insulin on cardiac output are minimal, and in the case of diabetes, that changes in vascular tone may be the result of hyperglycemia more than hypoinsulinemia. Nuutila et al. (45) showed that the insulin-induced increase in blood flow occurs much later than the effect of insulin on glucose uptake. More recently, Gudbjornsdottir et al. (46) measured the permeable surface area of insulin and glucose to assess the importance of the vasodilatory effects of insulin and noted that the permeable surface area for glucose was markedly increased by oral glucose, whereas additional vasodilation exerted by high insulin concentrations may not be physiologically relevant for capillary delivery of either glucose or insulin in muscle. In this respect, although blood flow was not measured in our studies, we do not believe that hyperinsulinemia would have altered blood flow significantly during our acute study to amend the interpretation of our results.
In conclusion, we have shown that insulin at high doses can stimulate the HPA axis both centrally and peripherally, effects that are independent of hypoglycemia. Indeed, although insulin activates pituitary-adrenal function to a similar degree in both normal and diabetic rats, the molecular mechanisms that underlie regulation of this response are different in the two groups. More importantly, in normal rats, although hyperinsulinemia per se stimulated the HPA axis, the response was much greater when hyperinsulinemia was associated with hypoglycemia or when the rats were exposed to restraint stress. In diabetic rats, the HPA responses to hyperinsulinemia, hyperinsulinemia, hypoglycemia, and restraint were equal, but were greatly diminished compared with those in normal rats. Based on these results, we propose that the peak response to stress is diminished by diabetes. Moreover, it appears that the ability to grade the HPA response to various forms of stress, which is observed in normal rats, is lost in diabetes and this may be related to defects in the corticosteroid receptor system. This latter finding may have important implications in the way patients with diabetes adapt to stressful stimuli, especially to the metabolic challenge of hypoglycemia.
Acknowledgments
We thank Elena Burdett and Debra Bilinski for their excellent technical assistance with this study.
References
Cameron OG, Kronfol Z, Grenden JF, Carroll BJ 1984 Hypothalamic-pituitary-adrenocortical activity in patients with diabetes mellitus. Arch Gen Psychiatry 41:1090–1095
Cameron OG, Thomas B, Tiongco D, Hariharan M, Greden JF 1987 Hypercortisolism in diabetes mellitus. Diabetes Care 10:662–664
Chan O, Chan S, Inouye K, Vranic M, Matthews SG 2001 Molecular regulation of the hypothalamo-pituitary-adrenal axis in streptozotocin-induced diabetes: effects of insulin treatment. Endocrinology 142:4872–4879
Roy M, Collier B, Roy A 1990 Hypothalamic-pituitary-adrenal axis dysregulation among diabetic out-patients. Psychiatry Res 31:31–37
Shapiro ET, Polonsky KS, Copinschi G, Bosson D, Tillil H, Blackman J, Lewis G, Van Cauter E 1991 Nocturnal elevation of glucose levels during fasting in noninsulin-dependent diabetes. J Clin Endocrinol Metab 72:444–454
Chan O, Chan S, Inouye K, Shum K, Matthews SG, Vranic M 2002 Diabetes impairs hypothalamo-pituitary-adrenal (HPA) responses to hypoglycemia, and insulin treatment normalizes HPA but not epinephrine responses. Diabetes 51:1681–1689
Chan O, Inouye K, Vranic M, Matthews SG 2002 Hyperactivation of the hypothalamo-pituitary-adrenocortical axis in streptozotocin-diabetes is associated with reduced stress responsiveness and decreased pituitary and adrenal sensitivity. Endocrinology 143:1761–1768
Baskin DG, Figlewicz LD, Seeley RJ, Woods SC, Porte Jr D, Schwartz MW 1999 Insulin and leptin: dual adiposity signals to the brain for the regulation of food intake and body weight. Brain Res 848:114–123
Havel PJ, Hahn TM, Sindelar DK, Baskin DG, Dallman MF, Weigle DS, Schwartz MW 2000 Effects of streptozotocin-induced diabetes and insulin treatment on the hypothalamic melanocortin system and muscle uncoupling protein 3 expression in rats. Diabetes 49:244–252
Fruehwald-Schultes B, Kern W, Bong W, Wellhoener P, Kerner W, Born J, Fehm HL, Peters A 1999 Supraphysiological hyperinsulinemia acutely increases hypothalamic-pituitary-adrenal secretory activity in humans. J Clin Endocrinol Metab 84:3041–3046
Fruehwald-Schultes B, Kern W, Born J, Fehm HL, Peters A 2001 Hyperinsulinemia causes activation of the hypothalamus-pituitary-adrenal axis in humans. Int J Obes Relat Metab Disord 25(Suppl 1):S38–S40
Shi ZQ, Rastogi KS, Lekas M, Efendic S, Drucker DJ, Vranic M 1996 Glucagon response to hypoglycemia is improved by insulin-independent restoration of normoglycemia in diabetic rats. Endocrinology 137:3193–3199
Dimitrakoudis D, Ramlal T, Rastogi S, Vranic M, Klip A 1992 Glycaemia regulates the glucose transporter number in the plasma membrane of rat skeletal muscle. Biochem J 284:341–348
Inouye K, Shum K, Chan O, Mathoo J, Matthews SG, Vranic M 2002 Effects of recurrent hyperinsulinemia with and without hypoglycemia on counterregulation in diabetic rats. Am J Physiol 282:E1369–E1379
Simpson IA, Vannucci SJ 2002 Glucose transport into brain: effects of hypoglycemia. Diabetes Nutr Metab 15:281–284
Matthews SG, Challis JR 1995 Regulation of CRH and AVP mRNA in the developing ovine hypothalamus: effects of stress and glucocorticoids. Am J Physiol 268:E1096–E1107
Paxinos G, Watson C 1997 The rat brain in stereotaxic co-ordinates. 3rd ed. San Diego: Academic Press
Lightman SL, Young III WS 1988 Corticotrophin-releasing factor, vasopressin and pro-opiomelanocortin mRNA responses to stress and opiates in the rat. J Physiol 403:511–523
Matthews SG 1998 Dynamic changes in glucocorticoid and mineralocorticoid receptor mRNA in the developing guinea pig brain. Brain Res Dev Brain Res 107:123–132
Miesfeld R, Rusconi S, Godowski PJ, Maler BA, Okret S, Wikstrom AC, Gustafsson JA, Yamamoto KR 1986 Genetic complementation of a glucocorticoid receptor deficiency by expression of cloned receptor cDNA. Cell 46:389–399
Herbert V, Lau KS, Gottlieb CW, Bleicher SJ 1965 Coated charcoal immunoassay of insulin. J Clin Endocrinol Metab 25:1375–1384
Peuler JD, Johnson GA 1977 Simultaneous single isotope radioenzymatic assay of plasma norepinephrine, epinephrine and dopamine. Life Sci 21:625–636
Scribner KA, Walker CD, Cascio CS, Dallman MF 1991 Chronic streptozotocin diabetes in rats facilitates the acute stress response without altering pituitary or adrenal responsiveness to secretagogues. Endocrinology 129:99–108
Chan O, Inouye K, Akirav EM, Park E, Riddell MC, Vranic M, Matthews SG 2002 Hyperglycemia impairs the hypothalamo-pituitary-adrenal (HPA) response to insulin-induced hypoglycemia in streptozotocin-diabetic rats. Stress 5(Suppl):27
Rastogi KS, Cooper RL, Shi ZQ, Vranic M 1997 Quantitative measurement of islet glucagon response to hypoglycemia by confocal fluorescence imaging in diabetic rats: effects of phlorizin treatment. Endocrine 7:367–375
Taborsky Jr GJ, Ahren B, Havel PJ 1998 Autonomic mediation of glucagon secretion during hypoglycemia: implications for impaired -cell responses in type 1 diabetes. Diabetes 47:995–1005
Brown MR, Fisher LA, Spiess J, Rivier C, Rivier J, Vale W 1982 Corticotropin-releasing factor: actions on the sympathetic nervous system and metabolism. Endocrinology 111:928–931
Kern W, Fittje A, Fohr W, Kerner W, Born J, Fehm HL 2000 Increase in systolic blood pressure and catecholamine level during hyperinsulinemia in a placebo-controlled euglycemic clamp in healthy subjects. Exp Clin Endocrinol Diabetes 108:498–505
Rashid S, Shi ZQ, Niwa M, Mathoo JM, Vandelangeryt ML, Bilinski D, Lewis GF, Vranic M 2000 ?-Blockade, but not normoglycemia or hyperinsulinemia, markedly diminishes stress-induced hyperglycemia in diabetic dogs. Diabetes 49:253–262
Sipols AJ, Baskin DG, Schwartz MW 1995 Effect of intracerebroventricular insulin infusion on diabetic hyperphagia and hypothalamic neuropeptide gene expression. Diabetes 44:147–151
Dirlewanger M, Schneiter PH, Paquot N, Jequier E, Rey V, Tappy L 2000 Effects of glucocorticoids on hepatic sensitivity to insulin and glucagon in man. Clin Nutr 19:29–34
Kim EM, Grace MK, Welch CC, Billington CJ, Levine AS 1999 STZ-induced diabetes decreases and insulin normalizes POMC mRNA in arcuate nucleus and pituitary in rats. Am J Physiol 276:R1320–R1326
Fujikawa T, Soya H, Fukuoka H, Alam KS, Yoshizato H, McEwen BS, Nakashima K 2000 A biphasic regulation of receptor mRNA expressions for growth hormone, glucocorticoid and mineralocorticoid in the rat dentate gyrus during acute stress. Brain Res 874:186–193
De Kloet ER, Vreugdenhil E, Oitzl MS, Joels M 1998 Brain corticosteroid receptor balance in health and disease. Endocr Rev 19:269–301
Hochberg Z, Dickerman Z, Kaufman H, Laron Z 1980 Evaluation of the renin-aldosterone system during hypo- and hyperglycemia in children and adolescents. Horm Res 12:16–21
Ramirez G, Ganguly A, Bruegemeyer C, Gomez-Sanchez C 1989 Aldosterone response to insulin-induced hypoglycemia in hemodialysis patients. Am J Nephrol 9:215–221
Baskin DG, Sipols AJ, Schwartz MW, White MF 1993 Immunocytochemical detection of insulin receptor substrate-1 (IRS-1) in rat brain: colocalization with phosphotyrosine. Regul Pept 48:257–266
Yamashita S, Melmed S 1986 Effects of insulin on rat anterior pituitary cells. Inhibition of growth hormone secretion and mRNA levels. Diabetes 35:440–447
Unger JW, Lange W 1997 Insulin receptors in the pituitary gland: morphological evidence for influence on opioid peptide-synthesizing cells. Cell Tissue Res 288:471–483
Unger JW, Betz M 1998 Insulin receptors and signal transduction proteins in the hypothalamo-hypophyseal system: a review on morphological findings and functional implications. Histol Histopathol 13:1215–1224
Zhao G, Sutliff RL, Weber CS, Wang J, Lorenz J, Paul RJ, Fagin JA 2001 Smooth muscle-targeted overexpression of insulin-like growth factor I results in enhanced vascular contractility. Endocrinology 142:623–632
Hasdai D, Rizza RA, Holmes Jr DR, Richardson DM, Cohen P, Lerman A 1998 Insulin and insulin-like growth factor-I cause coronary vasorelaxation in vitro. Hypertension 32:228–234
Hilzenrat N, Yaari A, Maislos M, Sikuler E 2001 Splanchnic and systemic hemodynamic effects of sustained euglycemic hyperinsulinemia in rats. Scand J Clin Lab Invest 61:383–388
Williams SB, Goldfine AB, Timimi FK, Ting HH, Roddy MA, Simonson DC, Creager MA 1998 Acute hyperglycemia attenuates endothelium-dependent vasodilation in humans in vivo. Circulation 97:1695–1701
Nuutila P, Raitakari M, Laine H, Kirvela O, Takala T, Utriainen T, Makimattila S, Pitkanen OP, Ruotsalainen U, Iida H, Knuuti J, Yki-Jarvinen H 1996 Role of blood flow in regulating insulin-stimulated glucose uptake in humans. Studies using bradykinin, [15O]water, and [18F]fluoro-deoxy-glucose and positron emission tomography. J Clin Invest 97:1741–1747
Gudbjornsdottir S, Sjostrand M, Strindberg L, Wahren J, Lonnroth P 2003 Direct measurements of the permeability surface area for insulin and glucose in human skeletal muscle. J Clin Endocrinol Metab 88:4559–4564(Owen Chan, Karen Inouye, )