| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Studies |
Department of Medicine, Neuroendocrine Research Unit, University of Queensland, Queensland, Australia
Address all correspondence and requests for reprints to: Associate Professor Richard V. Jackson, Neuroendocrine Research Unit, University Department of Medicine, Greenslopes Private Hospital, Greenslopes, Brisbane, Queensland 4120, Australia.
| Abstract |
|---|
|
|
|---|
Aspirin significantly reduced the cortisol response to AVP [mean peak increase from basal, 221.1 ± 20.1 vs. 165.4 ± 22.5 nmol/L (P = 0.0456); mean integrated response, 11,199.3 ± 1,560.0 vs. 6,162.3 ± 1,398.6 nmol·min/L (P = 0.0116) for placebo aspirin/AVP and aspirin/AVP, respectively]. The ACTH response was reduced, but did not reach statistical significance [mean peak increase from basal, 7.5 ± 2.2 vs. 4.3 ± 0.3 pmol/L (P = 0.0563); mean integrated response, 142.6 ± 36.0 vs. 96.2 ± 8.7 pmol·min/L (P = 0.12) for placebo aspirin/AVP and aspirin/AVP, respectively].
PGs may influence ACTH secretion by being stimulatory or inhibitory to the HPA axis at different levels, such as hypothalamic or pituitary. Which effect predominates in vivo during dynamic activation of the axis may depend on the level at which the secretory stimulus acts. We showed that when normal male volunteers were treated with the PG synthesis inhibitor, aspirin, they had a blunted HPA axis response to the pituitary corticotroph stimulator, AVP.
| Introduction |
|---|
|
|
|---|
In vitro studies demonstrate that PGs influence ACTH secretion. PGE2 inhibits ACTH release in response to arginine vasopressin (AVP) and CRH from rat pituitary cells (1, 7, 8). Conversely, cyclooxygenase inhibitors (indomethacin, diclofenac, and flurbiprofen) enhance ACTH immunoreactivity after stimulation by AVP or ovine CRH in rat pituitary quarters incubated in vitro (7).
We recently reported augmentation of ACTH secretion in response to naloxone stimulation after the prior administration of aspirin to normal human subjects (9). Naloxone blocks an inhibitory opioidergic tone on central noradrenergic pathways, thus enabling CRH release: this mechanism of action is supported by several animal and human studies (10, 11, 12, 13, 14, 15). Aspirin inhibits the enzyme PG synthetase, which catalyzes the first two reactions in the cyclooxygenase pathway (16). Aspirin has no direct effect on human adrenal steroidogenesis or basal secretion of ACTH or cortisol (9, 17). CRH and AVP are the two major hypothalamic hormones that regulate ACTH secretion via distinct corticotroph receptors (18, 19). AVP is a nonapeptide coexpressed with CRH in the parvicellular neurons of the hypothalamic paraventricular nucleus (18, 20, 21). Potentiation occurs between these two ACTH secretagogues, resulting in an important synergistic response (11, 18, 21, 22). Hypoglycemia in humans stimulates catecholamine secretion and concurrent release of CRH and AVP from hypothalamic neurons (23). Inconsistent results have been obtained from previous human studies using indomethacin and sodium or acetyl salicylate in conjunction with insulin-induced hypoglycemia as the stimulus for ACTH release (17, 24, 25).
In this placebo-controlled study in seven normal volunteers, we examined the effects of a single, clinically relevant dose of aspirin on ACTH and cortisol responses to exogenous AVP. The methodology used was similar to that of our previous naloxone study (9). This is the first reported human study to examine the effect of cyclooxygenase inhibition on AVP-stimulated ACTH and cortisol release.
| Subjects and Methods |
|---|
|
|
|---|
Seven healthy male volunteers, aged 1830 yr (mean, 24 yr), participated in this study. All had normal medical histories, physical examinations, serum chemistries, static anterior pituitary function tests, full blood counts, electrocardiograms, and urinalysis. Each subject underwent four separate tests, separated by at least 1 week: AVP alone (P/AVP), aspirin alone (ASP/P), the combination of aspirin and AVP (ASP/AVP), and the combination of placebo drugs (P/P).
Study design
All studies were randomized, placebo controlled, and single blinded and were performed in the midafternoon, a time of low basal secretory activity for ACTH and cortisol. The subjects, all nonsmokers, had been free of medication for at least 4 weeks and abstained from caffeine and alcohol for at least 24 h before each study. They were fasted for 3 h before insertion of a forearm venous cannula with a three-way tap to allow blood sampling and volume replacement with isotonic saline. Before each test, either two 300-mg soluble aspirin tablets (Disprin, Rickitt and Colman, West Ryde, Australia) or two placebo tablets of similar appearance were given at -225 min, i.e. 45 min before cannula insertion. Basal sampling was commenced at -45 min and continued at 15-min intervals until the administration of iv AVP (0.0143 IU/kg over 90 s; aqueous Pitressin, Parke-Davis, Carringbah, Australia) or normal saline as placebo at 0 min. An additional nine blood samples were collected over the next 2 h.
Blood pressure (BP) and heart rate (HR) were recorded every 15 min by a Dinamap (Critikon, Tampa, FL) automatic monitor, and an observer who noted the subjects condition and the occurrence of any side-effects was in attendance throughout each test. The study protocols were approved by the University of Queensland human ethics committee and the Greenslopes Hospital human ethics committee.
Hormone assays
ACTH was measured in unextracted plasma by RIA (9), using anticorticotropin serum, IgG-ACTH-1 (IgG Corp., Nashville, TN), which is directed at the ACTH-(5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18) sequence. All assays were performed in duplicate. The routine detectable concentration of ACTH was 1.1 pmol/L, and the inter- and intraassay coefficients of variations were 7.8% and 3.7%, respectively, at 7.7 pmol/L (n = 7).
Cortisol was measured in extracted plasma by a previously described high pressure liquid chromatography method (9), using prednisolone as the internal standard, and extraction with ether-dichloromethane (60:40). All assays were performed in duplicate. Recovery was 95100% for cortisol and prednisolone. The limit of detection of the assay was 30 nmol/L. The inter- and intraassay coefficients of variation at 165 nmol/L were 6.2% and 4.5%, respectively (n = 7).
Plasma AVP levels were measured by RIA using a double antibody technique as previously described (11).
Statistical methods
Results are expressed as the mean ± SE. Statistical significance was taken as P < 0.05. Comparisons of mean hormone levels between and within tests were performed by between- and within-subject ANOVA with repeated measures and Duncans multiple range test. Derived parameters [mean peak change and mean integrated area under the hormone-time curve (AUC)] were compared by two-way ANOVA with planned comparisons. The BP and HR data were analyzed by ANOVA. All statistical analyses were performed on the software package LSS (Statsoft, Tulsa, OK).
| Results |
|---|
|
|
|---|
Significant increases above basal (at 0 min) in ACTH occurred
within the P/AVP and ASP/AVP tests after AVP administration, but not in
the ASP/P and P/P tests. Cortisol increased significantly above the
basal level within the P/AVP and ASP/AVP tests, but not in the P/P
tests. In the ASP/P tests, cortisol decreased significantly from the
basal value at 75 min (P = 0.0343), reflecting normal
circadian changes with low cortisol secretion in the late afternoon. In
the P/AVP and ASP/AVP tests the elevated plasma ACTH levels were
maintained until 30 min, then returned to the basal level. The plasma
cortisol levels for AVP alone remained elevated until 120 min; however,
for ASP/AVP, they returned to basal by 45 min. Mean peak plasma AVP
levels in the ASP/AVP (299.3 ± 33.6 pmol/L) and P/AVP (282.7
± 15.2 pmol/L) were considerably greater than the threshold for ACTH
stimulation (135 pmol/L) previously reported in normal humans (24, 25).
There were no significant differences between these two tests with
respect to the AVP levels (P = 0.53 for mean peak
level; P = 0.97 for AUC; Fig. 1
).
|
A highly significant test by time interaction effect occurred for both ACTH (F = 15.58; df 36,376; P < 0.0001) and cortisol (F = 9.49; df 36,264; P < 0.0001) in the overall hormone responses to the four tests. However, no statistically significant differences in ACTH or cortisol levels between the P/AVP and ASP/AVP tests occurred at any individual time point. Statistically significant differences in cortisol levels occurred between the ASP/P and P/P tests (from 4590 min). A high mean basal cortisol level in the P/P tests was due to one of the subjects having a disproportionately high basal cortisol level on the test day. There also was a decrease from basal in the ASP/P test over a similar time period, reflecting circadian change in cortisol secretion.
The mean peak ACTH change from basal (at 0 min) was reduced by
prior administration of aspirin; however, this did not quite reach
statistical significance (P = 0.0563). The mean
integrated AUC in response to AVP did not change significantly with
aspirin administration (P = 0.12). In contrast,
statistically significant reductions occurred in the corresponding
cortisol parameters with the addition of aspirin to AVP
(P = 0.0456 and P = 0.0116,
respectively; Fig. 2
).
|
| Discussion |
|---|
|
|
|---|
We recently reported that the ACTH and cortisol responses to naloxone (acting via CRH release) were augmented by the prior administration of the same dose of aspirin as that used in this study (9). AVP stimulates corticotrophs directly via specific V3 receptors providing a pharmacologically distinct stimulus (18). In the current study we have clearly demonstrated a different and opposite effect from that obtained when aspirin is given before naloxone. Insulin-induced hypoglycemia has been the main technique used to stimulate ACTH release in previous studies examining the effect of cyclooxygenase inhibitors on the human HPA axis in vivo (17, 24, 25).
An iv infusion of sodium salicylate (40 mg/min for 2 h) significantly increased both ACTH and cortisol responses to hypoglycemia (24). However, 4 days of pretreatment with sustained release acetylsalicylic acid (aspirin) in a dose of 800 mg, four times a day, resulted in reduced ACTH and delayed cortisol responses (25). Oral indomethacin in a dose of 300 mg daily for 4 days delayed the onset of the ACTH response, but increased its magnitude and blunted the cortisol response (25). Lower doses of oral indomethacin (25 mg, four times a day for 13 days) delayed and decreased the ACTH response, but did not alter the cortisol response (17). PGE2 infusion (10 µg/min) potentiates ACTH secretion in response to CRH stimulation in normal men (28). This contrasts with the effect of aspirin pretreatment on naloxone stimulation, where augmentation of the ACTH response occurred (9). We also studied the effect of oral aspirin on naloxone-induced ACTH secretion in patients with myotonic dystrophy (DM), a multisystem disorder with an identified abnormal gene thought to encode a cAMP-dependent protein kinase (29). Patients with DM have hyperresponsive ACTH secretion in response to naloxone alone (30), which is paradoxically inhibited by prior aspirin administration, in contrast to the augmented response seen in normal humans (9, 31).
There are a number of potential explanations for the conflicting results obtained by the reported in vivo human studies. The stimuli used to elicit ACTH secretion are variable; insulin-induced hypoglycemia, CRH, AVP, and naloxone have been used. The level of the HPA axis at which these stimuli act to elicit corticotroph ACTH secretion may alter the effect of administered PGs or cyclooxygenase inhibitors. The response of corticotroph subpopulations to ACTH secretagogues and specific intracellular pathways involved in signal transmission also may have distinct and different interactions with PGs.
The varying results obtained from studies using insulin-induced hypoglycemia may reflect the recruitment of many different ACTH secretagogues by this powerful stress stimulus. In normal men, hypoglycemia elicits significant increases in plasma ACTH and cortisol, with concomitant increases in plasma AVP, CRH, adrenaline, and noradrenaline (23). In conscious, unrestrained sheep, relative levels of CRH and AVP in hypophysial portal blood varied with the degree of hypoglycemia achieved after insulin infusion (32). Naloxone elicits CRH release by blocking an inhibitory opioidergic tone on central noradrenergic pathways (10, 11). AVP and CRH stimulate corticotrophs directly via distinct receptors (18, 19). PGs may have different interactions with CRH- and AVP-mediated corticotroph stimulation, and the variable effect of cyclooxygenase inhibitors on hypoglycemia-induced HPA activation may reflect the various amounts of hypothalamic CRH and AVP secreted in response to this potent stimulus.
Data from in vitro studies support the hypothesis that PGs have distinct secretagogue-dependent interactions. PGE2 potentiates AVP-stimulated ACTH secretion from cultured fetal sheep pituitary cells (33). However, studies using rat anterior pituitary fragments have shown the opposite effect (1, 7). Neurotransmitter regulation of the HPA axis may be species specific (34), with differing responses to the acute administration of opioids and varying potency of the two main secretagogues, CRH and AVP (18, 35, 36). Hence, the role played by PGs in modulating ACTH secretion in response to different exogenous secretagogues may change depending on the species studied. Our current study demonstrates reduced HPA axis activation when aspirin is administered before an iv bolus of AVP, indicating that endogenous pituitary PGs in humans may potentiate AVP-induced ACTH secretion, as observed in sheep corticotrophs (33).
Endogenous PGs may play a dual role in the regulation of the HPA axis, with different effects on the hypothalamus and anterior pituitary. In vivo studies of both anesthetized and conscious, freely moving rats suggest that endogenous PGs stimulate hypothalamic CRH release and inhibit ACTH secretion from the pituitary (37, 38, 39, 40, 41, 42). The inhibitory effect of pituitary PGs may predominate in vivo (43). The augmented ACTH response when aspirin is administered before naloxone is consistent with blockade of an inhibitory effect of pituitary PGs in ACTH secretion. Alternatively, in humans, PGs may be inhibitory at the hypothalamic level and stimulatory on corticotrophs. This would explain the opposite responses we have demonstrated using naloxone (acting via hypothalamic CRH) and AVP (providing direct corticotroph stimulation).
PGE2 infusion potentiates ACTH secretion in response to CRH stimulation in normal men (28). This could be explained by increased corticotroph sensitivity to CRH, perhaps by up-regulation of receptors or alterations in local inhibitory factors (36). Alternatively, PGE2 may increase corticotroph sensitivity to basal AVP secretion, resulting in an increased synergistic response when CRH is administered. Subpopulations of corticotrophs may have different interactions with administered PGs. Studies performed on individual dispersed rat pituitary cells using the reverse hemolytic plaque assay have demonstrated functionally distinct classes of corticotrophs that respond, via an unknown mechanism, differentially to CRH and AVP (44). The effects of PGs, either stimulatory or inhibitory, could be influenced by secretagogue activation of specific corticotroph populations in the normal human pituitary.
CRH and AVP mediate ACTH release via different second messenger systems (45). CRH activates G protein-linked adenylate cyclase, leading to cAMP formation and protein kinase A activation, thus stimulating extracellular calcium influx via L-type voltage-sensitive channels (45, 46). Currently identified prostanoid receptors are also coupled to G proteins (47). Their second messenger systems predominantly involve alterations in adenylate cyclase activity and inositol trisphosphate-independent changes in intracellular calcium concentrations (47). A recent study demonstrated a dose-dependent stimulation of cAMP and PGE2 levels in pregnant myometrial membranes in response to human CRH, suggesting a significant linkage of the CRH receptor to both adenylate cyclase and the cyclooxygenase pathway (48). AVP activates phospholipase C, generating inositol trisphosphate and diaglycerol, hence mobilizing intracellular calcium and activating protein kinase C-induced extracellular calcium influx (45, 49). PGs may have specific interactions with these intracellular pathways of signal transmission. Patients with DM have an abnormal gene predicted to encode a cAMP-dependent protein kinase. We have previously reported significantly increased ACTH responses to exogenous and endogenous CRH stimulation in these patients (30, 50, 51). The ACTH response to exogenous AVP is delayed, but of normal magnitude (52). In comparison with normal subjects, administration of aspirin before naloxone stimulation in DM patients results in paradoxical inhibition of ACTH secretion (31). We propose that in DM, the function of the cAMP-dependent protein kinase A second messenger system is defective, resulting in an abnormal effect of cyclooxygenase inhibition on naloxone-stimulated ACTH release (31). This is consistent with the hypothesis that distinct PG interactions occur with cAMP-dependent and independent signal transduction pathways.
Aspirin is a commonly and increasingly prescribed drug. We have previously demonstrated a potentiating effect of PG inhibition by aspirin on naloxone-induced ACTH and cortisol release (9). In this study we have shown that acute administration of aspirin to normal humans results in decreased HPA axis activation by exogenous AVP. In these dynamic physiological studies we are unable to determine why PG inhibition affects CRH- and AVP-stimulated HPA axis activation in opposite ways. It may be due to the level of the HPA axis at which the stimulus is applied (hypothalamic vs. pituitary) and/or the nature of the corticotroph response to different secretagogues. Complex intracellular signal transmission with cAMP-dependent vs. cAMP-independent pathways may also be involved. Potentially important clinical questions arise from these data. No human studies have examined the effect of chronic aspirin therapy (such as in low dose cardiac/stroke prophylaxis) on ACTH and cortisol secretion in response to pharmacologically distinct stimuli. This is an exciting and important area for future research.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Current address: Developmental Endocrinology Branch, National
Institute of Child Health and Human Development, National Institutes of
Health, Building 10, Room 10N262, 9000 Rockville Pike, Bethesda,
Maryland 20892-1862. ![]()
Received June 12, 1996.
Revised November 8, 1996.
Accepted November 20, 1996.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. D. Snoep, M. M.C. Hovens, S. M. Pasha, M. Frolich, H. Pijl, J. T. Tamsma, and M. V. Huisman Time-Dependent Effects of Low-Dose Aspirin on Plasma Renin Activity, Aldosterone, Cortisol, and Catecholamines Hypertension, November 1, 2009; 54(5): 1136 - 1142. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Gravel and M. M. Vijayan Salicylate Disrupts Interrenal Steroidogenesis and Brain Glucocorticoid Receptor Expression in Rainbow Trout Toxicol. Sci., September 1, 2006; 93(1): 41 - 49. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. H. Messerli Aspirin: A Novel Antihypertensive Drug?: Or Two Birds With One Stone? J. Am. Coll. Cardiol., September 20, 2005; 46(6): 984 - 985. [Full Text] [PDF] |
||||
![]() |
R. D. Van Anholt, F. A. T. Spanings, W. M. Koven, O. Nixon, and S. E. W. Bonga Arachidonic acid reduces the stress response of gilthead seabream Sparus aurata L. J. Exp. Biol., September 1, 2004; 207(19): 3419 - 3430. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. D. van Anholt, T. Spanings, W. Koven, and S. E. W. Bonga Effects of acetylsalicylic acid treatment on thyroid hormones, prolactins, and the stress response of tilapia (Oreochromis mossambicus) Am J Physiol Regulatory Integrative Comp Physiol, November 1, 2003; 285(5): R1098 - R1106. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Eldar-Geva, C. Wood, N. Lolatgis, L. Rombauts, G. Kovacs, J. Fuscaldo, and A.O. Trounson Cumulative pregnancy and live birth rates in women with antiphospholipid antibodies undergoing assisted reproduction Hum. Reprod., June 1, 1999; 14(6): 1461 - 1466. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |