Background
Cushing’s syndrome (CS) can be divided into adrenocorticotropin (ACTH)-dependent and ACTH-independent disease [
1]. The latter can in rare cases result from ACTH-independent macronodular adrenocortical hyperplasia (AIMAH). AIMAH is characterized by multiple bilateral nodules consisting of hyperplastic adrenal cells that lead to non-ACTH-dependent (over)production of cortisol.
The hypercortisolism in AIMAH patients is caused by an exaggerated or ectopic response to stimulation by hormonal signals. Receptors for these hormones, eutopically or ectopically expressed on adrenocortical cells and activated by endogenous hormones, stimulate intracellular pathways leading to (sub-)clinical CS [
2]. Aberrant hormonal responses and the presence of hormone receptors in AIMAH have been well documented for glucose-dependent insulinotropic polypeptide (GIP) receptor [
3‐
5], α4-, β1- and β2-adrenergic receptor (AR) [
6,
7], arginine-vasopressin (AVP) type 1a and 2 receptors (AVPR1A/2) [
8,
9], luteinizing hormone receptor (LHR) [
10,
11] and serotonin (5-HT) type 4 receptor (5-HT4R) [
12,
13]. Other possible aberrantly expressed receptors include the AngII type I receptor (AT1R) [
14] and glucagon receptor [
15,
16].
Diagnostic protocols for AIMAH include administration of the various hormones or stimuli to patients [
2]. The relevance of these effects has been shown by the results of administration of antagonists to the expressed receptors [
6,
10,
17], confirmation of stimulatory effects of hormones on primary AIMAH cells
in vitro and/or the detection of the hormone receptors on AIMAH cells. Furthermore, these effects have been described in adrenocortical adenomas and carcinomas, which can also overexpress receptors responsive to endocrine and/or paracrine signals [
17].
Whereas the presence of aberrant receptors has been firmly established, little is known on the cause of such altered receptor expression and on downstream signals coupling receptor activation to stimulation of cell growth and steroidogenesis [
18,
19]. Although most patients present sporadically, some AIMAH cases occur in families [
20], suggesting a possible genetic background. Investigations in murine models have shown that adrenal overexpression of GIP [
21] and LH receptors [
22] can lead to AIMAH and adrenal CS. This implies that the aberrant expression of these GPCRs may play a role in adrenal cell hyperplasia apart from modulating cortisol production. Most evidence on AIMAH pathophysiology in humans has been collected from case reports, small case series and reviews; only three centers have reported on larger groups of 16, 18 and 32 AIMAH patients, respectively [
7,
16,
19]. Furthermore, no large series have systematically related clinical data to corresponding
in vitro findings.
The current study evaluated the presence of aberrant hormonal responses in patients with AIMAH in order to discover novel pathways involved in the pathophysiology of this rare disease. We performed in vivo and in vitro stimulation tests with ACTH and multiple ligands for hormone receptors in the largest group of AIMAH patients described so far.
Discussion
AIMAH is a rare disease associated with the presence of aberrantly expressed eutopic and ectopic receptors on adrenocortical cells [
2]. In the current study, ACTH still formed the most potent stimulus for steroidogenesis in AIMAH. ACTH stimulated cortisol production
in vivo and
in vitro in almost all patients. One patient showed only 10% increase in cortisol levels following ACTH. Unfortunately, repetitive testing was not available and incorrect administration or sample handling cannot be excluded in this case. The induction of cortisol by ACTH was comparable in primary cultures of AIMAH and non-AIMAH origin. We have previously demonstrated that MC2R levels, which are controlled by ACTH signaling through cAMP [
27,
28], are equal in AIMAH and other adrenal tissues [
29]. Although plasma ACTH levels are low or undetectable in AIMAH patients, other GPCR-coupled pathways in AIMAH cells could stimulate cAMP formation and thereby ensure MC2R expression.
The most prevalent exaggerated responses to hormonal stimulation in AIMAH patients were to AVP and upright posture, confirming previous findings [
16]. Several studies have now demonstrated the presence of the eutopically expressed V
1 receptor in AIMAH tissues [
9,
30]. The (near) undetectable levels of V
2 and V
3 receptors combined with the absence of effects of desmopressin, a selective V
2 agonist, plead against significant roles of these latter receptors in AIMAH. Although no overall significant differences in
AVPR1A mRNA levels between AIMAH and other adrenal cells were detected [
9,
30], individual patients could still overexpress the V
1 receptor [
31‐
33]. AVP induced cortisol and mRNA expression of StAR and four steroidogenic enzymes uniformly in AIMAH and non-AIMAH cells, suggesting that AVP has a physiological effect on adrenocortical steroidogenesis. However, the selective stimulation of CYP11B1, a key enzyme in cortisol synthesis, in AIMAH by AVP indicates a novel molecular mechanism underlying the coupling between the V
1 receptor and steroidogenesis in AIMAH. This suggestion is supported by the association between
AVPR1A levels and the AVP-induced mRNA expression of
CYP11B1. This correlation between receptor levels and cortisol stimulation is not present for MC2R and ACTH [
29]. More efficient coupling of the V1 receptor to 11β-hydroxylase could form a cause of AIMAH that is characterized by normal GPCR levels. These findings in AIMAH might open up new opportunities for medical treatment with selective V
1 receptor antagonists [
34]. Adrenocortical carcinomas seem to have an impaired response to AVP, possibly due to decreased expression of the type V
1 receptor.
The upright posture test was positive in half of the patients studied
in vivo. Of the 13 patients with a positive response to upright posture, 8 reacted also positively to AVP administration. The other 5 patients could have reacted aberrantly to surges in catecholamines or AngII [
2]. Interestingly, in the short-term experiments we found a difference in responsiveness to catecholamines between AIMAH and non-AIMAH cells. In contrast to the observation with AVP, this would suggest the presence of ectopic adrenergic receptors. Previous studies revealed that this could be related to adrenocortical expression of β
1-, β
2- or α
4-adrenergic receptors [
6,
7].
The type 4 serotonin receptor (5HT4R) is expressed in the adrenal gland and its activation can affect cortisol production [
35,
36]. In the present study, 34% of patients had a >25% stimulation of serum cortisol levels following metoclopramide, an agonist of 5-HT4R, which was less than the 56% observed in the previous study [
16].
In vitro, this response was also found in AIMAH samples as well as in controls. Moreover, there was no significant difference in response between AIMAH and non-AIMAH cells
in vitro, making it questionable whether the response to 5-HT in AIMAH patients is truly aberrant.
Other hormonal stimuli can lead to a stimulation of serum cortisol in a minority of AIMAH cases. For LHRH, TRH, glucagon and GIP we have found that this is the case in 10-22% of patients. Adrenal LHR or GIPR expression was previously confirmed in 6 patients [
5,
11,
33,
37,
38]. The association between cortisol induction following ACTH and after the standard mixed meal could be caused by food-induced stimulation of ACTH [
39,
40]. Unfortunately, we have no data on postprandial ACTH levels in these patients. The inverse correlation between morning cortisol and cortisol induction by a mixed meal reflects higher cortisol increments in patients with low basal cortisol levels. The cause of this finding is unknown, but might reflect higher adrenal sensitivity to GIP at lower cortisol levels.
In vitro, we found no major effects of the other hormonal stimuli on cortisol production besides in individual cultures.
The clinical description of AIMAH cases has often been coupled with
in vitro investigations on patient tissue samples. Several AIMAH patients with discrepancies between clinical and experimental hormonal responses have been described [
41,
42], which is in contrast to the majority of AIMAH studies in which identical
in vitro and
in vivo hormonal effects are obtained (reviewed in [
2]). We have evaluated
in vitro responses in 17 primary cultures of AIMAH tissues and detected a poor overall correlation between clinical and experimental responses to individual stimuli. Possible causes include publication bias of those patients in whom effects could be replicated
in vitro, hormonal effects through the pituitary, the concentrations of stimuli used
in vitro or the experimental set-up. Since ACTH levels were currently only measured in the
in vivo response tests relating to AVP, effects of the other hormonal stimuli through pituitary secretion of ACTH cannot be excluded and might also cause dissociation between clinical and experimental effects. With respect to the latter cause, we also found clear differences and an overall lack of association between short-term and long-term effects of the stimuli on cortisol concentrations. These conclusions should however be drawn with caution due to the small sample sizes of cultures in some of the short- and long-term experiments. Besides the rarity of the disease, the low percentage of patients being operated because of clinical disease and the multitude of testable hormonal stimuli hamper large
in vitro studies with all possible secretagogues in AIMAH patients. Also in this study limited availability of cells did not allow for systematic
in vitro evaluation of all possible stimuli.
Responses to the aberrant hormonal stimuli were not associated with
in vivo hormonal activity since patients with clinical and subclinical CS showed comparable results to stimuli. It should be emphasized that the definition of subclinical CS is controversial as some clinical symptoms of CS can be recognized in patients with subtle cortisol overproduction. In addition, no cut-off values of diagnostic tests have been established that define subclinical hypercortisolism. The findings on
in vivo responsiveness to hormonal stimuli are a confirmation of an earlier study [
16], in which it was postulated that clinical and subclinical CS represent a continuum of disease rather than two separate entities. This hypothesis is supported by the smaller adrenal sizes detected in subclinical CS patients; these patients could progress into clinical CS when adrenal size and coupled steroidogenic capacity increase.
Exaggerated responses to hormonal stimuli have been reported for multiple cases and case series in adrenal hyperplasia and adenomas, although the responses in healthy individuals have not been investigated for all hormones. It is therefore uncertain to what extent
in vivo responses are the results of aberrant expression patterns of hormone receptors. Previous [
36,
43] and current
in vitro data obtained in the non-AIMAH tissues suggest significant effects of AVP and possibly 5-HT. In our analysis we found no differences in responsiveness to hormonal stimuli among ACTH-dependent hyperplasia, adrenal adenoma and carcinoma samples, although numbers were small. Systematic
in vivo comparisons for responses between healthy individuals and AIMAH patients might improve the definition of an aberrant response. The arbitrarily used criterion of >50% elevation in serum cortisol may also be modified if effects in healthy individuals would be identified.
Competing interests
The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.
Authors’ contributions
JH, LJH, PMvK & JS performed the experiments. JH, LJH, PMvK, RRdK, FHvN, CvE, WWdH, MvA, JWdG, TPL and RAF were involved in patient care and/or tissue collection. JH, LJH, WWdH, FHdJ and RAF designed the study. All authors had the opportunity to revise the manuscript and agree to its final content.