In this issue of the Journal, Tatsi et al. from the Stratakis laboratory at the National Institutes of Health (NIH) report a novel mutation in NR3C1, the gene encoding the glucocorticoid receptor (GR), causing severe glucocorticoid resistance. The patient in question is a 17-year-old young woman, who had initially presented at 2 years and 10 months with hypertension, and 24-h urinary cortisol more than 30-fold higher and plasma adrenocorticotrophic hormone (ACTH) seven-fold higher than the upper limits of normal. She was initially treated with labetalol, spironolactone and potassium supplementation; given the clinical diagnosis of glucocorticoid resistance syndrome (GRS) dexamethasone (1 mg/day) was added at the age of 4 years, and was subsequently adjusted to control blood pressure (BP). Her case was initially reported when she was eight, as a then novel mutation in helix 10 of the glucocorticoid receptor .
Her initial presentation at the NIH was at the age of 11 years and 9 months, when her dexamethasone dosage had reached 12 mg/day. On computerized tomography she was shown to have bilateral adrenal hyperplasia and near complete epiphyseal fusion. Her dexamethasone dose was increased to 14 mg/day, providing better control of BP, improvement of acne and hirsutism, and normalization of her menses. At her latest follow-up, her plasma ACTH and morning cortisol levels were within the normal range; in addition, her leucocytes were subjected to whole exome sequencing. This level of sequencing, unavailable to the previous investigators, provided the key finding of the day. The patient proved to be a complex heterozygote, with two different mutations in NR3C1, with the previously described  p.R714Q mutation on one DNA strand, and on the other a novel mutation in exon 2, p.E198X. The novel mutation was also present in the mother, who is clinically normal, and the other variant is described as de novo; given that both parents’ DNA was also sequenced, paternity was presumably confirmed.
Given the well deserved status of the NIH, and the clear and unambiguous nature of the report, no commentary could be a critique: what it needs to do is to pose questions about GRS, the assumptions underlying its treatment. It may also be useful to recall the (patho) physiology of cortisol and its two receptors, GR and mineralocorticoid receptors. Questions might also be posed about the accepted roles of dexamethasone. Aldosterone is commonly considered the ‘cognate’ receptor for mineralocorticoid receptors (which it is not, unless ‘cognate’ is confined to epithelia), and dexamethasone the ‘cognate’ receptor for GR, which must await its endogenous production.
With progesterone receptors and androgen receptors (AR) mineralocorticoid receptors and GR form a closely related subfamily in the 48 (human) or 49 (mouse) family of nuclear receptors. A primordial corticoid receptor is the ancestor of the mineralocorticoid receptor/GR/progesterone receptor/AR subfamily, of which the mineralocorticoid receptor is the first to branch off , followed by the GR/progesterone receptor/AR. Mineralocorticoid receptors are expressed in a variety of organs in cartilaginous and bony fish, well before aldosterone synthase which marked the transition from obligate aqueous terrestrial animals. The ‘cognate’ ligand for aqueous mineralocorticoid receptors is contested – cortisol, deoxycorticosterone (DOC), progesterone, 19-norprogesterone – with cortisol probably the least unlikely. Human mineralocorticoid receptors have equivalent high affinity for cortisol, corticosterone, DOC, progesterone and aldosterone. Physiologically, aldosterone is the privileged agonist, despite its very low concentration, in epithelia, vascular wall and the nucleus tractus solitarius. In these tissues mineralocorticoid receptors are ‘protected’ by coexpression of the enzyme 11-beta hydroxysteroid dehydrogenase II, which converts cortisol to receptor-inactive cortisone.
What is not well recognized are two facets of glucocorticoid physiology – first, that more than 90% of human mineralocorticoid receptors are constitutively occupied but not ‘activated’ by cortisol, and secondly that the affinity of cortisol for mineralocorticoid receptors is more than 10-fold higher than for GR; an unusual, but potentially useful, way of looking at things is that mineralocorticoid receptors are ‘high affinity glucocorticoid receptors’. There is an unspoken convention that high affinity is good, and somehow low affinity is inferior. In fact, it's horses for courses: if progesterone receptors had the same low affinity for progesterone as acetylcholine receptors, the placenta would need to be the size of a 20-ft3 refrigerator: if acetyl choline receptors had the same high affinity for progesterone a hummingbird could beat its wings twice a minute .
Back to glucocorticoid receptors, and the GRS. To date, over 450 mutations in NR3C1 have been described ; mercifully, perhaps, only ∼25 have been shown to alter activity. Classically, and in GRS, the mutation(s) blunt, or on occasion, abolish GR activation: very recently, a mutation in GRβ, which differs from GRα in having a truncated C-terminus, was shown to underlie a ‘supersensitive’ GR, presumably in the absence of its normal inhibitory activity . The N-terminal noncoding region of the gene encoding GRα, in common with that in other nuclear receptors (e.g. mineralocorticoid receptors) is differentially expressed in different tissues; the underlying mechanisms and possible impacts of such activity to date are only lightly explored.
There are a number of questions arising from the findings. First, although the previously identified p.R714Q mutation leads to changes in the ligand binding domain (LBD)and reduced binding of endogenous and exogenous glucocorticoids, the newly discovered mutation p.E198X effectively kills the receptor. The patient's mother shares this mutation, presumably with a normal opposite DNA strand, and is reported as clinically normal: it would thus appear that p.E198X is not affecting wild type GRα. Although the authors quote Nader et al. as reporting ‘…rather mild functional effects of this mutation on the GR’, it would seem that the mild effects represent the halving of the normal number of GR in affected patients. The patient under review, with one dead GR and one with the p.R714Q mutation, is anything but mildly affected.
Second, the increase in dexamethasone from 12 to 14 mg/day seems an unusually fine trigger for the normalization of acne, hypertension, menstrual irregularity etc. One possibility is that the levels of adrenal androgens were substantially higher at the patient's first evaluation at NIH than subsequently. Very recently it has been shown that androgen signalling modulates GR transcriptional output  – agonists increase, antagonists attenuate. In the presence of higher androgen levels 12 mg/day may have been more or less enough; with lower androgen levels 14 mg/day may be needed.
The third, more global question is whether or not dexamethasone is necessarily the optimum therapeutic modality. The authors justify dexamethasone on the basis that it ‘…does not exhibit mineralocorticoid or other adrenocortical steroid (i.e. androgenic) effects’. This is strictly true, but what it does not address is that dexamethasone has middle-level affinity for mineralocorticoid receptors. Bruce McEwen used 10 nmol/l [3H] dexamethasone to gauge the density of GR and mineralocorticoid receptors in the central nervous system (CNS), to label essentially all the GR and most of the mineralocorticoid receptors. It is clearly not a mineralocorticoid receptor agonist, as the authors state, but it may have effects, benign and otherwise. Glucocorticoids affect ACTH secretion not only via pituitary GR, but also via hypothalamic mineralocorticoid receptors: whether or not these very high dexamethasone levels act as antagonists, excluding cortisol from hypothalamic mineralocorticoid receptors, remains moot.
Finally, the p.714Q mutation is noted by the authors to lead to reduced affinity of the GR for glucocorticoids by changing the structure of the LBD of the protein. Whether or not dexamethasone is currently the best synthetic glucocorticoid to reduce ACTH and cortisol excess in patients with heterozygous GR mutations could be addressed by in-vitro studies on their leucocytes – or cells transfected with the mutant allele. Like other steroid receptors, GR have subtly different binding profiles for a wide range of ligands. Among the scores of synthetic glucocorticoids – a cabinet full of compounds stretching back over 50 years – there may be agonists, reasonably selective, which better accommodate and activate the mutant LBD. In an era of personalized medicine, for GRS patients this may not be too much to ask.
One area in which more data are not provided in the report is that of BP. It is presumed to have been the genesis of the CNS vasculopathy, but no values at any of the four visits are listed in Table 1. A single aldosterone value – below detection limits – is shown, for the initial visit; if DOC levels were measured – and they should have been, if they were not – they presumably account for the BP elevation. There is no indication from the absolutely normal plasma electrolyte levels on all visits of abnormal epithelial actions of elevated cortisol; the default explanation would thus appear to be that of direct vasoconstrictor effects of high levels of DOC in response to ACTH excess, as is the case in ectopic ACTH secretion, and mitigated when ACTH levels are restored into the normal range.
In addition to the patient's core disorder – that of GR with complex heterozygote mutation – the authors also note microinfarcts in the basal ganglia, left thalamus and pons, presumably the result of uncontrolled hypertension. Incidentally found was a rare variant in the Six3 gene, of unknown significance: the patient's mother, clinically normal, showed the same mutation. In-silico analysis of the variant reported that it is benign/tolerated/neutral, as might be anticipated.
In summary, the authors report the ultimately successful management of a young woman with heterozygous mutations in her glucocorticoid receptors. Clearly appropriately treated on current criteria, her condition raises unanswered questions in terms of pathophysiology, and the possibility of more targeted modalities for ACTH suppression.
Conflicts of interest
There are no conflicts of interest.
1. Tatsi C, Xekouki P, Noti O, Bachrach B, Belyavskaya E, Lyssikatos C, Stratakis CA. A novel mutation in the glucocorticoid receptor gene as a cause of severe glucocorticoid resistance complicated by hypertensive encephalopathy. J Hypertens
2. Nader N, Bachrach BE, Hurt DE, Gajula S, Pittman A, Lescher R, Kino T. A novel point mutation in helix 10 of the human glucocorticoid receptor causes generalized glucocorticoid resistance by disrupting the structure of the ligand binding domain. J Clin Endocrinol Metab
3. Kassahn KS, Ragan MA, Funder JW. Mineralocorticoid receptors: evolutionary and pathophysiological considerations. Endocrinology
4. Funder JW. Receptors, hummingbirds and refrigerators. News Physiol Sci
5. Kino T. Single neucleotide variations of the human GR gene manifested as pathologic mutations or polymorphisms. Endocrinology
6. Santen RJ, Jewell CM, Yue W, Heitjan DF, Raff H, Katen KS, Cidlowski JA. Glucocorticoid receptor mutations and hypersensitivity to endogenous and exogenous glucocorticoids. J Clin Endocrinol Metab
7. Spaanderman DCE, Nixon M, Buurstede JC, Sips H, Schilperoort M, Kuipers E, et al. Androgens modulate glucocorticoid receptor activity in adipose tissue and liver. J Endocrinol