Baker Medical Research Institute Melbourne, Victoria, Australia 8008
Address correspondence and requests for reprints to: John W. Funder, Baker Medical Research Institute, Monash University, P.O. Box 6492, St. Kilda Road Central, Melbourne, Victoria, Australia 8008.
|
Introduction |
|---|
 Top Introduction References |
|---|
There are a number of reasons for the Cinderella status of MR. First, unliganded MRs are extraordinarily labile in broken-cell systems, for reasons that remain unclear; until the advent of molybdate in homogenization buffers, all studies on MR involved incubation of tissue slices with [3H]aldosterone, of fragrant memory (1). Secondly, G.D. Searle (Skokie, IL) introduced the aldosterone antagonist spironolactone (Aldactone) very early, well before MR had been characterized. Not unnaturally, then, other receptors in the steroid-thyroid-retinoid orphan superfamily received more attention toward the development of a tamoxifen or RU486.
Third, across a range of species, MR have equivalent intrinsic affinity for aldosterone and cortisol, and slightly (x3) higher affinity for corticosterone, the physiological glucocorticoid in rats and mice (2, 3). This is in itself unsettling, given our prejudice that receptors should be selective for "their cognate ligand," to say nothing of the much higher circulating levels of the physiological glucocorticoids. Finally, in contrast with the receptors for other steroid hormones, we do not know which genes are transcriptionally regulated by activated MR in vivo in epithelial aldosterone target-tissues or in other organs (e.g. brain, heart) in which MR-mediated effects have been demonstrated.
In the paper in this issue by Patel, Sim, and Challis, Fig. 3A provides indirect but compelling evidence that the enzyme 15-OH PGDH is directly MR-regulated in the placenta. While the data reported reflect enzyme activity over 24 h of steroid exposure rather than run-on hnRNA studies, the authors have parallel data for 15-OH PGDH mRNA levels. Final proof that the enzyme is transcriptionally regulated by MR will come from the sorts of studies listed above, time-course experiments in placentae or cotransfected cells, and/or experiments in which 15-OH PGDH mRNA levels are determined with protein synthesis blocked by cycloheximide.
What Fig. 3A shows is that when the enzyme 11ß hydroxy-steroid dehydrogenase 2 (11ßHSD2) is blocked in placental syncytiotrophoblasts, cortisol lowers 15-OH PGDH activity with an IC50 of 0.1 nmol/L. That this effect is unequivocally via placental MR is shown by the remainder of the data in Fig. 3. In the absence of carbenoxolone (CBX) (i.e. when 11ßHSD2 is not blocked and is able to convert cortisol to cortisone), cortisol lowers 15-OH PGDH activity with an apparent IC50 of approximately 20 nmol/L, more than 2 orders of magnitude higher than when CBX is present. When dexamethasone (DM) is used, carbenoxolone is without discernible effect, so that the curves for inhibitory action overlie one another, again with an IC50 of approximately 20 nmol/L.
Although dexamethasone has become the benchmark agonist for
studies on glucocorticoid receptor (GR) activation, a note of caution
needs to be sounded given the IC50 of 
20 nM,
surprisingly high for DM at GR; it is not beyond the bounds of
possibility (and it is certainly compatible with the data in Fig. 3)
that the effect of both cortisol and DM seen is uniquely via MR, for
which DM has considerably lower affinity than cortisol (2), but which
can clearly be activated by DM in transfection studies (4). The issue
is one that can be resolved relatively rapidly by the use of a
selective GR agonist (e.g. RU28362) or by studies in the
presence of an MR antagonist (RU28318). Given the autocrine effect of
progesterone in elevating 15-OH PGDH levels previously noted by the
authors, studies with the GR, PR antagonist RU486 may prove difficult
to interpret. The balance of probability is that placental 15-OH PGDH
can be regulated by an activated MR or GR, just as transepithelial
sodium transport can be (5, 6); but if it does prove MR-specific, from
RU28362/RU28318 studies, then in many ways so much the better.
In epithelial tissues the pivotal role of 11ßHSD2 in excluding the physiological glucocorticoids from MR is widely appreciated, both from the initial experimental studies on the rat (7, 8) and from subsequent studies on the syndrome of apparent mineralocorticoid excess (AME), in which loss-of-function mutations in 11ßHSD2 allow cortisol inappropriately to occupy MR (9). The resulting sodium retention is evidence that in epithelial MR cortisol, and in the rat corticosterone (5), acts as an agonist. What is perhaps not as widely appreciated is that activation of renal GR, by physiological glucocorticoids, by dexamethasone or by RU28362 (5, 6) is similarly followed by a urinary electrolyte effect indistinguishable from that of aldosterone, cortisol, or corticosterone via MR. This not only underlines the crucial specificity-conferring effect of 11ßHSD2 in epithelia in excluding glucocorticoids from both MR and GR, but also suggests that, in such tissues the genomic effects may be mediated by relatively nonspecific hormone response elements, of the familiar 15-mer palindromic type, which have been shown to bind both MR and GR in transfection and gel shift studies.
In nonepithelial tissues MR by definition play a different physiological role from that of mediating the effect of aldosterone on unidirectional transepithelial sodium transport. There are also operational differences in the mechanisms involved in MR action between epithelial and nonepithelial tissue, at the level of both the MR themselves and the response elements with which they interact. In studies on both blood pressure elevation [via MR in the AV3V region of the brain (10)] and on cardiac fibrosis [via cardiac MR (11)], the physiological glucocorticoids antagonize rather than mimic the effects of aldosterone in tissues in which MR are unprotected by 11ßHSD2. Secondly, activation of GR by selective agonists in the AV3V region does not mimic (or block) the effect of MR activation by aldosterone, in clear contrast with the promiscuity of the corticosteroid effects in epithelia. The factors binding to and subtly modifying MR in nonepithelial tissues have not been explored in depth; even less is known of the response elements operant in nonepithelial tissues that discriminate between activated MR and activated GR, in contrast with epithelia.
Those tempted to dismiss nonepithelial MR as "not the main game" should think about the syndrome of glucocorticoid suppressible hyperaldosteronism (12). This autosomal dominant syndrome reflects the activity of a chimeric gene, with the 5' regulatory and tissue-locating sequences of 11 hydroxylase and the 3' business end of aldosterone synthase. Such a chimeric gene produces a hybrid enzyme, expressed throughout the adrenal cortex, synthesizing aldosterone in response to ACTH and out of the normal Na+/K+ angiotensin/aldosterone control loop. The reason why such a chimera can occur is that the genes for the two enzymes, aldosterone synthase and 11ß hydroxylase (the signature enzyme for a glucocorticoid), are 94% identical and lie next to one another on chromosome 8q22, allowing unequal crossing over at meiosis if the two strands are misaligned.
Contrast this with MR (chromosome 4) and GR (chromosome 5), which share 90% amino acid identity over their DNA-binding domains, 57% in the ligand binding domains, and essentially none elsewhere. The most plausible scenario for 11 hydroxylase/aldosterone synthase is that the two genes, highly homologous and located in tandem, represent a relatively recent gene duplication event, certainly more recent than the divergence of MR and GR. In evolutionary terms aldosterone is a Johnny-come-lately, and the receptors we term MR branched off much earlier. The findings of the past decade, on hypertension and cardiac fibrosis mediated by nonepithelial MR, represent evidence for the pathophysiological importance of such receptors, over and above their acknowledged role in sodium transporting epithelia.
For over 30 years there has been a spectacularly unsuccessful search for aldosterone-induced proteins in epithelial tissues, or MR-regulated proteins in MR-rich nonepithelial tissues such as the hippocampus. Historically, attention focussed on key enzymes in the tricarboxylic acid cycle such as citrate synthase, crucial for energy provision required for transport, or generic "permeases." More recently, Na+/K+ ATPase subunits, or subunits for epithelial sodium channels (ENaC), have been obvious candidates. All have been shown not to be regulated at the genomic level, either by direct studies or by inference from MR knockout mice (13). Rapid, nongenomic effects of aldosterone have been reported—on intracellular ionic flux (14), on PKC activity (15), or Na+/K+/2Cl- cotransporter (16)—but to date not a single gene transcription of which is unequivocally directly regulated by activated MR in any tissue, epithelial or nonepithelial. It is into this vacuum that 15-OH PGDH has fallen.
The ramifications, if and when it is confirmed to be a direct transcriptional regulatory effect, are likely to be substantial. First up, concentrate on the placenta; if it is operationally an epithelial tissue, then prima facie what Patel et al. have shown is that it is an aldosterone target tissue. If this is the case, a whole new physiology will open up, and with it the possibility of a whole new pathophysiology, particularly given the salt, water, and blood pressure enigma represented by pre-eclampsia. Then think of the classic epithelial aldosterone target tissues and the will-of-the-wisp involvement of prostaglandins in sodium homeostasis; it is possible that 15-OH PDGH expression is aldosterone-regulated uniquely in placenta, but it is surely worth a look in kidney and colon, sweat gland and salivary gland. Finally, the enigma of the extraepithelial actions of aldosterone, increasingly recognized to be of considerable pathophysiologic importance, may be at least partly opened up with 15-OH PDGH as the lever.
In our familiar representation of Cinderella, her dainty foot is encased in a glass slipper, extravagantly twinkling to distract the more observant young reader from the look of anticipation on the Prince’s face. The slipper made of glass (en verre) is an inspired piece of ur-Disney, total suspension of disbelief: it may be less dangerous if sourced from Dow Corning, but one dance and you’re crippled for life. It also represents a marvellous mistranslation; for French children the slipper is "en vair," i.e. made of moleskin, and what is lost in twinkle is more than made up in comfort and practicality. So while sometimes things get lost in translation, on other occasions things get transformed: and in the translation from placenta to elsewhere we need to countenance the possibility that Cinderella’s slipper may be moleskin, Baccarat, or both.
Received December 11, 1998.
Accepted December 12, 1998.
|
References |
|---|
|
TopIntroductionReferences |
|---|