| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Prince Henry’s Institute of Medical Research, Clayton, Victoria 3168, Australia
Correspondence: Address all correspondence and requests for reprints to: John W. Funder, Prince Henry’s Institute of Medical Research, P.O. Box 5152, Clayton, Victoria 3168, Australia. E-mail: john.funder{at}phimr.monash.edu.au
 |
Abstract |
|---|
|
I. Introduction |
|---|
|
II. The Line in the Sand |
|---|
In retrospect, there are a number of facets of these studies worthy of note. First, they showed 22Na+ efflux to be biphasic—an actinomycin D-insensitive, ouabain-insensitive rapid response and an-ouabain-sensitive, actinomycin D-inhibited (and thus presumably genomic) delayed response. Second, the authors injected nearly physiological doses (10 µg/kg) of aldosterone into adrenalectomized rats and then killed them at various time points (15 min to 5 h) after injection; importantly, the tissues were studied approximately 2 h after the rats were killed, which included a 90-min 22Na+ loading period at 37 C; the effects may have been rapid, but they were also persistent. Third, both the nongenomic and genomic effects of aldosterone were progressively blocked by increasing doses of spironolactone or the open E-ring, water-soluble MR antagonist RU28318; the specific glucocorticoid receptor agonist RU26988 did not modify 22Na+ flux. Fourth, and in retrospect again importantly, the MR antagonists were administered in vivo by gavage 1 h before aldosterone was injected. Finally, in a parallel series of in vitro studies, aldosterone was shown to have a half maximal effect on both rapid (15 min) and delayed (120 min) 22Na+ flux at approximately 10–9 M; the delayed but not the rapid effect of aldosterone in vitro was actinomycin D inhibitable; and both rapid and delayed effects were progressively inhibited by preincubation with RU28318. No in vitro studies with spironolactone were reported; in many subsequent studies, for reasons that remain unclear, RU28318 and potassium canrenoate, but not spironolactone or canrenone, have been shown to inhibit the nongenomic effects of aldosterone.
On the basis of these studies, Moura and Worcel (1) concluded that both rapid nongenomic and genomic effects of aldosterone were mediated via classical MR. This conclusion was discounted throughout the 1990s in favor of a putative membrane receptor for aldosterone mediating its rapid effects, and it was not until recently that additional evidence for rapid nongenomic effects mediated via classical MR has been reported in a number of experimental systems. That said, classical MR may not account for all the rapid effects of aldosterone at physiological doses, and other recent studies have been interpreted as further evidence for a distinct membrane receptor (see Section IX).
|
III. Rapid Actions and Membrane Receptors |
|---|
A subsequent paper by Wehling et al. (3) is less circumspect; the title is "Membrane Receptors for Aldosterone: A Novel Pathway for Mineralocorticoid Action." In this and subsequent studies over the next couple of years, Wehling and his colleagues (4, 5, 6, 7, 8) attempted to marshal evidence for a nonclassical receptor being responsible for the rapid nongenomic action of aldosterone. Using aldosterone-3-(0-carboymethyl)-oximino-2-[125I]iodohistamine as tracer, the authors demonstrated very low levels of nondisplaceable binding (plateau, 25 cpm/mg protein) to HML membranes, with deoxycorticosterone acetate showing a 1000-fold lower affinity in competition assays and with corticosterone showing even lower potency. From the data obtained in this study, the authors proposed a binding affinity for aldosterone of Kd (37 C) 0.16 nM, with approximately 100 binding sites per cell. The disconnect between the 1000-fold lower apparent affinity of such sites for deoxycorticosterone and its potency close to that of aldosterone in terms of nongenomic effects was not canvassed.
When these studies were extended (6) to a potential epithelial target tissue for aldosterone (pig kidney membranes), the findings were even less convincing. Nondisplaceable binding was very high (of the order of 90% of total binding at 0.5 nM tracer concentration), and specific binding was even lower than for HML (of the order of 5 cpm/mg protein, with a tracer specific activity of 2000 Ci/mmol). In competition studies, deoxycorticosterone was reported to have approximately 10-fold lower affinity than aldosterone, in contrast with the approximately 1000-fold lower affinity in HML. The utility of the tracer and/or the rigor of the assays are not supported by the displacement curves for nonradioactive aldosterone; 10 pM aldosterone reduces the binding of 300 pM tracer to 70%, and 100 pM aldosterone reduces it to 30%.
When the HML binding studies were extended (7) by cross-linking the tracer to the putative receptor and SDS-PAGE, a single 100-cpm peak was found at approximately 50 kDa, and smaller (50 cpm) peaks were found at approximately 100 and 200 kDa. Although these latter are confidently labeled dimeric and tetrameric receptor protein, for reasons that are not canvassed, these higher molecular mass peaks are identical in the presence or absence of 1 µM nonradioactive aldosterone, making their unquestioned ascription as receptors even more problematic.
Last, in terms of putative membrane receptors, are the studies from the Wehling laboratory using relatively high specific activity (
120 Ci/mmol) tritiated aldosterone as tracer (8), rather than the iodinated derivative previously employed. In contrast with HML and porcine kidney, the membrane binding sites for aldosterone were of much lower affinity (>10 nM, >100 nM); maximal binding capacity, on the other hand, is considerably higher at 700 pmol/mg membrane protein.
Taken together, these studies do not provide consistent or convincing data toward the characterization of a membrane receptor for aldosterone. There is no question that studies from the Wehling laboratory in the early to mid 1990s were crucial in the now general acceptance of rapid nongenomic effects of aldosterone in a variety of cells [HML, endothelial cells, vascular smooth muscle cells (VSMC)]. There is equally no question that rigorous examination of binding studies over the past 15 yr has shown the putative membrane receptors to vary widely between target tissues in terms of concentration, which is to be expected, but also in terms of affinity and specificity, which is not.
The major driver for the existence of a putative membrane receptor for aldosterone was the demonstration that in a variety of effector systems aldosterone, deoxycorticosterone, and 9
-fludrocortisone were active as rapid nongenomic agonists, whereas cortisol and corticosterone were not; in addition, spironolactone and canrenone appeared inactive in the in vitro systems tested. The case for a membrane receptor was—and remains—circumstantial, a default position rather than one supported by convincing positive evidence. Over the last 5 yr, however, there has emerged considerable evidence in various tissues for rapid nongenomic effects of aldosterone being mediated via classical intracellular MR. Such a demonstration does not exclude effects via a membrane receptor; on the other hand, there is now good evidence, to be discussed below, for nongenomic effects to be ascribed at least in part to classical MR-mediated pathways.
|
IV. Rapid Nongenomic Effects of Aldosterone on the Kidney |
|---|
In subsequent studies on rapid nongenomic effects of aldosterone on ion flux in rat cortical collecting tubule segments in vitro, Fujii et al. (10) showed that aldosterone stimulated 86Rb uptake, as a measure of the K+ transporting ability of the Na+/K+ ATPase. The effect was clearly demonstrable at 30 min, half maximal at 1 nM aldosterone, and at 2 h was partly but incompletely inhibited by cycloheximide or actinomycin D. The authors took pains to exclude an effect on 86Rb flux secondary to increased Na+ influx and canvass the possibility of an effect secondary to an aldosterone-induced change in intracellular pH. In the same vein, a series of studies from Oberleithner’s laboratory (11, 12) showed that aldosterone acted within minutes on cellular pH and plasma membrane potassium conductance in various cellular preparations, with the plasma membrane Na+/H+ exchanger being an important target for this rapid effect of aldosterone.
Subsequently, the same laboratory made a series of contributions to our understanding of the mechanisms of rapid aldosterone action on cultured Madin-Darby canine kidney (MDCK) cells. Within 1 min of the introduction of physiological concentrations of aldosterone (EC50, 0.1 nM), a 3-fold rise in intracellular [Ca2+] was seen, as well as a [Ca2+]-dependent, ethylisopropanol amiloride (EIPA)-inhibitable increase in intracellular pH. When extracellular [Ca2+] was omitted, a Zn2+-dependent, aldosterone-induced fall in intracellular pH was seen (13). Taken together, the interpretation of these data was that aldosterone rapidly raises net entry of Ca2+ into the cell and plasma membrane proton conductance, as prerequisites for increasing plasma Na+/H+ activity, which in its turn modulates K+ channel activity. The aldosterone-induced proton conductance increase was subsequently further characterized under Na+-free (external) conditions (14) and shown to be spironolactone-insensitive and able to be blocked by protein kinase C (PKC) inhibitors, attenuated by pertussis toxin, and mimicked by phorbol esters. On the basis of these studies, the authors propose that aldosterone produces G protein-dependent PKC stimulation that activates proton conductance and thereby Na+/H+ exchange.
Similar findings on a mouse cortical collecting duct M-1 cell line were subsequently reported by Harvey and Higgins (15). They showed, in addition to the MDCK findings, that the aldosterone-induced Ca2+ influx was actinomycin D insensitive but abolished by the PKC inhibitor chelerythrine, that cortisol was without effect even at high concentrations in terms of raising intracellular [Ca2+], and that (nonphysiological, 10 nM) levels of estradiol and progesterone also induced PKC-dependent increases in intracellular [Ca2+].
Further studies from the Gekle laboratory (16) explored additional details of the rapid action of aldosterone in MDCK cell activation. First, they showed that nanomolar concentrations of aldosterone induced rapid phosphorylation of ERK1/2 and that the specific ERK1/2 inhibitor UO126 prevented aldosterone-induced activation of Na+/H+ exchange and increase in intracellular pH, reasonably interpreted as evidence for the necessary involvement of the MAPK in the rapid nongenomic signaling pathway in response to aldosterone in the kidney (16). This concept was extended by the demonstration that aldosterone action involves epidermal growth factor receptor phosphorylation and that inhibition of epidermal growth factor receptor kinase by tyrophostin AG-1478 abolished aldosterone-induced signaling (17).
The involvement of ERK1/2 in rapid nongenomic aldosterone action has also been shown by Good et al. (18, 19) in studies on the renal medullary thick ascending limb (MTAL). In their preliminary studies (18), these authors showed that aldosterone rapidly inhibits MTAL HCO3– absorption with an IC50 of 0.6 nM, an action unaffected by actinomycin D, cycloheximide, or spironolactone, and not mimicked by cortisol or corticosterone in the presence or absence of carbenoxolone. In subsequent studies (19), the same authors showed that a similar inhibition of HCO3– absorption could be produced by nanomolar 1,25-dihydroxyvitamin D3, additive to that seen with aldosterone; unlike the aldosterone effect, that of D3 was not blocked by the ERK1/2 inhibitor UO126, perhaps not in itself surprising given that the effect of D3 is at the genomic level.
An additional rapid effect of aldosterone has recently been shown for the inner medullary collecting duct (IMCD), the major site of vasopressin action in the kidney. In IMCD segments from intact rat kidneys, aldosterone (EC50, 1.2 nM) produced a dose-dependent 4-fold stimulation of cAMP generation within 4 min to levels equivalent to those seen with vasopressin, but not blocked by V1 or V2 receptor antagonists (20). In outer medullary collecting ducts isolated from intact mouse kidneys, aldosterone has been reported to raise vascular H+-ATPase approximately 2-fold within 15 min via a PKC-dependent mechanism (21); in contrast with most other reports of nongenomic aldosterone action, however, 1 nM was ineffective, with 10 nM routinely used, and no further dose-response data were shown.
In a recent, challenging paper on the effects, genomic and nongenomic, of aldosterone, Le Moellic et al. (22) use the RCCD2 (rat cortical collecting duct) cell line. They show that the early (up to 2.5 h, so early rather than rapid) aldosterone-induced increase in short circuit current is not blocked by the closed E-ring MR antagonist RU26752 and similarly is not blocked by cycloheximide. The effect is mediated via PKC on the basis of inhibitor studies and is reported as being able to be activated by BSA-aldosterone conjugates, interpreted as evidence for a membrane site of action. Two observations made by the authors might be construed as giving pause for such an interpretation. First, both aldosterone and BSA-aldosterone provoke phosphorylation of the classical MR; second, BSA-aldosterone on a molar basis is at least 100-fold less potent than aldosterone. Perhaps the most telling finding of the paper is that PKC blockade inhibits not only the early but also the late genomic response to aldosterone, pointing the way to a possible pathway of physiological integration of nongenomic and genomic effects.
Finally, almost 50 yr after Ganong and Mulrow (9), the wheel comes full circle. In the intact rat, rather than the anesthetized dog, Ashton’s laboratory (23) has shown that aldosterone infused at a robust but reasonable dose produced an increase in sodium excretion within 15 min, independent of changes in glomerular filtration rate, urine flow or pH, or of K+, Cl–, or HCO3– excretion. Extrapolating from their earlier studies on IMCD (20), the authors suggest the possibility of increased cAMP generation and activation of the cystic fibrosis transmembrane conductance regulator Cl– channel to drive IMCD Na+ secretion, as an explanation consistent with their unexpected and otherwise counterintuitive observation.
In summary, although there are a number of studies that clearly show acute nongenomic effects on the kidney, in vivo and in vitro, it is not yet possible to construct a coherent physiology for such effects on renal function. Many, but not all, of the demonstrated effects are seen at subnanomolar, i.e., circulating concentrations; an acute elevation of intracellular [Ca2+] and pH appears to be a likely early mediator, as does PKC stimulation; the lack of effect of cortisol and corticosterone in the presence of carbenoxolone (18) might be construed as evidence against an effect via classic MR, except that MTALs do not appear to express significant levels of 11ß-hydroxysteroid dehydrogenase (11ßHSD) type 2; and the acute natriuretic effect of aldosterone in vivo (9, 23) is both counterintuitive and difficult to reconcile with the demonstrated genomic antinatriuretic effects of aldosterone in the kidney.
|
V. Rapid Nongenomic Effects of Aldosterone on the Colon |
|---|
Whereas these initial studies (25) were carried out on rat colonic epithelium, further studies from the same laboratory used the human T84 colonic epithelial cell line (26). In these cells, aldosterone was shown to induce a rapid increase in intracellular sodium in approximately 2 min, independent of emptying of intracellular calcium stores by thapsigargin and not mimicked by cortisol. Again, no dose-response curves could be obtained for the three mineralocorticoids tested, and in this preparation the effects of DOCA at 0.1 nM to 1 µM were equivalent to those of aldosterone and fludrocortisone. Although recorded basal intracellular [Ca2+] levels varied substantially between individual cells, the extent of the stimulation (
10-fold) in intracellular calcium was of the same order as that for the elevation of PKC in the rat (25).
Very similar findings were seen in normal human colon cells from surgical specimens (27). Aldosterone activated PKC in colonic cytosols and raised intracellular [Ca2+] in whole cells, with both blocked by the PKC inhibitor chelerythrine chloride. Again aldosterone, fludrocortisone, and DOCA were equivalently active, elevating PKC levels 3- to 4-fold, with no dose-response curve over the concentration range 0.1–100 nM; cortisol over the same range showed values 1.2- to 1.3-fold control. The same laboratory then explored the downstream effectors in the pathway of rapid aldosterone effects by Ussing chamber studies and fluorescence microscopy (28). The PKC-driven, Ca2+-mediated increase in Na+/H+ exchange activity was shown to up-regulate an ATP-dependent K+ channel and inhibit a Ca2+-dependent channel, with the former operating to drive sodium absorption and the latter necessary for cAMP- and Ca2+-dependent chloride secretion. All of these effects were shown to be insensitive to spironolactone, actinomycin D, and cycloheximide, and on the basis of their findings the authors propose a priming role for the acute action of aldosterone on crypt cells via the differential effect on KATP and KCa channels.
In independent studies on rat colonic cells (29), using confocal laser imaging and a pH-sensitive fluorescent dye, intracellular alkalinization was seen within 1 min of application of aldosterone, now with a classical dose response curve over the concentration range 0.01–100 nM and an EC50 of 0.8 nM. In a series of inhibitor studies (chelerythrine chloride, forskolin, quinacrine, and piroxicam), the alkalinization was shown to involve G proteins, PKC, and prostaglandins; not unexpectedly, the effect was not blocked by actinomycin D, cycloheximide, or spironolactone and was reduced to near [Na+]-free medium levels by EIPA.
In subsequent studies from the Harvey laboratory (30), the claim was made for the rapid effect of aldosterone in the colon (activation of Na+/H+ exchange, K+ recycling, PKC and PKC-dependent Ca2+ entry via L-type Ca2+ channels) to reflect a direct stimulation of PKC by aldosterone. In studies on the PKC isoforms (human, recombinant) plus appropriate substrate peptide, cofactors, etc., Harvey et al. showed a modest (25%) increase over basal in PKC activity at 10 pM aldosterone and equivalent values at 1 nM, with an increase to 50% over basal at 10 nM; no effect of aldosterone on PKC
, PKC
, or PKC
activity was seen. Estradiol raised PKC and PKC (but not PKC or PKC) activity in vitro with a similarly gentle dose-response curve, a 25% increase at 100 pm and 1 nM, and a further increase in PKC at the even more nonphysiological level of 10 nM estradiol. On the basis of these findings, Harvey et al. proposed, "Therefore, PKC is a candidate nongenomic receptor for aldosterone," and "Thus, the PKC isoform may be a nongenomic receptor for estradiol" (30).
Finally, Bowley et al. (31) have explored the rapid effects of aldosterone (EC50 < 1 nM) to inhibit intermediate conductance basolateral K+ channels in human colonoscopy specimens using patch clamp techniques and RT-PCR. Aldosterone appeared both to decrease individual channel activity (in single channel patches) and to decrease the apparent number of channels per patch, suggesting that it may stimulate channel recycling into cytoplasmic vesicles as well as having a direct inhibitory effect. The 27pSK+ channel was provisionally identified as inhibited by aldosterone on the basis of the detection of the cognate KCNN4 mRNA by RT-PCR.
In summary, there appears to be good evidence for nongenomic effects of aldosterone on the colon from a number of reports that are difficult to reconcile in terms of both receptors and downstream events. Again, PKC activation, activation of Na+/H+ exchange, and increased intracellular [Ca2+] appear credible actors; on the other hand, plateau effects of aldosterone, fludrocortisone, and DOCA from 0.01 to 100 nM (25) are as difficult to reconcile with physiology as is a direct effect on PKC (30). Finally, as for the kidney, it is not obvious how the demonstrated acute nongenomic effects and those of aldosterone at the genomic level are to be construed into a coherent account of mineralocorticoid action on the colon.
|
VI. Rapid Nongenomic Effects of Aldosterone on the Vascular Wall |
|---|
In further studies from the same group, the rapid effects of aldosterone on the production of diacylglycerol via PKC were determined by enzymatic and immunoblot assays (33). Aldosterone (EC50 <1 nM) doubled diacylglycerol levels within 15 min, an increase blocked by the PKC inhibitors neomycin or U-73122; within 5 min of exposure to aldosterone, measured PKC levels in the cytosolic fraction fell by approximately 25%, and those in the membrane fraction rose by 70%. The 22Na+ studies were subsequently extended by determining the intracellular pH after exposure to aldosterone by the use of the pH-dependent fluorescein analog 2'-7'-bis (carboxyethyl)-5(6)-carboxyfluorescein acetoxymethyl ester. Aldosterone induced a rapid increase in intracellular pH with an EC50 of approximately 0.1 nM, with cortisol inactive and the effect blocked by the specific Na+/H+ exchange inhibitor HOE694 (8).
In porcine aortic endothelial cells, Wehling’s laboratory (34) reported that aldosterone (EC50, 1 pM) raised intracellular [Ca2+] by 50% within 1–5 min, with cortisol three orders of magnitude less potent (EC50, 1 nM). The effects were not inhibited by spironolactone, were abolished by thapsigargin, and were very much blunted by incubation in Ca2+-free medium. The EC50 values reported for both aldosterone and cortisol are such that the receptor involved would always be saturated at circulating concentrations of either or both steroids, making a physiological modulatory role problematic.
Subsequent studies from the Oberleithner laboratory (35) showed that aldosterone at 0.1 nM produced a rapid (5 min) 28% increase in bovine aortic endothelial cell volume, as measured by atomic force microscopy. The effect could be blocked by amiloride and was evanescent in that after 30 min, volume returned to baseline levels despite the continuing presence of aldosterone. No dose-response data were reported, so that sensitivity comparison between porcine and bovine cells was not possible.
The paper by Alzamora et al. (36) marked a sea-change in our understanding of rapid nongenomic effects of aldosterone. These authors took strips of human vessels obtained at cesarean section and documented the activity of both 11ßHSD-1 and 11ßHSD-2 in the vascular wall. They then measured the rapid effect of aldosterone on Na+/H+ exchange in VSMC by the use of a pH sensitive fluorescein dye. Aldosterone (1–10 nM) progressively elevated intracellular pH, an effect blocked by EIPA but not spironolactone, and not mimicked by cortisol. To this point there was nothing to challenge the received wisdom. The next two studies, however, very substantially question that the rapid, nongenomic effects of aldosterone are via a nonclassical membrane receptor. First, whereas the closed E-ring MR antagonist spironolactone was without inhibitory effect, the open ring, water-soluble antagonist MR823188 completely abolished the effect, harking back to Moura and Worcel (1). All that this may mean is that a nonclassical receptor binds RU28318 but not spironolactone; alternatively, it may be that the nongenomic effect is indeed via the classical MR, and that in vitro, in certain systems, closed ring MR antagonists are acutely ineffective.
That this latter is more likely to be the case is shown by the final series of experiments. As noted above, cortisol was without agonist effect when administered alone. When, however, the vessels were pretreated with the 11ßHSD inhibitor carbenoxolone, cortisol became a MR agonist, with a dose-response curve identical to that of aldosterone, and blocked by the addition of either EIPA or RU28318. Given that 11ßHSD is tethered in the intracellular endoplasmic reticulum, where it protects MR from inappropriate activation by cortisol, a parallel effect for an intracellular enzyme in protecting a putative plasma membrane receptor is at best counterintuitive. In brief, these studies do not exclude aldosterone-sensing mechanisms other than classical MR in VSMC or any other tissue. What they do is to justify a working hypothesis that the rapid nongenomic effects of aldosterone in VSMC can be mediated via a classical MR protected by 11ßHSD, and that spironolactone (and other closed-ring antagonists; see below) are for some reason acutely inactive in vitro.
In addition to their Na+/H+ exchanger study (36), Alzamora et al. have shown that aldosterone can produce a rapid inhibition of Na+/K+ ATPase activity in aortic rings, as gauged by ouabain-sensitive 86Rb/K uptake (37). This effect was maximal at 20 min, apparently evanescent in that values returned to baseline within 2 h despite the continuous presence of aldosterone; unaffected (at 20 min) by cycloheximide, actinomycin D, or the sodium ionophore monensin; but abrogated by eplerenone, rapamycin, colchicine, or the PKC inhibitor bisindolylmaleimide. The authors interpret these data as evidence for an acute aldosterone effect including dissociation of heat shock complexes from MR, PKC activation, and the involvement of microtubules on the basis of the inhibitor studies. They also acknowledge the bidirectionality of the effects of aldosterone on Na+/K+ ATPase (acute inhibition, chronic stimulation), which has also been documented in cardiomyocytes (38), but did not extend their actinomycin D/cycloheximide studies to 120 min, which may have revealed a more persistent nongenomic effect.
In an in vivo study on human forearm vascular reactivity, Romagni et al. (39) performed a placebo-controlled, double-blind crossover study in young (25- to 35-yr-old) normal volunteers. Importantly, blood flow was recorded in both forearms before, during, and after infusion of 25 pmol aldosterone (or placebo) over 10 min via the brachial artery in the nondominant arm. Forearm blood flow was reduced in the aldosterone-infused arm within 4 min, to a nadir of 35% basal levels at 12 min, and with an immediate and progressive return to baseline within 30 min of the infusion ending. No change was seen in the control arm, and no differences in plasma aldosterone concentrations were seen between levels before infusion and after infusion. Peak levels in forearm blood can be calculated to be of the order of 1 nM, at the high end of the range seen in sodium deficiency or hemorrhage, and thus consistent with no measurable elevation of systemic aldosterone levels.
The effects of aldosterone on vascular tone became an area of considerable debate at this time. Gunaruwan et al. (40) reported outcomes clearly different from those of Romagni et al. (39); in contrast with Romagni et al., however, their doses ranged from 10–100 ng/min, more than one to two orders of magnitude higher than in the Romagni study. Schmidt et al. (41), using an even higher dose (500 ng/min), showed an increase in forearm blood flow in the aldosterone-infused forearm, in contrast with the increase in systemic vascular resistance reported by the same author when the same dose of aldosterone is administered systemically (42). These inconsistencies may reflect a variety of differences between the studies, not least of which are the potential pitfalls in using doses of aldosterone far higher than physiological in in vivo (or in vitro) studies.
Support for vasorelaxant effects of aldosterone at very different (subpicomolar) concentrations came from the studies of Uhrenholt et al. (43) on rabbit preglomerular afferent arterioles. The effect of aldosterone was significant at 5 min, maximal at 20 min, and thereafter apparently evanescent; it was blocked by spironolactone but not RU486 or actinomycin D, and abolished by LY 29002 (implicating phosphatidylinositol-3 kinase) or by geldanamycin (suggesting that heat shock protein 90 liberated from MR, together with protein kinase B, activates endothelial nitric oxide synthase). Apparently conflicting results came from Arima et al. (44), again in studies on rabbit afferent arterioles. Over a modest concentration range (10–10 to 10–8 M), aldosterone caused a 20–30% vasoconstriction at all doses, augmented by endothelial disruption or nitric oxide synthesis inhibition and attenuated by pretreatment with chelerythrine or thapsigargin. The authors interpreted their findings as evidence for endothelial-derived nitric oxide modulating vasoconstriction induced by aldosterone via activation of both PKC and IP3 pathways.
In mesenteric vessels from Sprague-Dawley rats, further studies from the Marusic laboratory (45) detailed the mechanism of rapid vasoconstrictor action of aldosterone. At 10 nM, aldosterone caused rapid constriction of resistance vessels, and potentiated the vasoconstrictor effects of phenylephrine, rapidly elevating intracellular pH and [Ca2+] and decreasing protein kinase B phosphorylation. Vasoconstriction in response to aldosterone was abolished by inhibition of PKC or of Na+/H+ exchanger-1 and by eplerenone, and inhibitors of ERK1/2 phosphorylation had no effect on aldosterone-induced vasoconstriction.
In summary, the picture for nongenomic effects of aldosterone on the vasculature is not dissimilar to that for kidney or colon. There is ample evidence for acute nongenomic effects, and for the involvement of PKC activation, increased Na+/H+ exchanger activity, and intracellular [Ca2+]; there are clear inconsistencies between studies (e.g., Ref. 30 vs. Refs. 40 and 41 ; Ref. 43 vs. Ref. 44), and between doses of aldosterone used in vitro (10–12 to 10–8 M) and in vivo (1–500 ng/min by intrabrachial infusion). In contrast, however, with kidney and colon, a possible coherent physiology can be proposed, wherein aldosterone acts via 11ßHSD-protected MR to acutely constrict vessels, as part of a homeostatic response to postural change/volume loss, in concert with sympathetic discharge and angiotensin action on the vessel wall.
|
VII. Rapid Nongenomic Effects of Aldosterone: Clinical Studies |
|---|
In a subsequent study using lower doses (0.5 or 0.05 mg) of aldosterone, the major difference between aldosterone-infused and control groups was seen between 330 and 390 min after injection, with much smaller differences (for 0.5 mg only) at 6 and 30 min (42). The late effect was proposed as a nongenomically mediated (early onset within 15 min) increase in vagal tone during the aldosterone periods, with the apparent latency of more than 5 h not addressed. The aldosterone levels after 0.05 mg, which could not be shown to have rapid effects, peaked at 2 ng/ml. i.e., 7 nM, still considerably above physiological. Finally, again with 0.5-mg bolus doses in subjects pretreated with a ß-blocker (esmolol) or a ß-agonist (dobutamine), the group showed a 4.1% increase in mean blood pressure with esmolol, a 1.6% decrease with dobutamine, and no effect of phenylephrine, with significant differences between esmolol and dobutamine pretreatment over the first 12 min after injection (48). Taken together, these studies offer scant support as yet for a physiological cardiovascular role for the acute nongenomic effects of aldosterone.
|
VIII. Rapid Nongenomic Effects of Aldosterone on the Heart |
|---|
An even more counterintuitive effect of spironolactone was shown in isolated working rat heart (Langendorff) studies by Barbato et al. (50). Spironolactone and aldosterone both increased contractility 40–50% at 10 nM concentration within 2–4 min; in dose-response studies, spironolactone proved to be almost one order of magnitude more potent, and importantly the effects of aldosterone and spironolactone were additive, suggesting an effect via distinct and synergistic pathways. The EC50 for spironolactone was of the order 0.3 nM, possibly offering an explanation for the otherwise puzzling therapeutic effect of MR blockade at very low doses in the Randomized Aldactone Evaluation Study (spironolactone, 26 mg/d) (51) and the Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study (eplerenone, 43 mg/d) (52).
In further in vitro studies, Mihailidou and colleagues (38, 53, 54, 55) have explored the rapid nongenomic effects of aldosterone in patch clamp studies on rabbit cardiomyocytes. Aldosterone at 1–10 nM increases Na+/K+/2Cl– cotransporter activity, and the resultant increase in intracellular [Na+] results in increased Na+ pump activity, reflected in a 10-fold increase in current across the plasma membrane within 15 min. When Na+ influx is absent or blocked, aldosterone rapidly lowers Na+/K+ ATPase activity, consistent with the previous studies from the Marusic laboratory (37). Both Na+/K+/2Cl– cotransporter activation and Na+/K+ ATPase inhibition involve phosphorylation, directly or indirectly via PKC. The rapid effects of aldosterone are unaffected by actinomycin D or by canrenone or spironolactone acutely; they are, however, blocked by the open E-ring potassium canrenoate and by spironolactone administered to the rabbits before killing. Most recently (56), cortisol has been shown to be without agonist effect at 100 nM in normal cardiomyocytes, reminiscent of the Alzamora et al. studies (36) on VSMC, but to stoichiometrically block the effect of coadministered 10 nM aldosterone, consistent with its equivalent affinity for MR. Given that the cardiomyocyte does not express 11ßHSD2, and thus that MR will be constitutively glucocorticoid-occupied, the physiological role of such nongenomic (and possible genomic) effects of aldosterone on the heart are unclear. The finding of higher levels of left ventricular hypertrophy in patients with primary aldosteronism than in patients with essential hypertension and equivalent blood pressure levels, and of the cardiac fibrosis and failure in transgenic mice expressing 11ßHSD2 in cardiomyocytes (57), suggest that pathophysiological roles for aldosterone via MR in the heart are indeed worthy of further investigation.
|
IX. Rapid Nongenomic Effects of Aldosterone: Receptor Mechanisms Revisited |
|---|
|
X. Envoi |
|---|
|
Footnotes |
|---|
Abbreviations: DOCA, Deoxycorticosterone acetate; EIPA, ethylisopropanol amiloride; HML, human mononuclear leukocyte(s); 11ßHSD, 11ß-hydroxysteroid dehydrogenase; IMCD, inner medullary collecting duct; IP3, inositol 3-phosphate; MDCK, Madin-Darby canine kidney; MR, mineralocorticoid receptor(s); MTAL, medullary thick ascending limb; PKC, protein kinase C.
|
References |
|---|
activation, mineralocorticoid receptor phosphorylation, and cross-talk with the genomic response. J Am Soc Nephrol 15:1145–1160. Biochem Biophys Res Commun 213:123–129[CrossRef][Medline]
PKC. Endocrinology 145:773–780
Top
Abstract
I. Introduction