(Hypertension. 1999;34:1123-1128.)
© 1999 American Heart Association, Inc.
Scientific Contributions |
From the Molecular Medicine Centre, University of Edinburgh, UK (S.A.L-M., S.M.N., J.M., C.J.K.), and Faculty of Pharmacy, Centre National de la Recherche Scientifique ESA 5014, Lyon, France (M.F., M.L., J.S.).
Correspondence to Dr C.J. Kenyon, Molecular Medicine Centre, Western General Hospital, Edinburgh, EH4 2XU, UK. E-mail cjk{at}srv0.med.ed.ac.uk
| Abstract |
|---|
|
|
|---|
Key Words: glucocorticoids mineralocorticoids corticosterone cortisol cortisone renal function
| Introduction |
|---|
|
|
|---|
An important factor in the control of corticosteroid metabolism, particularly in relation to the balance between mineralocorticoid and glucocorticoid hormones, is the enzymatic interconversion of biologically active corticosterone (rodent) and cortisol (humans) to the inactive 11-ketone metabolites, 11-dehydrocorticosterone and cortisone, respectively.3 4 Two distinct isozymes of 11ß-hydroxysteroid dehydrogenase (11ß-HSD1 and 11ß-HSD2) catalyze this reaction. 11ß-HSD2 favors dehydrogenation and inactivates corticosterone and cortisol. The mineralocorticoid receptor, which binds aldosterone and corticosterone with similar affinities, shows specificity for aldosterone only when 11ß-HSD2 is present. Patients or transgenic mice with 11ß-HSD2 deficiency are characterized by hypertension, hypokalemia, and suppressed plasma renin activity and aldosterone concentration.4 5
11ß-HSD1 is expressed in a wide range of glucocorticoid-sensitive tissues, including the liver, kidney, and vasculature. In many intact cells it appears to act predominantly as a reductase, thereby promoting glucocorticoid hormone action by activating 11-dehydro metabolites.6 7 8 However, in cell-free extracts and in at least 1 cell line,9 11ß-HSD1 functions as a dehydrogenase. Clearly, the cellular context plays a major part in determining the overall direction of the reaction.
In the present study we consider whether the activities of 11ß-HSD enzymes differ between LH and LL rats. To assess 11ß-HSD activity in vivo, patterns of urinary steroid excretion were compared in LH and LL rats after injections of cortisol and cortisone. Activity in vitro was measured in liver and kidney microsomes. Urinary volume and electrolyte excretion after steroid injections were used to indicate mineralocorticoid and glucocorticoid responsiveness. Finally, 11ß-HSD mRNA expression was compared in LH and LL tissues by in situ hybridization.
| Methods |
|---|
|
|
|---|
For in vivo studies of renal function and steroid metabolism, individual rats were kept in metabolism cages with free access to food and water. Rats were injected subcutaneously on separate days with saline and then with 1 or 0.5 mg/kg of cortisone or cortisol, respectively, with an interval of 48 hours between steroid injections. Urine was collected for 24 hours after each injection. After preliminary extraction with Sep-Pak C18 cartridges, urinary free steroids were measured by modified radioimmunoassays for cortisol,10 cortisone,11 and corticosterone.12 Urinary sodium and potassium were measured by flame photometry (IL Photometer, model 243).
11ß-HSD mRNA Expression
In situ hybridization techniques were used to measure mRNA
expression, with liver and kidney tissues for 11-ßHSD1 and kidney
tissues for 11ß-HSD2. Tissues for these studies were collected
snap-frozen. Restriction enzymes and RNA polymerases were from Promega
Corporation. A rat HSD1 cDNA13 was linearized with Sty I.
Antisense and sense transcripts were generated with the use of T3 and
T7 RNA polymerase, respectively. A rat HSD2 clone14 was
linearized with appropriate restriction enzymes. Sense and antisense
transcripts were generated from the SP6 and T7 promoter, respectively.
Radioactively labeled RNA transcripts were synthesized and tissue
sections were prepared for autoradiography as
previously described.15
Hybridization signals in the kidney were quantified in key areas of expression. Hepatic mRNA expression was quantified in both the periportal and perivenous regions of the liver. A computer-aided image analysis system (Viewsonic 17PS, Zenith Data Systems) was used to quantify mRNA expression. No significant levels of hybridization with the sense probes were observed.
11ß-HSD Enzyme Activity
Kidneys were sectioned (50 µm) in ice-cold Krebs-Ringer
solution (0.1 mol/L NaCl, 2.5 mmol/L KCl, 2.5 mmol/L
CaCl2, 1.2 mmol/L
KH2PO4, 1.2 mmol/L
MgSO4, and 25 mmol/L
NaHCO3). For each section, cortical (including
inner cortex and outer medulla) tissues were dissected from inner
medullary tissue, and each region was then processed separately.
Tissues were homogenized (Ystral GMbH D-7801) in 4 volumes
of Ringers solution. Microsomal fractions were prepared by
centrifugation (16 000g for 20 minutes,
then centrifugation of the supernatant at
106 000g for 1 hour at 4°C). Dehydrogenase activity in
microsomes (0.25-mL aliquots) was determined as described by Low et
al16 by measuring the conversion of 12 nmol/L
3H-corticosterone
(1,2,6,7-3H-corticosterone, specific activity 86
Ci/mmol; Amersham Life Sciences Ltd) to radioactive
11-dehydrocorticosterone in the presence of
a range of concentrations of cold corticosterone (0 to 10
µmol/L) and either NADP or NAD as cofactor (0.2 mmol/L). After
incubation (37°C, 10 minutes), reactions were terminated by adding 2
mL ethyl acetate. Steroids extracted into ethyl acetate were separated
by high-performance liquid chromatography with
online ß-counting. The percent conversion of
3H-corticosterone to
11-dehydrocorticosterone with the various
concentrations of cold corticosterone was used to estimate maximal
velocity (Vmax) and
Km values.
Statistical Analysis
Results are expressed as mean±SE. Urinary electrolyte and
steroid values were compared by a Mann-Whitney test. In situ
hybridization results and enzyme Vmax
values were compared by ANOVA.
| Results |
|---|
|
|
|---|
40 mm Hg greater than that in LL
rats (119±2 mm Hg). The effects of cortisol and cortisone
injection on urinary volume and sodium and potassium excretion by LH
and LL strains at 4 and 24 hours are shown in Figure 1. Baseline urinary volumes and sodium
and potassium excretion were not different between strains except that
at 24 hours, urinary volume appeared slightly elevated in LL rats. As
expected,17 cortisol injection caused an increase in
urinary volume, with natriuresis and kaliuresis at 4 hours. At 24 hours
only a natriuretic response was seen, which was less in LL
than in LH rats. Cortisone markedly increased urinary volume in LL but
not in LH rats at 24 hours and did not affect electrolyte
excretion.
|
Urinary Steroid Excretion
Corticosterone but not cortisol is the major
endogenous glucocorticoid hormone in rodents. LH and LL
rats were injected with cortisol and cortisone to assess 11ß-HSD
activities independent of changes in endogenous steroid
metabolism. The possibility that injected steroids might
either decrease urinary corticosterone by negative feedback control of
corticosterone synthesis or increase urinary corticosterone excretion
by acting as competitive inhibitors at sites of
endogenous steroid metabolism was also
considered.
Urinary corticosterone values were generally greater for LH than for LL
rats, particularly for samples collected after control and cortisol
injections (Figure 2). Cortisol and
cortisone increased both LL and LH urinary corticosterone values
(P<0.01), which suggests that, at the doses given, both
steroids act as competitive inhibitors of
endogenous steroid metabolism. Cortisol values
in LH and LL samples after saline injections were <1% of those of
corticosterone. Cortisol injection increased cortisol excretion
20-fold in both LH and LL rats. The high ratio of urinary
cortisone:cortisol after control and cortisol injections reflects
greater dehydrogenase activity, which might be interpreted as higher
11ß-HSD2 than 11ß-HSD1 activity. After cortisol injection, there
were no strain differences of urinary cortisone:cortisol ratios or
cortisone excretion, implying that 11ß-HSD2 activities are similar in
LL and LH rats.
|
Increased urinary cortisol after cortisone injection indicates reductase activity, the direction of metabolism favored by 11ß-HSD1 in vivo, although, overall, the high ratio of cortisone:cortisol confirms that 11ß-HSD2 activity is greater than 11ß-HSD1 activity. However, after cortisone treatment, the ratio was less in LL than in LH rats, suggesting reduced 11ß-HSD1 activity in LH rats. This is supported by evidence of 2-fold greater levels of cortisone in the urine of LH than LL rats, although this difference was not statistically significant.
11ß-HSD Enzyme Assays
11ß-HSD activities were measured in microsomes of liver
and of kidney cortex with NADP as cofactor.
Vmax values were lower in LH than LL
samples (Figure 3), but
Km values (µmol/L) were not different (LL
liver, 5.3±0.8; LL kidney, 7.2±1.1; LH liver, 5.8±0.4; LH kidney,
6.4±1.5). No strain difference of 11ß-HSD activity
(Vmax) was seen in microsomes of the kidney
medulla when NAD was added (LL, 1.36±0.06; LH, 1.84±0.22 pmol
corticosterone per milligram protein per hour; n=7);
Km values for corticosterone also were not
different (LL, 0.14±0.01; LH, 0.17±0.01 µmol/L).
|
11ß-HSD mRNA Expression
In situ hybridization studies confirmed that 11ß-HSD1 mRNA is
expressed throughout the liver, with no differences between periportal
and perivenous regions in either strain. Expression of 11ß-HSD1 mRNA
appeared less in LH than in LL tissues
(Table), although the difference was not
significant (P<0.1).
|
Sites of 11ß-HSD1 and 11ß-HSD2 mRNA expression in the kidney are mutually exclusive (Figure 4). 11ß-HSD1 is localized to the outer medulla/inner cortex, whereas 11ß-HSD2 expression is in an inner band of the medulla; neither is highly expressed in the outer cortex or the papilla. Both 11ß-HSD1 and 11ß-HSD2 mRNA appeared lower in LH kidney.
|
| Discussion |
|---|
|
|
|---|
50% greater. These Vmax values
represent tissue enzyme levels and are not an indication of the
preferred direction of the reaction. Second, the ratio of cortisone to
cortisol after injection of cortisone was higher in LH than in LL
urine. Because, in vivo, 11ß-HSD1 and 11ß-HSD2 favor reductase and
dehydrogenase activities respectively, higher urinary
cortisone:cortisol values could, in theory, reflect increased
11ß-HSD2 activity. This seems unlikely because differences were not
seen after saline or cortisol injections; additionally, the increase in
ratio is associated with greater amounts of unmetabolized cortisone.
Since it appears that cortisol and cortisone injections interfere with
corticosterone metabolism, the reciprocal effect of
endogenous steroid levels on the metabolism of
exogenous cortisol and cortisone should also be considered. However, it
is difficult to show whether higher endogenous steroids or
intrinsic differences in steroid metabolic pathways are
primary or secondary causes of increased excretion of unmetabolized
cortisol and cortisone by LH rats. Third, 11ß-HSD1 impairment would
account for differences in renal responses between strains. Cortisone
only caused a diuresis in LL rats, implying that LH rats were
not exposed to active glucocorticoid hormone. Fourth, expression of
11ß-HSD1 mRNA in liver and kidney cortex is reduced in LH compared
with LL tissues. There appeared to be little or no difference in 11ß-HSD2 activities between strains. Neither the urinary excretion of cortisol and cortisone (after cortisol injection) nor the renal activity of 11ß-HSD2 differed between strains. Evidence of a difference in the pattern of electrolyte excretion after cortisol injection was inconclusive. Conventionally, mineralocorticoid hormones cause an antinatriuresis and kaliuresis, whereas glucocorticoids, as found in the present study, cause a natriuresis and kaliuresis.17 18 A smaller natriuretic response was observed in LL rats at 24 hours, but there was no difference in potassium excretion. The relative natriuresis seen in LH rats may relate to inherent differences of blood pressure between strains rather than altered 11ß-HSD activity. Expression of 11ß-HSD2 mRNA in the renal medulla of LH rats was 10% less than that of LL. Because patients with the syndrome of apparent mineralocorticoid excess with a complete lack of 11ß-HSD2 activity exhibit marked sodium retention, the slightly greater natriuresis in LH rats after cortisol injection is incompatible with reduced 11ß-HSD2 mRNA expression.
Other studies of 11ß-HSD enzymes in rat models of genetically determined hypertension have been reported. Increased corticosterone levels in Milan hypertensive rats were associated with decreased hepatic 11ß-HSD1 activity and mRNA.19 There were no differences in renal activities or expression levels. The association between high corticosterone levels and 11ß-HSD1 deficiency is not unexpected. It has been argued that adrenocortical secretory activity and plasma corticosterone concentrations are elevated in 11ß-HSD1 knockout mice to compensate for reduced regeneration of corticosterone from 11-dehydrocorticosterone in glucocorticoid target tissues.5 While this latter study shows that 11ß-HSD1 deficiency alone does not cause hypertension, one might speculate that raised corticosterone, acting through either glucocorticoid receptors or mineralocorticoid receptors in tissues unprotected by 11ß-HSD2, could be a contributory factor when acting in concert with other genetic factors.
Differences in 11ß-HSD activity have been noted in mesenteric arteries20 and kidney and liver fractions of spontaneously hypertensive rats compared with Wistar-Kyoto rats.21 Although a clear distinction between 11ß-HSD1 and 11ß-HSD2 was not made in these studies, both indicated that reductase rather than oxidase activity is favored in spontaneously hypertensive rats. Again, it is possible that the 11ß-HSD phenotype is associated with increased adrenocortical activity.12 The functional significance of 11ß-HSD activity in vascular tissues is interesting. A bidirectional 11ß-HSD1 appears to be expressed in vascular smooth muscle,7 22 whereas 11ß-HSD1 and 11ß-HSD2 are expressed in vascular endothelial cells.23 Glucocorticoids are known to control vascular reactivity,24 and there are reports that inhibition of dehydrogenase activity potentiates the effects of corticosterone on noradrenaline-induced vasoconstriction in aortic rings.25 26 However, it is not clear whether 11ß-HSD1 or 11ß-HSD2 regulates corticosterone activity in vascular tissues or even whether this process is involved in blood pressure control. The recent studies of Takeda et al27 28 have suggested that decreased expression and activity of 11ß-HSD2 in Dahl salt-sensitive rats is a factor in blood pressure determination, although, in other respects, this model of hypertension has little in common with the syndrome of apparent mineralocorticoid excess.
One possibility that should be considered is that altered 11ß-HSD activity is a consequence of hypertension or some other common phenotype. It is perhaps significant that a number of strains of rat, including LH, Milan hypertensive, obese Zucker, Dahl salt-sensitive, and certain strains of spontaneously hypertensive rats, all exhibit evidence of elevated plasma insulin levels in relation to glucose29 30 31 32 33 as well as evidence of impaired 11ß-HSD activity.19 21 27 34 Insulin has been shown to downregulate 11ß-HSD mRNA expression and activity.6 35
In summary, we have clear evidence that 11ß-HSD1 activity is lower in LH than in LL rats, whereas 11ß-HSD2 is relatively unaffected. Raised plasma insulin may be the cause of lower 11ß-HSD activity. Since 11ß-HSD1 acts as a reductase in many tissues, the conversion of inactive 11-dehydrocorticosterone to active corticosterone is impaired, thereby reducing glucocorticoid activity including, perhaps, negative feedback control of hypothalamo-pituitary-adrenal activity. Increased urinary corticosterone excretion in LH rats may reflect compensatory increases in adrenocortical steroidogenesis. It is unlikely that reduced 11ß-HSD1 in LH rats directly causes hypertension, but it may be of secondary importance. Raised corticosterone secretion could potentiate responsiveness to vasoconstrictors in vascular tissues. Alternatively, if 11ß-HSD1 were to act as a dehydrogenase rather than as a reductase in some tissues,9 then lower 11ß-HSD1 might potentiate corticosterone actions.
| Acknowledgments |
|---|
Received March 22, 1999; first decision March 31, 1999; accepted June 23, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
F. Aguilar, M. Lo, B. Claustrat, J. M. Saez, J. Sassard, and J. Y. Li Hypersensitivity of the Adrenal Cortex to Trophic and Secretory Effects of Angiotensin II in Lyon Genetically-Hypertensive Rats Hypertension, January 1, 2004; 43(1): 87 - 93. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Hypertension Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1999 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |