Effects of genetic variation in the aldosterone synthase (CYP11B2) gene on enzyme function.

OBJECTIVE: Evidence suggests that high levels of aldosterone lead to hypertension and increased risk of cardiovascular disease. Around 15% of patients with essential hypertension have a raised aldosterone to renin ratio (ARR) suggesting that aldosterone production is inappropriately high in relation to its principal agonist angiotensin II. This may be due to increased activity of aldosterone synthase caused by genetic variation in the CYP11B2 gene. We screened the coding region of human CYP11B2 for genetic variants and tested their effects on function in vitro. PROTOCOL: Normotensive subjects (n = 69) were screened for sequence variants in the coding region of CYP11B2 by single-stranded conformation polymorphism (SSCP) analysis and sequencing. The effects of nonsynonymous variants on enzyme activity were assessed in JEG-3 cells transiently transfected with wild-type or variant expression plasmids. The conversion of the substrate 11-deoxycorticosterone (DOC) to corticosterone (B) and aldosterone was measured. RESULTS: Twenty variants were detected in CYP11B2 and eight analysed functionally (Arg87Gly, Asn281Thr, Gly288Ser, Lys296Asn, Asp335Asn, Gln404Arg, Ala414Pro and His439Tyr). Corticosterone synthesis was unaltered and aldosterone synthesis reduced in variant Arg87Gly; Asn281Thr increased corticosterone and decreased aldosterone production; Gly288Ser increased corticosterone production and abolished aldosterone production; Lys296Asn reduced both corticosterone and aldosterone production; Asp335Asn increased corticosterone synthesis but did not affect aldosterone production. Variants Gln404Arg, Ala414Pro and His439Tyr showed increases in both corticosterone and aldosterone synthesis compared to the wild-type. CONCLUSION: The study confirms the genetic variability of the CYP11B2 gene and provides us with additional valuable structure-function information.

An influence of variation in the aldosterone synthase gene (CYP11B2) on corticosteroid responses to ACTH in normal human subjects.

OBJECTIVE: Previous evidence suggests that the efficiency of 11beta-hydroxylase is at least partly heritable and also that it may be mildly impaired in essential hypertension. In both cases, assessment of activity was based on the response of 11-deoxycorticosterone (DOC) and 11-deoxycortisol to ACTH. The gene (CYP11B1) coding for this enzyme is highly homologous with and lies a relatively short distance downstream from the gene coding for aldosterone synthase (CYP11B2) on chromosome 8. Two polymorphisms of CYP11B2 have been described. The first involves a change of -344C to T in a putative steroidogenic factor-1 (SF-1) binding site and the other, the intron conversion, an exchange of intron 2 for that of CYP11B1. These polymorphisms are in linkage dysequilibrium. Their effects on 11beta-hydroxylation were studied. METHODS AND RESULTS: Normal subjects (n = 135) were genotyped and those homozygous for either or both the polymorphisms were given ACTH (250 microg, i.v.). Plasma was sampled before and 30 minutes after administration. Basal concentrations of DOC, corticosterone, 11-deoxycortisol and cortisol and responses of corticosterone and cortisol to ACTH were not affected by genotype. However, the responses of DOC (P = 0.002 and P = 0.001, respectively) and 11-deoxycortisol (P = 0.025 and P = 0.002, respectively) were significantly greater in subjects homozygous for SF-1 T and/or intron conversion than in those homozygous for SF-1 C and/or normal intron. CONCLUSIONS: These results indicate different 11beta-hydroxylase efficiencies. Thus, variation in CYP11B2 appears to affect the product of CYP11B1. The mechanism is unclear. The close proximity of the two genes may lead to competition for transcription factors or specific differences in intron 2 may affect transcription. Alternatively, the polymorphisms may be acting as markers for adjacent functional genetic variations.

Familial pattern of corticosteroids and their metabolism in adult human subjects--the Scottish Adult Twin Study.

Corticosteroids are important in the regulation of normal physiology and are key factors in regulating cardiovascular physiology and disease, the development of which is known to have a genetic component. However, there is little information on the extent to which plasma and urine steroid levels are determined by familial and genetic factors. We have examined basal and ACTH-stimulated plasma steroid levels and 24-h corticosteroid metabolite excretion rates in 146 pairs of adult twins [75 monozygotic (MZ); 71 dizygotic (DZ)]. Intraclass correlation coefficients were measured for all variables; several plasma steroid measurements were strongly related in both (MZ) and (DZ) twins, consistent with a familial pattern. These included basal levels of 11-deoxycortisol and aldosterone. ACTH-stimulated plasma aldosterone levels were also significantly correlated, to a significant degree, in both MZ and DZ twins. The index of 11beta-hydroxysteroid dehydrogenase activity (tetrahydrocortisol + allotetrahydrocortisol/tetrahydrocortisone) and of the more specific index of activity of the type 2 isoform of this enzyme (urine free cortisol/cortisone) also correlated, to a similar degree, in DZ and MZ twins. In contrast, for the basal and ACTH-stimulated plasma concentrations and 24-h urine excretion rates of several corticosteroids, there was evidence of significant heritability (H2), in that correlation in MZ twins was greater than in DZ. For example, basal plasma corticosterone concentrations (B) (H2 = 0.44), basal and stimulated 11-deoxycorticosterone concentrations (DOC) (H2 = 0.44 and 0.41, respectively), stimulated 11-deoxycortisol concentrations (H2 = 0.53), and the index of 11beta-hydroxylase activity DOC/B (H2 = 0.49) were all significantly heritable. For the urinary variables, 24-h tetrahydrodeoxycortisol (H2 = 0.59) and free aldosterone (H2 = 0.56) were significantly heritable. Our data provide the first evidence that plasma and urine levels of important glucocorticoids and mineralocorticoids show a strong familial pattern, and in some instances, there is evidence of a genetic component to this. This suggests that corticosteroids have a plausible role in essential hypertension that has a similar heritable component.

Aldosterone excretion rate and blood pressure in essential hypertension are related to polymorphic differences in the aldosterone synthase gene CYP11B2.

Significant correlation of body sodium and potassium with blood pressure (BP) may suggest a role for aldosterone in essential hypertension. In patients with this disease, the ratio of plasma renin to plasma aldosterone may be lower than in control subjects and plasma aldosterone levels may be more sensitive to angiotensin II (Ang II) infusion. Because essential hypertension is partly genetic, it is possible that altered control of aldosterone synthase gene expression or translation may be responsible. We compared the frequency of 2 linked polymorphisms, one in the steroidogenic factor-1 (SF-1) binding site and the other an intronic conversion (IC), in groups of hypertensive and normotensive subjects. In a larger population, the relationship of aldosterone excretion rate to these polymorphisms was also evaluated. In 138 hypertensive subjects, there was a highly significant excess of TT homozygosity (SF-1) over CC homozygosity compared with a group of individually matched normotensive control subjects. The T allele was significantly more frequent than the C allele in the hypertensive group compared with the control group. Similarly, there was a highly significant relative excess of the conversion allele over the "wild-type" allele and of conversion homozygosity over wild-type homozygosity in the hypertensive group compared with the control group. In 486 subjects sampled from the North Glasgow Monitoring of Trends and Determinants in Cardiovascular Disease (MONICA) population, SF-1 and IC genotypes were compared with tetrahydroaldosterone excretion rate. Subjects with the SF-1 genotypes TT or TC had significantly higher excretion rates than those with the CC genotype. The T allele was associated with higher excretion rates than the C allele. However, no significant differences were found in excretion rate between subjects of different IC genotype. Urinary aldosterone excretion rate may be a useful intermediate phenotype linking these genotypes to raised BP. However, no causal relationship has yet been established, and it is possible that the polymorphisms may be in linkage with other causative mutations.

Glucocorticoid receptor polymorphism in genetic hypertension.

The Milan hypertensive strain of rat (MHS) displays abnormalities in both renal function and adrenocortical activity. While the pressor role of the former has been studied in detail, the role of the latter has not yet been clearly evaluated. In the present study, glucocorticoid receptor (GR) binding characteristics in liver cytosol from adult MHS and Milan normotensive controls (MNS) have been investigated. Dexamethasone, aldosterone and corticosterone were bound with lower affinity to cytosol of MHS rats compared with that of MNS rats. This pattern of binding could explain the raised plasma corticosterone concentrations and adrenocortical hypertrophy previously noted in MHS. The coding sequence of MHS and MNS GR genes have been determined. The MHS gene differed in four respects from that of MNS: three silent point mutations and a polymorphic microsatellite region in exon 2. The latter polymorphism has been used in cosegregation studies of F2 hybrids of MHS x MNS. The MHS GR genotype was associated with hypercalciuria and lower blood pressure in female rats and lower body weight in male rats. Although the effect on blood pressure is small, it is consistent with the affinity data. MHS GR genotype cosegregated with lower blood pressure in F2 rats and displayed a lower affinity in binding studies. In conclusion, GR polymorphism may be responsible for differences of adrenocortical function between MHS and MNS. This may lead to a reduction in the blood pressure difference between the two strains.

Glucocorticoid receptor polymorphism, skin vasoconstriction, and other metabolic intermediate phenotypes in normal human subjects.

Genetic variation of the glucocorticoid receptor (GR) locus is associated with differences in blood pressure. To define the intermediate phenotypes associated with this variation, we investigated the biochemical and clinical significance of a BclI restriction fragment length polymorphism of the GR locus in 64 normal male volunteers. Blood samples were genotyped as either AA (homozygous large allele; n = 6), Aa (heterozygous; n = 51), or aa (homozygous small allele, n = 7). Four primary glucocorticoid variables were measured including GR binding characteristics and glucocorticoid-sensitive lysozyme release of leukocytes in vitro and the blanching response of forearm skin to budesonide. A large number of secondary variables (urinary and plasma steroid measurements, blood pressure and indices of body fat metabolism, and routine biochemical and hematological measurements) were also considered. In vivo sensitivity to budesonide was greater in AA than aa individuals (mean +/- SE EC50 values: 13 +/- 5 and 42 +/- 10 ng; P < 0.01). In contrast, leukocytes of AA subjects tended to have lower affinity and reduced sensitivity for dexamethasone, although these effects were not statistically significant. Based on urinary steroid measurements, 11 beta-hydroxysteroid dehydrogenase activity [ratio of tetrahydrocortisol (THF) to tetrahydrocortisone (THE) metabolites] was not affected by genotype. The relative activities of 5 alpha- and 5 beta-reductase activity (allo-THF/THF + THE) appeared lower in AA than aa subjects (0.22 +/- 0.04 cf. 0.33 +/- 0.06; P < 0.005) but were not judged to be significantly different when corrected for multiple comparisons. Single and multivariate analyses were carried out to determine which variables influence GR binding characteristics and glucocorticoid responsiveness and to see whether cardiovascular risk factors (blood pressure and body fat) were influenced by glucocorticoid-dependent functions. Only 15-20% of the variations in the dissociation constant (Kd) and maximum binding capacity (Bmax) were influenced by other variables; plasma cholesterol was the most important for affinity and plasma sodium concentration for binding capacity. Multivariate analysis showed that several factors including GR genotype and urinary cortisol account for 10% of the variation of in vivo responses to glucocorticoid hormones; plasma calcium concentration was the only variable that contributed to in vitro sensitivity of leukocytes to dexamethasone. Glucocorticoid-dependent responses were of negligible importance in determining blood pressure or percentage body fat within the narrow physiological ranges of the present study. We conclude that GR genotype affects steroid sensitivity in a tissue-specific manner because of altered GR function or possibly because of linkage to a locus that controls hormone access to the receptor by influencing steroid metabolism.

In vivo and in vitro effects of carbenoxolone on glucocorticoid receptor binding and glucocorticoid activity.

Carbenoxolone potentiates the mineralocorticoid activity of endogenous glucocorticoid hormones by inhibiting the enzyme 11 beta-hydroxysteroid dehydrogenase, which converts cortisol and corticosterone to inactive 11-oxo-derivatives. We addressed the question of whether glucocorticoid activity is also affected by carbenoxolone. Using a rat model involving low dose corticosterone treatment, we found that carbenoxolone neither potentiated nor inhibited the modest increases in blood pressure or reductions in weight gain caused by steroid treatment. Other indices of glucocorticoid activity including white blood cell number, thymus weight, and down regulation of the glucocorticoid receptor were unaffected. In vitro studies with liver and kidney cytosol preparations indicated that carbenoxolone did compete for 3H-dexamethasone binding sites. Carbenoxolone was 5-10 times more effective than glycyrrhetinic acid, 20-30 thousand times less effective than dexamethasone, and is therefore, approximately 1000 times less effective than corticosterone. Analysis of dexamethasone-binding curves indicated a single class of receptor. We conclude that carbenoxolone at the dose tested does not have intrinsic glucocorticoid activity in vivo, nor does it modulate the activities of corticosterone. Carbenoxolone binds weakly to the glucocorticoid receptor. It is not clear whether this weak affinity accounts for some or any of the direct in vitro effects of high concentrations of carbenoxolone that others have described.

Effects of the glucocorticoid antagonist RU486 in spontaneously hypertensive and Sprague Dawley rats.

The effect of a low dose of the gluco-corticoid receptor antagonist, RU486, was tested in spontaneously hypertensive (SHR) and in Sprague Dawley (SD) rats. In SD rats, RU486 (50 micrograms/day, 18 days) significantly increased growth rate and thymus weight, probably by antagonising the actions of endogenous glucocorticoid hormones. Blood pressure and plasma corticosterone levels were not affected by RU486 at this dose. In leucocytes, RU486 treatment in vivo reduced the number of glucocorticoid binding sites but did not affect binding affinity. In contrast, growth rate and thymus weight were not altered in SHR when treated with the same dose of RU486 for a longer period (60 days). Blood pressure was unaffected as were leucocyte glucocorticoid receptor characteristic. Glucocorticoid receptor binding characteristics for RU486 were similar for liver cytosol of SD and SHR. We conclude that the apparent difference in sensitivity to RU486 between strains is probably not due to differences in interaction of the antagonist with the glucocorticoid receptor but may be caused by differences in pharmacokinetics of RU486 between strains.

Adolescents' perceived ability to say "no" to unwanted sex.

PIP: The authors examined the degree to which adolescents believe they can resist a dating partner's verbal pressure to have sex when they themselves really do not want to. 2472 10th-grade students of mean age 15.5 years from eight public high schools in Dade County, Florida, participated in the study. 49% were male, 27.8% White, 38.2% Hispanic, and 27.7% Black. 50.1% reported having ever experienced sexual intercourse. The female students were found to be more likely than males to believe that they could refuse unwanted sex. No consistent differences were found along ethnic and racial lines. The multivariate analysis identified the following as predictors of the ability to refuse sex: a less-permissive attitude toward sex, the low importance of peer influence, and, for females, a generalized sense of self-efficacy.^ieng

Differences in temperature-sensitive receptor binding of glucocorticoids in spontaneously hypertensive and normotensive Wistar-Kyoto rats.

Glucocorticoid receptor binding was compared in liver cytosol preparations from spontaneously hypertensive rats (SHR) and Wistar-Kyoto (WKY) rats using homologous displacement of [3H]dexamethasone. At 5 degrees C, there was no difference in receptor binding affinity or concentration between strains for dexamethasone, corticosterone or aldosterone. At 37 degrees C, affinity for dexamethasone was lower than at 5 degrees C for both rat strains and decreased with time. However, at this higher temperature, binding affinity in the SHR preparation was consistently higher than in the WKY preparation. The WKY preparation had a higher receptor concentration. The rate of dissociation of the [3H]dexamethasone-receptor complex prepared at 5 degrees C and then incubated at 37 degrees C was rapid but not different between strains. A possible explanation of these results is that the relationship of the heat shock proteins to the receptor heterocomplex is different between strains. Evidence exists of a genetic difference in Hsp 70 between SHR and WKY rats, although its cosegregation with blood pressure has not been established.

Evidence of abnormalities in corticosteroid secretion leading to volume-dependent hypertension in Milan rats.

We examined corticosteroid secretory patterns and their relation to altered salt and water metabolism in Milan hypertensive and normotensive rats. Hypertensive rats had significantly higher blood pressures, exchangeable sodium (hypertensive, 41.2 +/- 0.3 mmol.kg-1; normotensive, 38.4 +/- 0.03 mmol.kg-1, P < .001), plasma volume (hypertensive, 5.39 +/- 0.12 mL.100 g-1; normotensive, 4.84 +/- 0.10 mL.100 g-1, P < .001), and plasma concentrations of atrial natriuretic peptide (hypertensive, 38.8 +/- 4.0 pg.mL-1, normotensive, 22.4 +/- 3.1 pg.mL-1, P < .02). These features coincide with those of mineralocorticoid-induced hypertension. Adrenal venous secretory rates (picomoles per minute) of corticosterone (hypertensive, 1696 +/- 202; normotensive, 873 +/- 139), 18-hydroxycorticosterone (hypertensive, 49.7 +/- 8.3; normotensive, 25.7 +/- 3.3), and aldosterone (hypertensive, 1.16 +/- 0.17; normotensive, 0.52 +/- 0.08) were higher in the hypertensive than the normotensive strain, but that of 11-deoxycorticosterone (DOC) (hypertensive, 94.4 +/- 14.9; normotensive, 114.3 +/- 33.9) was similar in the two strains. The corticosterone-DOC, 18-hydroxycorticosterone-DOC, and aldosterone-DOC ratios were higher in the hypertensive than the normotensive strain (P < .02), but the 18-hydroxycorticosterone-corticosterone and aldosterone-18-hydroxycorticosterone ratios were not. These results indicate increased activity of the "late" aldosterone biosynthetic pathway in the hypertensive compared with the normotensive strain caused by an increased conversion rate of DOC to corticosterone. The comparison of corticosterone secretion between the two strains indicates that 11 beta-hydroxylase rather than aldosterone synthase activity is more active in the hypertensive than the normotensive rats.

Inhibition of lysozyme synthesis by dexamethasone in human mononuclear leukocytes: an index of glucocorticoid sensitivity.

Glucocorticoids inhibit translation of the lysozyme gene. This effect may be the basis of an improved method of measuring glucocorticoid responsiveness in human tissues. We have compared lysozyme synthesis in various types of white blood cells and examined the specificity of inhibitory responses to various steroid hormones. The dose-related effects of the glucococorticoid receptor antagonist RU486 on dexamethasone responses were also assessed. Glucocorticoid receptor binding in mononuclear leukocytes (HML) was characterized by homologous displacement of [3H]dexamethasone and compared with the dose-related inhibitory effect of dexamethasone on lysozyme synthesis. Lysozyme activity was measured photometrically as the ability to cause lysis of Micrococcus lysodeikticus in the medium. The greatest effect of dexamethasone was observed after 72 h of culture. Qualitatively similar effects of dexamethasone were observed on cell lysozyme content and lysozyme activity in the medium, but for convenience, activity in medium, rather than cell content, was measured in subsequent assays. Lysozyme activities in various cell types prepared from the blood of healthy volunteers were ranked as follows: polymorphonuclear cells > monocytes > mononuclear cells > lymphocytes. However, dexamethasone inhibited lysozyme synthesis to a similar degree for all types. As mononuclear cells are more conveniently prepared in greater yield compared with other cells, this HML fraction formed the basis of a method of assessing glucocorticoid responsiveness and sensitivity. Lysozyme activity from HML was not significantly affected by incubation with 1 mumol/L estradiol, progesterone, dehydroepiandrosterone, or aldosterone. Dexamethasone and cortisol at 1 mumol/L both inhibited release by 45-50%. Although RU486 when added alone partially inhibited lysozyme activity, the same concentration (1 mumol/L) antagonized glucocorticoid responses and shifted the IC50 and threshold values for the effect of dexamethasone from 1.2 nmol/L to more than 1 mumol/L and from less than 1.0 to 19 nmol/L, respectively. The equilibrium dissociation constants (Kd) for dexamethasone binding to the glucocorticoid receptor ranged from 2.8-12.5 nmol/L and were positively correlated with dexamethasone IC50 values for lysozyme synthesis (r = 0.57; P = 0.002). In conclusion, the inhibition of lysozyme synthesis by dexamethasone in human mononuclear cells is a convenient and specific method of measuring responsiveness to glucocorticoids.

The role of glucocorticoid activity in the inheritance of hypertension: studies in the rat.

Young (3-week-old) spontaneously hypertensive rats (SHR) had significantly higher basal plasma corticosterone levels than WKY rats and maximum responses to ACTH were also higher. In isolated adrenocortical cells from these rats, corticosterone production was also more responsive to ACTH in SHR. There was no significant difference in aldosterone production. Mononuclear leucocytes from older (10-week-old) SHR had a higher affinity for dexamethasone but a smaller number of binding sites per cell. The SHR therefore has higher circulating glucocorticoid levels and the target cells have a higher apparent affinity for this agonist. However, the target cells also have a smaller binding capacity. The precise resultant effect of these changes on glucocorticoid activity will require additional studies on specific glucocorticoid-dependent variables.

Effect of the benzodiazepines diazepam, des-N-methyldiazepam and midazolam on corticosteroid biosynthesis in bovine adrenocortical cells in vitro; location of site of action.

Diazepam and midazolam inhibited cortisol and aldosterone synthesis in bovine adrenal cells in vitro. The biologically active metabolite des-N-methyldiazepam did not. Midazolam was a more potent inhibitor (IC50: 6 micrograms/ml) than diazepam (IC50: 13 micrograms/ml) in ACTH-stimulated cells. Both compounds inhibited steroidogenesis at several points in the biosynthetic chain; the greatest effects were on 17 alpha hydroxylation and 21 hydroxylation. Diazepam had a relatively greater effect on 17 alpha hydroxylation; midazolam on 21 hydroxylation. Both were less potent inhibitors of 11 beta hydroxylation and had little apparent effect on side chain cleavage. Thus microsomal hydroxylation is more vulnerable to benzodiazepines than mitochondrial hydroxylation. It is suggested that the drugs act by competing with steroid mixed function oxidases for cytochrome P-450. The plasma concentrations required for these effects are high in relation to therapeutic levels but may be achieved, for example, during acute infusions or when they are used in combination with imidazole drugs such as cimetidine.

The relative importance of glucocorticoids and mineralocorticoids in the development of adrenal regeneration hypertension in rats.

The adrenocortical tissue which regenerates after adrenal enucleation, and contralateral uninephrectomy and adrenalectomy, resembles histologically zona fasciculata tissue which normally synthesises glucocorticoids. However, increases in blood pressure after enucleation (adrenal regeneration hypertension-ARH] were preceded by a rise in exchangeable body sodium similar to that found with mineralocorticoid-induced hypertension (e.g. DOC/salt rat model). Glucocorticoid involvement in ARH rats was tested, firstly by infusing dexamethasone into control and ARH rats to see whether ACTH suppression would lower blood pressure by reducing adrenocortical activity and, secondly, by infusing dexamethasone into rats with intact adrenals to see whether conditions for ARH (i.e. uninephrectomy and/or saline consumption) pre-disposed rats to the hypertensinogenic properties of glucocorticoids. Low-dose dexamethasone infusions (10 micrograms/day for 5 days) in ARH rats did not affect blood pressure but in control animals caused a significant (P less than 0.01) increase from 128 +/- 3 to 151 +/- 5 mmHg. Corticosterone, 18-hydroxycorticosterone and deoxycorticosterone plasma concentrations were suppressed in both groups by dexamethasone treatment; plasma renin concentrations were lower in ARH rats than in controls. Uninephrectomy or 1% NaCl as drinking fluid did not affect the blood pressure rise induced by sc infusion of 10 micrograms dexamethasone/day for 14 days in rats with intact adrenals. The temporal relationship between blood pressure changes and exchangeable body sodium in ARH rats resembles that in mineralocorticoid-induced hypertension. Glucocorticoid, unlike mineralocorticoid, induced hypertension is not affected by a reduction in renal mass or increased sodium intake.

Inhibition of corticosteroid 11 beta hydroxylation in bovine zona fasciculata cells by the potassium entry blocker 4-aminopyridine.

The potassium entry blocker, 4-aminopyridine has been used to evaluate the importance of this cation in several tissues including the adrenal cortex. Since it inhibited corticosteroidogenesis, an important role for potassium entry in the control of this process has been implied. In the present in vitro study, 4-aminopyridine inhibited cortisol and corticosterone production and, at high concentrations, that of 11-deoxycortisol and 11-deoxycorticosterone. However, at lower concentrations, 11-deoxycorticosteroid production increased suggesting that 4-aminopyridine acts as an inhibitor of 11 beta hydroxylation. Its slight resemblance to metyrapone may suggest that it acts, at least in part, by competing for cytochrome P450. However, its use as a potassium inhibitor in studies of the adrenal cortex may give ambiguous results.

The effect of 5-fluorouracil on rat adrenal function.

This study was undertaken to determine the effects of 5-fluorouracil on corticosteroidogenesis and adrenal size in the rat. The rats were injected intraperitoneally with 5-FU (12.5 mg kg-1 or 25 mg kg-1), normal saline or actinomycin D (0.02 mg kg-1), daily for 10 days. High-dose 5-FU induces adrenal hyperplasia, in association with mild impairment of corticosteroidogenesis (manifest by lower corticosterone and higher ACTH levels). The response to low-dose 5-FU and actinomycin D is of lesser adrenal hyperplasia, relative to high-dose 5-FU (P less than 0.05) and elevated corticosterone levels. There may be a dose-related effect on the suppression of corticosteroidogenesis in the rat by 5-FU.