Rafn Benediktsson

Glucocorticoid exposure in utero: new model for adult hypertension

Hypertension is strongly predicted by the combination of low birthweight and a large placenta. This association could be due to increased fetal exposure to maternal glucocorticoids. Fetal protection is normally effected by placental 11 beta-hydroxysteroid dehydrogenase (11 beta-OHSD), which converts physiological glucocorticoids to inactive products. We found that rat placental 11 beta-OHSD activity correlated positively with term fetal weight and negatively with placental weight. Offspring of rats treated during pregnancy with dexamethasone (which is not metabolised by 11 beta-OHSD) had lower birthweights and higher blood pressure when adult than did offspring of control rats. Increased fetal glucocorticoid exposure secondary to attenuated placental 11 beta-OHSD activity may link low birthweight and high placental weight with hypertension.

Dysfunction of placental glucocorticoid barrier : link between fetal environment and adult hypertension ?

Abstract Low birthweight is associated with the subsequent development of common disorders of adult life, especially hypertension; maternal malnutrition has been suggested as the cause. We suggest an alternative aetiology—increased fetal exposure to maternal glucocorticoids. This hypothesis is supported by our findings that in rats decreased activity of the enzyme that acts as a placental barrier to maternal glucocorticoids (11 β-hydroxysteroid dehydrogenase) is associated with low birthweight. Furthermore, increased exposure of the fetus to exogenous glucocorticoids leads to low birthweight and subsequent hypertension in the offspring. Glucocorticoids acting during critical periods of prenatal development may, like other steroid hormones, exert organisational effects or imprint patterns of response that persist throughout life. Thus, the lifetime risk of common disorders may be partly determined by the intrauterine environment.

Placental 11β‐hydroxysteroid dehydrogenase: a key regulator of fetal glucocorticoid exposure

OBJECTIVE Placental 11β‐hydroxysteroid dehydrogenase (11β‐HSD), which converts active cortisol to inactive cortisone, has been proposed to be the mechanism guarding the fetus from the growth retarding effects of maternal glucocorticoids; however, other placental enzymes have also been implicated. Placental 11β‐HSD is unstable in vitro, and enzyme activity thus detected may not be relevant to the proposed barrier role. We have therefore examined placental glucocorticoid metabolism in dually perfused freshly isolated intact human placentas.

11β-Hydroxysteroid dehydrogenases: Key enzymes in determining tissue-specific glucocorticoid effects

Recent studies have demonstrated that the interconversion of active and inactive glucocorticoids plays a key role in determining the specificity of the mineralocorticoid receptor and controlling local tissue glucocorticoid receptor activation. Two distinct isoforms of the enzyme 11 beta-hydroxysteroid dehydrogenase (11 beta-HSD) have been identified. 11 beta-HSD1 is NADPH-dependent and at its major site of action (the liver) is a reductase, converting cortisone to cortisol (11-dehydrocorticosterone to corticosterone in the rat). 11 beta-HSD2 is NAD-dependent, is present in tissues such as the kidney and placenta, and converts cortisol to cortisone (corticosterone to 11-dehydrocorticosterone in the rat). Congenital or acquired deficiency of 11 beta-HSD2 produces the syndrome of apparent mineralocorticoid excess (SAME) in which cortisol gains access to the unprotected nonspecific mineralocorticoid receptor. The congenital deficiency is associated with mutations in the gene encoding the kidney isoform of 11 beta-HSD2; the acquired form results from inhibition of the enzyme by licorice, carbenoxolone, ACTH-dependent steroids in the ectopic ACTH syndrome, and possibly circulating inhibitors of the enzyme. This paper focuses on recent evidence, which suggest that low levels of placental 11 beta-HSD2 result in increased exposure of the fetus to maternal glucocorticoid and low birth weight. In animal studies using the rat we have shown that birth weight is correlated positively and placental weight negatively with the level of placental 11 beta-HSD. Thus animals with low birth weight and large placentae were those likely to be exposed to the highest level of maternal glucocorticoid. In man a similar relationship was found with birth weight being significantly correlated either with placental 11 beta-HSD activity or with the extent of cortisol inactivation by isolated perfused placental cotyledons. Administration of dexamethasone (which is poorly metabolized by placental 11 beta-HSD2) to pregnant rats resulted in decreased birth weight and the development of hypertension in the pups when adult. The same results were obtained when pregnant rats were given carbenoxolone, an inhibitor of placental 11 beta-HSD2. Low protein diet during pregnancy in the rat resulted in low birth weight of the pups, increased placental weight but decreased placental 11 beta-HSD activity, and adult hypertension. Thus increased glucocorticoid exposure of the fetus secondary to a failure of the normal inactivation of maternal glucocorticoid by the placental may be an important mechanism linking changes in the in utero environment and common adult diseases.

Placental 11β-hydroxysteroid dehydrogenase and the programming of hypertension

Abstract Excessive foetal exposure to glucocorticoids retards growth and “programmes” adult hypertension in rats. Placental 11β-hydroxysteroid dehydrogenase (11β-HSD), which catalyses the conversion of corticosterone and cortisol to inert 11 keto-products, normally protects the foetus from excess maternal glucocorticoids. In both rats and humans there is considerable natural variation in placental 11β-HSD, and enzyme activity correlates with birth weight. Moreover, inhibition of placental 11β-HSD in the rat reduces birth weight and produces hypertensive adult offspring, many months after prenatal treatment with enzyme inhibitors; these effects are dependent upon maternal adrenal products. These data suggest that placental 11β-HSD, by regulating foetal exposure to maternal glucocorticoids, crucially determines foeto-placental growth and the programming of hypertension. Maternal protein restriction during pregnancy also produces hypertensive offspring and selectively attenuates placental 11β-HSD activity. Thus, deficiency of the placental barrier to maternal glucocorticoids may represent a common pathway between the maternal environment and foeto-placental programming of later disease. These data may, at least in part, explain the human epidemiological observations linking early life events to the risk of subsequent hypertension. The recent characterization, purification and cDNA cloning of a distinct human placental 11β-HSD (type 2) will aid the further study of these intriguing findings.

Congenital and acquired syndromes of apparent mineralocorticoid excess

The enzyme 11 beta-hydroxysteroid dehydrogenase (11 beta-OHSD) interconverts cortisol and cortisone. Congenital deficiency of the renal isoform of the enzyme results in hypertension, hypokalemia and suppression of the renin-angiotensin-aldosterone system--the apparent mineralocorticoid excess syndrome (AME). In these patients cortisol acts as a potent mineralocorticoid. Suppression of plasma cortisol with dexamethasone results in natriuresis, potassium retention and reduction in blood pressure. Ingestion of excess liquorice or taking carbenoxolone produces an acquired form of AME. The active component of liquorice is glycyrrhetinic acid (GE) and carbenoxolone is the hemisuccinate derivative. Both GE and carbenoxolone are potent inhibitors of 11 beta-OHSD. In vitro studies have shown that 11 beta-OHSD is present in aldosterone-selective tissues and acts as an autocrine mechanism which prevents cortisol from gaining access to the non-specific mineralocorticoid receptor (MR). Congenital or acquired absence of this enzyme allows cortisol to bind to MR resulting in AME. 11 beta-OHSD also appears to be important in controlling cortisol access to glucocorticoid receptors. Variable placental 11 beta-OHSD may alter foetal exposure to maternal cortisol and affect growth as indicated by the correlation between foetal weight and placental 11 beta-OHSD. Thus the tissue-specific distribution, ontogeny and modulation of this enzyme allows it to dictate glucocorticoid effects in addition to its key role in ensuring the specificity of the MR.

Fetal osteocalcin levels are related to placental 11β‐hydroxysteroid dehydrogenase activity in humans

OBJECTIVE Overexposure to glucocorticoids in utero reduces birth weight and, In animals, leads to persistent hypertension in the offspring. The fetus is normally protected from maternal glucocorticoids by placental 11β‐hydroxysteroid dehydrogenase (11β‐HSD) which catalyses the conversion of Cortisol to Inert cortisone. In adult humans, osteocalcin is a sensitive marker of glucocorticoid exposure. The aim of this study was to determine whether cord blood osteocalcin levels were related to the ability of placental 11β‐HSD to inactivate maternal Cortisol.

Lack of effect of nicotine or ethanol on the activity of 11β-hydroxysteroid dehydrogenase type 2

Low birth weight in combination with a large placenta predicts human hypertension. The pathophysiological link remains unclear, but glucocorticoid excess impairs fetal growth and leads to offspring hypertension. A key controller of fetal glucocorticoid exposure and local tissue availability is 11beta-hydroxysteroid dehydrogenase type 2 (11beta-HSD2). The activity of placental 11beta-HSD2 correlates with fetal growth in animals and humans. Ethanol abuse and smoking are known to retard fetal growth which may relate to altered glucocorticoid action or dynamics. This study has examined whether nicotine or ethanol modulate glucocorticoid action in the placenta or fetus by inhibiting 11beta-HSD2, using clonal cell cultures, freshly isolated dually perfused intact human placentas and placentas from in vivo treated rats. No significant effect on the activity of 11beta-HSD2 by pathophysiologically relevant nicotine or ethanol concentrations was observed. The mechanism of action of nicotine and ethanol relevant to reduced fetal growth requires further study.

Cellular selectivity of aldosterone action: role of 11 beta-hydroxysteroid dehydrogenase

Mineralocorticoid receptors in the distal nephron have no intrinsic specificity for mineralocorticoids over glucocorticoids (cortisol in humans; corticosterone in rodents), but are protected from glucocorticoids by the enzyme 11 beta-hydroxysteroid dehydrogenase, which inactivates these steroids to cortisone and 11-dehydrocorticosterone, respectively. Recent work has demonstrated that the enzyme is expressed as multiple tissue-specific isoforms, some of which catalyse the reverse conversion of cortisone to cortisol. These isoforms may allow 11 beta-hydroxysteroid dehydrogenase to modulate access of ligands to glucocorticoid and mineralocorticoid receptors, as well as to amplify and attenuate tissue responses. 11 beta-hydroxysteroid dehydrogenase-mediated protection of mineralocorticoid receptors fails in congenital 11 beta-hydroxysteroid dehydrogenase deficiency and after inhibition of the enzyme by liquorice. In these circumstances, cortisol-dependent mineralocorticoid excess and hypertension ensue. Recent studies suggest that similar deficiencies of 11 beta-dehydrogenase activity may contribute to pathophysiology in common clinical syndromes, illustrating the potential significance of this novel mechanism for development of hypertension.

Essential hypertension : Should we operate?

In the last year, a hypothesis invoking the central nervous system (CNS) in the pathogenesis of essential hypertension has been reinforced by two publications (Naraghi et al., 1994; Akimuraet al., 1995). Curiously, these seem not to have caused much of a stir in the medical community, although at first sight the implications appear to be that many patients with essential hypertension could potentially be cured by a single operation. We have known for a long time that the CNS plays an important regulatory role in cardiovascular responses and the key role of the rostral ventrolateral medulla (RVLM) in blood pressure control has recently been emphasized (Reis et al., 1994). In 1881 it was shown (Naunyn & Schreiber, 1881) that acute elevation of cerebrospinal fluid (CSF) pressure resulted in significant arterial hypertension. This, the so called Cushing response, is so named after animal experiments performed in the beginning of the century (Cushing, 1902) and was initially thought to be the result of ischaemia in the medulla oblongata, leading to a massive sympathetic discharge. It has since been shown that the Cushing response can be reproduced by mechanical or electrical stimulation of a precisely defined area in the floor of the fourth ventricle (RVLM) (Doba & Reis, 1972), which relays to the spinal intermediolateral column. In the RVLM, 2-adrenergic receptors are highly expressed. These are the target of old-fashioned centrally acting antihypertensive drugs, although so are recently described imidazoline receptors, which may yet prove key in central blood pressure regulation (Head, 1995). The question now raised is whether the Cushing response operates in chronic essential hypertension. Jannetta and colleagues (1985a) reported 53 patients operated on for trigeminal neuralgia or hemifacial spasm, who also had hypertension. They incidentally noted compression, by arterial looping, of the left lateral medulla at the exit of the IXth and Xth cranial nerves in 51 of these patients. In 36 of the 42 who underwent surgical decompression, hypertension normalized or improved post-operatively. Whether relief of the primary condition was in fact the cure for hypertension is open to speculation, but this led to animal experiments, including studies in non-human primates, reproducing neurovascular compression of the left medulla and resulting in hypertension (Jannetta et al., 1985b; Yamamotoet al., 1991). Two studies further supported this hypothesis. First, a retrospective examination of angiographs from 107 patients with essential hypertension found 80% to have an artery that crossed the root entry zone of left cranial nerves IX and X, as compared to only 35% of a matched normotensive control group (Kleineberget al., 1992). Secondly, at post-mortem, all of 24 patients with essential hypertension but none of 10 patients with renal hypertension or 21 normotensive controls had evidence of medullary compression (Naraghi et al., 1992). Both of the two latest additions to the literature (Naraghi et al., 1994; Akimuraet al., 1995) used magnetic resonance imaging (MRI) to examine the postulated posterior fossa microvascular abnormalitiesin vivo. Naraghi and colleagues (1994) studied brain-stem vascular anatomy in 24 patients with essential hypertension, 14 patients with renal hypertension and 14 normotensive individuals. The groups were well age matched and the duration of hypertension in both the renal and essential hypertension groups was similar. The actual blood pressure level or medication were not detailed, nor were we informed about the referral or selection procedure. Although renovascular disease was determined, the possibility of vascular disease elsewhere seems to have been overlooked. Twenty (83%) of those with essential hypertension, 2 (14%) of the renal hypertension controls and 1 (7%) of the normotensive controls had neurovascular compression on the left, while on the right side such compression was observed in 17, 29 and 14% of the groups, respectively. It is worth noting that the two observers (blind re. diagnosis) disagreed about the presence of neurovascular compression in 13% of the essential hypertension group, but nonetheless the findings are striking. The second study (Akimura et al., 1995), involved 32 highly selected patients with essential hypertension, 6 patients with secondary hypertension and 18 normotensive controls. The blood pressures and ages were similar in the two hypertensive groups, while the control group was significantly younger. Again the findings were striking. Neurovascular compression (by the vertebral artery, posterior inferior cerebellar artery or both), on the left was observed in 25 (78%) of the essential hypertensives, compared to 1 (17%) with secondary hypertension and 3 (17%) of the normal controls. As in the study by Naraghi et al. (1994), a striking asymmetry was observed with 13% of the essential hypertensives, none of those with secondary hypertension and 6% of the normotensive controls exhibiting right-sided neurovascular compression. Both groups (Naraghi et al., 1994; Akimuraet al., Commentary