| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology, Division of Medical Sciences, University of Birmingham, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, United Kingdom
Correspondence: Address all correspondence and requests for reprints to: Paul M. Stewart, M.D., FRCP FMedSci, Professor of Medicine, Division of Medical Sciences, University of Birmingham, Queen Elizabeth Hospital, Edgbaston, B15 2TH, United Kingdom. E-mail: p.m.stewart{at}bham.ac.uk
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
The last decade has seen an exponential increase in research focusing on 11ß-HSD1, principally because of its putative role in human obesity and insulin resistance, but also in other diseases in which glucocorticoids have historically been implicated (osteoporosis, glaucoma). These clinical studies have been underpinned by studies in vitro and the manipulation of enzyme expression in vivo using recombinant mouse models. Finally, the molecular basis for the putative human 11ß-HSD1 knockout—cortisone reductase deficiency (CRD)—has recently been described, an observation that also answers a long-standing conundrum relating to the enzymology of 11ß-HSD1.
|
II. Cortisol Metabolism and History of 11ß-HSD1 |
|---|
|
TopAbstractI. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
Cortisol is the principal circulating glucocorticoid in man and is secreted under the control of the hypothalamo-pituitary-adrenal (HPA) axis. Aldosterone, which is secreted from the outer zona glomerulosa predominantly under the control of the renin-angiotensin system through angiotensin II, is the principal mineralocorticoid in man. Glucocorticoids are secreted in relatively high amounts [cortisol, 15 mg/d (4, 5), corticosterone, 2 mg/d (6)] and mineralocorticoids are secreted in low amounts [aldosterone, 150 µg/d (7)].
More than 90% of circulating cortisol is bound, predominantly to the 
2-globulin, cortisol-binding globulin. Only the free fraction is biologically active; the excretion of this "free" cortisol through the kidneys is termed urinary free cortisol (UFF) and represents only 1% of the total cortisol secretion rate. The circulating half-life of cortisol varies between 70 and 120 min. The major steps for cortisol metabolism are:
1. The interconversion of the 11-hydroxyl (cortisol, Kendall’s compound F) to the 11-oxo group (cortisone, compound E) through the activity of 11ß-hydroxysteroid dehydrogenase (EC 1.1.1.146) (8, 9). The metabolism of cortisol and cortisone then follow similar pathways.
2. Reduction of the C4–5 double bond to form dihydrocortisol (DHF) or DHE followed by hydroxylation of the 3-oxo group to form tetrahydrocortisol (THF) and tetrahydrocortisone (THE). The reduction of the C4–5 double bond can be carried out by either 5ß-reductase or 5-reductase to yield, respectively, 5ß-THF (THF) or 5-THF (allo-THF) (10). In normal subjects, the 5ß-metabolites predominate (5ß-THF:5-THF, 2:1). THF and THE are rapidly conjugated with glucuronic acid and excreted in the urine (11).
3. Additional reduction of the 20-oxo group by either 20- or 20ß-HSD to yield - and ß-cortols and cortolones from cortisol and cortisone, respectively. Reduction of the C-20 position may also occur without A ring reduction giving rise to 20- and 20ß-hydroxycortisol (12).
4. Hydroxylation at C-6 to form 6ß-hydroxycortisol.
5. Cleavage of THF and THE to the C19 steroids 11-hydroxy- or 11-oxo-androsterone or etiocholanolone.
6. Oxidation of the C-21 position or cortols and cortolones to form the extremely polar metabolites cortolic and cortolonic acids (13).
Approximately 50% of secreted cortisol appears in the urine as THF, allo-THF, and THE; 25% as cortols/cortolones; 10% as C19 steroids; and 10% as cortolic/cortolonic acids. The remaining metabolites are free, unconjugated steroids (cortisol, cortisone, 6ß- and 20/20ß-metabolites of cortisol and cortisone) (14).
Although some cortisone may be secreted by the adrenal (15), circulating cortisone concentrations are principally dependent upon "oxidative" 11ß-HSD2 and are approximately one fifth those of cortisol (
60 nmol/liter). However, because of the lower binding affinity to cortisol-binding globulin, free cortisone concentrations are similar to those of free cortisol (16, 17).
The biological activity of any glucocorticoid relates, in part, to the presence of a hydroxyl group at position C-11 of the steroid structure. Cortisol and the principal glucocorticoid in rodents, corticosterone, are active steroids whereas cortisone and 11-dehydrocorticosterone, possessing a C-11 keto group, are inactive. Thus, any tissue expressing 11ß-HSDs can regulate the exposure of that tissue to "active" glucocorticoid.
In retrospect, the first appreciation of 11ß-HSD activity came through the discovery by Kendall (18) of cortisone and elucidation of its potent antiinflammatory activity in patients with rheumatoid arthritis. Unbeknownst to him at the time, he had discovered an "inactive" hormone; bioactivity was dependent upon 11ß-HSD activity in the liver, activating cortisol from cortisone. Subsequently, cortisol was characterized as the active ligand, and shortly thereafter the first description of tissues converting cortisol to cortisone was published. These early studies demonstrated significant amounts of 11ß-HSD activity in human placenta (19), kidney (20), and liver (21), although the "set point" of the enzyme varied, with oxidative activity (F to E) predominating in the placenta and kidney and reductive (E to F) in the liver. Isotopic studies (22) and clinical studies measuring F/E levels in patients with renal disease (23, 24) confirmed that the kidney was an important site for cortisol to cortisone conversion. Selective venous catheterization studies indicated significantly lower circulating F/E ratios in renal venous blood compared with peripheral values. In contrast, circulating F/E ratios were much higher in hepatic venous blood, confirming that the liver predominantly converts cortisone to cortisol (25). This is explained by the activity of two distinct isozymes of 11ß-HSD, a predominantly reductive nicotinamide-adenine dinucleotide phosphate reduced (NADPH)-dependent type 1 or "liver" isozyme and a NAD-dependent oxidative type 2 or "renal" isozyme.
The contributions of these isozymes to global cortisol-cortisone interconversion can be assessed clinically through gas chromatographic/mass spectrometry analysis of the principal urinary cortisol and cortisone metabolites. Thus, the ratio of UFF/urinary free cortisone (UFE) accurately reflects the activity of renal 11ß-HSD2. If this ratio is normal, then the ratio of urinary tetrahydrocortisols (5-THF or allo-THF and 5ß-THF) to tetrahydrocortisone (THE) is a useful marker of 11ß-HSD1 activity (26). Other workers extend this ratio to include all cortisol and cortisone metabolites (Fm/Em ratio: THF, allo-THF, cortols, cortisol/THE, cortolones, and cortisone) (27, 28, 29, 30).
|
III. Short-Chain Dehydrogenases/Reductases (SDRs) and Enzymology of 11ß-HSD1 |
|---|
|
TopAbstractI. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
At present, around 3000 members of this family have been identified through the occurrence of several distinct sequence motifs highlighted in Table 1
(35, 37, 38, 39, 40, 41). These primary structure features form essential parts of the nucleotide cofactor-binding region (Rossmann-fold) and the active site (37, 40).
|
G helix), and regions in the C-terminal segment (44, 45). The crystal structures of 27 members of the family have been reported, and their atomic coordinates have been deposited in the Protein Data Bank. All share a nearly superimposable protein-folding arrangement of -helices and ß-strands ( – ß – – ß x 2) to form a Rossmann fold for cofactor binding (46), although this similarity is not present in the substrate-binding pocket (47). A comparison between the conformations of five SDR crystal structures (bacterial 3-, 20ß-HSD, human 17ß-HSD1, bacterial 7-HSD, mouse dihydropteridine reductase, and mouse lung carbonyl reductase) revealed that although there are only 11 fully conserved amino acid residues common to the five structures, the three-dimensional conformation is highly conserved (48). The -helix F interface of human 11ß-HSD1 has been modeled on the crystal structure of Streptomyces hydrogenans 20ß-HSD and has identified similar residues in type 1 that are important in the stabilization of the dimer of 20ß-HSD (49).
B. 11ß-HSD1 enzymology
1. Kinetic analyses.
From the pioneering studies of White and colleagues (50) and Monder and co-workers (51, 52) an 11ß-HSD was purified from rat liver, and an antiserum was raised against the protein and used to clone a rat cDNA although the cDNA sequence was subsequently updated in 2002 (53). This enzyme is microsomal (54), and activity is nicotinamide-adenine dinucleotide phosphate (NADP) dependent; in the cell-free system it behaved mainly as a dehydrogenase, and no reductase activity was detected in the purified preparation. Subsequently, this enzyme was named 11ß-HSD1. Homogeneous enzyme gave rectilinear Eadie plots and Michaelis-Menten (Km) constants of 1.83 ± 0.06 µM for corticosterone and 17.3 ± 2.24 µM for cortisol. First-order rate constants were one order of magnitude higher for corticosterone than cortisol, but maximal velocities were similar (52). Subsequently, cDNAs and proteins were published for the human (55), mouse (56), squirrel monkey (57), sheep (58), rabbit (54), pig, cow, and guinea-pig (59, 60) 11ß-HSD1. Human liver 11ß-HSD1 was eventually purified in an active form and was postulated to exist as a dimer (61). The value of the Km determined for 11ß-HSD1 dehydrogenase activity is puzzling given that it is more than two orders of magnitude higher than the circulating level of free cortisol (1–100 nM). Maser et al. (61, 62) discovered an unusual kinetic mechanism of action of the human liver 11ß-HSD1. They determined that this isoform exhibits Michaelis-Menten kinetics with respect to cortisol but cooperative kinetics with cortisone. In this way, 11ß-HSD1 could operate at both nanomolar and micromolar substrate concentrations. However, using recombinant purified guinea pig and human proteins, no evidence for cooperative kinetics has been found (60). Mouse liver 11ß-HSD1 has been shown to accept nicotinamide-adenine dinucleotide (NAD) as well as NADP as cofactor (63, 64). Guinea pig liver 11ß-HSD1 has been shown to have equal affinity for cortisone and cortisol with apparent Km value in intact cells for both substrates being 3 µM (59) and the purified protein exhibiting a Km value for cortisone of 0.8 µM (60). A summary of the published kinetic analyses of 11ß-HSD1 in differing species is given in Table 2.
|
In every case, however, when cells are disrupted or the enzyme purified, reductase activity is lost. This striking change in directionality between intact cells and homogenates seems to reflect the specific intracellular localization of 11ß-HSD1 within the lumen of the endoplasmic reticulum (ER), where neighboring enzymes may be powerful generators of the reduced cosubstrate NADP phosphate (NADPH). Indeed, studies using purified human enzyme have shown that the equilibrium constant for the E to F direction (defined as the concentration of products divided by concentration of reactants) is 0.03. Given that a figure of 1 would represent the exact equilibrium position, a value of 0.03 indicates a strong preference toward dehydrogenase (F to E) activity (82). Reductase activity can be regained from tissue homogenates and purified enzyme, upon inclusion of a NADPH regeneration system employing the cytosolic enzyme glucose-6-phosphate dehydrogenase (82, 83). This suggests that reductase activity predominates in intact cells due to a high level of NADPH present within the ER lumen. Recently, it has been shown that the enzyme hexose-6-phosphate dehydrogenase (H6PDH) serves this crucial role in generating NADPH levels in the ER (84) (Fig. 1A) (see Section VIII.B).
|
), particularly within the cofactor-binding region (GASKGIG) and the catalytic site (YSASK). The 11ß-HSD1 protein has a single hydrophobic N-terminal extension preceding the cofactor-binding domain, suggesting that this region anchors the protein in microsomes. The precise topology of 11ß-HSD1 was demonstrated using 11ß-HSD1 constructs with attached FLAG epitopes at the N- and C-terminal regions (85). The protein was shown to be intrinsic to the membrane of the ER, having a short five-amino acid region on the cytosolic side of the membrane, followed by a single transmembrane domain (Fig. 1B
) and the majority of the enzyme residing in the lumen of the ER. Chimeric proteins, where the N-terminal regions from 11ß-HSD1 and 11ß-HSD2 were exchanged, led to inverted orientation within the ER. Both chimeric proteins were inactive (85). Within the single N-terminal transmembrane region, the charge distribution of two positively charged lysine residues on the cytoplasmic side and two negatively charged glutamate residues suggests these are crucial residues in the orientation of 11ß-HSD1 in the ER membrane. Mutation analysis of Lys5 residue suggests that it is critical in the determination of 11ß-HSD1 topology and that its charge and specific side chain are both important (85). The importance of the transmembrane domain upon 11ß-HSD1 activity has been studied, but with conflicting results. An N-terminally truncated variant of rat 11ß-HSD1 was expressed in COS cells and reported to be inactive (86, 87). However, this construct encoded a protein that had lost more than just the transmembrane helix, and may, therefore, have lost vital parts of the enzymatic domain. In addition, these expression studies were performed in COS and Chinese hamster ovary cells, where the truncated protein would have been targeted (because of the lack of signal sequence) to the cytosol and not the ER. The lumen of the ER promotes the formation of disulfide bonds, and studies have indicated that there are important intrachain disulfide bonds within the 11ß-HSD1 protein (54). Extraction of the enzyme from the cells by sonication or the addition of fusion proteins to the C terminus (88) possibly disrupting the conformation of the protein structure to the detriment of enzyme activity (47) could also account for the reported requirement of the N-terminal region for activity. However, analysis of 11ß-HSD1 constructs expressed in Escherichia coli demonstrated enzyme activity from an N-terminus-deleted construct, with activity levels even higher than those observed for the full-length 11ß-HSD1 construct (82).
|
C. Substrate specificity and inhibitors of 11ß-HSD1
Numerous studies have been directed toward understanding the effects of various steroid moieties upon 11ß-HSD1 activity because any factors that inhibit metabolism of the 11ß-hydroxyl group will increase glucocorticoid potency. Most studies appear to have been performed with tissue extracts containing 11ß-HSD1 (9). In essence, a substrate for 11ß-HSD1 possesses a flat A/B ring junction (5), with the 5ß conformation disallowed; bulky groups on the -surface inhibit binding, although the effect of -halogens appears to be inductive rather than steric; an aromatic A ring is forbidden, and steroids with bulky groups at C-21 are not substrates. Data indicate the importance of the structural conformation of the A and B rings because modifications to these can confer specific inhibitory properties on some steroidal compounds (92). C ring deoxysteroids, such as chenodeoxycholic acid (CDCA), can also be inhibitors for 11ß-HSD1 (93). Many mammalian steroid dehydrogenases, including 11ß-HSD1, have been implicated in the detoxification of molecules in addition to roles in steroid metabolism (94, 95, 96). In addition to its important endocrinological involvement in glucocorticoid metabolism, 11ß-HSD1 mediates the phase I biotransformation of several carbonyl group-bearing foreign compounds, including xenobiotics (64), drugs (97, 98), insecticides (99, 100), and carcinogens (101, 102). The reductive metabolism of xenobiotic compounds such as metyrapone, p-nitroacetophenone, and p-nitrobenzaldehyde (64) allows the formation of a hydroxyl group rendering the toxic substrate more hydrophilic and more likely to be conjugated by glucuronidation or sulfation facilitating excretion (103). 11ß-HSD1 exhibits other protective roles with the inactivation of carcinogen, nitrosamine 4-methylnitrosamino-1-(3-pyridyl)-1-butanone (NNK), to its secondary metabolite, 4-methylnitrosamino-1-(3-pyridyl)-1-butanol (102), the metabolism of the antineoplastic agent, oracin, into its active metabolite, 11-dihydrooracin (97, 104), the conversion of the nonsteroidal antiinflammatory prodrug, 5,5-dimethyl-3-(3-fluorphenyl)-4-(4-methysulfonyl)phenyl-2-(5H)-furanone-lactol, to its active lactone form (98), and the detoxification of antiinsect agents (azole analogs of metyrapone) (100).
Inhibitors may have properties different from these. An exhaustive list of inhibitors has been compiled and includes steroids with C-21 and 2-methyl substituents (9). The most commonly used inhibitor for in vitro studies and of clinical relevance are the licorice derivatives, glycyrrhizic acid, its hydrolytic product glycyrrhetinic acid, and the hemisuccinate derivative carbenoxolone (CBX). Glycyrrhetinic acid is a potent inhibitor of 11ß-HSD1 (both competitive and inhibiting 11ß-HSD1 mRNA levels) (105, 106), and, in addition, inhibits 11ß-HSD2 with an inhibition constant (Ki) of 5–10 nM (73, 107).
Far fewer steroids have been shown to be inhibitors of 11-oxidoreduction, and obligatory functional groups have not been assigned. Reduction at C-20 eliminates inhibitory activity, but the specific configuration of side chains is not critical as androgens are also potent inhibitors (108). Because the protein sequence of 11ß-HSD1 is not identical between species, subtle differences in protein conformation may lead to differences in substrate or inhibitor efficacy. Indeed, CBX displays little inhibition of ovine 11ß-HSD1 (109), although it inhibits both oxoreductase and dehydrogenase activities in human liver microsomes (93). Clinical studies in subjects consuming glycyrrhetinic acid vs. CBX also suggest that CBX is an inhibitor of 11ß-HSD1 in vivo (110, 111).
Prednisolone and prednisone are substrates for 11ß-HSD1 (112, 113). 9-Fluorinated steroids, such as dexamethasone, are metabolized by 11ß-HSD2 (114) but may also be regenerated by 11ß-HSD1 (115). The inhibitory effects of progesterone, glycyrrhetinic acid, and related compounds on 11ß-HSD1 have been reported, and 5 - and 5ß-adrenocorticoids inhibit 11ß-HSD1. Bile acids are potent inhibitors with lithocholic acid exerting the strongest effect (116). In intact cells 11-hydroxyprogesterone is a more potent inhibitor of 11ß-HSD1 than glycyrrhetinic acid or 11ß-hydroxyprogesterone (117, 118).
D. Selective inhibitors
To date, there are few inhibitor compounds reported to be specific for 11ß-HSD1. As mentioned earlier, a variety of bile acids have inhibitory effects on 11ß-HSD1 with lithocholic acid and CDCA reported as the most potent. However, only CDCA has been shown to be selective for 11ß-HSD1 oxoreductase and dehydrogenase activities (93). The antidiabetic arylsulfonamidothiazole compounds have been shown to inhibit 11ß-HSD1 both in vivo and in vitro (119, 120, 121). The diethylamide derivative was shown to inhibit human 11ß-HSD1 with an IC50 of 52 nM, and an N-methylpiperazinamide form (BVT.2733: 3-chloro-2-methyl-N-{4-[2-(methyl-1-piperazinyl)-2-oxoethyl]-1,3-thiazol-2-yl} benzenesulfonamide) was shown to be specific for the mouse enzyme (IC50 of 96 nM). In the hyperglycemic mouse strain KKA(y), the compound BVT.2733 lowered hepatic phosphoenol pyruvate carboxykinase (PEPCK) and glucose-6-phosphatase mRNA, blood glucose, and serum insulin concentrations (121), raising the possibility that inhibition of 11ß-HSD1 might be used therapeutically to treat patients with insulin resistance (see Section VII.B.3).
|
IV. Molecular Biology of 11ß-HSD1 |
|---|
|
TopAbstractI. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
). Subsequently, 11ß-HSD1 cDNAs have been cloned for a number of species including sheep (58), squirrel monkey (57), mouse (56), baboon (123), and guinea pig (59). Interestingly, and possibly unique among mammalian species, is the Australian koala, which appears to be devoid of 11ß-HSD1 activity in its liver. A study suggests that this may be due to the absence of a gene encoding 11ß-HSD1 activity homologous to that of other known species (124). Alternate 11ß-HSD1 mRNA transcripts, as a result of differential promoter usage and alternate splicing mechanisms, have been demonstrated. In rat kidney, liver, and lung, a transcript initiated from an intron 1 promoter uses methionine 27 in exon 2 as a new start codon, maintaining the reading frame, and has been designated 11ß-HSD1B (86, 87). However, expression of the truncated enzyme did not produce a soluble protein in its native form in cells (87) but was found to be active once released from the ER membrane when overexpressed in yeast (125). However, the precise role of this truncated form is not clear. Additional studies have also revealed a third putative 11ß-HSD1 congener in the sheep arising as the result of the deletion of exon 5 (126). The reading frame is maintained from exon 4 into exon 6, with the loss of 48 amino acids, which includes the catalytic domain (126). Again, no functional significance has been attributed to this transcriptional variant. The three proteins are now referred to as 11ß-HSD1A, 11ß-HSD1B, and 11ß-HSD1C, respectively. The message for 11ß-HSD1B is restricted to the kidney in the rat and parallels the developmental expression of 11ß-HSD1A mRNA (127). At present, there is no evidence that 11ß-HSD1C exhibits enzymatic activity toward glucocorticoids. However, the association of 11ß-HSD1A with carbonyl reductase activity in mouse liver suggests that it may act on substrates including xenobiotics (128, 129). Of note, an expressed sequence tag expressed in human pregnant uterus has been described that represents 11ß-HSD1C; however, the significance and validity of this finding remains unclear (130).
B. Human HSD11B1 gene
Hybridization of the human 11ß-HSD1 cDNA to a human-hamster hybrid panel localized the single HSD11 gene to chromosome 1 (subsequently refined to chromosome 1q32.2–41). Genomic clones of HSD11 were isolated from a chromosome 1-specific library, again using the cDNA as a probe (55). The human gene has been designated HSD11B1 and consists of six exons (182, 130, 111, 185, 143, and 617 bp, respectively) and five introns (776, 767, 120, 25,300, and 1,700 bp, respectively) (Fig. 3).
|
To date, there are few reports describing polymorphism in and around the HSD11B1 locus. A scan of the GenBank single-nucleotide polymorphism (SNP) database (db SNP at http://www.ncbi.nim.nih.gov/SNP/) reveals a number of documented sequence variations detected primarily through human genome-sequencing projects (and our unpublished observations). All but one polymorphism is located in noncoding regions of the gene; 31 SNPs are within intron 4, one SNP is located in the 3'-untranslated region, and seven SNPs are located within 2 kb of the mRNA transcript (three in 5'-regions of the gene and four in 3'-gene regions). An adenine insertion has been detected in intron 3. The coding region SNP is a synonymous C to T change in exon 5 that has no effect upon the encoded amino acid (Ser204Ser). Also, two polymorphic CA repeat microsatellites located at opposite ends of the 25-kb intron 4 of the HSD11B1 gene have been isolated (124). A deletion of 11 bp in intron 1 [position 441–451 of GenBank accession no. M76661 (Exon 1)] was detected in one study. This polymorphism does not alter splicing and does not affect donor or acceptor splice sites (130). Analysis of the polymorphism showed that the 11 bases appear to belong to a tandem repeat consisting of two contiguous repetitions of the same 11 nucleotides, and thus could easily be deleted due to mispairing during replication.
Recently, two further polymorphisms have been identified within intron 3 of HSD11B1 that are in complete linkage disequilibrium: an A insertion (83557), and 40 bp downstream a T to G substitution (83597). Functional analyses showed that these polymorphisms reduce transcriptional activity of HSD11B1 by 2.5-fold in luciferase reporter assays, suggesting that this region of the gene acts as an intronic enhancer of HSD11B1 expression (84).
The rat 11ß-HSD1 promoter has been cloned from genomic DNA (132). An initial study utilizing this sequence demonstrated a single major promoter in the rat liver, but two further promoters are used in the kidney (132). Analysis of the promoter revealed the presence of a CCAAT sequence at –73 to –69 (transcription start site is +1), and the lack of a TATA box. Several putative transcription factor-binding sites were identified including several glucocorticoid response element consensus half-sites, hepatocyte nuclear factor 1, hepatocyte nuclear factor 3, and CAAT/enhancer binding proteins (C/EBP) sites. Also a (CT)26 microsatellite is present at –462 (133).
Recent studies upon the rat 11ß-HSD1 promoter showed that it is predominantly regulated by the C/EBP family of transcription factors, mainly C/EBP. C/EBP coordinately regulates a series of genes concerned with the metabolism of fuels (134), and C/EBP is regulated by glucocorticoids in a tissue-specific manner (135). In liver, basal C/EBP levels are high, ensuring high levels of 11ß-HSD1 transcription, and hence high glucocorticoid levels (80). The 11ß-HSD1 promoter exhibits an unusually large number of these sites, having at least 10, with most genes containing two or three sites. Williams et al. (133) developed a series of reporter plasmids containing increasing sized promoter regions (from –88 to –3.5 kb/+49), transfected into HepG2 cells (human hepatoma cell line). These experiments identified a repressor element between –812 and –754. C/EBP inducibility of the 11ß-HSD1 promoter was most prominent between –579 and –88.
DNase 1 protection analysis identified 11 sites of nuclear protein interaction with the 11ß-HSD1 promoter, and 10 of these can be occupied by C/EBP-related proteins (133). Two of these regions, FP1 and FP2, span the transcription start site between –88 and +76. Mutation analysis of four footprinted sites proximal to the transcription start site FP1, FP2, FP3, and FP4 showed that they are required for full C/EBP inducibility and basal transcription. Mutation of FP2 actually decreased basal transcription levels, suggesting that C/EBP may act here as an initiator (Inr)-binding protein. EMSA analysis confirmed that C/EBP binds to at least two of the footprinted sites (133). Analysis of 11ß-HSD1 liver RNA expression in C/EBP knockout mice confirmed these findings: 11ß-HSD1 was dramatically reduced in these mice compared with wild-type littermates (136).
Similar analysis of the related transcription factor, C/EBPß, showed it to be a weak activator of 11ß-HSD1 transcription, and C/EBPß knockout mice (137) show increased hepatic 11ß-HSD1 mRNA expression (133).
The importance of C/EBP in the regulation of 11ß-HSD1 transcription has been shown only for the rat promoter thus far, and it will be interesting to note whether this translates to the human promoter sequence and whether other transcription factors play a similarly important role.
The analysis of the HSD11B1 gene in patients with CRD and the application of the polymorphisms in linkage and association studies are detailed in Sections VIII and IX.
C. Recombinant models of 11ß-HSD1
To determine the role of 11ß-HSD1 in vivo, transgenic mice with a null HSD11B1 gene have been generated by the replacement of the genomic region containing exons 3 and 4 with a neomycin-resistance cassette via homologous recombination in mouse 129 embryonic stem cells (138). The resulting knockout mice were fertile, had regular litter size with pups of normal birth weight, and postnatal development with normal morphological appearance. There was no deviation from Mendelian inheritance of alleles, and therefore no embryonic lethality associated with this knockout was assumed. No mRNA from 11ß-HSD1 homozygous mutant mice, and approximately 50% mRNA from heterozygous mice was detected by Northern analysis compared with wild type, confirming the true ablation of this gene. In homozygous mutant mice hepatic 11ß-HSD activity was less than 5% of wild type. Wild-type and knockout mice were adrenalectomized and implanted with 11-dehydrocorticosterone pellets. Wild-type mice readily converted 11-dehydrocorticosterone to corticosterone, whereas corticosterone levels in knockout mice remained undetectable, demonstrating that 11ß-HSD1 is the only 11-oxoreductase (at least in the mouse) able to generate active glucocorticoid from inert 11-ketosteroids. 11ß-HSD –/– mice also displayed adrenal hyperplasia due to reduced negative feedback on the HPA axis causing increased ACTH-stimulated corticosterone secretion. The expression and activity of the 11ß-HSD2 enzyme appeared to be unaffected in this model, suggesting no compensatory mechanisms.
An important experiment, which tested the hypothesis that increased 11ß-HSD1 activity within adipose tissue may be implicated in obesity and the metabolic syndrome, was the creation of transgenic mice overexpressing the enzyme (139). This was achieved through the fusion of 5.4 kb of the aP2 promoter/enhancer, which is an adipocyte-specific promoter, and a 1.6-kb fragment of rat 11ß-HSD1 cDNA, followed by an SV40 consensus polyadenylation signal. This construct was microinjected into the pronucleus of fertilized FVB mouse eggs. Successful targeting of transgene expression was determined by RNase protection assay using a probe able to differentiate between transgene-derived and endogenous 11ß-HSD1 mRNA from various adipose tissues and showed relative equivalence in expression in adipose tissue from sc, epididymal, mesenteric, and interscapular brown adipose tissue depots. The transgene was not expressed in nonadipose tissue such as brain and liver of transgenic mice. All mice studies were performed on inbred strains of male FVB mice in which transgene mRNA expression was increased 7-fold compared with endogenous mRNA. 11ß-HSD1 enzyme activity was increased almost 3-fold in adipose tissue, comparable to ob/ob mice or that seen in obese humans (140), demonstrating that the extent of transgenic amplification of 11ß-HSD1 activity is physiologically relevant. Transgenic mice under nonstressed conditions had similar serum corticosterone concentrations as controls, whereas concentrations in adipose tissue were elevated up to 30% higher compared with wild-type mice, reflecting local increased activation of glucocorticoid via 11ß-HSD1.
The detailed phenotype of these animals is discussed in Section VII. However, to further define the role of 11ß-HSD1 upon homeostasis, it may be necessary to develop more refined recombinant mouse models. The use of Cre-LoxP technology in the generation of a conditional HSD11B1 allele, whereby particular gene promoters of Cre recombinase can ablate gene function in a spatial and temporal manner, would provide information on the relative contributions of specific tissues such as liver and adipose to the global effects of 11ß-HSD1 enzyme activity on murine physiology.
|
V. Localization and Ontogeny of 11ß-HSD1 |
|---|
|
TopAbstractI. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
provides a comprehensive list of the tissue-specific distribution of 11ß-HSD1 in different species from which it can be seen that 11ß-HSD1 is expressed in many tissues throughout the body. Expression often occurs in what have traditionally been regarded as glucocorticoid target tissues. Highest levels of expression are seen in the liver, gonad, adipose tissue, and brain, and these are discussed in more detail in Section VII.
|
Within the rat fetal brain, 11ß-HSD1 is undetectable until late gestation. Expression then begins to appear in the hippocampus, precerebellar area, and medulla. Subsequently, a more generalized increase in expression occurs, and at birth expression is highest in the thalamus, neocortex, hypothalamus, pituitary, periaqueductal gray area, spinal cord, and hippocampus (143). However, although expression has been detected using in situ hybridization, there is little detectable enzyme activity (143). In contrast, in primary cultures of rat fetal hippocampal neurons, significant amounts of 11ß-HSD1 reductase activity can be detected (72). Within the postnatal rat cortex and hippocampus, 11ß-HSD1 activity decreases until postnatal d 10 and then increases. Conversely, in the cerebellum a peak of activity is reached by d 10 and then gradually falls until adult levels are achieved (d 15) (144).
In the mouse, 11ß-HSD1 expression is undetectable but rises dramatically after birth until sexual maturity, after which it declines (145).
Ontological expression has also been studied in sheep. In sheep liver 11ß-HSD1 expression is present by midgestation. Levels remain constant until immediately before birth when they increase more than 2-fold. Reductase activity always exceeds dehydrogenase activity (146). In the pars distalis of the sheep pituitary, 11ß-HSD1 expression can be detected by midgestation and, although levels do not change in late gestation, they increase dramatically after birth (147). Both reductase and dehydrogenase activities are present, although dehydrogenase activity predominates. However, in a further study, levels of activity in liver and pituitary were similar in tissues from late gestational fetuses and adults (148). In the pars intermedia, 11ß-HSD1 is only detectable at term, and levels do not change after birth (147). In the ovine placenta 11ß-HSD1 immunoreactivity is observed in the fetal trophoblast cells. Dehydrogenase activity exceeds reductase, and activity decreases in late gestation (149).
Studies in humans are very limited. In fetal lung tissue homogenates, small amounts of dehydrogenase activity have been detected. In the neonatal and infant period, reductase activity is present although this is lost on progression through childhood (150). The significance of these results and, in particular, the source of dehydrogenase activity are uncertain, being published before the elucidation of the two 11ß-HSD isozymes. However, in primary cultures from explants of midgestation human fetal lung, both reductase and dehydrogenase activity are observed (151), suggesting the presence of 11ß-HSD1. In other tissues, however, 11ß-HSD1 could not be identified at least at midgestation (152).
The activity of 11ß-HSD1 in childhood is not well characterized. Cortisone acetate therapy in neonates with congenital adrenal hyperplasia is ineffective up to 2 months of age, reflecting a lack of 11ß-HSD1 reductase (principally hepatic) activity (153). In children aged 4 or 5 yr, activity, as measured by urinary corticosteroid metabolites, is similar in boys and girls (154). In boys, activity remains relatively constant up to, and during, puberty. However, in girls activity decreases around the time of puberty (154), and it is possible that it is at this time point that the well-described sexually dimorphic pattern of activity is obtained. In most studies in both elderly normal individuals (155) and in GH-deficient, hypopituitary patients (28), 11ß-HSD1 activity is higher in men. In rats, although not in mice (56), a similar sexual dimorphic pattern is observed (156). The explanation for this difference in rats is believed to lie in patterns of GH secretion. Hypophysectomized male rats given a continuous GH infusion (as observed in female rats), suppress 11ß-HSD1 in the liver to levels observed in females (156, 157). The explanation in humans is not clear although it seems unlikely to involve regulation by estrogen as differences persist into the postmenopausal period (155). However, one additional study (158), although demonstrating sex differences in A-ring reductase activity, has failed to confirm this relationship.
|
VI. Regulation of 11ß-HSD1 Expression |
|---|
|
TopAbstractI. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
summarizes the studies that have analyzed the regulation of 11ß-HSD1 and details the tissues and species studied. To summarize, glucocorticoids, C/EBP, peroxisome proliferator-activated receptor-
agonists, and some proinflammatory cytokines (TNF, IL-1ß) increase 11ß-HSD1 expression, whereas GH (acting via IGF-I) and liver X receptor agonists inhibit expression. The effect of other factors, including sex steroids, insulin, and thyroid hormone, vary from tissue to tissue and between species (Table 4).
|
|
VII. Role of 11ß-HSD1 in Normal Physiology and Pathophysiology in Peripheral Tissues |
|---|
|
TopAbstractI. IntroductionII. Cortisol Metabolism and...III. Short-Chain...IV. Molecular Biology of...V. Localization and Ontogeny...VI. Regulation of 11ß...VII. Role of 11ß-HSD1...C. Fetoplacental tissuesD. Cardiovascular systemE. GonadF. Central nervous system...G. BoneH. EyeI. Malignant tissuesJ. Immune tissuesK. Other tissuesVIII. CRDIX. HSD11B1 Linkage and...X. ConclusionsReferences |
|---|
11ß-HSD1 is expressed in cultured rat glomerular mesangial cells where it is up-regulated by IL-1ß and TNF and may modulate the antiinflammatory effects of glucocorticoids at this site (164). Renal 11ß-HSD1 is down-regulated in a heritable model of polycystic kidney disease, the cpk mouse (165). The relevance of these observations to human renal physiology or pathophysiology is uncertain.
11ß-HSD activity was demonstrated in the human colon in the early 1980s (166); expression is confined to nonepithelial cells within the lamina propria of the rat colonic mucosa (167). The function of the 11ß-HSD1 enzyme at this site is unknown. A nuclear receptor with a high affinity for 11-dehydrocorticosterone has been postulated to be present within the rat colon (168), and it is possible that 11ß-HSD activity may modulate ligand exposure to such a receptor.
In the skin, 11ß-HSD1 is expressed in the epidermis, and whereas the directional activity of the enzyme at this site has not been established, the potency of topically applied hydrocortisone (as assessed by the skin vasoconstrictor assay) can be increased by glycyrrhetinic acid administration (169). In vitro, reductase activity predominates in human skin fibroblasts, and this is increased by glucocorticoids and inhibited by insulin (68).
B. Liver and adipose tissue
1. Liver.
11ß-HSD1 is expressed in the rodent and human liver, and, in man, the activity of this enzyme confers biological potency upon orally administered cortisone. In the human liver, 11ß-HSD1 is localized centripetally with maximum expression around the central vein (170). Whereas the reductase activity of 11ß-HSD1 appears to be unstable in homogenates in vitro, primary cultures of rat and human hepatocytes indicate exclusive 11-oxoreductase activity (67, 171). In the intact perfused rat liver, activity is predominantly, although not exclusively, reductase. Interestingly, simultaneous perfusion with CBX fails to inhibit activity. However, 7 d of pretreatment with oral CBX decreased reductase activity significantly (172).
In rats (157, 172, 173, 174), but not mice (56), 11ß-HSD1 expression is 18-fold higher in males compared with females (157), an observation that can be explained by the sexual dimorphic pattern of GH secretion (156) (see Section V.B). Estrogens and insulin reduce 11ß-HSD1 expression in the rodent liver (174), but a series of growth factors including TGFß, basic fibroblast growth factor, epidermal growth factor, and hepatocyte growth factor are without effect (171). In the rat 2S FAZA hepatoma cell line, reductase activity is also inhibited by insulin and IGF-I and stimulated by dexamethasone (175). The promoter region of the rat 11ß-HSD1 gene has been cloned and is positively regulated by C/EBP (133) and, to a lesser extent, by C/EBPß. T4 appears to regulate hepatic 11ß-HSD1 mRNA and activity levels (176), although varying effects have been reported in different tissues in rodents and man (171, 176, 177). In man, hyperthyroidism moves the set point of F to E conversion toward E, and studies suggest that this requires a functional thyroid hormone receptor rather than being due to a direct effect of thyroid hormone per se on 11ß-HSD1 (178).
In sheep liver microsomes, metyrapone inhibits 11ß-HSD1 reductase activity (179), and this may provide a further explanation for its inhibitory effects on adrenal steroidogenesis.
Chronic liver disease is associated with deranged cortisol metabolism. Urinary steroid profiles performed on patients with both alcoholic and nonalcoholic chronic liver disease indicate a marked increase in the THF+allo-THF/THE ratio, suggesting either a reduction in renal 11 ß-HSD2 or an increase in hepatic 11 ß-HSD1 oxoreductase activity (180). However, in rats with cirrhosis, both hepatic 11 ß -HSD1 and renal 11 ß-HSD2 were reduced, and this could be explained by the inhibitory action of bile salts (181).
2. Adipose tissue.
Significant expression of 11ß-HSD1, but not 11ß-HSD2, has been found in human adipose tissue (69, 182). Activity is predominantly reductase in nature and is induced by glucocorticoids and proinflammatory cytokines (183, 184, 185). Activity and expression are significantly higher in omental compared with sc preadipocytes (69, 70). The enzyme is induced upon adipocyte differentiation in human adipose tissue cultures (stromal cells to adipocytes). In stromal cell cultures, this is more related to a "switch" in enzyme set point from dehydrogenase (stromal cells) to reductase (adipocytes) without any significant change in 11ß-HSD1 mRNA levels (78). This may be explained upon induction of H6PDH across differentiation (84) (see Section VIII). With the known effect of glucocorticoids on adipose tissue function and distribution, it has been postulated that the enhanced conversion of E to F within omental adipose tissue plays an i
