| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6084
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
-Hydroxysteroid Dehydrogenases -HSD cDNAs -HSD
genes -HSD deficiencies -Hydroxysteroid Dehydrogenases -HSD cDNAs -HSD
genes |
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
Because HSDs catalyze bidirectional reactions, it has been difficult to
understand how they can be involved in both the synthesis and
inactivation of steroid hormones. cDNA cloning indicates that each HSD
exists in multiple isoforms, which show tissue specificity in
expression; this, coupled with the properties of each isoform
(reductase or dehydrogenase), can determine the role of the enzyme in
steroid hormone action. Advances in cDNA isolation have led to the
concept that HSDs belong to at least two distinct protein phylogenies:
the short-chain dehydrogenase/reductase family (SDR; formerly
short-chain alcohol dehydrogenase) whose members include the
3ß-HSD/ketosteroid isomerase (3ß-HSD/KSI), 11ß-HSD, and 17ß-HSD
(1, 2), and the aldo-keto reductase (AKR) superfamily (3, 4, 5, 6) whose
members include 3-HSD and 20-HSD. Three-dimensional structures
now exist for mammalian HSDs that belong to each of the two families
and serve as templates for structure-function studies on HSDs within
each protein family. These represent the first available structures for
mammalian enzymes involved in steroidogenesis and steroid metabolism.
Because of their central role in steroid hormone action, HSDs as a group of enzymes are considered therapeutic targets. Several reviews (7, 8) have been written on the design of HSD inhibitors, and that material will not be reviewed here. It is sufficient to say that specificity in drug design may depend on the successful targeting of a tissue-specific isoform.
|
II. 3ß-Hydroxysteroid Dehydrogenase/Ketosteroid Isomerase (3ß-HSD/KSI) |
|---|
|
TopAbstractI. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
4-3-ketosteroids. The enzyme catalyzes two reactions,
the dehydrogenation of 3ß-equatorial hydroxysteroids and the
subsequent isomerization of the 5-3-ketosteroid products
to yield the ,ß-unsaturated ketones. The majority of steroid
hormones, with the exception of estrogens and the 5-reduced
androgens, contain this functional group. Since these latter hormones
are themselves derived from 4-3-ketosteroids,
3ß-HSD/KSI is required for all steroid hormone biosynthesis. It
should be noted that, in the placenta and corpus luteum, 3ß-HSD/KSI
is responsible for the final steps in progesterone biosynthesis
necessary for the maintenance of pregnancy (Fig. 1
). In
the adrenal cortex the 3ß-HSD/KSI plays a pivotal role in the
formation of glucocorticoids and mineralocorticoids.
|
B. Cloning and expression of the 3ß-HSD/KSI cDNAs
The purification of the dual functional 3ß-HSD/KSI from human
placenta microsomes (14) led to the development of a polyclonal
antibody, which enabled the screening of a human placental cDNA
gt-11 expression library (9). This was a remarkable achievement
because numerous investigators had failed to solubilize and purify the
3ß-HSD/KSI activities from either adrenal or placental microsomes.
Because the cDNA for human placental 3ß-HSD/KSI was the first to be
isolated, it was termed type I 3ß-HSD/KSI. The cDNA for the type I
enzyme encodes a protein of 372 amino acids with a calculated molecular
weight of 42,216. The type I enzyme was transiently expressed in
mammalian cells and catalyzed the conversion of dehydroepiandrosterone
and pregnenolone to produce androstenedione and progesterone,
respectively (10, 11). In addition, when transfected cell homogenates
were supplemented with NAD+, they efficiently oxidized
5-androstane-3ß,17ß-diol to 5-dihydrotestosterone (5-DHT),
and upon addition of NADH the homogenates reduced 5-DHT back to the
diol. These data ended an ongoing and lengthy debate by establishing
unequivocally that a single polypeptide chain could catalyze both the
dehydrogenase and isomerase reactions. The availability of this cDNA
clone led to the isolation of a cDNA for human adrenal or type II
3ß-HSD/KSI (11). Human type I and type II 3ß-HSD/KSI share in
excess of 90% sequence identity. Transient expression of these two
isoforms revealed that the catalytic efficiency
(Vmax/Km) of type I is 5.9-, 4.5-, and 2.8-fold
higher than that of the type II 3ß-HSD using pregnenolone,
dehydroepiandrosterone, and dihydrotestosterone (reverse direction) as
substrates, respectively (11). The higher Michaelis-Menten constant
(Km) values observed for type II 3ß-HSD, which is
expressed mainly in steroidogenic tissues, could be related to the
higher levels of endogenous substrates present in these tissues (15).
Subsequently, four subtypes of rat 3ß-HSD/KSI have been cloned (type
I through type IV); because numbers were assigned in order of
discovery, the rat isoforms do not correspond to the human isoforms by
number (16, 17, 18). Rat type I and type II 3ß-HSD/KSI share 93.8%
sequence homology, yet the type I enzyme has a relative specific
activity that is 64- and 46-fold higher than the type II enzyme for
pregnenolone and dehydroepiandrosterone, respectively. The lower
activity of the type II enzyme is due to a change in four amino acids
in a membrane-spanning region predicted to exist by hydropathy plots
(16). Construction of chimeric 3ß-HSDs, in which the
membrane-spanning domain was reinserted into the type II enzyme,
restored the specific activity of the protein (16). Rat type III
3ß-HSD/KSI appears to be found exclusively in the liver where it
appears to function predominantly as a liver-specific 3-ketoreductase
to inactivate steroid hormones (17). Rat type IV 3ß-HSD/KSI is a
novel form found in the skin but, remarkably, is the predominant form
found in rat placenta. Ribonuclease (RNase) protection studies have
provided a thorough analysis of the tissue distribution of these
isoforms (18). Such studies have shown that rat type IV is equivalent
to human type I 3ß-HSD/KSI (placental form), and rat types I and II
are related to human type II 3ß-HSD/KSI (adrenal form). Since the
isolation of the human and rat clones, cDNA clones encoding for the
murine adrenal (19), bovine ovary (20), and macaque ovary (21) have
been isolated. Initial pair-wise comparison of these sequences indicate
that they share in excess of 80–90% homology with each other but
little or no homology with other HSDs (Table 1).
|
Interest in the regulation of the ovarian 3ß-HSD/KSI exists because it catalyzes the final step in progesterone biosynthesis. Down-regulation of this enzyme may contribute to declining progesterone levels, which is a sign of luteolysis. In rat ovary, 3ß-HSD/KSI mRNA and enzyme activity are decreased by ovine PRL and up-regulated by human CG (hCG) (24, 25). These responses are physiologically relevant because PRL stimulates luteolysis whereas hCG is luteotrophic. Changes in mRNA were confirmed by in situ hybridization studies using luteinized porcine granulosa cells. The inhibitory effect of PRL on 3ß-HSD expression and activity is correlated with a progressive decrease in serum progesterone concentration while serum pregnenolone levels are elevated. These findings indicate that inhibition of 3ß-HSD/KSI gene expression in the corpus luteum occurs early in the luteolytic process induced by PRL and could well play a role in PRL-induced luteolysis.
In luteinized porcine granulosa cells, gonadotropins, as well as agents that increase intracellular cAMP accumulation (cholera toxin, forskolin, (Bu)2cAMP), increased 3ß-HSD mRNA levels (25). In contrast, activation of the protein kinase C pathway induced cAMP accumulation but led to a marked inhibition of the stimulatory effect of hCG, LH, forskolin, and cholera toxin on 3ß-HSD mRNA levels (24). The cross-talk that occurs between these two signaling pathways to regulate 3ß-HSD/KSI expression in granulosa cells remains to be elucidated.
In the rat testis, 3ß-HSD/KSI is required for androgen biosynthesis. Not surprisingly, hCG, acting as a mimic of LH, caused a 3.0-, 19.7-, and 11.5-fold increase in steady state levels of testicular 3ß-HSD mRNA, immunoreactive 3ß-HSD/KSI protein, and enzyme activity, respectively, in hypophysectomized male rats (15).
In rat placenta, the temporal modulation of type I, II, and IV 3ß-HSD/KSI has been measured during the second half of gestation. It will be recalled that rat type IV 3ß-HSD/KSI is most similar to human placental type I 3ß-HSD/KSI. It was found that type I and II 3ß-HSD mRNA levels are modulated in a closely parallel manner, peaking sharply on day 15 but falling abruptly thereafter. On the other hand, type IV 3ß-HSD mRNA levels remained elevated and were stable from days 10–17 of gestation, suggesting that this form of the enzyme is responsible for the substantial increase in 3ß-HSD and hence progesterone production that occurs in midpregnancy (15).
Collectively, the studies on the regulation of rat tissue-specific isoforms of 3ß-HSD/KSI indicate that expression is under the control of trophic hormones and that alterations in expression occurs in response to the need to change circulating steroid hormone levels. The ability of the appropriate trophic hormone to increase 3ß-HSD/KSI mRNA levels and protein and enzyme activities on demand in the adrenal, ovary, and testis points to common mechanisms of regulation that involve cAMP.
To gain insight into the regulation of the human 3ß-HSD/KSI genes,
type I and type II genes have been cloned (26, 27). Each gene is 7.8 kb
in length and consists of four exons and three introns. Both genes are
localized to chromosome 1p13.1 (Fig. 2). Both genes
encode for transcripts that are 1.7 kb in length. Because the type I
gene is expressed at high levels in syncyctial trophoblasts and the
type II gene is expressed at high levels in the adrenal cortex, gene
regulation studies were performed in human choriocarcinoma (JEG-3)
cells and human adrenocortical tumor cell lines H295 cells,
respectively.
|
In human adrenal cells, activators of the protein kinase
A-signaling pathway [e.g. forskolin,
(Bu)2cAMP] enhance the production of
dehydroepiandrosterone and androstenedione with a concomitant increase
in steady state levels of type II 3ß-HSD and 17-hydroxylase
(CYP17). In contrast, phorbol ester treatment dramatically elevated
3ß-HSD mRNA but attenuated 17-hydroxylase mRNA. The 5'-flanking
region of the 3ß-HSD type II gene was found to contain a consensus
sequence for the orphan nuclear receptor steroidogenic factor 1 (SF-1).
The functionality of the SF-1 site was tested in
promoter-chloramphenicol acetyl transferase constructs in which the
SF-1 site was either present or deleted. It was found that the phorbol
ester response had an absolute requirement for the SF-1 element, and
that the response was further enhanced if the cDNA encoding for SF-1
was cotransfected (29).
In an attempt to identify cis-regulating elements that may control tissue-specific expression of the type I and type II 3ß-HSD/KSI genes in the placenta and adrenal, respectively, it was found that a strong positive regulatory element exists in the first intron of the type II gene. This 3ß 1-A element was able to drive transcription of a reporter gene in both placenta cells (JEG-3) and adrenal cells (SW13) and was found to bind four proteins including SP-1 (30). Therefore, this intronic element appears to be important for gene transcription but is not responsible for the tissue-specific expression of the type II gene (30).
D. 3ß-HSD deficiencies
Classic 3ß-HSD/KSI deficiency can cause the salt-wasting and
non-salt-wasting forms of congenital adrenal hyperplasia due to
cortisol deficiency with or without aldosterone deficiency,
respectively. It is estimated that defects in this enzyme are
responsible for 10% of the patients presenting with congenital adrenal
hyperplasia and that the remainder have defects in either the steroid
21-hydroxylase (CYP21) or 17-hydroxylase/20-lyase genes (CYP17). In
its most severe form, 3ß-HSD deficiency results in the blockade of
steroidogenesis in both the adrenal and gonads, and as a result
there is an elevated steroid-5-ene to steroid-4-ene ratio. This
manifests itself in males as pseudohermaphroditism and in females as
virilization at birth. Since the human type II 3ß-HSD gene is the
predominant form in these tissues, it is not surprising that the
defects occur in this gene product. Simard and colleagues (31, 32, 33)
analyzed 12 male and four female patients who had defects in this gene.
It was found that there were 14 unique point mutations in the type II
3ß-HSD gene. In the most severe form of the disease, which is
salt-wasting, the mutations detected were as follows: W171Stop
(nonsense), which may coexist with an insertion of a single C between
codon 186 and 187, which leads to a truncated protein at 202; deletion
of K273 (frameshift) and point mutations Q142K; Y253N; G15D; L108W; and
P86L (missense) (34, 35, 36). In the more moderate form of the disease, the
non-salt-wasting disease, the following point mutations (missense) were
detected: A245P; G129R; N100S; Y254D; L173R; and A82T (31, 37, 38).
Several of these mutations were introduced into the type II
3ß-HSD/KSI cDNA, and the resultant mutant proteins were expressed in
COS-1 cells. Kinetic analysis of the mutants revealed that of the point
mutations present in the salt-wasting form, the G15D and the G15A
mutants showed a 4-fold decrease in catalytic efficiency for the
conversion of pregnenolone to progesterone, and the L108W and P186L
mutants resulted in a decrease in catalytic efficiency of 40-fold
(Table 2). The reduction in catalytic efficiencies
observed with the G15D and G15A mutants has been partially explained by
the increase in Km for NAD+. This glycine
residue is predicted to form part of a Gly-XXX-Gly-X motif found in the
Rossmann fold for cofactor binding (39)(see Sections VII and
VIII). Interestingly, the N100S mutant found in the
non-salt-wasting form of the disease resulted in a 30-fold decrease in
catalytic efficiency. The similar decreases in catalytic efficiency
observed with mutations found in the salt and non-salt wasting forms of
the disease indicate that they alone cannot explain the genetic lesions
responsible for the differences in the two diseases.
|
|
III. 17ß-Hydroxysteroid Dehydrogenases |
|---|
|
TopAbstractI. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
). When these actions occur in androgen or
estrogen target tissues, 17ß-HSD subtypes are in a unique position to
control ligand occupancy of the respective steroid hormone receptor and
could work as molecular switches. Such actions of 17ß-HSD may control
the growth of hormonally dependent cancers of the prostate and mammary
gland (15, 43, 44). There is thus a need to develop isozyme-specific
17ß-HSD inhibitors, and this area has been reviewed recently (8).
B. Cloning and expression of the 17ß-HSD cDNAs
Different isozymes have been designated for 17ß-HSD (types I-IV)
and are named according to the chronological order in which their cDNAs
were cloned. The type I 17ß-HSD is the soluble form originally
purified from human placenta by Jarabak et al. (41, 42); it
converts estrone to 17ß-estradiol using NADPH as cofactor and is also
known as estrogenic 17ß-HSD. This is the principal activity of this
enzyme since it displays a 100-fold higher affinity for C18 than for
C19 steroids. It functions as a homodimer and displays modest 20-HSD
activity. Its availability led to the development of a polyclonal
antibody that was used to screen a bacteriophage expression library.
The cDNA clone for human type I 17ß-HSD was isolated by three groups
(45, 46, 47) and encodes a protein of 327 amino acids. It contains no
obvious membrane-spanning domains. When the recombinant enzyme was
transiently expressed in cultured mammalian cells, it almost
exclusively catalyzed the reduction of estrone to 17ß-estradiol as
predicted by the earlier biochemical studies (48). The enzyme has also
been overexpressed in Baculovirus-infected Sf9 insect cells,
and the resultant recombinant protein retained the properties of the
purified enzyme (49, 50).
Type II 17ß-HSD is a microsomal form and uses NAD+ as
cofactor. It catalyzes the oxidation of testosterone and
17ß-estradiol to form androstenedione and estrone, respectively. Its
principal function is to inactivate circulating androgens and
estrogens. It is located in the liver, small intestine, secretory
endometrium, and placenta, which would be consistent with its role in
steroid hormone inactivation (40). The cDNA for type II 17ß-HSD was
isolated by expression cloning of a cDNA library from human prostate
mRNA using cytomegalovirus promoter constructs (51). The cDNA encodes a
protein of 387 amino acids, which contained an amino-terminal signal
anchor motif and a carboxy-terminal endoplasmic reticulum retention
motif that would locate the enzyme in the membrane. The presence of a
hydrophobic amino acid terminus in type II 17ß-HSD distinguishes the
enzyme from the type I 17ß-HSD, which has a hydrophilic amino acid
terminus. When the cDNA for type II 17ß-HSD was expressed in human
embryonal kidney 293 cells, it also displayed modest 20-HSD activity
and converted 20-hydroxyprogestrone to progesterone (51), a reaction
that produces an active progestin.
Type III 17ß-HSD, also known as androgenic 17ß-HSD, is found in the microsomes of the testis where it reduces androstenedione to testosterone (40). It prefers NADPH as cofactor. Using an expression-cloning strategy, the cDNA for human testis type III 17ß-HSD was obtained (52). The cDNA encodes a protein of 310 amino acids. Analysis of the amino acid sequence for the type III enzyme shows a hydropathy profile that does not support the presence of any membrane-spanning regions, but the short hydrophobic N- and C-termini may associate the protein with the endoplasmic reticulum. Pairwise comparisons between the three 17ß-HSD isozymes show that the amino acid identities between the type I, type II, and type III proteins are approximately 23% (51, 52), indicating that these enzymes are members of the same family (53).
Recently, a type IV 17ß-HSD has been found in human, mouse, and pig. It is similar to the type II isoform in that it is NAD+ dependent and is principally involved in the oxidation and therefore inactivation of estrogens and androgens. In porcine gonads, immunofluorescence detected the type IV 17ß-HSD in granulosa cells as well as the Leydig cells and Sertoli cells in the testis. The cDNA encodes an 80-kDa protein featuring domains not present in other 17ß-HSDs (54). The 80-kDa protein is cleaved at the N terminus to yield a 32-kDa fragment with 17ß-HSD activity. Both the 80-kDa and the N-terminal 32-kDa fragment have 17ß-HSD activity (55). The central domain (324–596 amino acids) is 40% identical to the C-terminal domain of the fatty acid hydratase/dehydrogenase of yeast and catalyzes ß-oxidation of fatty acids. The C-terminal domain is 39% identical with sterol carrier protein. Type IV 17ß-HSD is found in a number of human cells including those from the breast (MCF-7 and T47D) and liver (HepG2) (56). The peroxisomal location of this isozyme suggests that the protein may have other as yet undefined roles in lipid and steroid metabolism (57).
C. Structure, regulation, and tissue-specific expression of the
17ß-HSD genes
Two major mRNA species for type I 17ß-HSD, 2.2 kb and 1.3 kb in
length, have been identified in human placenta. S1 nuclease analysis
indicates that in placenta the major transcript is 1.3 kb in length and
starts nine nucleotides upstream from the start codon. The 2.2-kb
transcript is a minor species and contains approximately 971
nucleotides upstream from the same in frame start codon (58). Type I
17ß-HSD mRNA is distributed in placenta, liver, ovary, endometrium,
prostate, testis, and adipose tissue. Immunochemical analyses have
confirmed the presence of the type I enzyme in the syncytiotrophoblast
of human placenta (59), the granulosa cells of human ovary (60), the
epithelial cells of human breast (43), and the endometrium (61). In
studies on estrogen-sensitive human breast cancer cell lines (ZR-75–1,
MCF-7, T-47D, and R-27), type I 17ß-HSD was elevated and favored
formation of 17ß-estradiol. In estrogen receptor-negative cell lines,
the activity and expression were considerably lower (44, 62) and
estrone formation was favored. Data such as these indicate that type I
17ß-HSD may be an important component in estrogen-dependent growth of
mammary tissues and a target for drug development.
The regulation of the type I 17ß-HSD 1.3- kb transcript has been measured in JEG-3 cells. The mRNA for 17ß-HSD was elevated with an analog of cAMP and a protein kinase C agonist, phorbol-12-myristrate-13-acetate. The effects of cAMP were mimicked with forskolin and isobutyl methyl xanthine (a phosphodiesterase inhibitor) but were not abolished by a cAMP-dependent protein kinase inhibitor. By contrast, the effects of protein kinase C were abolished with a diacylglycerol kinase inhibitor. This indicates that type I 17ß-HSD is regulated by a nonclassic cAMP-dependent mechanism that remains to be elucidated (28, 63).
The expression pattern of the type II 17ß-HSD mRNA in human tissues supports its role as a steroid hormone inactivator. The 1.5-kb transcript is most highly expressed in liver, followed by the placenta, small intestine, and endometrium. The glandular epithelium of the endometrium almost exclusively expressed the type II 17ß-HSD, and it is positively regulated by progestins to inactivate estrogens (64).
The 1.3-kb mRNA encoding type III 17ß-HSD has been detected only in the testis, which is consistent with its role in the formation of testicular androgens (52).
To understand the basis of the tissue-specific and regulated
17ß-HSD gene expression, the human type I, II, and III 17ß-HSD
genes have been cloned (52, 53, 58, 65). The type I gene consists of
seven exons and five introns that span 6.2 kb and encode a transcript
of 2.2 kb with a long 5'-untranslated region (UTR) (Fig. 2). A highly
homologous pseudogene (previously designated 17ß-HSD I) has also been
cloned, but there is evidence that the pseudogene is not expressed (46, 63, 66). The pseudogene contains eight exons and five introns. The two
genes appear in tandem and together span more than 21 kb. These genes
were mapped to the long arm of chromosome 17 and more specifically to
the region q11-q12 or q12–21 (46, 63, 66). It is now known that the
gene for human type I 17ß-HSD is closely linked to the susceptibility
gene for hereditary breast and ovarian cancer, the BRCA1
gene on chromosome 17q21 (46). Both 17ß-HSD genes contain a promoter
region with a TATA box as well as GC and inverse CAAT boxes. The
5'-flanking regions contained consensus sequences for
cis-acting elements that may function as regulators of
17ß-HSD gene expression. These sequences included estrogen,
progesterone, and glucocorticoid response elements and a cAMP response
element (67). No functional studies on the 17ß-HSD type I gene
promoter have been reported.
The type II 17ß-HSD gene is located on chromosome 16q24. The gene was isolated from a human leukocyte genomic DNA library (65). The type II gene contains seven exons and spans 40 kb. This gene gives rise to two alternatively spliced mRNAs, but only the major transcript (1.45 kb) is functional and encodes for the 387-amino acid protein. The cap site is between 179 and 167 nucleotides upstream from the ATG start codon (65).
In contrast, the type III 17ß-HSD gene (androgenic 17ß-HSD) is found on chromosome 9q22. This gene contains 11 exons and is in excess of 60 kb in length (52, 53). Although the lack of chromosomal synteny between the type I-III 17ß-HSDs and the low degree of sequence homology suggested that they arose by convergent evolution, they appear to be SDR family members (53).
D. 17ß-HSD deficiency
Defects in 17ß-HSD were originally described by Saez and
colleagues (68), who reported familial male pseudohermaphroditism due
to a deficiency in testicular 17-ketoreductase. The characteristic
phenotype is a 46,XY individual with testes and male Wolffian
duct-derived urogenital structures (e.g. epididymus, vas
deferens, and seminal vesicles), but with external female genitalia.
These individuals are usually classified as females at birth but are
genetic males and develop a male phallus at puberty. It is recognized
that the 17ß-HSD deficiency, which results in pseudohermaphroditism,
results from autosomal recessive mutations in the type III 17ß-HSD
gene (androgenic form) (69, 70, 71). This defect underscores the importance
of the type III 17ß-HSD in androgen formation. Examination of 14 male
pseudohermaphrodites showed that a number of unique mutations occurred
within the type III 17ß-HSD gene. These mutations included 10 point
mutations (missense): S65L, R80Q, Q176P, V205E, A203V, F208I, E215D,
S232L, M235V, and P282L (53, 72); three splice-junction abnormalities;
and one small deletion 777 that results in a frame shift and
slightly truncated protein. Mutations were also found in introns 3 and
8, which disrupted the splice-acceptor sites. The point mutations in
the open-reading frame were introduced into the cDNA for type III
17ß-HSD, and their effect on enzyme activity was measured after
transient transfection into human embryonal kidney 293 cells. It was
found that eight of the nine missense mutant enzymes were devoid of
catalytic activity, establishing that these mutations were the cause of
the deficiency. The exception was the R80Q mutation, which had one
fifth the specific activity of the wild type enzyme, and the decrease
in activity was found to result from a reduced affinity for NADPH. The
residual activity observed was consistent with the low formation of
testosterone observed in the affected patient.
|
IV. 11ß-Hydroxysteroid Dehydrogenases |
|---|
|
TopAbstractI. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
).
|
B. Cloning and expression of the 11ß-HSD cDNAs
Rat liver 11ß-HSD was one of the first mammalian HSDs to be
cloned (79) and was expressed using a vaccina virus infection protocol
(80). The cDNA contained a 861-bp open reading frame that encodes a
protein of 287 amino acids. Sequence comparison indicated that there
was low but significant identity to ribitol dehydrogenase, the product
of the Nod G gene, and Drosophila alcohol
dehydrogenase, which are recognized as being members of the SDR family
(79). It is now known that two isoforms of 11ß-HSD exist and both
belong to this superfamily.
The type I 11ß-HSD refers to the form that was originally purified and cloned from rat liver. It is a glycoprotein and the molecular mass of the mature protein is 34 kDa. It catalyzes oxidation and reduction using NADP(H) as cofactor, and it is expressed in a variety of rat and human tissues including liver, lung, testis, colon, and kidney. A remarkable feature of this enzyme is that, when it is purified from rat liver, it functioned only as a dehydrogenase. However, the recombinant enzyme, expressed using vaccina virus, exhibited both oxidation and reduction of steroids (80). These observations led to the speculation that the level of glycosylation determined directionality of enzyme catalysis (80). When tunicamycin, an inhibitor of glycosylation, was added to the vaccina virus expression system, enzyme activity was reduced by more than 50%. This represents one of the first examples in which the function of a mammalian HSD may be regulated by posttranslational modification. It was originally proposed that the type I 11ß-HSD was responsible for conferring mineralocorticoid specificity; however, immunocytochemical studies within the rat kidney indicate that it is localized in the proximal tubular elements and not colocalized with the type I mineralocortioid receptor in the distal tubular elements (81). Studies on the hepatic form of 11ß-HSD in the squirrel monkey also indicate that the type 1 11ß-HSD does not protect against AME. This enzyme shares 75% sequence identity with the rat type 1 11ß-HSD but is only expressed in the liver and fibroblasts. Since squirrel monkeys have highly elevated cortisol levels but do not display Cushing’s disease or AME, an enzyme other than the hepatic 11ß-HSD must protect the mineralocorticoid receptor from cortisol occupancy (82).
In contrast, a type II 11ß-HSD was found in the kidney and placenta microsomes of sheep and humans. Clones encoding the kidney isozyme have been isolated from sheep and human kidney cDNA libraries (83, 84). The human type II 11ß-HSD is 41 kDa in length and consists of 405 amino acids. It is NAD+ specific and it has a lower Km (10–100 nM) for steroid substrate than the type I enzyme (75). It catalyzes only 11ß-dehydrogenation, i.e. it inactivates cortisol. The type II 11ß-HSD and the type II 17ß-HSD share 35% sequence identity and suggest that both these dehydrogenases belong to the SDR superfamily.
The pattern of diformazan disposition (a dye that produces a blue color after the reduction of cortisol by cofactor in the presence of enzyme) colocalized the type II 11ß-HSD activity with the mineralocortioid receptor (85). It is now clear that it is the type II enzyme that is important in regulating occupancy of the mineralocorticoid receptor (75, 76, 85).
C. Structure, regulation, and tissue-specific expression of the
11ß-HSD genes
At least four different mRNA species hybridize with the type I
11ß-HSD cDNA in rat kidney (86), indicating that one or more
alternative forms of 11ß-HSD exist that may be products of either
alternative splicing of the pre-mRNA or the use of a different cap
site. The multiple RNA transcripts for type I 11ß-HSD can be
generated by either differential promoter usage by the type I 11ß-HSD
gene (86) or by differential polyadenylation. Isoforms have been termed
type IA and type IB 11ß-HSD. Type IB is exclusively expressed in the
kidney medulla and cortex, and interest originally existed in this form
because this enzyme had the potential to colocalize with the
mineralocorticoid receptor (87). The type IB 11ß-HSD mRNA encodes for
a protein in which the first 26 amino acids from the N terminus are
missing (87). This truncated enzyme is encoded by a transcript that
originates from intron-1 of the type I gene. This truncation results in
the loss of a hydrophobic polypeptide chain that may be important in
intracellular transport of the enzyme. Expression of the type IB
11ß-HSD cDNA failed to yield active enzyme using either cortisol or
corticosterone as substrates and suggests that the truncated region is
essential for enzyme activity. This truncation stops just before the
NADP+ binding site, which is predicted to be located at the
N terminus.
Using the cDNA for type I 11ß-HSD and polyclonal antibody, it was found that this enzyme was expressed in tissues that were not mineralocorticoid target tissues (e.g. liver, lung, skin, and testis) (88). In addition, no type I 11ß-HSD was found in several mineralocorticoid target segments in the kidney (e.g. cortical collecting ducts and collecting tubules) (85, 88, 89). This profile of tissue-specific expression further confirms that type I 11ß-HSD does not protect against mineralocorticoid activity of the glucocorticoids.
The human type I 11ß-HSD gene has been cloned and is located on
chromosome 1 and contains six exons that span a total length of 9 kb
(90) (Fig. 4). The human type II 11ß-HSD gene consists
of five exons and four introns and is 6.2 kb in length and, by contrast
to the type I gene, is located on chromosome 16q22 (91). The
5'-flanking regions of these genes have not been analyzed for the
presence of cis-acting elements, and functional studies on
these promoters have not been described.
|
,20ß-HSD; both these enzymes
are members of the short-chain dehydrogenase/reductase family and are
inhibited by carbenoxolone (a derivative of glycyrrhetinic acid, the
active ingredient of licorice) (see Sections VII and
VIII). Deletion of Y232 is anticipated to result in
inactive enzyme since this residue is part of a Tyr-X-X-X-Lys catalytic
motif that is conserved (see Sections VII and
VIII). Similarly, R208 and R213 are predicted to make up
portions of the steroid-binding pocket, and changes in
Km for steroid substrate are anticipated. |
V. 3-Hydroxysteroid Dehydrogenases
|
|---|
|
TopAbstractI. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
-HSDs work in concert with the 5- and
5ß-reductases to reduce 3-ketosteroids to produce the 5,3- and
5ß,3-tetrahydrosteroids. Although these reactions are responsible
for the metabolism of the majority of steroid hormones, they are not
without consequence. In the prostate, 3-HSD works as a molecular
switch and regulates occupancy of the androgen receptor. In this tissue
the enzyme interconverts 5-DHT [dissociation constant
(Kd) for the androgen receptor of 10-11
M] to 3-androstanediol [Kd for the
androgen receptor of 10-6 M], thereby
reducing the affinity of the steroid ligand for the androgen receptor
by 5 orders of magnitude (96, 97, 98, 99) (Fig. 3). In the dog prostate,
evidence has mounted that the enzyme works predominantly as a
dehydrogenase and works in concert with 5-reductase to maintain high
levels of 5-DHT, which is required for both the normal and abnormal
growth of the prostate (100, 101). Inhibitors of prostatic 3-HSD
would block the interconversion of 5-DHT to 3-androstanediol and
may be useful in preventing the build-up of 5-DHT. Such inhibitors
could be useful in combination with Finasteride for the treatment of
benign prostatic hyperplasia and prostatic cancer.
In the brain, 3-HSD converts 5-reduced steroids (e.g.
5-dihydroprogesterone) to yield tetrahydrosteroids (e.g.
5,3-tetrahydroprogesterone), which are potent allosteric
effectors of the 
-aminobutyric acid (GABA)a receptor
(Fig. 3). These steroids do not bind to the GABAa receptor
by themselves but, in the presence of GABA, nanomolar concentrations of
these steroids increase the affinity of the receptor for this
neurotransmitter. As a consequence, these neurosteroids increase
GABA-dependent chloride conductance and have anxiolytic and anesthetic
properties (102, 103, 104). The 3-hydroxysteroid, alphaxalone, was
developed as an anesthetic based on this mechanism of action (105).
Thus, 3-HSD regulates occupancy of steroid hormone receptors,
whether they are members of the nuclear receptor superfamily, such as
the androgen receptor, or a membrane-bound chloride-ion gated channel
such as the GABAa receptor.
B. Cloning and expression of the 3-HSD cDNAs
The first mammalian 3-HSD to be purified was from rat liver
cytosol. This enzyme was obtained in milligram amounts (106) and led to
the development of a polyclonal antibody (107). Immunoscreening of a
gt-11 expression library led to the isolation of a full-length cDNA
clone that encodes a protein of 322 amino acids (3). The fidelity of
the clone was confirmed by its overexpression in E. coli
(108). The same cDNA was cloned and expressed independently by several
groups (109, 110, 111). The cDNA for rat liver 3-HSD was found to share
no sequence identity with other HSDs that had been cloned at that time,
including 3ß-HSD/KSI (9, 10, 11, 16), 11ß-HSD (79), and 17ß-HSD
(45, 46, 47). Instead, a search of GenBank revealed that it shared high
sequence identity with aldose-reductases from rat and bovine lens (112, 113), human placental aldose reductase (114, 115), prostaglandin
F2 synthase (116), and 
-crystallin from frog lens
(117). These proteins all share in excess of 58% sequence identity and
are known members of the AKR superfamily (see Section VII).
After this initial work a number of potential human 3-HSD cDNAs were
cloned, including human liver DD1, DD2, and DD4, where DD refers to the
associated dihydrodiol dehydrogenase activity. These enzymes display
greater than 70% sequence similarity with rat liver 3-HSD.
Expression of these clones reveals that DD1 acts predominantly as a
20-HSD rather than a 3-HSD and differs from DD2 in only seven
amino acids (118, 119). Recombinant DD2 corresponds to the human bile
acid-binding protein (119, 120). In contrast, DD4 appears to be
identical to chlordecone reductase (119). Chlordecone reductase is now
known to be human type I 3-HSD (121). Neither DD2 nor DD4 functions
as the human ortholog of the rat liver 3-HSD. DD2, like the rat
liver enzyme, binds bile acids with high affinity but differs in that
it is not an efficient 3-HSD. In contrast, DD4 is an efficient
3-HSD but does not bind bile acids well. Recently, a type II
3-HSD has been cloned from human liver (121). Recombinant type I and
type II 3-HSDs differ in their Km values for 5-DHT.
The type I enzyme has a Km of 1.2 µM, and the
type II enzyme has a Km of 19.2 µM. Type I
and type II 3-HSDs display high sequence identity. The availability
of these clones has also led to the cloning and expression of two
different 3-HSDs from human prostate libraries (122, 123). The
prostatic 3-HSD cDNA isolated by Lin et al. (122) was
originally expressed as a his-tag protein in E. coli and was
devoid of 3-HSD activity, even though it displayed high sequence
identity with the rat and human liver isoforms. Subsequently,
expression without the his-tag has led to the overexpression of the
human prostatic 3-HSD activity in E. coli. Kinetic
characterization of the recombinant enzyme reveals that it favors
5-DHT formation (H.-K. Lin and T. M. Penning, unpublished results).
The prostatic 3-HSD cDNA clone isolated by Dufort et al.
(123) was transiently expressed in human embryonal kidney 293 cells and
also had demonstrable 3-HSD activity and weak 20-HSD activity.
This clone has 86% sequence identity with human liver type II
3-HSD. Which of the two cDNAs represents the major human prostatic
3-HSD is unknown. No attempt has been made to determine which
prostate cells (epithelial vs. stromal cells) express
3-HSD, and no RNase protection studies have been performed to
determine their relative abundance across tissues.
C. Structure, regulation, and tissue-specific expression of the
3-HSD genes
Using rat liver 3-HSD cDNA as a probe, expression of a 2.7-kb
transcript was found to be limited to peripheral tissues such as the
liver, lung, and small intestine (124). When poly(A)+ RNA
was used, expression was also detected in the testis, mammary gland,
and ovary. In the ovary two smaller transcripts, 1.2 and 1.6 kb, were
observed. 3-HSD is a member of the AKR superfamily and, because
members share high-sequence identity, probes based on open-reading
frames may result in cross-hybridization and complicate data
interpretation concerning tissue-specific expression.
In rat liver, 3-HSD displays high constitutive expression and
represents 0.5–1.0% of the soluble protein. It also displays sexual
dimorphic expression; mRNA, enzyme protein, and enzyme activity are all
elevated in adult female rat liver, and expression appears to be under
the control of estrogens. Maximal expression is seen after 21 days in
males but will decline unless estrogens are administered (125). This
sexual dimorphism has been observed primarily in rat liver for other
enzymes involved in steroid hormone metabolism. Feminization of steroid
hormone metabolism involves elevation of microsomal 5-reductase and
decreased 6ß- and 16-hydroxylase levels and has been shown to be
governed by gonadal regulation of GH secretion (126, 127, 128). In these
studies estrogens stimulated the constant secretion of GH, and
androgens stimulated the pulsatile release of GH (128). The mechanism
by which constant exposure to GH controls the coordinate expression of
select steroid-metabolizing enzymes is unknown. In hypophysectomized
female and male rats, levels of 3-HSD mRNA, enzyme protein, and
enzyme activity were decreased 8-fold. However, the administration of
either GH or estrogens alone failed to elevate levels to those seen in
intact animals. The data would suggest that GH is insufficient by
itself to cause increased expression of the 3-HSD gene. One
explanation for this phenomenon is that in rat liver the estrogen
receptor is regulated by GH (129), and the liver may need priming by GH
before it will respond to estrogens.
The genes for rat 3-HSD (130), human type I 3-HSD (chlordecone
reductase/DD4) (121), and human type II 3-HSD (121) have been cloned
and show similar organization (Fig. 5). They contain at
least nine exon-intron boundaries that are highly conserved within
members of the AKR superfamily (131, 132). The rat 3-HSD gene is
more than 47 kb long, whereas the type I and type II human 3-HSD
genes are 20 and 16 kb in length, respectively (121). The human type I
3-HSD is located on chromosome 10 at p14 (133). No location of the
human type II 3-HSD gene has been reported.
|
-HSD gene (130). The bulk of the 2.7-kb
transcript encodes for a long 3'-UTR which may be an important
determinant in mRNA stability (134, 135). Estimates of the t1/2
of rat liver 3-HSD mRNA are of the order of 12 h. Exons
encoding for the long 3'-UTR present in the mRNA have not been isolated
(Fig. 5). Primer extension analysis has located the transcription start
site at -54 or -55 bp upstream (+1). Upstream from this site is a
TATTA box that presumably binds RNA polymerase in forming the basal
transcriptional complex. A genomic clone containing 9.5 kb of the
5'-flanking region of the gene has been isolated, and 2.5 kb
immediately upstream from the start site have been sequenced. Analysis
of this region for transcription factor consensus sequences shows the
presence of a large number of imperfect steroid hormone response
elements, including estrogen response elements, progesterone response
elements, and glucocorticoid response elements. Interspersed between
these elements are several AP-1, AP-2, and Oct sites. It has been
suggested that these elements may comprise a steroid response unit.
Nested deletions of the 5'-flanking region of the gene fused to the
reporter gene chloramphenicol acetyltransferase (CAT) have been
generated and cotransfected into human hepatoma (HepG2) cells to
identify the cis-elements involved in regulating
constitutive expression (130). With respect to the transcription start
site (+1), it was found that -199 to +55 bp contained the basal
promoter, and gave a level of CAT expression similar to that observed
in a pSV40-CAT positive control. Between -499 and -199 bp, a proximal
enhancer was identified that elevated CAT activity 1.8- to 2.5-fold;
between -755 and -498 bp, a silencer was observed that prevented CAT
expression. Further upstream a powerful distal enhancer was located
between -4.0 and -2.0 kb, which elevated CAT expression 40-fold and
is thought to control high liver-specific expression. An examination of
the silencer or negative response element (NRE) indicated that it
contained two unique Oct sites, one on either strand. Band shift and
supershift assays using a radiolabeled NRE and anti-Oct monoclonal
antibodies provided evidence for the binding of Oct transcription
factors to the NRE and suggest that Oct factors may be repressors of
the 3-HSD gene (130).
The function of the steroid response unit on the 5'-flanking region of
the 3-HSD gene has also been examined (136). It was found that in
isolated rat hepatocytes, dexamethasone was able to increase steady
state 3-HSD mRNA levels, in a time- and concentration-dependent
manner, and that this effect was blocked by antiglucocorticoids such as
RU486. Nuclear-run off assays established that the effect of
dexamethasone was at the level of new gene transcription (136). It was
Tomkins (137, 138) who first implicated hepatic 3-HSD as playing a
major role in glucocorticoid metabolism. These findings suggest that
glucocorticoids may increase their own metabolism by inducing 3-HSD
gene expression. This may comprise a positive feedback loop to
eliminate these steroids.
D. 3-HSD deficiencies
Although, no 3-HSD deficiencies have been described, a clue may
exist in glaucoma patients. These patients are sensitive to increases
in intraocular pressure induced by the administration of
corticosteroids (139). Studies on the metabolism of glucocorticoids in
peripheral blood monocytes obtained from glaucoma patients indicate
that 5ß-reductase activity is elevated, and 3-HSD activity is
attenuated (139). This would cause a build up of
5ß-dihydrocorticoids, which are known to increase intraocular
pressure (140). Whether this is a mere correlation or a casual
relationship, resulting from a genetic defect, requires further work.
|
VI. 20-Hydroxysteroid Dehydrogenases
|
|---|
|
TopAbstractI. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
-HSD; this enzyme was
originally characterized and purified from rat ovary (141, 142, 143). In the
ovary it plays an important role in converting progesterone (a potent
progestin) into 20-hydroxyprogesterone (a weak progestin) (Fig. 6). In steroid target tissues, it would regulate the
amount of progestin that can bind to the progesterone receptor, and in
this regard may be important in the placenta and endometrium. Early
studies on the luteolytic effects of prostaglandin F2 on
rat ovary established that 20-HSD activity is induced 150-fold.
Because of the ability of this enzyme to metabolize progesterone to its
inactive metabolite, 20-hydroxyprogesterone, the underlying
induction mechanism may contribute to luteolysis (144).
20-Hydroxyprogesterone is unable to sustain pregnancy and, in the
rat, the increase in luteal 20-HSD activity contributes to the lower
circulating progestin levels associated with the termination of
pregnancy. The possibility exists that inhibitors of the human isoform
may help maintain pregnancy before the luteal-placental shift in
steroidogenesis.
|
-HSD activity. In the ovary 20-HSD catalyzes the inactivation
of progesterone as described. But this should be distinguished from the
enzyme found in other steroidogenic tissues (testis and adrenal) where
its principal substrates may be either 17-hydroxypregnenolone or
17-hydroxyprogesterone. By converting the C20 ketone to an alcohol,
the 17,20-lyase would be deprived of substrate, and this would prevent
the conversion of C21 to C19 steroids. In the human placenta, 20-HSD
activities are catalyzed by type I or type II 17ß-HSD.
B. Cloning and expression of the 20-HSD cDNAs
cDNAs for rat and rabbit ovarian 20-HSD have been cloned (5, 6). These enzymes display high-sequence identity with the mammalian
3-HSDs and are members of the AKR superfamily. Recently, bovine
testicular 20-HSD has been cloned and has been found to be identical
to aldose reductase (145). The testicular enzyme will only catalyze the
reduction of 17-hydroxypregnenolone and 17-hydroxyprogesterone
and will not utilize progesterone as a substrate. The mRNA for this
enzyme is also found in the adrenal cortex. Both type I and type II
17ß-HSD also display 20-HSD activity (51, 146). It is proposed
that the type I enzyme (which is estrogenic and a reductase)
inactivates progesterone whereas the type II enzyme (which inactivates
estrogens and acts as a dehydrogenase) converts
20-hydroxyprogesterone to progesterone. This switching mechanism may
contribute to the estrogen- and progestin-dependent regulation of
endometrial growth.
C. Structure, regulation, and tissue-specific expression of the
20-HSD genes
In rat granulosa cells, GnRH or FSH alone increases
20-HSD activity whereas PRL decreases the activity of the enzyme
(147). The availability of the cDNA for rat ovarian 20-HSD, which
detects a 1.2-kb transcript in corpora lutea, has led to an extension
of these studies. mRNA levels were markedly reduced by PRL treatment
and appear to be under tight control by PRL (148). Tissue distribution
studies indicate that the expression of the 1.2-kb transcript is
specific for the rat corpus luteum, it being almost nondetectable in
the uterus, kidney, lung, and heart. Its tissue-specific expression and
tight control by gonadotrophs confirm the potential importance of this
enzyme in luteolysis.
The structure of genes encoding ovarian 20-HSD has not been
reported. By contrast, the gene structures of several AKRs have been
described (131, 132). These have a genomic organization similar to that
described for rat 3-HSD and contain the nine highly conserved
exon-intron boundaries. A similar genomic organization is anticipated
for 20-HSDs that are AKRs.
|
VII. HSDs Belong to Two Protein Families |
|---|
|
TopAbstractI. IntroductionII. 3{beta}-Hydroxysteroid...III. 17{beta}-Hydroxysteroid...IV. 11{beta}-Hydroxysteroid...V. 3{alpha}-Hydroxysteroid...VI. 20{alpha}-Hydroxysteroid...VII. HSDs Belong to...VIII. Structure/Function of HSDsIX. ConclusionsReferences |
|---|
) (149, 150, 151, 152). SDR family members are nonmetalloenzymes
that have monomers of 25–28 kDa. The monomer molecular masses of the
type I-IV 17ß-HSDs and the type II 11ß-HSD are somewhat larger.
These proteins often function as multimers, and they share
approximately 25% sequence identity. Despite this relatively low
homology, family members share identical protein folds (1, 2). Part of
the protein fold includes an arrangement of -helix and ß-strands
(ß--ß--ß)2 to produce a Rossmann fold for
cofactor binding. Binding of the cofactor, NAD(P)(H) (all nicotinamide
adenine dinucleotides) is in the N-terminal part of the molecules,
where a common Gly-XXX-Gly-X-Gly motif occurs. In addition, they all
share a conserved consensus sequence Tyr-X-X-X-Lys, which has
been implicated by site-directed mutagenesis to play a role in
catalysis (Table 2) (153, 154, 155). Originally, 3ß-HSD/KSI was not
considered part of this family but these enzymes contain both the
Gly-XXX-Gly-X-Gly and the Tyr-X-X-X-Lys motifs. Other SDR family
members include porcine testicular 20ß-HSD (156), 3,20ß-HSD from
Streptomyces hydrogenans (157), 15-hydroxyprostaglandin
dehydrogenase (158), Drosophilia alcohol dehydrogenase
(154), ribitol dehydrogenase from Klebsiella aerogenes
(159), carbonyl reductase (160), the ACT111 gene product of
Streptomyces coelicolor (161), and AP27, an adipocyte
protein of unknown function (162). There are currently 57 members in
the protein family, and two crystal structures exist for HSDs that
belong to this family; these are for 3,20ß-HSD from
Streptomyces hydrogenans and the human placenta type-1
17ß-HSD (163, 164, 165, 166).
|
