Chapter 1 Steroids
Chapter 1 Steroids
Chapter 1 Steroids
21
18
12
11
1
2
3
HO
A
4
19
10
5
13
14
20
22
23
24
17
16
15
26
25
27
Figure 1 shows the structure of cholesterol. The carbons are numbered according to
the convention used for steroid nomenclature. The rings are assigned letters from
left to right for reference in discussion of the steroid backbone.
Cholesterol is a hydrophobic molecule; in fact, it is essentially insoluble in water. In
the body, the majority of the cholesterol is associated with cell membranes, where it
has an important role in maintaining membrane fluidity. During transport (in HDL
and LDL) and storage (in intracellular lipid droplets), however, the 3 position
1
O
O
As we will see in greater detail in the next chapter, all steroid hormones are derived
from cholesterol. The production of the hormones involves a number of precise
modifications to the cholesterol structure, with different series of modifications
occurring in different pathways. These modifications include attack at the 11, 16,
17, 18, 19, 20, and 21 positions, conversion of the 3-hydroxyl to a ketone, and
isomerization of the 5-6 double bond to the 4-5 position (Figure 3). A number of
additional modifications, not shown here, occur during conversion of the steroid
hormones to inactive metabolites. Each set of alterations to the steroid backbone
alters the affinity of the steroid for a given steroid receptor.
21
20
18
11
19
17
16
HO
5
Figure 3. Locations of attacks on the cholesterol backbone during conversion to
steroid hormones.
In organic chemistry, sites of unsaturation (i.e. double bonds) are often referred to
using the Greek letter . Thus, steroids containing the 5-6 double bond, such as
cholesterol, are designated 5 steroids; those with a 4-5 double bond are called 4
steroids.
As we will see in Chapter 9, Vitamin D 3 and its metabolites are also derived from
cholesterol. The numbering used for the metabolites of Vitamin D is the same as
that used for all of the other cholesterol metabolites; however, some aspects of the
nomenclature can be somewhat confusing because during the synthesis of Vitamin
D3 the 9-10 bond of the B-ring is cleaved and the A-ring is flipped around the 6-7
bond. The active derivative of Vitamin D3 , 1,25-dihydroxy-Vitamin D 3 , is produced
2
CH3
HO
H
H
The names we will use for the biologically relevant steroids imply the correct
stereochemistry unless specifically noted otherwise (e.g. in order to be cholesterol
the steroid must have a 3-hydroxyl, and the branched 8-carbon side-chain at the
17 position). However, there is a chemical nomenclature for each steroid that
uniquely denotes the structure for that compound. This nomenclature is based on
describing the modifications to one of four possible backbones (Figure 5).
Cholestane
(C27)
Pregnane
(C21)
Androstane
(C19)
Estrane
(C18)
All physiological steroids are derivatives of one of these four backbone structures.
The differences among these basic structures are relatively minor. The differences
between cholestane (which has 27 carbons), pregnane (which has 21 carbons), and
3
androstane (which has 19 carbons) are limited to the length of the side-chain at the
17 position. Estrane differs from androstane in that it lacks the 19 methyl group.
To generate the names of the actual compounds, it is simply necessary to decide
which backbone is correct, and then make a note of the modifications. For example
the chemical name of cholesterol is 5-cholestene-3-ol. In other words it is a C27
steroid (cholestane) with a 5-6 double bond (5-cholestene), and a 3-hydroxyl (3-ol).
This same nomenclature is used for all physiological steroids. Further examples of
this nomenclature are given in Table I.
Table I.
Steroid Nomenclature Examples
Backbone
Cholestane
Pregnane
Androstane
Estrane
Trivial Name
Chemical Name
Cholesterol
5-cholestene-3-ol
Cortisol
Aldosterone
Progesterone
Pregnenolone
4-pregnene-11,17,21-triol-3,20-dione
4-pregnene-11,21-diol-3,18,20-trione
4-pregnene-3,20-dione
5-pregnene-3-ol-20-one
Androstenedione
Testosterone
DHEA
Dihydrotestosterone
4-androstene-3,17-dione
4-androstene-17-ol-3-one
5-androstene-3-ol-17-one
androstane-17-ol-3-one
Estradiol
Estrone
Estriol
1,3,5(10)-estratrien-3,17-diol
1,3,5(10)-estratrien-3-ol-17-one
1,3,5(10)-estratrien-3,16,17-triol
During the initial purification and characterization of the adrenal steroids in the 1930s, the
structures of the compounds were not known, and, in fact, it was not known that these compounds
were steroids. The researchers involved gave the compounds identifying letters for reference
purposes. The table gives the code letters most commonly associated with the steroids. This table is
of more than simply historical interest, since these letter codes are still often used as abbreviations
in the literature. The structures of all of these steroids are shown in Chapter 2.
Compound
B
E
F
S
Current Name
corticosterone
cortisone
cortisol (hydrocortisone)*
deoxycortisol
Cortisol
Aldosterone
20-ketone
18-aldehyde
OH
11-hydroxyl
OH
O
O
HO
HO
OH
17 position
not modified
3-ketone
17-hydroxyl
O
4
Figure 6. The structures of the primary human glucocorticoid steroid hormone,
cortisol (4-pregnene-11,17,21-triol-3,20-dione), and of the primary human
mineralocorticoid, aldosterone (4-pregnene-11,21-diol-3,18,20-trione).
17-hydroxyl
Progesterone
17-hydroxyl
Testosterone
O
Estradiol
OH
Missing
19 carbon
OH
HO
Aromatic A-ring
Figure 7. The structures of the sex steroids: the progestin progesterone (4-pregnene3,20-dione), the androgen testosterone (4-androstene-17-ol-3-one), and the estrogen
estradiol (1,3,5(10)-estratriene-3,17-diol).
HO
OH
HO
OH
HO
Cell
Key:
R
Steroid
S
S
R
R
Unoccupied Receptor
Activated Steroid
Receptor Complex
mRNA
Nucleus
New Proteins
EFFECTS
In Figure 8 the unoccupied receptor is shown as being located both in the cytoplasm
and in the nucleus. Most of the available evidence suggests that the receptors are
predominantly present within the nucleus in both the presence and absence of
hormone. The older model, in which the binding of hormone was thought to result in
a translocation of the receptor from the cytoplasm to the nucleus, is probably
incorrect, at least for the majority of the receptor types. However, due to the
conflicting data concerning this issue, I have presented both models in this drawing.
7
It should be noted that there is evidence for cell surface receptors for some of the steroid hormones,
and for some steroids that had not previously been considered to be hormones. In some experiments,
responses to hormonal administration have been observed on short time scales (i.e. a few seconds to a
few minutes). These responses are too rapid to be mediated by alterations in gene transcription, and
may be mediated by cell surface receptors working through one of the classical second messenger
pathways. Although most of the current evidence suggests that the major actions of the steroid
hormones are mediated by the intracellular receptors, this may be due, at least in part, to the fact
that the cell surface receptors have not yet been isolated and characterized.
The steroid hormone receptors are generally present in small quantities (a few
hundred to a few thousand molecules per cell). They have high affinity for their
ligands (dissociation constant usually less than 1 nanomolar). They function as
ligand activated transcription factors, specifically activating a small number of
genes (less than 50, and possibly less than 10 genes per cell).
The Steroid Hormone Receptor Superfamily
All of the characterized steroid hormone receptors are members of a large
evolutionarily-related superfamily of proteins. The family also includes receptors for
thyroid hormone and some metabolites of Vitamins A and D. This family of proteins
is also called the nuclear receptor superfamily, because the actions of these proteins
are thought to occur also a result of gene transcription modulatory effects exerted
within the nucleus. Members of the superfamily are thought to be present in most if
not all multicellular animals.
Table II shows the names of some of the human protein known to be members of the
superfamily, the size of the receptor protein (in amino acids), and the chromosomal
location of the gene. The receptor genes are widely scattered throughout the genome
(note that the androgen receptor is on the X chromosome!), and vary in size from
427 amino acids for the Vitamin D receptor to 984 amino acids for the
mineralocorticoid receptor.
Note the Type I and Type II glucocorticoid receptors. The GR Type I is listed as the
mineralocorticoid receptor; this is only partially true, because it also binds
glucocorticoids with high affinity (in fact, with 10-fold higher affinity than GR Type
II receptors), and probably mediates glucocorticoid action in some tissues. This
illustrates an important point: there is no label on the cloned gene; the name of the
ligand is not spelled out in the nucleotide sequence! The names of the receptors are
related to their major observed functions, but under some circumstances it is
possible for a receptor to respond to other ligands, both physiological and nonphysiological.
Type II lists thyroid hormone twice. These are the two thyroid hormone receptor
genes; in fact, there are actually at least four different thyroid hormone receptor
proteins as a result of differential splicing of the gene transcripts to form the final
mRNA. The presence of multiple forms of the receptor protein is probably also the
case for most if not all of the other receptors as well. The receptor sizes listed in
8
Tables II and III are for one representative isoform of the gene product.
Table II.
Members of the Steroid Hormone Receptor Superfamily
Receptor
Chromosome
Thyroid hormone-
Thyroid hormone-
490
456
17
3
Vitamin D
427
12
Retinoic Acid-
Retinoic Acid-
Retinoic Acid-
462
448
454
17
3
12
Retinoid-X-
Retinoid-X-
Retinoid-X-
462
533
454
9
6
1
Estrogen-
Estrogen-
595
477
6
14
777
984
Androgen
919
Progesterone
934
11
A total of at least six different genes code for receptors for the morphogen retinoic
acid. It is thought that the physiological ligand for the Retinoic Acid Receptors is
all-trans-retinoic acid, while that of the Retinoid-X Receptors is 9-cis-retinoic acid
(Figure 9). Neither form of retinoic acid is usually considered to be a classical
hormone, but the control of retinoic acid levels and the scope of their physiological
function are currently poorly understood.
COOH
9-cis-retinoid acid
COOH
A number of other members of the receptor superfamily have been identified in humans; however,
the physiological ligands of these orphan receptors are not known. These orphan receptors are all
thought to act as transcription factors. Some may be constitutively active, and some may have their
activity modulated by mechanisms other than ligand binding (e.g. phosphorylation). In some cases,
the proteins are known to be ligand-activated, but the ligands are either unknown or only been
tentatively identified. For example, all three of the known peroxisome proliferator-activated receptor
(PPAR) gene products bind some xenobiotics, but have not been found to bind any endogenous
Variable
domain
DNA
binding
421
Hormone
Binding
486 528
777
One region, a short segment of about 70 amino acids, is the part of the protein that
specifically binds DNA, and is the most highly conserved part of the protein (see
Table II). An approximately 250 amino acid region at or near the C-terminus of the
protein is the hormone binding domain. Finally, there is a third region at the Nterminus (the variable domain) that is by far the least conserved, in either length or
amino acid sequence. This last region is responsible for mediating some of the
transcriptional effects of the protein. Table III shows the percent amino acid
sequence identity of the different regions relative to the corresponding part of the
glucocorticoid Type II receptor.
Table III.
Features of Selected Receptor Proteins
DNA Binding
Domain
Receptor
Glucocorticoid (GR Type II)
Mineralocorticoid
(GR Type I)
Progesterone
Androgen
Estrogen-
Thyroid hormone-
Retinoic acid-
Vitamin D
Size of
Size
%
variable (amino Identity
domain acids)
421
66
100
603
66
94
567
559
180
102
88
24
66
66
66
67
66
66
10
90
91
53
47
45
42
Hormone Binding
Domain
Size
(amino
acids)
250
251
%
Identity
255
243
250
225
265
236
55
52
30
17
15
<15%
100
57
Notice the differences in length of the variable domain (this domain has very little
sequence similarity among the different receptors). Also, notice the similarities of
the size and the sequence identities of the DNA binding domain. The most divergent
receptors still have greater than 40% identity in the DNA binding domain; due to
this high degree of conservation, this region is used as a marker for identifying
members of the superfamily.
The receptors are divided into three sub-families by similarities of protein sequence
and by some functional aspects: The GR Type I and II, progesterone, and androgen
receptors form one family; the estrogen receptor forms a family of least two genes
for the same ligand (as well as some closely related orphan receptors); and the
Vitamin D, thyroid hormone, and the two types of retinoic acid receptors comprise a
third family.
Molecular Mechanism of Steroid Hormone Receptor Action
It has not yet been possible to determine the three-dimensional structure of one of
the steroid hormone receptors, and there are many gaps in our understanding of
mechanism by which the interaction of the steroid hormone with its receptor elicits
a biological effect. The following discussion presents a working model for steroid
hormone action at the molecular level. The model is certainly incomplete, is
somewhat simplified, and may be incorrect in some details, but represents a picture
of the current consensus opinion.
The receptors for retinoic acid, thyroid hormone, and Vitamin D appear to be tightly associated with
DNA in both the presence and absence of hormone. For the thyroid hormone receptor, in particular,
the effect of hormone appears change the nature of the receptor activity from a negative effect to a
positive effect (see Chapter 7).
In contrast, the receptors for the steroid hormones are, at most, loosely associated with the DNA (or
the chromatin) in the absence of hormone. In addition, ligand-free steroid receptors are probably
bound to several other proteins, including hsp90 (hsp = heat shock protein, 90 refers to the fact that
this particular heat shock protein has a molecular weight of 90,000) and hsp70. The hsp complex is
thought to both stabilize the receptor and prevent the unoccupied receptor from affecting gene
transcription. The hsp complex proteins dissociate from the receptor upon ligand binding.
The receptors for retinoic acid, thyroid hormone, and Vitamin D are thought not to form these
complexes.
The model shown in Figure 11 uses the estrogen receptor as an example; all of the
steroid hormone receptors are thought to work by similar mechanisms, although the
ligands are different. If an antagonist (shown in Figure 11 as 4-hydroxy-tamoxifen,
but the same principle applies to all steroid hormone antagonists) binds to the
receptor, nothing happens. This nothing can be profound therapeutically, because
as long as an antagonist is bound to the receptor, the receptor cannot bind an
agonist, and therefore a hormone, even if present, has no effect. On the other hand,
if an agonist binds (shown as estradiol in Figure 11), the receptor undergoes an
event called activation or transformation. The nature of this event is not yet
11
clear, but it probably involves a conformational change within the hormone binding
domain. The activated receptor binds a specific type of enhancer DNA sequence
called a hormone response element (HRE). (Note that an HRE is a region of DNA
just like any other except for the specific sequence -- the real DNA double helix does
not have a box labeled HRE!) The receptor dimer can also interact with other
transcription factors (shown in Figure 11 as irregular polygons with various
patterns). The transcription factors that comprise this complex are not yet fully
characterized, and probably vary in different cell types and/or for different genes. It
is thought that the presence of the activated receptor stabilizes the complex formed
by the other transcription factors and stimulates the binding of RNA polymerase,
and as a result, initiation of RNA synthesis.
Unoccupied
Receptor
OH
ER
4-hydroxy
Tamoxifen
Estradiol
HO
HO
HO
ER
Antagonist
Bound
Receptor
OH
OH
ER
Agonist
Bound
Receptor
HO
HRE
ER*
Transformed
Receptor
RNA
Polymerase
TATA
mRNA
The full length receptor proteins are somewhat difficult to work with. However, individual domains
are often more tractable than the full length protein. While the three-dimensional structure of the
complete receptor protein has not yet been determined, the structure of the isolated DNA-binding
domain has been solved for several receptors. In all cases this highly conserved region of the protein
has been shown to fold around two zinc ions into a structure that can bind specific sequences of DNA
with high affinity.
Like the DNA-binding domain, the hormone binding domain for some receptors has been produced
by recombinant DNA technology; this part of the protein appears to exhibit all of the features
required for ligand discrimination. In the last few years, hormone-binding domain structures have
been solved for several of the receptors.
The model in Figure 11 implies that steric hindrance by the antagonist results in a different
conformational change from that induced by an agonist ligand. This is based on comparison of
estradiol-bound and antagonist-bound estrogen receptor- hormone binding domain structures, and
is currently believed to be a general phenomenon for the steroid receptor superfamily proteins.
12
from estrogen receptor- and estrogen receptor-, and the heterodimer). These
complexes probably have different effects on different genes; a cell could tailor its
response to estrogen by producing only one form, or by altering the relative amounts
of the two forms produced.
5) Multiple forms of the receptor: There are two genes for thyroid hormone receptors.
Each gene produces at least two different proteins due to differential splicing of the
exons during mRNA maturation. As a result there are at least four different
thyroid hormone receptors; in fact, one of these proteins is incapable of binding
hormone. By controlling the relative amounts of the different forms of the receptor a
cell can vary its response to a given level of thyroid hormone. Similar phenomena
have been observed for several other members of the steroid receptor superfamily.
6) Phosphorylation: Most, if not all of the receptors have been shown to be
phosphorylated. In some cases this inactivates the receptor; in others it may be
required for full activity. Since both protein kinases and phosphatases are under the
control of other hormones, modulation of the phosphorylation state of the receptor
allows the response to a steroid hormone to be blocked, attenuated, or enhanced by
other hormones.
7) Alteration of the hormone: An additional mechanism by which a cell can modulate
its response to a hormone is for the cell to modify the hormone. This can take two
forms: conversion of an inactive compound into an active hormone, or inactivation of
an active hormone prior to receptor binding. An example of the latter is elaborated
in the Clinical Correlation section (below).
The important concept to remember at this point is that the response to a given
hormone depends on a number of other factors. The response cannot be considered
in isolation; it depends on the cell type, on the presence or absence of other proteins,
and on the levels of other hormones to which the cell can respond, or to which the
cell has recently responded.
Clinical Correlation
Diseases, and especially genetic disorders, involve defects in normal biochemical
pathways. One example of a defect in steroid hormone action is a rare disorder
called Syndrome of Apparent Mineralocorticoid Excess. AME is characterized by
hypertension and other symptoms of high mineralocorticoid levels, but presents
with low levels of aldosterone (and all other physiological mineralocorticoids).
Individuals ingesting large amounts of licorice have similar symptoms, as do
individuals given carbenoxolone (an anti-ulcer drug, now rarely used due to its
hypertensive side effects).
In order to understand the basis for the disorder, it is necessary to consider a few
facts regarding mineralocorticoid physiology. In in vitro experiments, cortisol, the
major human glucocorticoid, binds the mineralocorticoid receptor with equal affinity
to that of aldosterone, the major human mineralocorticoid. (As mentioned in an
earlier section, the mineralocorticoid receptor is also called the Type I
glucocorticoid receptor, because the Type I GR has a higher affinity for cortisol
14
than does the Type II GR.) In normal individuals, the concentration of cortisol free
in circulation is always much (~100-fold) higher than that of aldosterone. You might
predict from these facts that the mineralocorticoid receptor is misnamed, and never
responds to aldosterone. However, there is a specific response to aldosterone.
How can these observations be explained? There must be some mechanism to allow
mineralocorticoid-responsive cells to screen out the vast excess of cortisol so that
they can specifically respond to aldosterone. Since the usual method of doing this
(i.e. having a specific receptor for the desired ligand) doesnt apply, the cells must do
something else.
A few years ago this something else was discovered: mineralocorticoid-responsive
tissues were found to contain an enzyme that inactivates cortisol, but not
aldosterone. For both GR Type I and Type II, steroid ligands that contain a ketone
at the 11 position, such as cortisone, are inactive. An enzyme, 11-hydroxysteroid
dehydrogenase (11-HSD), reversibly interconverts cortisol and cortisone.
Aldosterone, unlike cortisol, contains an 18-aldehyde; in solution, this forms a cyclic
hemiacetal with the 11-hydroxyl (Figure 12). The cyclic hemiacetal form of
aldosterone is not a substrate for 11-HSD, and therefore aldosterone is not
inactivated by 11-HSD.
OH
HO
Aldosterone
(aldol form)
OH
O
HO
Nonenzymatic
Aldosterone
(hemiacetal form)
OH
OH
O
O
HO
OH
11-HSD
OH
Cortisol
Cortisone
Why does an individual with AME respond as if the levels of mineralocorticoid were
high? After all, aldosterone levels are unmeasurably low! The reason is that the cell
doesnt know (or care) that we call cortisol a glucocorticoid -- the cell simply
responds to the high levels of a circulating compound that is capable of binding to
and activating the GR Type I (mineralocorticoid receptor). This is an important
concept: activation of the receptor by a compound results in all of the consequences
of hormone action at that receptor, regardless of what we call the compound. Drugs
such as synthetic agonists use this as their mode of action. In some cases, drugs
may affect hormone action unintentionally; the hypertension induced by the antiulcer drug carbenoxolone (as a result of its inhibition of 11-HSD) is this type of a
side effect.
Why is the 11-HSD mechanism required (i.e. why isnt there a mineralocorticoid receptor that
doesnt bind cortisol)? The simple answer is we dont know. A separate mineralocorticoid response
may be a relatively late evolutionary development. There is some evidence that the fetus uses
11-HSD to reduce its exposure to cortisol until relatively late in fetal life (when cortisol is required
for development, especially in the lungs). It is possible that the differential functioning of the two
receptor types in some tissues is an evolutionary outgrowth of this ability to selectively inactivate
cortisol.
In some tissues, such as sweat glands, there is evidence that cortisol mediates the sodium retention
response. It appears, in fact, the separate actions of mineralocorticoids and glucocorticoids are
largely confined to a few selected tissues (e.g. kidney, colon, and parotid, and a few small areas of the
brain responsible for salt appetite), and that, with those exceptions, the primary adrenal hormone for
physiological purposes is cortisol. (This is even more apparent in adults, who have much lower
requirements for mineralocorticoids than do children; in fact, adrenalectomized adults treated with
cortisol may not require mineralocorticoid supplementation.) In tissues lacking 11-HSD but
containing Type I receptors (e.g. heart, liver, and hippocampus), cortisol appears to exert differential
effects via interaction with high (Type I) and low (Type II) affinity receptors; these effects may be
mediated by Type I/Type II heterodimer formation or through different HREs activated by the
different receptors. Thus, the designation mineralocorticoid receptor probably represents
recognition of a special case; physiologically (other than in a few tissues) there are actually two forms
of glucocorticoid receptor: Type I and II.
Type I and II receptors can be differentially affected by synthetic agonists or antagonists, which is
useful therapeutically.
16
hormones and their receptors in the etiology and maintenance of at least some types of cancer. In
addition, there is evidence that tumors derived from hormonally responsive tissue are usually, if not
always, initially hormone-dependent. According to this hypothesis, the 65-70% of breast cancer cases
that are not hormone-dependent upon detection have lost their requirement for estrogen. Based on
this hypothesis, a large (16,000 women) clinical trial is currently underway, testing the efficacy of
prophylactic treatment with tamoxifen (an estrogen receptor antagonist) in women at high risk for
breast cancer. This trial is somewhat controversial, because there is some evidence that tamoxifen
treatment during tumor initiation may select for more aggressive tumors, and because tamoxifen has
been implicated in the induction of uterine cancer, as well as having other side effects.
Cancer may be the result of expression of mutant forms and/or overexpression of wild-type forms of
the receptors. The classical definition of an oncogene is a coding segment acquired by a tumorigenic
retrovirus. By this definition v-erbA is the only known oncogene derived from the superfamily (v-erbA
is a mutated form of one of the thyroid hormone receptor genes), and v-erbA is, in fact, only directly
oncogenic in a few cell lines. By a wider definition, a number of the receptors can be designated as
potentially proto-oncogenic. A number of hepatic carcinomas have been shown to overexpress retinoic
acid receptor-, a receptor not normally found in hepatic cells, suggesting a role for this receptor in
carcinogenesis. A variant form of the estrogen receptor that lacks exon 5 (i.e. part of the hormone
binding domain) has been described in a number of putatively estrogen receptor-negative but
progesterone receptor-positive breast tumors (this is important because progesterone receptor gene
expression is normally considered to require estrogen action). The Exon 5 (-) ER is constitutively
active (albeit with only 10-15% of wild-type activity) in cultured cells, which may account for the
presence of PR in the breast cancer tissue. Furthermore, a constitutively active ER would be
expected, even in the absence of estrogen, to exhibit the mitogenic capabilities normally inherent in
its hormonally responsive wild-type homologue.
Constitutive activity, either as an activator or as a repressor, may be a common feature in
carcinogenesis resulting from some receptor-related oncogene products. The v-erbA protein is a
constitutive repressor. RAR- is a constitutive repressor in the absence of retinoic acid; alternatively,
retinoids have been shown to exhibit proliferative effects in epithelial cells, such as those from which
hepatic tumors are usually derived.
Acute promyelocytic leukemia was an essentially uniformly fatal disorder. A few years ago, the
disease was discovered to be associated with a chromosomal translocation between chromosomes 17
and 15 that fuses the genes for RAR- and a previously unknown nuclear protein called PML. The
PML-RAR fusion appears to inhibit differentiation of the mutant cells. Treatment of APL patients
with retinoids induces differentiation and cessation of growth of the tumor cells, and has resulted in
a dramatically reduced fatality rate, with complete remission in a majority of cases.
An additional potential mechanism for carcinogenesis involves the formation of covalent adducts
resulting in constitutively active receptors. Long-term exposure of estrogen-responsive tissue to the
estrogen metabolite 16-hydroxy-estrone may result in irreversible covalent modification of the
receptor. There is some evidence suggesting that some 16-hydroxy-estrone-protein adduct appears
to be tightly associated with the nucleus. If this protein is ER, and if the 16-hydroxy-estrone-ER is
constitutively active, this would suggest a mechanism for tumor induction. 16-hydroxy-estrone has
relatively low affinity for the estrogen receptor in vitro, but has been shown to be a potent estrogen
in bioassays, and some correlation between 16-hydroxy-estrone levels and breast cancer incidence
has been observed.
17
References
Evans (1988) The steroid and thyroid hormone receptor superfamily, Science 240,
889-894.
Godowski & Picard (1989) Steroid receptors: how to be both a receptor and a
transcription factor, Biochem. Pharmacol. 38, 3135-3143.
Sluyser (1990) Steroid/thyroid receptor-like proteins with oncogenic potential: a
review, Cancer Research 50, 451-458.
Carson-Jurica, Schrader, & OMalley (1990) Steroid receptor family: structure and
functions, Endocr. Rev. 11, 201-220.
Wahli & Martinez (1991) Superfamily of steroid nuclear receptors: positive and
negative regulators of gene expression. FASEB J. 5: 2243-2249.
Ort, et al. (1992) Phosphorylation of steroid hormone receptors. Endocr. Rev.
13:105-128.
Freedman (1992) Anatomy of the steroid receptor zinc finger region. Endocr. Rev.
13:129-145.
Funder (1993) Aldosterone action. Ann. Rev. Physiol. 55:115-130.
Rhodes & Klug (February, 1993) Zinc fingers. Scientific American 268: 56-65.
Wurtz, et al. (1996) A canonical structure for the ligand-binding domain of nuclear
receptors. Nature Struct. Biol. 3: 87-94.
Katzenellenbogen, et al. (1996) Tripartite steroid hormone receptor pharmacology:
interaction with multiple effector sites as a basis for the cell- and promoterspecific action of these hormones. Mol. Endocrinol. 10: 119-131.
Parker & White (1996) Nuclear receptors spring into action. Nature Struct. Biol. 3:
113-115.
Bamberger, et al. (1996) Molecular determinants of glucocorticoid receptor function
and tissue sensitivity to glucocorticoids. Endocr. Rev. 17: 245-261.
Revelli, et al. (1998) Non-genomic actions of steroid hormones in reproductive
tissues. Endocr. Rev. 19: 3-17.
18