Developmental Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892-1862
Address correspondence and requests for reprints to: George P. Chrousos, M.D., Developmental Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892-1862. E-mail: chrousog{at}mail.nih.gov
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
|
Starting in 1985, the receptors for each of the steroid hormones
were cloned and sequenced and found to belong to the Type 1 subclass of
classic nuclear receptors, which together with receptors of the Type 2
subclass (including the receptors for Vitamin D, thyroid hormone,
retinoids, rexinoids, and farsanoids) and an ever expanding list of
orphan receptors, constitute the superfamily of nuclear hormone
receptors (6, 7). Generally, nuclear receptors are homologous modular
proteins with a carboxyterminal ligand-binding domain (LBD), a middle
DNA-binding domain (DBD), and a variable amino-terminal domain (NTD)
(6, 7, 8) (Fig. 1). The latter is quite long and nonhomologous in steroid
hormone receptors. It contains a strong independent transactivation
domain (AF1 or
1) and is important for adding
specificity to receptor action. The DBD has two DNA-binding
"zinc-fingers" and contains also a dimerization and a nuclear
localization domain (NLS1). The LBD, in addition to binding the
hormone, has a second transactivation domain (AF2 or
2), a second nuclear localization sequence
(NLS2), a heat shock protein 90-binding domain, a corepressor domain
important for silencing of the receptor, and domains that interact with
other nuclear transcription factors, such as the cjun-cfos and nuclear
factor (NF)-
B heterodimers.
|
The ability of the ligand-bound steroid receptors to transactivate a
steroid-responsive gene depends on the presence of AF1- and
AF2-interacting, "bridging" nucleoproteins, the coactivators that
have chromatin-remodeling and other enzymatic activities (11, 12)
(Figs. 1 and 2
). The known
coactivators of steroid receptors belong to several families (Table 2
). The p160 family and the recently described
riboprotein coactivator steroid receptor activator (SRA) include
members whose activities are limited to nuclear receptors (11, 12, 13). The
CREB-binding protein (CBP)/p300 family of coactivators and the
CBP/p300-associated PCAF are important for other signal transduction
systems as well, including the protein kinase A-cAMP-CREB, the growth
factor-cfos/cjun, the growth factor/cytokine Jak-STAT, and the
cytokine-NF
B pathways. Because of their wider functions, CBP and
p300 have been also called cointegrators.
|
|
Nuclear nonligand-bound steroid receptors are kept inactive by
association to another group of nuclear proteins, the corepressors (11, 16) (Fig. 1 and Table 2
). The corepressor complex contains one or more
of several histone deacetylases (HDAC) that function as condensers of
the nucleosomes and, hence, as silencers of the transcriptional
activity of the steroid receptors (Fig. 2
). Collectively,
coactivators/cointegrators and corepressors have been called
coregulators of steroid receptors. Their use by steroid and other
nuclear receptors is characterized by varying degrees of relative
specificity and combinatorial flexibility. Also, their relative
abundance may vary from tissue to tissue and can be stoichiometrically
limiting for some but not other nuclear receptors and other factors in
a particular tissue. For example, the mutual inhibition of
glucocorticoid receptor and NF-
B seems to be due to competition for
common coactivators (17). On the other hand, the SRC-1 knockout mouse
suffers from a mild, subclinical form of multiple steroid hormone
resistance syndrome, apparently as a result of compensatory elevations
of SRC-2 replacing some of the missing SRC-1 activity (18). In
contrast, the CBP and p300 knockout mice seem to have major distinct
pathologies, CBP cannot fully compensate for the absence of p300, and
compound heterozygous mutants for CBP and p300 invariably die in
utero (19). These findings suggest that there is an absolute
requirement for a combined level of these two homologous coactivators
to allow normal animal development.
Because both coactivators and corepressors may occupy limiting quantities of components of the coactivator or corepressor complex that are essential for the positive or negative activity of a certain nuclear receptor, increasing their levels may respectively lead to squelching or potentiating the activity of this receptor, behaving thus paradoxically and antithetically, as corepressors or coactivators (20). Thus, the qualitative and quantitative mix of coactivators and corepressors in a cell can increase or decrease the sensitivity of this cell to one or more steroids and other nuclear hormones. A change in this mix could lead to hypersensitivity and/or resistance to steroid hormones with hormonal and/or tissue predilection.
The clinical and/or biochemical manifestations of the New World
nonhuman primate syndrome and the human multiple steroid resistance
syndrome described in this issue (Table 1) can both be explained by a
defect of a not necessarily the same coregulator molecule. A major
difference between the two is the fact that the former is a
physiological evolutionary quantitative trait associated with multiple
appropriate adaptations, whereas the latter is a pathological syndrome
with adverse clinical manifestations. Also, the former includes
resistance to all steroid hormones and vitamin D, whereas the latter
has clear manifestations of glucocorticoid, mineralocorticoid, androgen
and estrogen resistance, and possibly, albeit not yet definitely,
progesterone resistance, but no Vitamin D resistance.
The index case of the human syndrome could have been diagnosed with only isolated generalized glucocorticoid resistance on the basis of the clinical and biochemical findings (21). Yet, the physicians were alerted by the paradoxical absence of hyperandrogenic manifestations to pursue the evaluation further. Her mineralocorticoid and mild estrogen resistance could have been missed entirely were it not for the informed diagnostic persistence of the clinicians. The advanced bone age was most likely the result of the different impact of the putative defective coregulator on the cells of the growth plate, where apparently the elevated adrenal androgens were allowed to exert their effects. The absence of a high-circulating aldosterone, estradiol, and progesterone concentration may be due to, respectively, adequate effects of the elevated nonaldosterone adrenal mineralocorticoids on the kidney mineralocorticoid receptor (resulting in excessive salt retention and potassium loss) or estrogen and/or progesterone resistance limited to the uterus and not extended to the hypothalamic-pituitary unit. The milder form of multiple steroid resistance observed in the sister of the proposita could be due to variable penetrance of a similar defect, possibly as a result of genetic background differences in genes with epistatic effects on the functions affected. The absence of overt clinical or biochemical pathology in the parents of the affected sisters could be due to an adequately compensated heterozygotic defect.
The New World primate physiological and biochemical syndrome and the two pathological human multiple steroid resistance syndrome cases described by New et al. (5) are the first states in which a defective steroid receptor coregulator was suggested to be responsible for the clinical and/or biochemical picture (4, 5). Although many attempts have been made to explain the first by imputing steroid receptor binding inhibitory factors in the cytosol, increases in antagonistic steroid receptor isoforms, or presence of hormone-responsive element inhibitors, none of these have been universal to all affected steroid/sterol receptor systems and, thus, cannot explain the entire syndrome in a satisfactory fashion. The likelihood that a coregulator is affected in multiple steroid resistance is high, and the search for altered coregulators should continue in the nonhuman primates and the patients.
The Rubinstein-Taybi syndrome is a good example of a coregulator
disease, however, its primary manifestations are not related to the
steroid/sterol hormone signal transduction systems (22). On the other
hand, the presence of amplification of SRC-3 in certain human breast
and ovarian cancers suggests that this coactivator might be responsible
for excessive tumor growth, through hypersensitization of the estrogen
receptor signal transduction pathway (23). In this instance, a clonal
somatic mutation of the coactivator or of one of its regulatory
molecules may have led to breast or ovarian tumorigenesis, even though
the systemic secretion of estrogen may have been entirely normal in the
affected patients. Another imputed human glucocorticoid
hypersensitivity state due to steroid coactivator hyperfunction is
systemic infection with HIV-1, characterized by profound
immunosuppression with a shift of the T cell helper (h) phenotype from
Th1 to Th2, as well as by myopathy and muscle atrophy, both known
effects of glucocorticoid excess (24). It has been suggested that Vpr,
an accessory HIV-1 protein, may be in part responsible for these
manifestations by interacting with glucocorticoid receptors, host
coactivators and TAF, and by exerting marked glucocorticoid coactivator
activity in the immune system and muscle (Table 2).
Now that physicians have been alerted to the pathogenic potential of altered coregulators, it is quite likely that more steroid hormone-related syndromes with diverse, albeit logically explained, clinical pictures will be discovered. Some of these syndromes will be admittedly rare, but others may be subtle and common. A good place to start searching for alterations of steroid receptor coregulator molecules is in already recognized syndromes of steroid hormone resistance or hypersensitivity, in which the pathogenicity of the steroid receptor and other known noncoregulator postreceptor steps have been excluded. Importantly, however, the coregulators of steroid hormones should be examined as potential participants in the pathogenesis of common polygenic disorders, such as obesity and the insulin-resistance (visceral fat) syndrome, the polycystic ovary syndrome, hypertension, major depression, autoimmune disorders, infertility, and so on, in which the glucocorticoid, mineralocorticoid and sex steroid signal transduction systems may be pathophysiologically involved (25, 26).
![]() |
Footnotes |
---|
Received May 12, 1999.
Accepted May 21, 1999.
![]() |
References |
---|
![]() ![]() ![]() |
---|