Endocrine Reviews 20 (2): 156-188
Copyright © 1999 by The Endocrine Society
The Vitamin D Receptor and the Syndrome of Hereditary 1,25-Dihydroxyvitamin D-Resistant Rickets1
Peter J. Malloy,
J. Wesley Pike and
David Feldman
Department of Medicine (P.J.M., D.F.), Stanford University School
of Medicine, Stanford, California 94305-5103; and Department of
Molecular and Cellular Physiology (J.W.P.), University of
Cincinnati, Cincinnati, Ohio 45267
 |
Abstract
|
|---|
- I. The Syndrome of Hereditary 1,25-Dihydroxyvitamin D-Resistant Rickets
(HVDRR)
- A. Historical
- B. Clinical features of HVDRR
- C. Pathophysiology
- D. Alopecia
- E. 1,25-Dihydroxyvitamin D [1,25-(OH)2D] action and HVDRR
- II. Vitamin D Physiology
- A. Metabolism
- B. 1
-Hydroxylase deficiency
- III. 1,25-Dihydroxyvitamin D Action Mediated by the Vitamin D receptor (VDR)
- A. Historical aspects of VDR structure and function
- B. The domain structure of the VDR
- C. The regulation of gene expression by the VDR
- IV. Cellular Basis of HVDRR
- A. Studies in cultured skin fibroblasts
- B. Studies in other cells
- V. The VDR Gene and the Molecular Basis of HVDRR
- A. The VDR chromosomal gene
- B. The VDR gene promoter
- C. Polymorphisms of the VDR gene
- VI. HVDRR Mutations Causing the Ligand-Binding Positive Phenotype
- A. Initial description of DNA-binding domain (DBD) mutations
- B. Characterization of additional DBD mutations
- C. Structural analysis of DBD mutations
- VII. HVDRR Mutations Causing the Ligand-Binding Negative Phenotype
- A. Initial description of ligand-binding domain (LBD) mutations
- B. Characterization of additional LBD mutations
- C. Structural analysis of LBD mutations
- VIII. Additional Mutations In The VDR Gene
- A. Hinge region mutations
- B. Splice site mutations
- C. Major structural mutations
- D. Vitamin D resistance without a mutation in the VDR
- IX. HVDRR Mouse Model
- X. Treatment of HVDRR
- A. Vitamin D
- B. Calcium
- C. Prenatal diagnosis
- D. Spontaneous healing of rickets
- XI. Analysis, Summary, and Conclusions
|
I. The Syndrome of Hereditary 1,25-Dihydroxyvitamin D-Resistant
Rickets (HVDRR)
|
|---|
VITAMIN D, the primary regulator of calcium homeostasis in
the body, is particularly important in skeletal development and in bone
mineralization. The active form of vitamin D, 1,25-dihydroxyvitamin
D, [1,25(OH)2D] (this notation will be used to signify
either D2 or D3), mediates its actions by
binding with high affinity to specific vitamin D receptors (VDRs)
located in the nucleus of target cells. Hereditary vitamin D-resistant
rickets (HVDRR) is a rare genetic disorder caused by a generalized
resistance to 1,25(OH)2D action (1, 2, 3). Heterogeneous
mutations in the VDR that alter the function of the receptor are the
molecular basis of HVDRR. A variety of mutations have been identified,
some of which render the VDR nonfuctional, imparting a complete
hormone-resistant state, while other mutations reduce VDR activity,
causing a hyporesponsive state. In this review, we will describe the
clinical manifestations of HVDRR, the VDR and its gene, the mechanism
of vitamin D action, and the genetic defects in the VDR that result in
this hormone-resistant syndrome. Other recent reviews of these subjects
can be found in the recently published volume entitled "Vitamin D"
and the references therein (4).
A. Historical
In 1978, Brooks et al. (5) described a patient with
osteomalacia who exhibited hypocalcemia, hypophosphatemia, and
secondary hyperparathyroidism. Interestingly, the patient had markedly
increased serum levels of 1,25(OH)2D. Brooks et
al. (5) suggested that the rickets was due to impaired
responsiveness of target organs to 1,25(OH)2D. They termed
this disease vitamin D-dependent rickets type II (VDDR-II) to
distinguish it from a closely related syndrome known as vitamin
D-dependent rickets type I (VDDR-I), which is due to a defect in an
enzyme leading to the synthesis of 1,25(OH)2D (see
below, Section II.B). Later the same year, Marx et
al. (6) reported similar findings in two children, and the authors
again suggested that the disease was due to end-organ resistance to
1,25(OH)2D. Since these initial studies, there have been
many reports of patients with apparent target organ resistance to
1,25(OH)2D. Over the years a number of different terms have
been used to describe this syndrome. As noted, the original reports
referred to this entity as vitamin D-resistant rickets type II. The
disease also has been called pseudo-vitamin D deficiency rickets type
II (PDDR-II), calcitriol-resistant rickets, vitamin D-resistant
rickets, and hereditary hypocalcemic vitamin D-resistant rickets. We
prefer the designation hereditary vitamin D-resistant rickets (HVDRR)
as a simple and accurate description of this syndrome caused by genetic
resistance to vitamin D.
B. Clinical features of HVDRR
The major clinical findings in patients with HVDRR,
hypocalcemia and rickets, are due to defective intestinal calcium
absorption leading to impaired mineralization of newly forming bone and
preosseous cartilage. The rickets is often severe and is usually
exhibited within months of birth.
Patients suffer from bone pain, muscle weakness, hypotonia, and
occasionally convulsions from hypocalcemia. Children are often growth
retarded, and in some cases they develop severe dental caries or
exhibit hypoplasia of the teeth (7, 8, 9, 10, 11, 12, 13). Some infants have died from
pneumonia caused by poor respiratory movement due to severe rickets of
the chest wall (8, 11, 14). Many children with HVDRR have sparse body
hair, and some have total scalp and body alopecia, including eyebrows
and in some cases eyelashes (Fig. 1
). Alopecia will be discussed in
more detail below (Section I.D).
View larger version (94K):
[in this window]
[in a new window]
|
Figure 1. Children with HVDRR and alopecia. [Reprinted with
permission from J. F. Rosen et al.: J
Pediatr 94:729–735, 1979 (7 )].
|
|
An example of the typical serum biochemistry levels found in HVDRR
cases is shown in Table 1. The
abnormalities include low concentrations of calcium and phosphate and
elevated serum alkaline phosphatase activity. The hypocalcemia leads to
secondary hyperparathyroidism with elevated PTH levels and
hypophosphatemia. The 25(OH)D values are normal and, importantly, the
1,25-(OH)2D levels are elevated. This clinical feature
distinguishes HVDRR from 1-hydroxylase deficiency (VDDR-I or PDDR
discussed below, Section II.B) since the serum
1,25-(OH)2D values in the latter syndrome are depressed. In
the cases in which it has been measured, 24,25(OH)2D levels
have been normal or low (8, 11, 15, 16, 17, 18, 19, 20, 21). Unlike patients with
1-hydroxylase deficiency, most HVDRR individuals are resistant to
supraphysiological doses of all forms of vitamin D therapy (Table 1).
HVDRR follows an autosomal recessive pattern of inheritance. The
recessive nature of the disease is evident from the parents who are
heterozygous for the genetic trait but show a normal phenotype with no
symptoms of the disease and normal bone development. In many, if not
all cases, parental consanguinity is associated with the disease. Males
and females are equally affected and often a family has several
affected children. An extensive pedigree of seven related families with
HVDRR is illustrated in Fig. 2 (22).
View larger version (17K):
[in this window]
[in a new window]
|
Figure 2. Pedigree of large kindred with HVDRR. The families
represented are F11 (C1-C4), F34 (E1-E3), F35 (F1-F3), F36 (H1-H3), F37
(J1-J4), F38 (K1-K3), and F39 (L1). Solid symbols
indicate patients with HVDRR and half-solid symbols
indicate heterozygotes. Double lines denote
consanguineous marriages. (*) Indicates probable brothers in generation
I and III. [Reprinted with permission from P. J. Malloy et
al.: J Clin Invest 86:2071–2079, 1990 (22 ) by
copyright permission of The American Society of Clinical
Investigation.]
|
|
C. Pathophysiology
Among the many biological processes attributed to vitamin D,
maintenance of calcium and bone homeostasis is most apparent.
1,25-(OH)2D is essential for promoting calcium and
phosphate transport across the small intestine and into the
circulation, which is necessary for the normal mineralization of bone.
Approximately 50% of the total intestinal calcium absorption is
attributed to 1,25-(OH)2D action, while the remaining 50%
is due to passive absorption (23, 24). It is now well established that
the biological actions of 1,25-(OH)2D are mediated by the
VDR, a nuclear transcription factor that regulates gene expression in
1,25-(OH)2D-responsive cells. Since vitamin D regulates the
translocation of calcium and phosphate, interference with the
1,25-(OH)2D action pathway causes decreased mineral
transport and hypocalcemia. The hypocalcemia, in turn, results in
secondary hyperparathyroidism, which induces hypophosphatemia. The
calcium and phosphate deficiencies interfere with normal bone
mineralization, leading to rickets in children and osteomalacia in
adults. In HVDRR, 1,25-(OH)2D target organs such as the
intestine are resistant to hormone action, and therefore the intestine
is less efficient in promoting calcium and phosphate absorption into
the circulation. The vitamin D resistance is due to mutations in the
VDR that render the receptor nonfunctional or less functional than the
wild-type VDR.
D. Alopecia
A clinical feature that is generally found in patients with HVDRR
is alopecia totalis (Fig. 1
). The majority of HVDRR patients have
sparse body hair, and some exhibit total scalp and body alopecia (17, 25, 26). Children with extreme alopecia often lack eyebrows and in some
cases eyelashes. Hair loss may be evident at birth or it occurs during
the first few months of life. An analysis of HVDRR patients shows that
there is some correlation between the severity of rickets and the
presence of alopecia (26). Patients with alopecia generally have more
severe resistance to calcitriol than those without alopecia. In
families with a prior history of the disease, the absence of scalp hair
in newborns provides initial diagnostic evidence for HVDRR. The
mechanism causing alopecia is unknown, but VDRs are present in the hair
follicle (27, 28). Skin biopsy has revealed apparently normal follicles
with no hair shaft present. The lack of 1,25-(OH)2D action
during a critical stage of hair follicle development is the suspected
cause of alopecia.
E. 1,25-(OH)2D action and HVDRR
The biological actions of 1,25-(OH)2D in tissues and
cells are orchestrated through complex changes in gene expression (29, 30). These changes lead to cell-specific alterations in the level of
proteins directly responsible for a myriad of differentiated cell
functions, as well as in proteins that act as transcription factors or
as signaling molecules to regulate secondary and tertiary levels of
gene expression (31). In the latter case, these molecules may function
directly within the cell or indirectly via additional cellular
signaling pathways in either autocrine or paracrine fashion. As
indicated earlier, most, if not all, of the molecular actions of
1,25-(OH)2D in the nucleus are mediated by the VDR. The
classic role of 1,25-(OH)2D is to regulate mineral
homeostasis, achieved through its coordinated actions on intestine,
kidney, bone, and parathyroid gland (32, 33). It is not surprising,
therefore, that initial evidence for the existence of the VDR derived
from early investigations in these tissues (34, 35, 36, 37, 38, 39, 40, 41, 42). Interestingly, the
VDR is expressed in a wide variety of tissues, including kidney, skin,
liver, pancreas, muscle, breast, prostate, adrenal, thyroid, and cells
of mesenchymal or hematopoietic origin (27, 28, 43, 44, 45, 46, 47). Although the
VDR in these tissues appears to arise from the same chromosomal gene,
its role in cellular function is not homeostatic in nature but rather
pleiotropic. Whereas the classic actions of vitamin D are to regulate
calcium homeostasis, the expanded scope of vitamin D pleiotropic
actions include stimulation of differentiation, inhibition of cell
proliferation, and suppression of the immune response (45, 46, 47, 48). In
addition, the regulation of cellular proliferation and differentiation
by 1,25-(OH)2D appears to be a common feature in many
tissues examined, and it is likely that this regulatory feature is a
fundamental component of all biological responses to
1,25-(OH)2D. Notwithstanding the complexity and diversity
of biological responses elicited by 1,25-(OH)2D, the
profound skeletal abnormalities demonstrated in patients with HVDRR
emphasizes the fundamental and essential role of
1,25-(OH)2D in calcium homeostasis.
Although there are multiple pleiotropic tissue responses regulated by
1,25-(OH)2D, children with HVDRR appear relatively normal
except for the constellation of features that relate to their calcium
deficiency, rickets, and alopecia. VDRs have been found in endocrine
glands such as pituitary, pancreas, parathyroid, gonads, and placenta,
and 1,25-(OH)2D3 regulates hormone synthesis
and secretion from these glands (29, 45, 46, 49, 50, 51). VDRs have also
been found in hematolymphopoietic cells, and
1,25-(OH)2D3 regulates cell differentiation and
the production of interleukins and cytokines (52). Hochberg et
al. (53) examined hormone secretion in patients with HVDRR and
found no abnormalities in insulin, TSH, PRL, GH, and testosterone
levels in serum. Even et al. (54) showed that urinary cAMP
and renal excretion of potassium, phosphorous, and bicarbonate were
normal in HVDRR patients treated with PTH. However, PTH failed to
decrease urinary calcium and sodium excretion in these patients to the
extent found in controls. This suggests that 1,25-(OH)2D
may selectively modulate the renal response to PTH and facilitate the
PTH-induced reabsorption of calcium and sodium (54). Although minor
aberrations have been noted in the fungicidal activity of neutrophils
from HVDRR patients (55), the patients do not exhibit any clinically
apparent immunological defects. In the light of the diverse actions of
1,25-(OH)2D demonstrated in many nonosteogenic tissues, the
absence of related findings in children with HVDRR suggests that the
pleiotropic responses regulated by 1,25-(OH)2D in these
nonosteogenic tissues are redundant and that other factors or
compensatory mechanisms subsume the role of vitamin D in such a way
that abnormalities are not clinically manifested. Similarly, the VDR
knockout mouse displays the same phenotypic and physiological patterns
as patients with HVDRR (56, 57). The VDR knockout mouse model can be
used to analyze the abnormalities caused by the loss of VDR action in
detail not possible in the HVDRR patients (see Section IX).
|
II. Vitamin D Physiology
|
|---|
A. Metabolism
Vitamin D is a fat-soluble secosteroid that exists in two forms,
vitamin D3 (cholecalciferol) from animal sources and
vitamin D2 (ergocalciferol) from plant sources. Since
vitamin D is synthesized in the skin, it is a hormone and not a true
vitamin. In animals, the precursor (provitamin) molecule,
7-dehydrocholesterol, is cleaved between carbon-9 and -10 in the B
ring, using the energy derived from UV B rays of sunlight, which opens
the ring and creates the secosteroid structure (Fig. 3) (58). Vitamins D2 and
D3 are essentially biologically inactive and must be
converted to hydroxylated metabolites to gain hormonal activity. Upon
entering the circulation, vitamin D (D2 or D3)
binds to the vitamin D-binding protein, a 58-kDa plasma -globulin
(59, 60). In the liver, vitamin D is hydroxylated at the carbon-25
position to form 25-hydroxyvitamin D [25(OH)D] by the enzyme
25-hydroxylase (61). This enzyme, also known as CYP27, is a
multifunctional cytochrome P450 oxidase that also hydroxylates
cholesterol and bile acids. In the kidney the enzyme 25-hydroxyvitamin
D-1-hydroxylase (1-hydroxylase), also a P450 oxidase, adds a
second hydroxyl group to 25(OH)D to form 1,25-(OH)2D, the
hormonally active form of vitamin D.
View larger version (43K):
[in this window]
[in a new window]
|
Figure 3. Synthesis of 1,25-dihydroxyvitamin D3
and molecular mechanism of action. The steps leading to the synthesis
of the active form of vitamin D, 1,25-dihydroxyvitamin D3
[1,25-(OH)2D3], are illustrated. In the skin,
7-dehydrocholesterol is cleaved by UV light to form the prohormone
vitamin D3. In the liver this molecule is hydroxylated at
the 25 position to form 25-hydroxyvitamin D3, the major
circulating form of vitamin D. In the kidney a second hydroxyl group is
added to 25-dihydroxyvitamin D3 at the 1 position by
1-hydroxylase converting it to the hormonally active metabolite
1,25-dihydroxyvitamin D3. 1,25-Dihydroxyvitamin
D3 circulates in the blood mainly bound to the vitamin
D-binding protein (DBP) in equilibrium with a small amount of free
hormone. The free lipophilic ligand diffuses through the lipid bilayer
of the cell membrane and into the nuclear compartment, where it binds
with high affinity to the VDR. Ligand binding causes a conformational
change in the receptor that opens surfaces on the molecule, which allow
it to heterodimerize with RXR. Specific DNA binding to VDREs occurs via
the two zinc finger modules of the DBD of the receptors. Once bound to
DNA, the VDR-RXR heterodimer recruits additional coactivators. The
VDR-RXR-coactivator complex interacts with the general transcription
apparatus (GTA) to initiate gene transcription. The newly synthesized
proteins determine the physiological response to the hormone.
|
|
The level of renal 1-hydroxylase activity is tightly regulated
by PTH (62). Patients with hyperparathyroidism have elevated serum
levels of 1,25-(OH)2D, whereas patients with
hypoparathyroidism show reduced levels of the hormone (62, 63). Other
factors, including phosphate and 1,25-(OH)2D and possibly
phosphatonin, also regulate 1-hydroxylase activity. High phosphate
levels in the serum lead to a suppression of 1-hydroxylase activity
while low phosphate levels tend to increase the amount of enzyme
activity. 1,25-(OH)2D regulates its own production, both by
suppressing PTH secretion and feedback inhibition of renal
1-hydroxylase activity (62). In the feedback inhibition loop, low
1,25-(OH)2D levels lead to increased 1-hydroxylase
activity and high 1,25-(OH)2D levels inhibit the enzyme
activity in the kidney. 1-Hydroxylase activity is also influenced by
the concentration of calcium in the serum. The calcium acts indirectly
by regulating PTH levels via the calcium-sensing receptor, thereby
determining 1-hydroxylase activity.
The catabolism of 1,25-(OH)2D involves a series of
enzymatic steps, the first of which is catalyzed by the enzyme
25-hydroxyvitamin D-24-hydroxylase (24-hydroxylase) (64).
24-Hydroxylation generates the trihydroxy form of vitamin D,
1,24,25(OH)3D, a biologically less active form of
1,25-(OH)2D, that is further subjected to additional
hydroxylations and oxidations leading to the production of calcitroic
acid, which is excreted in the urine (65, 66). Interestingly,
1,25-(OH)2D induces 24-hydroxylase activity in a number of
cells and, therefore, regulates the rate of its own metabolism. The
genes encoding the rat and human 24-hydroxylase (CYP24) have been
cloned (67, 68, 69), and a vitamin D-response element (VDRE) (see
below, Section III.C) has been identified in the regulatory
region of these genes (70, 71). Regulation of 24-hydroxylase gene
expression or enzyme activity is a very useful marker of
1,25-(OH)2D responsiveness in the evaluation of HVDRR (16).
In humans, the 24-hydroxylase gene is found on chromosome 20 at 20q13
(69).
In addition to hydroxylating 1,25-(OH)2D, 24-hydroxylase
can convert 25(OH)D to 24,25(OH)2D. The biological role of
24,25(OH)2D has not been established with certainty.
However, recent studies with a 24-hydroxylase knockout mouse model
strongly suggest that 24,25(OH)2D may have some biological
activity distinct from 1,25-(OH)2D, especially on cartilage
cells (72).
B. 1-Hydroxylase deficiency
We will discuss this disease entity briefly since it has some
elements in common with HVDRR and must be distinguished from it (Table 2). Decreased production of
1,25-(OH)2D is found in patients with 1-hydroxylase
deficiency. In 1961, Prader et al. (73) described a patient
with a genetic form of vitamin D-resistant rickets that ultimately was
due to a deficiency of renal 1-hydroxylase. This disease has been
known as vitamin D-dependent rickets type I (VDDR-I) and also as pseudo
vitamin D deficiency rickets (PDDR) (74). Genetic linkage studies
indicate that the mutation causing 1-hydroxylase deficiency is
linked to chromosome 12 at 12q14 (75, 76, 77). Recently, several groups
have cloned the 1-hydroxylase gene (78, 79, 80, 81, 82, 83), and it has been
confirmed that the locus on chromosome 12 is the 1-hydroxylase gene
and that VDDR-I (or PDDR) is due to mutations in the gene encoding the
1-hydroxylase. In several cases, mutations in the 1-hydroxylase
gene have been elucidated (78, 83).
1-Hydroxylase deficiency is an autosomal recessive disease that is
manifested at an early age, presenting with hypotonia, muscle weakness,
growth failure, and rickets. Hypocalcemia, elevated PTH levels,
increased alkaline phosphatase activity, and low urine calcium
excretion are also found. Tetany and convulsions may occur with severe
hypocalcemia. These symptoms are also characteristic of HVDRR. Patients
with 1-hydroxylase deficiency have normal or elevated 25(OH)D levels
but low 1,25-(OH)2D levels due to various defects in the
1-hydroxylase enzyme. Patients with this condition are
treated with physiological doses of
1,25-(OH)2D3 (0.25–2 µg/day) that bypass the
defective enzyme and restore serum calcium concentrations to
normal. The low serum levels of 1,25-(OH)2D and the
therapeutic response to physiological doses of exogenous
1,25-(OH)2D3 distinguish 1-hydroxylase
deficiency from HVDRR (Table 2).
|
III. 1,25-Dihydroxyvitamin D Action Mediated by the Vitamin D
Receptor (VDR)
|
|---|
A. Historical aspects of VDR structure and function
A number of observations have established the central role played
by the VDR in the biological activities of
1,25-(OH)2D3. Initial support evolved from
early studies which revealed that the vitamin D ligand (or its
precursor) interacted with a protein found in the nucleus (34, 35, 36, 84, 85). This protein displayed properties consistent with those of
receptor molecules known to exist for other steroid hormones. The
protein was expressed in low copy number exclusively in target tissues
and exhibited both high affinity and selectivity for
1,25-(OH)2D3 (32) as well as the capacity to
bind to DNA (86). This latter observation was not only conceptually
important in the characterization of the VDR, but pivotal in the
purification of the protein, leading to the eventual development of
useful anti-VDR antibodies (87, 88). Confirmation that the VDR was
indeed a nuclear receptor emerged definitively as a result of the
molecular cloning of the chicken gene in 1987 (89) and the human (90)
and rat (91) genes shortly thereafter. Examination of the deduced
primary sequence of the VDR cDNA revealed that it is structurally and
functionally homologous to an emerging family of genes that encoded
receptors for the sex and adrenal steroids, thyroid hormone, and
retinoic acid (92, 93). This superfamily of nuclear receptor and
transcription factor genes, derived from both vertebrate and
invertebrate sources, presently includes more than 60 members (94). It
is comprised of the known nuclear ligand-activated receptors, as well
as an ever growing number of receptors for which ligands have not yet
been identified. While many of these orphan receptors may not require
ligands for activation, investigation of several of these receptors led
to the identification of at least four new lipophilic ligands, which
include 9-cis retinoic acid, prostaglandin J2, and certain
farnesol intermediates (94), as well as several derivatives of
cholesterol (95).
B. The domain structure of the VDR
1. Overview. The VDR exhibits a modular domain structure
generally similar to that of other members of the nuclear receptor gene
family (see Fig. 4). Since the mutations within the VDR gene, which
will be detailed subsequently in this review, are distributed across
several of its domains, we will describe the structural organization of
the functional protein. While the structural domains were initially
deduced from the gene sequence, it is important to note that functional
correlations with these domains have evolved from extensive examination
of receptor activities of natural mutations in patients with HVDRR as
well as site-directed mutagenesis of the cDNA and recombinant
expression in host mammalian cell types. Recent crystallographic
studies for several of the receptors support the predicted
structure-function relationships (see below).
View larger version (14K):
[in this window]
[in a new window]
|
Figure 4. Schematic illustration of VDR mRNA and protein.
The linear diagram above the protein rectangle documents
an mRNA species of 4800 nucleotides that is derived from the
chromosomal gene and is composed of spliced exons designated 1–9. The
boundaries of these exons within the mRNA and their corresponding
locations within the VDR protein are indicated. The VDR protein is
comprised of 427 aa and five general domains, regions designated A/B,
C, D, and E whose approximate boundaries are illustrated. The
shaded regions within the protein exhibit strong
homology with other members of the nuclear receptor gene family. These
regions of homology include the amino-terminal DBD (C domain,
gray) and the carboxy-terminal LBD (E domain), which
contains three internal regions of homology (black).
Subregions E1 and AF-2 represent helices within the E domain that
participate in transactivation.
|
|
With respect to the nuclear receptor gene family, the common structural
domains are designated A–F (Fig. 4)
(96). While a considerable degree of sequence and structural homology
exists across the superfamily for several of the receptor domains,
others exhibit little or no sequence homology and in some instances
domains may be abbreviated or absent (92). This diversity is manifest
most strongly in regions A/B, D, and F. Segment A/B includes all
residues that extend amino terminal to the DNA-binding domain (DBD) of
the receptor. Its size is highly variable, ranging from hundreds of
amino acids (aa) in the progesterone receptor (PR), for example, to
approximately 24 aa in the VDR. Region C comprises the highly conserved
DBD of the receptor gene family and represents the hallmark feature of
this group of proteins, which will be described in detail below. The D
domain, which is least conserved among the nuclear receptors, appears
to serve as a hinge between the DBD and the E region. The D region
within the VDR is 50 residues longer than that found in the other
steroid receptors and is likely the result of an additional exon within
the VDR chromosomal gene (97). Consistent with other receptors, this
region in the VDR exhibits very limited species conservation,
particularly at the level of the inserted exon. The hinge region is
also a site of serine phosphorylation (98, 99) and may play additional
functional roles. The E region encodes the ligand-binding domain (LBD)
of the hormone-activated receptors and exhibits several important
functional activities, which will be discussed below. The small F
region is not conserved within the nuclear receptor family and is, in
fact, absent in the VDR as well as in several other members of the
family.
2. DBD. The DBD of the VDR (aa residues 24–90)
contains nine highly conserved cysteine residues and consists of two
similar motifs each comprised of a zinc-coordinated finger structure
(Fig. 5). Each zinc atom is tetrahedrally
coordinated through four of the highly conserved cysteine residues that
serve to stabilize the finger structure itself. These finger modules
are structurally unrelated to the multiple zinc fingers found in
transcription factor IIIA wherein the zinc atom is coordinated
through two cysteines and two histidines (100, 101). Interestingly,
although the two zinc modules of the VDR appear highly related
structurally, they are not topologically equivalent (100). This lack of
equivalency is consistent with the fact that each module serves a
different function within the protein. The amino-terminal module,
comprised of an -helix known as the P box (aa residues 41–46),
functions to direct specific DNA-binding in the major groove of the
DNA-binding site. The carboxy-terminal module, on the other hand, which
also contains an -helix known as the D box (aa residue 61–65),
serves as a dimerization interface for interaction with a partner
protein (see below) (102, 103). This region may serve a lesser function
in the VDR as a result of the asymmetric nature of the interaction of
VDR with its partner protein, retinoid X receptor (RXR), when bound to
a cognate DNA response element. Indeed, modeling studies suggest that
several residues (aa residues 91 and 92) located in an extended
-helix immediately downstream of the second zinc finger (aa residues
90–101) and termed the T box, provide key interactions with partner
proteins. This T box region also likely makes minor groove contacts
with nucleotides located between the two DNA half-sites, thus
strengthening the interaction of the VDR with its DNA-binding elements.
Posttranslational modification of VDR by phosphorylation of Ser51
inhibits its ability to complex with the VDRE and may serve as a
negative regulator of VDR activity (104, 105). When the
three-dimensional structure of the VDR DBD emerges, either as a result
of nuclear magnetic resonance spectroscopy or x-ray crystallography
studies, our understanding of the structural organization of these
modules, as well as the mechanisms through which they function to
interact with DNA, will be enhanced. Recent successes using these
techniques with the estrogen receptor (ER), glucocorticoid receptor
(GR), retinoic acid receptor (RAR), and RXR DBDs have provided
significant insights into the structure of this domain (106, 107, 108, 109, 110, 111). As
will be discussed below, mutations in critical amino acids within both
zinc finger modules have rendered the VDR nonfunctional and have caused
HVDRR presumably by interfering with VDR binding to DNA.
3. LBD. The E region of the VDR represents a complex
multifunctional domain that retains strong dimerization and
transactivation potential, induced and manifested through the binding
of 1,25-(OH)2D3. Despite the fact that
mutations have been identified that compromise hormone binding, little
is known regarding the overall structure of the ligand-binding pocket
of the VDR. It is hypothesized that binding leads to conformational
changes that expose, enhance, or produce novel dimerization and/or
transactivation interfaces. Indirect evidence for
1,25-(OH)2D3-induced changes in conformation
were first provided by Allegretto and co-workers (112, 113), using
proteolytic digestion assays. Current efforts with more sensitive
proteolytic digestion assays confirm that, with the exception of helix
H12 (the helix structure is discussed below), the E domain of the VDR
acquires resistance to proteolysis after interaction with ligand, much
like that of other nuclear receptors (114). Whether ligand-induced
conformational changes are restricted to the E region of the VDR is
unknown but probably unlikely. Functionally, ligand conversion of the
VDR to the active form leads to an increased capacity to form protein
dimers with partner proteins (see below) and stimulates DNA binding
(115). Ligand binding also results in exposure of additional regions of
the VDR, which act to recruit proteins active in modifying chromatin,
such as SRC-1 (116, 117, 118) and SMRT (119, 120), or which
facilitate contact with such proteins as transcription factor IIB or
the TATA box binding protein-associated factors (TAFs) that are
associated with the core transcriptional machinery (121, 122). It is
likely that these latter interactions occur in both gene promoter- and
cell-specific ways to produce the observed tissue-selective actions of
1,25-(OH)2D3 that are well documented. As with
other nuclear receptors, these tissue-selective actions are
particularly manifested by the VDR when associated with synthetic
vitamin D analogs, suggesting that subtle differences in receptor
conformation are capable of producing profoundly different biological
consequences (114, 123). Mutations throughout the LBD may cause HVDRR
by a number of mechanisms. They may completely prevent ligand binding
or reduce affinity for 1,25-(OH)2D3.
Alternatively, mutations may alter VDR conformation, compromising its
ability to dimerize with RXR or interact with coactivators.
4. Structural modeling. Recently, the three-dimensional
structure of the LBD of RXR (124), RAR
(125), and thyroid
receptor (TR1) (126), all RXR partners, as well as ER (127) and PR
(127, 128), have been elucidated. We will focus our analysis of the LBD
on the RXR partners since they are more relevant to the VDR. RAR and
TR1 receptors were crystallized in the presence of ligand
(holodomains), whereas the RXR structure was determined in the
absence of ligand (apodomain). Twelve -helices (H1-H12) arranged as
an antiparallel -helical sandwich comprise the bulk of the LBD of
each of the receptors (129). It is likely that the VDR will be arranged
in a structurally similar manner, and a model of the VDR helical
structure based on the canonical structure of nuclear receptors is
shown in Figs. 6 and 7. These
three-dimensional structures support the idea that H9 and H10 are
essential for the formation of RAR, VDR, TR, or peroxisome
proliferator-activating receptor (PPAR) heterodimers with a common RXR
subunit. While functional studies support the requirement of these
helical sequences in heterodimer formation, H9 and H10 may be
insufficient for VDR and PPAR dimerization. This conclusion is
experimentally supported by a complete evaluation of the dimerization
properties of the carboxy-terminal E region of the VDR (130).
5. Transactivation. An additional function inherent to the E
region of the VDR is an activation function termed AF-2. The core of
this domain function is located at the extreme carboxy terminus of the
protein and is associated with a small subregion defined
crystallographically as H12 (117, 118, 124, 125, 126, 131). An important
observation regarding H12 is the clear repositioning of this -helix
back upon the hydrophobic core and upstream -helices in the E domain
in response to ligand binding (Fig. 7).
This repositioning likely "locks"
1,25-(OH)2D3 into its binding pocket and leads
to the formation of a complex high-affinity protein surface capable of
interacting with specific comodulators such as SRC-1 and others. This
requisite interaction between H12 and more upstream -helices likely
accounts for the observed loss of VDR transcriptional capacity after
mutation of residues located in either region of the receptor molecule
(132, 133). The three-dimensional structure of the VDR will, however,
have to be solved directly to confirm these predictions.
C. The regulation of gene expression by the VDR
The proof of the regulation of gene expression by VDR was
initially developed by studies of HVDRR where natural mutations in the
VDR gene prevented 1,25-(OH)2D3 induction of
target genes such as 24-hydroxylase (16, 18, 134). An understanding of
the molecular mechanisms through which
1,25-(OH)2D3 and its receptor modulate gene
expression is outlined in Fig. 3 and emerged initially through
examination of the promoters for genes known to be regulated by this
hormone. A major focus of these studies was to demonstrate that
1,25-(OH)2D3 could regulate promoter activity,
that specific DNA sequences within the promoter were required, and that
a functional VDR was an essential component of the transactivation
machinery. These principles of direct regulation of gene expression
by 1,25-(OH)2D3 were established using human
(135, 136, 137) and rat (138, 139, 140) osteocalcin gene promoters as models and
were strengthened through subsequent investigation of the osteopontin
gene (141), the calbindin genes (142, 143), and the 25-hydroxyvitamin
D3 24-hydroxylase genes (70, 144, 145). HVDRR mutant VDRs
were shown to be incapable of activating such promoter constructs, both
supporting the critical role of functional VDR in transactivation as
well as defining the defect causing HVDRR (146, 147, 148).
1. Direct regulation of transcription by 1,25-(OH)2D3. Initial investigations by McDonnell
et al. (149) demonstrated that
1,25-(OH)2D3 stimulated transcription from the
human osteocalcin gene promoter in a dose-dependent fashion consistent
with its actions on the expression of the endogenous gene. The
sensitivity of this promoter to 1,25-(OH)2D3
enabled subsequent definition of the DNA sequence element within the
promoter responsible for mediating hormone action (see below). Equally
important, the ability of 1,25-(OH)2D3 to
stimulate this transcriptional activation was dependent upon the
presence of intact VDR. Thus, while the osteocalcin promoter was
unresponsive to 1,25-(OH)2D3 when introduced
into a VDR-negative cell line or cotransfection with an HVDRR mutant
VDR, cointroduction of the wild-type VDR via a recombinant expression
vector restored hormonal responsiveness. The establishment of this VDR-
and hormone-dependent assay in cells enabled a determination of the
functional domains of the VDR. As will be seen in subsequent sections
of this review, this assay has become essential in establishing the
inactivating nature of numerous mutations found in the VDR chromosomal
gene in patients with HVDRR.
2. VDREs. Deletion analysis of the human osteocalcin promoter
led to the identification of a cis-acting element that was
located approximately 500 bp upstream of the transcriptional start site
and that mediated 1,25-(OH)2D3 induction (135).
This study and one by Ozono et al. (137) resulted in
definition of the first VDRE, a directly repeated hexanucleotide
sequence separated by 3 bp. Parallel studies using the rat osteocalcin
gene promoter led to similar conclusions regarding the organizational
motif of the VDRE (138, 139, 140). Definitive VDREs have since been
localized in the mouse osteopontin gene (141), the rat and human
24-hydroxylase genes (70, 144, 145), and the human p21 gene (150). As
seen in Fig. 8, each of these elements is
comprised of two directly repeated hexanucleotide half-sites and, like
the human osteocalcin VDRE, the half-sites are separated by a 3-bp
spacer. Studies on the mouse calbindin D9K (143) and the
rat calbindin D28K (142) genes have also revealed
apparent VDREs, although these sequences mediate rather weak
1,25-(OH)2D3 induction in their natural
promoter environments and do not exhibit structural similarity to the
more well characterized VDREs. One possibility is that the VDR binds to
these sites, perhaps in combination with a nonconventional protein
partner(s) (see below). The exact mechanism whereby the VDR interacts
with its VDRE, which is described in the next section, is particularly
relevant since a large majority of the mutations found within the VDR
gene that lead to HVDRR are found within the VDR DBD and their
mechanism is to interfere with DNA binding.
View larger version (42K):
[in this window]
[in a new window]
|
Figure 8. Model of the RXR/VDR heterodimer bound to a
consensus VDRE. RXR (R) and VDR (V) are illustrated bound to a directly
repeated VDRE located upstream of the start site of transcription. The
dark circle represents the
1,25-(OH)2D3 ligand, whereas P represents a
serine phosphorylation site at residue 208 in the human protein. The
nucleotide sequences of several known VDREs and the genes in which they
reside are documented below the figure. Each
hexanucleotide sequence is separated by a 3-bp spacer.
|
|
3. Interaction of VDR with VDRE DNA. The definition of VDREs
within vitamin D-sensitive gene promoters facilitated subsequent
examination of the interaction between the VDR and its DNA-binding
sites. An extensive battery of mutations introduced into the VDR and
tested for functionality have revealed that two domains within the VDR
are required for high-affinity DNA binding in vitro, the
DNA-binding or C domain and the ligand-binding or E domain.
Interestingly, the molecular basis for abrogation of DNA binding in
each of the two regions is fundamentally different. In the DBD,
missense mutations found in HVDRR cases or mutations generated by
site-directed mutagenesis that lead to a disruption of one or both of
the zinc-coordinated finger structures reduce or prevent direct
receptor-VDRE interaction and/or sensitize the receptor to subsequent
proteolytic degradation (151, 152, 153, 154, 155). In contrast, loss of DNA binding
via LBD mutations within the carboxy-terminal E domain result from the
disruption of functions unrelated to DNA binding per se but
essential to the formation of a high-affinity receptor DNA complex
capable of transactivation (130, 132, 149, 156). Mutations that
specifically lead to loss of hormone binding represent one such class.
Assuming that the defect is restricted to ligand interaction, this type
of mutation results in a protein capable of DNA binding but unable to
process the signal that initiates the DNA binding events and subsequent
transactivation. Several mutations of this class have been identified
in HVDRR and will be discussed below. A second series of mutations
disrupts the capacity of VDR to form dimers with partner proteins that
include, but are not restricted to, the receptor RXR (see below for
discussion). Examples of this type of mutation have been observed in
HVDRR (157), although it arises more commonly in thyroid hormone
resistance (158). Finally, mutations that lead to truncations within
the carboxy-terminal domain also abrogate DNA binding, but do so as a
result of the obvious loss of both dimerization and hormone-binding
capabilities.
4. Subunit structure of the active VDR heterodimer. The
observation that VDREs are comprised of two half-sites led to the
prediction that the VDR might bind to these sites as a functional
dimer. Indeed, the interaction of each of the classic steroid receptors
with their respective hormone response elements (HREs) are known to
occur via homodimerization. It was a surprise, therefore, to observe
that the VDR did not associate with DNA as a homodimer, but rather as a
heterodimer (Fig. 8). This finding, made by Liao et al.
(159) and Sone et al. (152, 160), demonstrated that DNA
binding of recombinant VDR produced in yeast or in vitro
could only be achieved after reconstitution with mammalian cellular
extracts. The factor supplied by the extract, whose presence was
confirmed by several additional groups (161, 162), was termed
nuclear accessory factor or NAF. NAF was found to be widely distributed
in cells and tissues and hypothesized to be an unknown nuclear
receptor. Similar heterodimeric activities were identified for the
thyroid hormone receptor (TR) and RAR. In 1991 and 1992, four
groups of investigators demonstrated that this activity derived from
three related members of the nuclear receptor gene family termed
retinoid X receptor (RXR, RXRß, and RXR) (163, 164, 165, 166). It was
concluded that NAF was a single manifestation of a mixture of one or
more of the RXRs expressed in any given cell type. True dimerization
between the VDR and RXR has been confirmed through definition of the
VDR dimerization domain (130). Utilizing an extensive series of
internal deletions of the VDR, two regions located within the E domain
were shown to be essential for interaction with RXR. These regions
coincide with two subregions within the E domain that are moderately
conserved within the entire nuclear receptor gene family. Perlmann
et al. (167) suggested recently that a small region of 40 aa
lying within the second E/F homology domain (corresponding to H9 and
H10 of the crystal structure of the RXR LBD) is sufficient within RXR
to form dimers with RAR and TR. This same region is apparently not
sufficient, however, to permit formation of RXR-VDR heterodimers,
suggesting that the domains responsible for interaction between RXR and
other signaling partners may be different.
Interestingly, in view of the asymmetric nature of the VDRE (two
directly repeated half-sites) and the heterodimeric nature of the VDR
modulatory unit (RXR and VDR), distinct polarity must also exist with
respect to the half-site-receptor interaction. Indeed, Jin et
al. (130) and Freedman and co-workers (153) demonstrated that RXR
binds to the upstream or 5'-half-element and that VDR binds to the
downstream half-element on consensus type VDREs (Fig. 8). Whether this
polarity is maintained on all VDREs remains to be proven. However, this
organization is consistent with that noted for both RXR-TR and RXR-RAR
heterodimers bound to their respective HREs (168, 169). Irrespective of
the mechanism, the existence of a general permissive partner protein
that functions as a central regulator for several endocrine systems
suggests the potential for considerable cross-talk between the systems.
5. The transactivation domain of the VDR. As indicated
earlier, two regions are believed to mediate the transactivation
capacity of the VDR, a carboxy-terminal AF-2 region comprised of H12,
which includes residues 416–424, and the E1 region located in the
midregion of the receptor comprising H3 and H4, which includes residues
from 232 to 272 (30) (Figs. 4 and 6). Additional helices downstream of
H4 may be involved as well. Based upon the crystal structure of apo-RXR
LBD and holo-RAR LBD (124, 125, 126), it is likely that interaction of
1,25-(OH)2D3 with the VDR leads to the folding
of H12 back against the hydrophobic core of the E domain and helices
therein. This repositioning completes the high-affinity binding site
for 1,25-(OH)2D3 and creates a complex surface
capable of additional protein-protein interactions (Fig. 7). Indeed,
mutations in each of these two regions can selectively abrogate ligand
binding, transactivation, or both.
Does the AF-2 domain of RXR contribute to transactivation by the VDR?
It is clear that RXR has the potential to function as a signaling
partner when complexed to orphan receptors, such as LXR and
NGF-IB (167, 170). Additional studies suggest that the AF-2 function of
the silent partner is required for activation by the signaling partner
(171, 172). This supports the idea that the AF-2s of the silent and the
signaling receptors both participate in a common protein surface
essential for activation of transcription. Perhaps activation of the
silent partner occurs via the process of dimerization with the
signaling partner and/or after DNA binding (173). Evidence is rapidly
accumulating for such regulatory events.
6. Transcriptional comodulators and cofactors. The
transactivation domain of most nuclear receptors interacts directly
with additional classes of proteins termed comodulators. These proteins
include the positive modulators or coactivators SRC-1 (116),
GRIP-1/TIF2 (174, 175), and ACTR (176) and the negative regulators or
corepressors NCoR (177) and SMRT (119, 120). Indeed, the AF-2 of
the VDR is found to be essential for interaction with SRC-1 (117, 118)
and likely is essential for other comodulator interactions as well
(Fig. 9). After recruitment of a
comodulator into the promoter region by a nuclear receptor, these
proteins both induce chromatin structural changes enzymatically or
recruit additional proteins capable of similar actions or both.
Coactivators appear to exhibit histone acetyltransferase activity (176)
and recruit additional histone acetyltransferases such as P/CAF
(178) and CBP/p300 (179), each of which functions to
enzymatically modify histones such that nucleosomes are destabilized
and transcription is facilitated. The ability of CBP to interact
simultaneously with a number of unrelated transcription factors also
suggests an additional function for these large proteins—that of
integrating multiple inputs on a gene promoter so as to produce a
coherent output (180, 181). In contrast, corepressors apparently served
to recruit deacetylases such as HDAC1 (182, 183, 184). Histone deacetylase
activity serves to stabilize chromatin and repress transcription (111).
In this paradigm, nuclear receptors appear to function, at least in
part, to recruit the enzymatic machinery necessary for carrying out the
changes in chromatin structure essential for the effective regulation
of transcription. Nuclear receptors are, however, capable of additional
important contacts. Two groups have shown that a domain within the VDR
hinge region makes important contacts with TFIIB (121, 122), a member
of the initiation complex per se. This interaction, one of
several made by nuclear receptors, appears to enhance transcription
through direct stabilization of the transcriptional machinery.
View larger version (31K):
[in this window]
[in a new window]
|
Figure 9. Hypothetical recruitment of comodulators to a
vitamin D-sensitive gene promoter. Entry of
1,25-(OH)2D3 into cells leads to the formation
of a heterodimer of RXR and VDR and subsequent DNA binding. DNA binding
of this heterodimer initiates recruitment of additional protein factors
to the oligomeric complex. These factors are involved in the
modification of chromatin structures through histone acetylation or
deacetylation and facilitate essential contact with the general
transcription apparatus (GTA). Proteins that may play a role include
the known comodulators, SRC-1, P/CAF, and CBP/p300, as well as
unknown proteins indicated as X.
|
|
As will be detailed in subsequent sections of this review,
mutations that interfere with various steps in the
1,25-(OH)2 D-VDR hormone action pathway lead to the
syndrome of HVDRR.
|
IV. Cellular Basis of HVDRR
|
|---|
A. Studies in cultured skin fibroblasts
The syndrome of HVDRR was first recognized as an entity in
1978/1979 (5, 6, 7, 185), and since that time a number of cases of vitamin
D resistance have been described, which are detailed in Table 3. The HVDRR cases are referred to
throughout this review as F1, F2, etc., where the F denotes the family
number (8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 25, 134, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209). The clinical findings in children
affected with HVDRR, along with their failure to respond to the
administration of physiological and even supraphysiological doses of
vitamin D, suggest that the disease is caused by end-organ resistance
to vitamin D. The identification and characterization of the VDR as the
mediator of 1,25-(OH)2D3 action and its
presence in the major target tissues of vitamin D action led
investigators to suspect that the cause of HVDRR was due to a genetic
defect in the VDR (6, 7, 8, 9). Studies of VDR from HVDRR cases began after
the demonstration that VDRs were present in skin (27, 43, 210) and
could be studied in cultures derived from human skin biopsies (211).
In 1982, Feldman et al. (16) examined the vitamin D system
in cultured skin fibroblasts from two siblings with HVDRR (family F11).
This study demonstrated that high-salt extracts of cultured fibroblasts
from HVDRR patients had undetectable levels of
[3H]1,25-(OH)2D3 binding. The
fibroblasts from the affected individuals were also resistant to high
concentrations of 1,25-(OH)2D3, since
24-hydroxylase activity, a well characterized biomarker for
1,25-(OH)2D3 responsiveness, could not be
induced after hormone treatment. This approach provided a model for
studying HVDRR subjects using cultured dermal fibroblasts to examine
both the ligand-binding properties of the VDR as well as cellular
responsiveness to 1,25-(OH)2D3 measured by the
induction of 24-hydroxylase. Subsequently, a number of other HVDRR
cases were examined using cultured skin fibroblasts (11, 188, 189, 190, 191) or,
in one case, cells derived from bone (192). In some patients the
fibroblasts lacked specific
[3H]1,25-(OH)2D3 binding (11, 188, 189, 190, 191, 192) similar to the cases reported by Feldman et al.
(16). On the other hand, some fibroblasts exhibited normal
[3H]1,25-(OH)2D3 binding.
Nevertheless, the HVDRR cells showed no response to
1,25-(OH)2D3 treatment (11, 18, 190, 192, 197, 203). From these early studies it was apparent that at least two
classes of VDR defects exist. Patients who had normal
[3H]1,25-(OH)2D3 binding were
categorized as "receptor-positive" or "ligand-binding positive"
and, in most cases, their defect was later shown to be in the DBD of
the VDR. Those cases that showed decreased or absent
[3H]1,25-(OH)2D3 binding were
denoted "receptor negative" or "ligand binding negative" and
the defect was later shown to be in the LBD of the VDR or due to absent
VDR.
Studies by Griffin and Zerwekh (190) and Liberman et al.
(191, 192) also used 24-hydroxylase activity to demonstrate
1,25-(OH)2D3 resistance. On the other hand,
Clemens et al. (189) showed that fibroblasts from HVDRR
patients were not growth-arrested after hormone treatment in contrast
to fibroblasts from healthy individuals that were growth-arrested.
These early observations showed that cells from HVDRR patients were
resistant to 1,25-(OH)2D3 and that a variety of
abnormalities in the VDR existed.
1. Ligand-binding negative phenotype. As the number of reports
on HVDRR increased, the heterogeneous nature of the
defects in the VDR became more apparent. Hochberg
et al. (17, 25) reported clinical findings in four patients
from two unrelated families of Arab origin (F11, F18) who exhibited
HVDRR and alopecia. Fibroblasts from three of these patients and
several of their parents, as well as an additional unrelated family
from Germany (F17), were studied by Chen et al. (134). In
these studies, HVDRR fibroblasts exhibited negligible
[3H]1,25-(OH)2D3-binding, and
hormonal treatment failed to induce 24-hydroxylase enzyme activity.
These cases were designated as representing the ligand-binding negative
phenotype. Interestingly, of the five parents in the study, only the
father in family F18 exhibited a phenotype theoretically expected of a
heterozygote. His fibroblasts contained half of the normal amount of
[3H]1,25-(OH)2D3-binding and also
showed a half-maximal response to 1,25-(OH)2D3.
The fibroblasts of the other four parents showed a normal complement of
VDR and a normal response to 1,25-(OH)2D3
treatment, which has been the usual pattern found in fibroblasts of
heterozygotes.
In 1982, after the development of a monoclonal antibody to the VDR (87, 212, 213), the presence of VDR in some patients with the ligand
binding-negative phenotype (11, 15) was assessed by Pike et
al. (214). Using a radioligand immunoassay (215), an
immunoreactive protein was detected in cell extracts from fibroblasts
of ligand binding-negative HVDRR patients (214). The authors speculated
that the defect in the VDR was due to a structural abnormality in the
LBD preventing [3H]1,25-(OH)2D3
from binding to the receptor and not from defective synthesis of the
VDR protein (214).
2. Ligand binding-positive phenotype. A ligand
binding-positive case was described by Hirst et al. (18) who
studied the VDR in fibroblasts from a Haitian family (F19) with HVDRR.
Cultured fibroblasts from two sisters with HVDRR were unresponsive to
1,25-(OH)2D3 treatment despite normal
[3H]1,25-(OH)2D3-binding and VDR
abundance. In addition, sucrose gradient sedimentation studies revealed
a VDR of normal size. However, the protein exhibited a decreased
ability to form aggregates in low-salt conditions as compared with
normal receptor. Using DNA-cellulose chromatography, the authors
demonstrated that the VDR from the HVDRR fibroblasts exhibited a
significant decrease in affinity for heterologous DNA. The normal
receptor eluted from the DNA-cellulose at 170–173 mM KCl,
while the mutant receptor eluted at 105–109 mM KCl. A
second HVDRR family (F31), studied by Malloy et al. (203),
exhibited a similar defect in the DNA binding properties of the
receptor. The VDR from the affected individuals displayed normal
[3H]1,25-(OH)2D3 binding and was
normal in size as shown by Western blotting. However, DNA-cellulose
chromatography clearly revealed that the VDR had a low affinity for DNA
eluting at 100 mM KCl in contrast to the wild-type VDR,
which eluted at 200 mM KCl. The patients from F19 and F31
families were therefore categorized as having the ligand
binding-positive phenotype and, in addition, had a VDR that exhibited a
low affinity for DNA. When the VDR from the parents’ cells were
subjected to DNA-cellulose chromatography, two forms of the receptor
were found. One receptor form eluted at 200 mM KCl,
indicating that it had a high affinity for DNA similar to the wild-type
receptor, while the other form eluted at 100 mM KCl,
demonstrating a low affinity for DNA similar to the HVDRR patient. This
was the first clear evidence showing the heterozygous state of the
HVDRR parents. It was suspected that the defects in these cases would
likely be due to point mutations in the DBD (18, 203), which was later
proved correct (146).
Liberman et al. (197) also described four cases (F1, F3, F5,
F20) of ligand binding-positive resistance to
1,25-(OH)2D3. Two of the cases (F5, F20)
exhibited VDRs with a low affinity for DNA similar to the F19 and F31
families. Gamblin et al. (194) examined
1,25-(OH)2D3 induction of 24-hydroxylase
activity in F5 and F20 fibroblasts and demonstrated complete hormone
resistance. In the other cases, F1 and F3, Liberman et al.
(197) demonstrated that the VDRs had a reduced ability to localize to
the nucleus despite showing a normal affinity for DNA. Gamblin et
al. (194) further showed that the F1 and F3 fibroblasts exhibited
24-hydroxylase activity when exposed to high concentrations of
1,25-(OH)2D3. Patients F1 and F3, whose
fibroblasts showed a response to high concentrations of
1,25-(OH)2D3 in vitro, also
exhibited a calcemic response to high doses of calciferols in
vivo.
Castells et al. (196) also described a ligand
binding-positive patient (F22) who had sparse hair, rickets, and high
circulating 1,25-(OH)2D3 levels. Studies of the
VDR from the patient’s fibroblasts showed that the receptor had a
decreased affinity for
[3H]1,25-(OH)2D3. The patient
responded with a marked improvement in the disease after treatment with
extremely high doses of 1,25-(OH)2D3,
apparently overcoming the low-affinity binding abnormality in the VDR.
B. Studies in other cells
In addition to cultured skin fibroblasts and bone cells, studies
of the VDR from patients with HVDRR have been carried out in a number
of other cell types including peripheral mononuclear cells (195),
phytohemagglutinin (PHA)-stimulated lymphocytes (200, 206), myeloid
progenitor cells (201), Epstein-Barr virus (EBV)-immortalized B
lymphoblasts (22, 146, 147, 203), and HTLV-1 virus-immortalized T
lymphoblasts (205). It is interesting to note that EBV-immortalized B
lymphoblasts, from normal subjects that express wild-type VDR, do not
exhibit induction of 24-hydroxylase activity or growth inhibition in
response to 1,25-(OH)2D3 (22). On the other
hand, PHA-stimulated lymphocytes and HTLV-1-immortalized T lymphoblasts
from normal subjects are capable of responding to
1,25-(OH)2D3 (205, 216). In PHA-stimulated
lymphocytes, the absence of an inhibitory effect of
1,25-(OH)2D3 on DNA synthesis or lack of
induction of 24-hydroxylase activity have been used as markers to
rapidly diagnose HVDRR (200, 206). In addition, Takeda et
al. (206) showed that PHA-stimulated lymphocytes from parents of
children with HVDRR expressed intermediate levels of 24-hydroxylase
when treated with 1,25-(OH)2D3.
|
V. The VDR Gene and the Molecular Basis of HVDRR
|
|---|
Investigations into the molecular basis of HVDRR began shortly
after the human VDR cDNA was described by Baker et al. (90).
Util
Top
Abstract
I. The Syndrome of...
-loop (loop 1–3)
flips under H6, which stabilizes H11 and H12 positions.