| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Departments of Medicine (D.M., X.B., D.P., A.C.K., D.G.), Dentistry, and Anatomy and Cell Biology (M.D.M.), McGill University, Montréal, Québec, Canada H3A 1A1
Address all correspondence and requests for reprints to: Dr. David Goltzman, Calcium Research Laboratory, Department of Medicine, Room H4.67, Royal Victoria Hospital, 687 Pine Avenue West, Montréal, Québec, Canada H3A 1A1.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Many insights into the pathogenesis of Hyp have been derived from studies of a murine homolog of the human disease observed in Hyp mice. Linkage studies in the mouse have mapped the Hyp mutation to a region of the mouse X-chromosome syntenic to the human Hyp locus. The Hyp mouse arose naturally as a spontaneous mutation and has since been bred on the C57BL/6J background. Like their human counterparts, this murine model has renal phosphate wasting, impaired mineralization, and growth retardation. The renal defect in phosphate reabsorption in Hyp mice is believed to reside at the level of the brush-border membrane of the proximal tubule (2). To evaluate whether the phosphate wasting resulted from a primary renal defect or from elaboration of a humoral factor that alters phosphate uptake in the renal proximal tubule, Meyer et al. (3) initially performed parabiosis experiments and suggested the presence of a circulatory phosphaturic substance that was subsequently termed phosphatonin (4). Nesbitt et al. (5) performed renal cross-transplantation between Hyp and normal mice. They found that when normal kidneys were transplanted into nephrectomized Hyp mice, the kidneys wasted phosphorus. When Hyp kidneys were transplanted into nephrectomized normal mice, the kidneys retained phosphorus normally. Thus, the defect in the Hyp mouse is neither corrected nor transferred by renal cross-transplantation, indicating that the phosphate transport defect in the Hyp mouse is not due to an intrinsic renal abnormality.
In view of the prominence of the skeletal phenotype in this disorder, considerable attention in recent years has been directed at the skeletal osteoblast. An intrinsic osteoblast defect was proposed by Ecarot et al. (6, 7, 8), who performed experiments transplanting periostea and osteoblasts into the gluteal muscles of normal and Hyp mice. When normal cells were transplanted into Hyp mice, mineralization was impaired. However, when Hyp cells were transplanted into normal mice, reduction, but not normalization, of mineralization was observed. These studies supported the hypothesis that there is a primary osteoblast defect in the Hyp mouse, but did not definitively exclude the possibility that a putative circulating factor in the Hyp mouse could lead to a prolonged defect in the osteoblast.
More recently, a positional cloning approach was used to identify PHEX (phosphate-regulating gene with homology to endopeptidases on the X-chromosome) as the candidate gene for Hyp (9). Human and mouse PHEX/Phex complementary DNAs have now been cloned and sequenced (10, 11, 12, 13, 14). Amino acid sequence comparisons have demonstrated structural homologies between PHEX/Phex. Whereas several investigations have examined PHEX/Phex expression at the tissue level using a variety of techniques (10, 11, 12, 15, 16, 17) and have shown that PHEX/Phex messenger RNA (mRNA) expression is predominantly in bone and teeth, the mechanism by which loss of PHEX/Phex function leads to disease is not immediately apparent.
The organic constituents of bone form a well organized extracellular matrix whose main component is type I collagen, but which also contains, on a molar basis, a roughly equal proportion of noncollagenous proteins (NCPs). It is thought that these proteins play critical roles in regulating mineralization and in facilitating the attachment of osteoblasts and osteoclasts to the bone matrix. Accordingly, investigation of the changes in NCP content and distribution may yield valuable insight into the pathogenesis of Hyp osteomalacia and bone mineralization in general. However, relatively few data are available regarding the abnormalities of NCPs in hypophosphatemia, and these stem largely from analysis of Hyp serum and Hyp-derived osteoblast cultures. To date, no abnormalities of NCP expression and distribution have been reported in intact Hyp bone.
In the present study to explore how the loss of Phex function contributes to the pathogenesis of osteomalacia, we have addressed several questions. 1) Which cells in intact normal bone synthesize and secrete Phex protein? 2) What is the relationship between the Phex abnormality and osteomalacia in Hyp mice? 3) Are there alterations in the expression and secretion of bone matrix proteins in situ in Hyp mice? 4) Besides the mineralization defect, are there osteogenic cell differentiation defects? To answer these questions, we have examined the abnormalities of mineralization, Phex and bone matrix protein expression, and extracellular localization occurring in Hyp mice in vivo and investigated whether there were any osteogenic cell differentiation and mineralization defects by using ex vivo bone marrow cell cultures.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animals
All mice used were of the C57BL/6J strain. Inbred Hyp mice were
initially obtained from The Jackson Laboratory (Bar
Harbor, ME). Normal and Hyp mice were bred and maintained on Purina
chow (Ralston Purina Co. of Canada Ltd., Montréal,
Canada) containing 0.5% calcium and 0.74% phosphorus. Normal males
(+/Y) and hemizygous Hyp males (Hyp/Y), aged 45–50 days, were used in
these experiments.
Histology
Six 40- to 45-day-old age-matched, male, wild-type and Hyp mice
were killed by cervical dislocation. The distal ends of femurs were
removed, dissected free of soft tissue, and fixed in PLP fixative (2%
paraformaldehyde containing 0.075 M lysine and 0.01
M sodium periodate solution) for 24 h at 5 C. The left
distal femurs were then decalcified in EDTA-glycerol solution (14.5 g
EDTA, 15 ml of glycerol, and 85 ml distilled water with solid sodium
hydroxide added until a final pH of 7.3 was reached) for 5–7 days at 5
C. Tissues were dehydrated in graded alcohols and embedded in low
melting point paraffin, after which 5-µm sections were cut on a
rotary microtome. The sections were immunostained as described below.
Undecalcified right distal ends of femurs were embedded in LR White
acrylic resin (London Resin Co. Ltd., London, UK). One-micron sections
were cut on an ultramicrotome. These sections were stained for mineral
with the von Kossa staining procedure and counterstained with toluidine
blue.
Bone marrow cell cultures
Tibiae and femurs of 40- to 45-day-old age-matched, male,
wild-type and Hyp mice were removed under aseptic conditions, and bone
marrow cells (BMC) were flushed out with DMEM containing 15% FCS, 50
µg/ml ascorbic acid, 10 mM ß-glycerophosphate, and
10-8 M dexamethasone.
Cells were dispersed by repeated pipetting, and a single cell
suspension was achieved by forcefully expelling the cells through a
22-gauge syringe needle. No selection for cells of the osteoblast
lineage was made. Total bone marrow cells (106)
were cultured in 36-cm2 petri dishes in 5 ml of
the above-mentioned medium. Such cultures have previously been shown to
contain osteogenic cell precursors, which in culture can produce
mineralized bone nodules (20, 21). The medium was changed
every 4 days, and cultures were maintained for 18 days. Conditioned
medium was harvested every 4 days from normal and Hyp mouse BMC
cultures. At the end of the culture period cells were washed with PBS,
fixed with PLP fixative, and stained cytochemically or
immunocytochemically. Total RNA was extracted for RT-PCR, and proteins
were extracted for Western blotting, as described below.
Cytochemical staining
Alkaline phosphatase (ALP) cytochemistry. Cells were
incubated for 15 min at room temperature in 100 mM
Tris-maleate buffer containing naphthol AS-MX phosphate (0.2 mg/ml;
Sigma) dissolved in ethylene glycol monomethyl ether
(Sigma) as a substrate and Fast Red TR (0.4 mg/ml;
Sigma) as a stain for the reaction product. After washing
with distilled water and air-drying, ALP-positive colony areas were
measured by image analysis as described below.
Alizarin Red S staining for calcium. Cells were exposed to a solution of Alizarin Red S, pH 6.2 (1 mg/mg), for 30 min at room temperature, after which the colonies were gently washed under running water and left to dry. Images of stained petri dishes were photographed (see below). The colonies were destained for further analysis (collagen staining) by briefly washing in 5% perchloric acid.
Picrosirius red staining for collagen. Cells were exposed to 1% sirius red in saturated picric acid for 30 min. Stained cell layers were washed in water thoroughly to remove any nonspecific staining and dried before images were taken.
Methylene blue staining for total colony assessment. Cells were first washed in borate buffer (10 mM; pH 8.8) before staining with 1% methylene blue (wt/vol) in borate buffer for 30 min at room temperature. Cells were then washed three times in borate buffer alone and left to dry. Images of stained petri dishes were taken.
Immunocytochemistry and immunohistochemistry
Cultured cells in petri dishes or paraffin sections of
decalcified bone were stained immunocyto/histochemically for Phex and
bone matrix proteins using the avidin-biotin-peroxidase complex (ABC)
technique (22). The cultured cells or dewaxed sections of
bone were first treated with 1% bovine testicular hyaluronidase
(Sigma) for 30 min at 37 C to increase antibody
penetration and access to epitopes. Primary antibody was applied to
tissues or cells overnight at room temperature. As a negative control,
the preimmune serum or TBS (50 mM Tris-HCl, 150
mM NaCl, and 0.01% Tween-20, pH 7.6) was substituted for
the primary antibody. After washing with high salt buffer (50
mM Tris-HCl, 2.5% NaCl, and 0.05% Tween-20, pH 7.6) for
10 min at room temperature followed by two 10-min washes with TBS, the
cells were incubated with secondary antibody [biotinylated rabbit
antigoat IgG, biotinylated goat antirabbit IgG, biotinylated goat
antimouse IgG, Fab special, (Sigma)]. Cells or sections
were then washed as before and incubated with the
Vectastain ABC-AP kit or the Vectastain Elite
ABC kit (Vector Laboratories, Inc. Ontario, Canada) for 45
min. After washing as before, red pigmentation to demarcate regions of
immunostaining was produced by a 10- to 15-min treatment with Fast Red
TR/Naphthol AS-MX phosphate (Sigma; containing 1
mM levamisole as endogenous ALP inhibitor), or gray
pigmentation was likewise produced using a Vector SG kit (Vector Laboratories, Inc.). After washing with distilled water, the
sections were counterstained with methyl green and mounted with
Kaiser’s glycerol jelly.
Computer-assisted image analysis
After immunostaining of three each of age-matched, male
wild-type and Hyp mice, images of primary spongiosa and cortical bone
from a single section from one femur were digitally recorded using a
rectangular template and three different fields. In the primary
spongiosa, each image was photographed from the edge of the metaphyseal
border of the growth plate (i.e. at the level of the zone of
vascular invasion). In cortical bone, images were taken from the
diaphyseal bone close to the metaphysis. All digital images were
captured at a magnification of x200, producing a field area of 0.4
mm2. For the cultured cells, dish images were
photographed with transmitted light over a light box. All images were
taken with a Sony digital camera, and processed using Northern Eclipse
image analysis software, version 5.0 (Empix Imaging, Inc., Mississauga,
Canada). For determining the area and density of the mineralized matrix
and bone matrix proteins in stained bone sections and positive colonies
in cultured cells, thresholds were set using green and red channels.
The thresholds were determined interactively and empirically on the
basis of three different images. Subsequently, this set threshold was
used to automatically analyze all recorded images of all sections that
were stained in the same staining session under exactly the same
conditions. The ratio of the area of the mineralized matrix and the
area of cytochemical or immunohistochemical stained regions was
calculated automatically by the software in each microscopic field.
Pixel counts (summary total gray) of immunoreaction products
(i.e. bone matrix proteins) were also calculated from
thresholding of each microscopic field (0.4 mm2);
these data reflect the relative amount of bone matrix proteins, as
measured by immunostaining intensity. The software therefore determines
summary total gray as the integrated immunostaining intensity over a
given tissue area.
Western blot analysis
Proteins were extracted from 18-day cultures and quantitated by
the protein assay kit (Bio-Rad Laboratories, Inc.,
Mississauga, Canada). Protein samples (100 µg) were fractionated by
SDS-PAGE and transferred to a polyvinylidene difluoride membrane.
Immunoblotting was carried out using the antibodies described above.
Bands were visualized using the ECL chemiluminescence detection method
(Amersham Pharmacia Biotech, Aylesbury, UK). Western
products were quantified by image analysis using NIH 1.61 image
software.
RT-PCR
Total RNA for RT-PCR was extracted from 18-day cultures by a
single step method using TRIzol reagent, reverse transcribed, and
amplified by PCR using QIAGEN One-Step TR-PCR kit
(QIAGEN, Mississauga, Canada) according to the
manufacturer’s instructions. The primer sets employed are depicted in
Table 1
.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Osteomalacia in Hyp mice as demonstrated by von Kossa
staining
Mineralized bone matrix was demonstrated in trabecular and
cortical bone in wild-type mice (Figs. 1, A and C, respectively). In Hyp mice, a disproportionately large amount
of unmineralized bone matrix was seen in trabecular and cortical bone
(Fig. 1, B and D, respectively, asterisks). Typical,
hypomineralized periosteocytic lesions were also observed in both
trabecular and cortical bone of Hyp mice (Fig. 1, B and D,
respectively, arrow, inset). Quantitative data
derived from image analysis showed that total mineral per unit area was
decreased by 82.6% in the primary spongiosa and by 50.7% in cortical
bone when comparing Hyp to wild-type mice.
|
, E and G), but not in Hyp mice (Fig. 1, F and H).
Immunohistochemical staining of Phex protein was also stronger in
osteocytes than in osteoblasts and localized predominantly to the
plasma membrane (Fig. 1E, inset).
Immunohistochemistry for type I collagen in primary spongiosa and
cortical bone
Immunolocalization of type I collagen in trabecular (Fig. 2, A and B) and cortical (Fig. 2, D and
E) bone was similar between wild-type and Hyp mice. Type I collagen was
detected by intense staining of bone matrix and by faint staining of
osteoblasts in both wild-type (Fig. 2, A and D) and Hyp (Fig. 2, B and
E) mice. No obvious differences were apparent in the distribution
pattern of type I collagen in cortical bone between wild-type (Fig. 2D)
and Hyp (Fig. 2E) mice. Likewise, a similar abundance of
immunodetectable type I collagen relative to total tissue area was seen
in the primary spongiosa between wild-type and Hyp mice. When
observations were made, relative to trabecular area, rather than
relative to the entire primary spongiosa, however, there was less
immunostaining in Hyp mice than in wild-type mice (confirmed by
quantitation data below). This reduction in the ratio of type I
collagen in the trabecular area in Hyp vs. wild-type mice
was attributable primarily to wider trabeculae and more trabecular area
in the Hyp mice than in wild-type mice. More remnant cartilage matrix
was evident within the larger trabeculae in Hyp mice compared with
wild-type mice, as evidenced by type II collagen immunostaining
(23), and in both Hyp mice and wild-type mice, little or
no immunostaining was observed for type I collagen in remnant cartilage
matrix within trabecular and cortical bone.
|
. No difference was seen in
cortical bone (Fig. 4).
|
|
, A and C). Although a similar staining
pattern was observed for osteocalcin in Hyp mice (Fig. 5, B and D),
there were some regions of cortical bone matrix that were strikingly
devoid of immunodetectable osteocalcin in the Hyp mice (Fig. 5D).
|
, E and G). Its localization in both
osteocytes and most regions of bone matrix was reduced in Hyp mice
(Fig. 5, F and H).
Vitronectin was found in bone matrix and some osteocytes at high
levels, whereas in other osteocytes immunostaining was weak in
wild-type mice (Fig. 5, I and K). Immunostaining for vitronectin
decreased in both osteocytes and most regions of bone matrix in Hyp
mice (Fig. 5, J and L). In wild-type mice, little or no staining for
osteocalcin, bone sialoprotein, or vitronectin was detectable in
remnant cartilage matrix within trabecular and cortical bone.
Quantitative data from image analysis for the immunoreactive
osteocalcin, bone sialoprotein, and vitronectin showed that although
their percent positive area and abundance were not altered
significantly in primary spongiosa relative to total tissue area, their
percent positive area relative to trabecular area decreased
substantially by 50.4%, 34.6%, and 54.2% in Hyp mice, respectively
(Fig. 3). In cortical bone their percent positive area and abundance
(summary total gray) were also decreased significantly in Hyp mice
(Fig. 4). Less intense staining of bone matrix proteins could reflect
decreased synthesis, secretion, and/or retention in the bone
matrix.
Immunohistochemistry for biglycan and fibrillin-1 in primary
spongiosa and cortical bone: up-regulated in Hyp mice
In wild-type mice, biglycan was detected by immunohistochemistry
predominantly in osteoid and some osteocytes and to a lesser extent in
some osteoblasts and subperiosteal cells (Fig. 6, A and C). Weak staining was observed
in bone matrix and remnant cartilage matrix in primary spongiosa and
cortical bone in wild-type mice (Fig. 6, A and C). In Hyp mice,
biglycan staining was increased in osteocytes, most regions of bone
matrix, and remnants of cartilage matrix (Fig. 6, B and D).
|
, E and G). Its localization in osteocytes, most regions of bone
matrix, and remnant cartilage matrix in primary spongiosa and cortical
bone was increased in Hyp mice (Fig. 6, F and H).
Quantitative data from image analysis of the immunoreactivity for
biglycan and fibrillin-1 showed that their percent positive area and
abundance were greatly increased in cortical and trabecular bone in Hyp
mice (Figs. 3 and 4).
Osteogenic cell differentiation and mineralization and protein and
gene expression of Phex protein and bone matrix proteins in ex vivo BMC
cultures
To explore whether changes in bone result from intrinsic
osteogenic abnormalities and/or an altered extracellular environment,
we employed ex vivo primary BMC cultures. Such cultures have
previously been shown to contain osteogenic cell precursors, which in
culture can produce mineralized bone nodules (20, 21).
Osteogenic cell differentiation and mineralization defects in Hyp
mice
We first assessed osteogenic cell growth, differentiation, and
mineralization in wild-type and Hyp mice by sequential cytochemical
staining of 18-day cultures for total colonies and ALP, calcium, and
collagen positive colonies. The quantitation of total colony area and
of ALP-, collagen-, and calcium-positive colonies was determined by
image analysis. Results showed that total colony area was unchanged;
however, areas of ALP-, collagen-, and calcium-positive colonies were
decreased significantly in Hyp mouse cultures compared with the
wild-type mouse cultures (Fig. 7A). When
the positive percentages of ALP-, collagen-, and calcium-positive
colonies relative to total colonies were calculated, all were lower
in the Hyp mouse cultures than in wild-type mouse cultures (Fig. 7B).
When the percentages of ALP-, collagen-, and calcium-positive colonies
in Hyp mice relative to wild-type mice were calculated, ALP- and
collagen-positive colony areas were decreased by 25.1% and 45.8%,
respectively. The mineralized area was decreased by 58.4% (Fig. 7C).
Those findings suggest that although cell growth was not impaired in
the cultures, osteogenic cell differentiation and mineralization were
retarded.
|
). Immunocytochemical results
showed that Phex protein was expressed only in wild-type
osteoblast-like cells, and predominantly on their cell membrane (Fig. 9A), but not in cells from Hyp mice (Fig. 9B). Western blotting confirmed the absence of Phex protein in Hyp
cells (Fig. 10).
|
|
|
) and Western blotting (Fig. 10).
The areas of calcified nodule formation and immunodetectable
osteocalcin were smaller in Hyp mouse cultures relative to wild-type
mouse cultures by double staining for calcium and osteocalcin (Fig. 9, I and J, respectively). Similar results were observed with calcified
nodule formation and type I collagen staining in Hyp vs.
wild-type cultures (Fig. 9, G compared with H). Immunodetectable bone
sialoprotein and vitronectin were decreased significantly in both
staining intensity and distribution in Hyp mouse cultures (Fig. 9, D
and F, respectively) compared with wild-type mouse cultures (Fig. 9, C
and E, respectively). In contrast, biglycan and fibrillin-1 were at
similar abundance in Hyp mice compared with those in wild-type mice by
immunocytochemistry (data not shown). Western blot analysis (Fig. 10)
indicated that osteocalcin, bone sialoprotein, and vitronectin were
down-regulated in cells from Hyp mouse BMC cultures compared with
wild-type mouse BMC cultures, whereas biglycan and fibrillin-1 were at
similar abundance. Gene expression studies confirmed a significant
decrease in type I collagen, osteocalcin, bone sialoprotein, and
vitronectin mRNA in the Hyp cells relative to wild-type cells, whereas
no significant change in biglycan and fibrillin mRNA expression was
observed (Fig. 8).
Effect of Hyp and wild-type conditioned medium on osteogenic cell
proliferation, differentiation, and mineralization of BMC cultures
To examine whether Hyp BMC cultures release possible humoral
factors that might act to induce a Hyp phenotype, BMC from wild-type
mice were cultured for 18 days with 10% conditioned medium harvested
from Hyp mouse or wild-type BMC cultures. The cultures were then
analyzed by sequential cytochemical staining for total colonies, ALP,
collagen, and calcium and quantitated by image analysis. The results
showed that total colony area was unchanged; however, areas of ALP-,
collagen-, and calcium-positive colonies were decreased significantly
in the cultures treated with conditioned medium from Hyp mouse BMC
cultures compared with those in cultures treated with conditioned
medium from wild-type mouse BMC cultures (Fig. 11A). When the positive percentages of
ALP-, collagen, and calcium-positive colonies relative to total
colonies were calculated, all were lower in the cultures treated with
conditioned medium from Hyp mouse BMC cultures than in those from
wild-type mouse BMC cultures (Fig. 11B). When the percentages of ALP-,
collagen, and calcium-positive colonies in Hyp mice relative to
wild-type mice were calculated, ALP-, collagen-, and calcium-positive
colony areas were decreased by 21.9%, 19.5%, and 25.0%, respectively
(Fig. 11C). The inhibition was not as dramatic as that seen when Hyp
mouse BMC cultures per se were analyzed (Fig. 7). Those
findings suggest that factors released by Hyp mouse BMC cultures may
result in partially retarded osteogenic cell differentiation and
mineralization in an autocrine or paracrine manner.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mineralization was decreased, as expected, in both the primary spongiosa and cortical bone of Hyp mice in situ. Diminished mineralization of Hyp mouse osteogenic BMC cultures was also observed ex vivo relative to that in wild-type mouse osteogenic BMC cultures despite similar levels of ambient phosphate. The in vitro mineralization defect was accompanied by a decrease in ALP and type I collagen in the cultures. Decreased type I collagen was also observed in situ within trabecular bone, and gene expression of type I collagen was diminished in osteogenic BMC in vitro. Impaired mineralization in vitro has also been observed in cultured immortalized osteoblasts prepared from Hyp mice, although expression of type I collagen was normal in that system (16). Nevertheless, both in vitro models demonstrate the presence of an intrinsic mineralization defect in Hyp mouse osteoblastic cells. Indeed, an intrinsic osteoblast defect resulting in impaired mineralization had previously been proposed by in vivo experiments in which Hyp osteoblastic cells were transplanted into normal mice in vivo (7).
Hypomineralized periosteocytic lesions were observed in the Hyp mouse bone in situ. Such perilacunar lesions have also been observed in the bone of children with Hyp and have healed incompletely after therapy with phosphate and 1,25-dihydroxyvitamin D3. Our demonstration that Phex protein is normally expressed in osteocytes, therefore, supports the thesis that these lesions are due to a defect in the Hyp osteocyte that lacks Phex protein and is consistent with the view that the mineralization defect in Hyp/Hyp is intrinsic to osteogenic cells that express Phex protein.
We next examined whether the elaboration of several bone matrix proteins was altered by osteogenic cells. Our studies show that in vivo in Hyp mouse bone, osteocalcin immunostaining was reduced in trabecular and cortical bone, and this is in keeping with the reduced level of osteocalcin previously found in bone powder of Hyp mice (24). Reduced osteocalcin protein and gene expression were also observed in our bone marrow cultures of Hyp mice in vitro. These studies contrast with gene expression studies of osteocalcin in cultures of immortalized osteoblasts (16) or in calvareal bone (24) from Hyp mice in which osteocalcin mRNA levels were unaltered or elevated, respectively. Such distinctions could represent differences in the stage of maturity of the osteogenic cells analyzed, the source of osteogenic cells studied by different workers, or the use of immortalization techniques. Regardless of these issues, whether selective osteocalcin deficiency can contribute to deficient bone mineralization in vivo remains uncertain (25, 26).
In vivo biglycan and fibrillin-1 immunostaining were increased in both trabecular and cortical bone. Biglycan, also know as proteoglycan-2, has been found in many connective tissues and in osteoid and osteocytic lacunae of bone as in our studies (27). In vivo biglycan-deficient mice and biglycan-deficient humans (in Turner’s syndrome or monosomy X) display growth retardation and decreased bone mass. In vitro, biglycan at high concentrations can also inhibit the growth and proliferation of mineral crystals (28). Fibrillin-1 is the principal component of the microfibrillar aggregates in the skeleton as well as in the eyes and vascular tissue (29). Our results show that fibrillin-1 is normally localized in osteoblasts and in some osteocytes and displays a patchy distribution in bone matrix of wild-type mice. In view of the fact that fibrillin-1-deficient mice display excess calcification in blood vessels (30), fibrillin-1 excess, as seen in the bone of Hyp mice, may contribute to the diminished mineralization observed in this entity.
Despite the increase in biglycan and fibrillin-1 protein deposition observed in Hyp mouse bone in vivo, no alteration in the levels of mRNA encoding these proteins was observed in osteogenic cells of Hyp mouse BMC cultures. In view of the fact that Phex appears to exert endopeptidase activity (10), Phex deficiency in osteogenic cells may contribute to delayed elimination of biglycan and fibrillin-1 in vivo and consequently facilitate their accumulation in the matrix of Hyp bone.
Finally, bone sialoprotein and vitronectin immunostaining were both decreased in Hyp mouse bone in vivo, and levels of their proteins were also diminished in Hyp mouse BMC cultures in vitro. Bone sialoprotein is relative unique to mineralizing tissue (31, 32) and is expressed only by those cells forming a mineralized matrix (33). Bone sialoprotein expression in cell culture is maximal during early matrix mineralization (33) and in solution studies acts as a hydroxyapatite nucleator (34). It has therefore been suggested that bone sialoprotein is involved in bone mineralization (35, 36, 37); however, the present study is the first to implicate bone sialoprotein in a mineralization defect in vivo.
Vitronectin is a multifunctional extracellular matrix glycoprotein that contains an RGD sequence and is involved in cell adhesion, spreading, and migration by interaction with integrins (38). Vitronectin-deficient mice appear to demonstrate normal development, so in its absence, vitronectin functions may be subserved by alternate constituents of the extracellular matrix. It is unclear, however, whether bone will develop normally when vitronectin deficiency is accompanied by a reduction or an excess of other bone matrix proteins.
Gene expressions of bone sialoprotein and vitronectin were diminished in the osteogenic Hyp mouse bone marrow cultures. These results point to a differentiation defect in these cells; whether this is due to a different marrow cell precursor population or alteration in the normal marrow cell population is at present unclear. However, this defect appears to accompany and presumably indirectly result from the mutation in the Phex gene.
In view of the fact that Nesbitt et al. (39) recently demonstrated that conditioned medium from Hyp osteoblasts displays phosphate transport inhibitory activity, presumably due to phosphatonin, we assessed the effect of conditioned medium from Hyp BMC to reproduce Hyp defects in wild-type BMC. The results indicated that under the experimental conditions used, partial defects could indeed be induced. Consequently, at least same of the abnormalities in these cells may be due to the action of a secreted factor, possibly synonymous with phosphatonin, which is acting in an autocrine or paracrine manner.
In summary, then, our studies have localized Phex protein to osteoblastic and especially osteocytic cells and suggest that the hypomineralized periosteocytic lesions in Hyp bone may be a local consequence of Phex deficiency in the osteocyte. A reduction in some bone matrix proteins, such as bone sialoprotein and vitronectin, can be observed both in Hyp bone in vivo and in Hyp BMC cultures in vitro, whereas elevations in other bone matrix proteins, such as biglycan and fibrillin-1, may be seen. This perturbation of the normal complement of bone matrix protein may contribute to the impaired mineralization observed in Hyp bone and could potentially account for the inhibitory effect on normal bone cell mineralization previously reported to be elicited by Hyp bone cells (7, 16). Based on analyses of osteogenic cells in vitro, these alterations in bone matrix protein concentrations may result from diminished gene expression or from altered posttranslational handling of these proteins. These studies demonstrate that osteogenic cell disturbances may occur independently of ambient phosphate concentrations and may therefore be of equal importance for the generation of the Hyp phenotype as the much studied changes in renal function in this disorder.
|
|
Footnotes |
|---|
2 Recipient of an Medical Research Council of Canada fellowship.
3 Scholar of the Fonds de la Recherche.
4 Medical Research Council Scientist.
Received June 28, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Martin, V. David, J. S. Laurence, P. M. Schwarz, E. M. Lafer, A.-M. Hedge, and P. S. N. Rowe Degradation of MEPE, DMP1, and Release of SIBLING ASARM-Peptides (Minhibins): ASARM-Peptide(s) Are Directly Responsible for Defective Mineralization in HYP Endocrinology, April 1, 2008; 149(4): 1757 - 1772. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Liu, W. Tang, J. Zhou, L. Vierthaler, and L. D. Quarles Distinct roles for intrinsic osteocyte abnormalities and systemic factors in regulation of FGF23 and bone mineralization in Hyp mice Am J Physiol Endocrinol Metab, December 1, 2007; 293(6): E1636 - E1644. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Gao, R. Tiwari-Pandey, R. Samadfam, Y. Yang, D. Miao, A. C. Karaplis, M. R. Sairam, and D. Goltzman Altered Ovarian Function Affects Skeletal Homeostasis Independent of the Action of Follicle-Stimulating Hormone Endocrinology, June 1, 2007; 148(6): 2613 - 2621. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xue, A. C. Karaplis, G. N. Hendy, D. Goltzman, and D. Miao Exogenous 1,25-Dihydroxyvitamin D3 Exerts a Skeletal Anabolic Effect and Improves Mineral Ion Homeostasis in Mice that Are Homozygous for Both the 1{alpha}-Hydroxylase and Parathyroid Hormone Null Alleles Endocrinology, October 1, 2006; 147(10): 4801 - 4810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Liu, J. Zhou, W. Tang, X. Jiang, D. W. Rowe, and L. D. Quarles Pathogenic role of Fgf23 in Hyp mice Am J Physiol Endocrinol Metab, July 1, 2006; 291(1): E38 - E49. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Imel and M. J. Econs Fibroblast Growth Factor 23: Roles in Health and Disease J. Am. Soc. Nephrol., September 1, 2005; 16(9): 2565 - 2575. [Full Text] [PDF]
|
|
|
|
|
|
Z. Tang, R. Yu, Y. Lu, A. F. Parlow, and J.-L. Liu Age-dependent onset of liver-specific IGF-I gene deficiency and its persistence in old age: implications for postnatal growth and insuli |
