| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Immundiagnostik GmbH, Wiesenstr. 4, 64625 Bensheim, Germany.
2
Department of Medicine, Division of Endocrinology and
Metabolism, University of Heidelberg Medical School, Bergheimerstr. 58,
69115 Heidelberg, Germany.
a Author for correspondence. Fax Int. +49-6251-39084.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
vß3 of
the transformed murine cell line ROS 17/2.8 (2). According
to human cDNA analysis, the molecular mass of the core protein is ~33
kDa (3). In a 4–20% sodium dodecyl sulfate (SDS)
gradient gel, BSP migrates as a 80-kDa band, as carbohydrates
contribute to nearly 50% of the molecular mass (1). The function of BSP is still not fully understood. BSP stimulates hydroxyapatite formation in vitro (4) and appears to mediate adhesion between cellular surfaces and extracellular matrix components via the RGD binding site. The expression of BSP is stimulated by dexamethasone and inhibited by calcitriol in vitro (5).
BSP has been found predominantly in bone. However, Northern analysis of total RNA has shown mRNA encoding for BSP also in epiphyseal cartilage and decidua, although these amounts were much smaller than those detected in skeletal tissues (3). Bone sialoprotein appears to be present mainly in cells derived from bone, such as osteoblasts, osteocytes, osteoclasts, and, to a lesser extent, hypertrophic chondrocytes (6). The only nonmineralizing tissue where BSP was also detected is the developing placental trophoblast (6).
The present paper describes the purification of BSP from human femoral bone, by wide-pore reversed-phase HPLC, and the development of a new RIA for the measurement of circulating BSP in human serum.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
extraction
Human femoral bone was obtained from total hip replacements. The
extraction was performed according to the method of Fisher et al.
(7). Bone was crushed and ground in a Retsch mill cooled
with liquid nitrogen. For each preparation, 100 g of bone powder
was washed for 30 min in water. The powder was then transferred to 3 L
of buffer containing 4 mol/L guanidine HCl, 50 mmol/L Tris pH 7.4, and
protease inhibitors (0.1 mol/L 6-aminocaproic acid, 5 mmol/L
benzamidine HCl, and 1 mmol/L phenylmethylsulfonyl fluoride) and
incubated for 48 h at 4 °C. The suspension was centrifuged at
3000g for 20 min, the supernatant was discarded, and the
bone proteins were extracted for 72 h with 0.5 mol/L sodium EDTA
in 3 L of the above buffer. The extract was centrifuged
(3000g, 20 min), the supernatant was filtered through a
glass fiber filter to remove lipids, and then concentrated to a volume
of 30 mL in a Millipore tangential ultrafiltration system with 10-kDa
filter cutoff.
gel filtration
A sample of 10 mL was centrifuged for 20 min at
10 000g, applied to a Sepharose 6B column (100 x 2.5
cm), and eluted with 4 mol/L guanidine HCl and 50 mmol/L Tris pH 7.4 at
a flow rate of 0.5 mL/min. The elution was monitored at 280 nm and
fractions of 5 mL were collected.
anion-exchange chromatography
The BSP-containing fractions obtained from the gel filtration,
identified by SDS-polyacrylamide gel electrophoresis (PAGE) and Stains
All, were pooled. The material was concentrated and the buffer was
exchanged to a starting buffer containing 7 mol/L urea and 50 mmol/L
sodium acetate pH 4.5 with a Millipore tangential ultrafiltration
system. The sample was centrifuged at 10 000g for 2 min,
loaded on a DEAE-Sephacel column (10 x 1 cm), and washed until
the absorbance at 280 nm returned to baseline. BSP was eluted with a
gradient ranging from 0 to 0.5 mol/L sodium chloride in starting
buffer, and fractions of 5 mL were collected.
reversed-phase hplc chromatography
Fractions of the anion-exchange step containing BSP were
concentrated in an Amicon stirring cell equipped with a YM 10 filter
(cutoff 10 kDa) to a volume of 200 µL and centrifuged. One hundred
microliters were chromatographed on a Latek Hypersil WP
300-C4 5 µm, 150 x 4 mm column. For the
elution of BSP, a gradient of buffer A (1.3 mL/L heptafluorbutyric
acid) and buffer B (1.3 mL/L heptafluorbutyric acid, 750 mL/L
CH3CN) was applied as follows: 0 min, 30% buffer B; 30
min, 100% buffer B; 35 min, 100% buffer B; 40 min, 30% buffer B. The
system was run with a flow rate of 1 mL/min and fractions of 1 mL were
collected. The elution was monitored at 280 nm.
sds-page
The fractions obtained by chromatography were monitored by
SDS-PAGE. Samples were diluted 10-fold with 980 mL/L ethanol and
proteins were allowed to precipitate for 3 h at -20 °C. The
samples were centrifuged (10 min, 10 000g) and washed with
ethanol. The pellet was dried for 1 h at 60 °C, resuspended in
loading buffer, and analyzed by a nonreduced linear 4–20%
polyacrylamide gel stained with Stains All according to Wallace and
Begovac (8).
western blotting
Reversed-phase HPLC-purified BSP and osteopontin, which was
purified according to the method of Fisher et al. (7),
were electrophoresed on a linear 4–20% polyacrylamide gel. The gel
was transblotted for 1 h at 500 V onto nitrocellulose according to
the method of Towbin et al. (9) with the following
changes: The last washing was performed by a buffer containing 10
mmol/L Tris pH 7.5, 14 g/L NaCl, and 0.1 g/L Tween 20 to reduce
negative bands on the blot. The chicken anti-human BSP was present in a
dilution of 1:100 and the peroxidase-conjugated rabbit anti-IgY was
diluted 1:1000.
automated amino acid sequencing
The N-terminal amino acid sequence of 5 pmol of BSP was determined
by automated Edman degradation on an Applied Biosystems Procise amino
acid sequencer.
immunization
A chicken was immunized with 100 µg of BSP-containing fractions
from the anion-exchange chromatography in 0.5 mL of 0.1 mol/L sodium
acetate, pH 6.3, emulsified in 0.5 mL of Freund's complete adjuvant
(Calbiochem) by intramuscular injection at multiple sites. Four weeks
later the chicken was boosted with 50 µg of human BSP in 0.5 mL of
0.1 mol/L sodium acetate (NaAc), pH 6.3, emulsified in 0.5 mL of
Freund's incomplete adjuvant (Difco Labs.). The antibodies were
isolated from the egg yolk according to the method of Polson et
al. (10).
preparation of 125i-labeled bsp
Reversed-phase HPLC-purified human BSP, 1 µg dissolved in 10
µL of 0.1 mol/L NaAc pH 6.3, was radioiodinated with 13 MBq (350
µCi) of Na125I according to the chloramine T method
described by Hunter and Greenwood (11). Radiolabeled BSP
was separated from Na125I by purification on a Sep-Pac
C18 cartridge (Millipore) as reported by
Schöneshöfer et al. (12). The
125I-labeled BSP was diluted in PPNE containing 20 g/L BSA
and 100 g/L polyethylene glycol (Mr 6000) to a
final concentration of 60 000 cpm/100 µL.
preparation of the calibration curve
The protein concentration of reversed-phase HPLC-purified human
BSP was determined by the Pierce protein assay calibrated with BSA. A
BSP calibration curve was then prepared by diluting BSP with PPNE
containing 20 g/L BSA to the following concentrations: 120, 60, 30, 15,
7.5, 3.75, and 1.875 µg/L.
assay procedure (bsp ria)
One-hundred microliters of calibrator, serum sample, or control
sera were incubated for 24 h with 100 µL of
125I-labeled BSP and 100 µL of the 1:200 diluted
chicken anti-human BSP antibody in RIA tubes (0.6 mL volume; Sarstedt).
All analyses were done in duplicate. For determination of nonspecific
binding, the antibody was replaced by PPNE containing 1 g/L BSA. The
bound and free ligands were separated by incubation with 100 µL of
donkey anti-chicken IgY for 30 min, followed by centrifugation at
1800g for 10 min. The supernatants were removed and the
pellets washed by addition of 250 µL of NaCl (0.9 g/L) with
polyethylene glycol 6000 (60 g/L) and subsequent centrifugation at
1800g for 10 min. The radioactivity in the precipitate was
counted in a 
-spectrometer for 1 min. All procedures were performed
at 4 °C. The calibration curve (Fig. 1
) was prepared with B/B0 vs concentration of BSP
(µg/L). The amount of BSP in patient sample was calculated by a
four-parameter curve-fitting algorithm.
|
stability of analyte and preliminary clinical evaluation
All venous blood samples were obtained in Vacutainer Tubes without
additive between the hours of 0800 and 1000. The material was
centrifuged at 1500g within 2 h of collection, and
serum aliquots were stored at -20 °C.
To evaluate the stability of BSP in serum during prolonged storage, serum aliquots of 1 mL were stored for up to 3 months at ambient temperatures of -20 °C, 4 °C, and 25 °C, respectively.
In a preliminary clinical evaluation, serum concentrations of
circulating BSP were also determined in healthy controls ages 20–80
years (n = 90), and in patients with Paget disease of bone (n
= 24), primary hyperparathyroidism (PHPT; n = 11), renal secondary
hyperparathyroidism (SHPT; n = 25), chronic renal failure without
SHPT (RF; n = 26), and alcoholic liver cirrhosis (LC; n =
35). Data on anthropometric and clinical chemistry variables are
summarized in Table 1
. In each group, the respective diagnosis was based on standard
clinical evaluation techniques, including history, physical
examination, plain radiograms, bone density, and laboratory
measurements (see Table 1
). Before sample collection, written informed
consent was obtained from each individual. The study was approved by
the local ethics committees and was performed in accordance with the
Declaration of Helsinki, amended by the 29th and 35th World Medical
Assembly.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). BSP was clearly separated from other proteins at a
concentration of 41% buffer B. The purity and identity of BSP was
shown by SDS-PAGE (Fig. 3
) and by N-terminal amino acid sequencing (Table 2
). The determined amino acid sequence is identical to the one
described by Fisher et al. (7).
|
|
|
preparation of 125i-labeled bsp and assay development
One microgram of human BSP purified by reversed-phase HPLC was
radioiodinated. The specific activity was 16.28 MBq/µg (440
µCi/µg). The iodinated BSP was diluted to 60 000 cpm/100 µL,
equal to 0.14 ng of 125I-labeled BSP per assay tube.
During the development of a specific RIA for human BSP, several variables had to be optimized. Thus, although assay kinetics were similar at 4, 25, and 37 °C, radioactive yield was lower at higher incubation temperatures. To maximize count rates, a temperature of 4 °C was therefore chosen for all incubations. Incubations of 24 h for the first incubation and 30 min for the second antibody precipitating step proved sufficient for the reactants to reach equilibrium.
assay characteristics
The specificity of the assay for BSP was demonstrated by testing
for cross-reactivity with other noncollageneous components such as
osteocalcin, osteonectin, and bone alkaline phosphatase. None of these
proteins showed any reactivity within the concentration ranges tested
(Fig. 4
). The specificity of the chicken anti-BSP was demonstrated by
Western blotting. No cross-reactivity was detected with human
osteopontin (Fig. 5
).
|
|
The linearity of the assay was determined by serial dilution of three serum samples. The correlation coefficients, determined by linear regression, were 0.9819 (sample 1: 91.7 µg/L), 0.9937 (sample 2: 70.6 µg/L), and 0.9933 (sample 3: 63.8 µg/L). The recovery was tested by supplementing a human serum sample and an EDTA-plasma sample with different concentrations of BSP from calibration solutions. The recovery was between 92% and 108% in serum but only between 18% and 26% in EDTA-plasma.
For testing accuracy, two control sera were tested 12 times in one assay. A within-run variation of 7.0% (mean 10.9 µg/L) and 6.1% (mean 38.8 µg/L) was found. Reproducibility was tested by measuring two control sera nine times on different days. A between-run variation of 9.2% (mean 11.0 µg/L) and 9.4% (mean 39.0 µg/L) was found.
The lowest detectable concentration was 0.7 µg/L, as defined by the concentration 3 SD above B0 (zero calibrator), which was measured 12 times.
stability of bsp in stored serum samples
The stability of BSP during prolonged storage was tested for
various intervals and at different temperatures (Table 3
). At -20 °C, serum BSP concentrations remained essentially
unchanged over 3 months. When serum samples were stored at 4 °C, no
significant change in BSP concentrations was observed during the first
5 days of storage. However, after 14 days and thereafter, increased
concentrations of BSP were measured. Also, storage of serum samples at
25 °C led to an increase in apparent BSP concentrations.
|
preliminary clinical evaluation
Measurements obtained from 90 healthy controls showed a mean serum
BSP concentration of 12.1 ± 5.0 µg/L (mean ± SD) with
individual values ranging between 6.2 and 22.4 µg/L. In patients with
Paget disease of bone, serum BSP concentrations were 32.3 ± 17.3
µg/L, whereas in patients with PHPT a mean concentration of 24.7
± 13.5 µg/L was noted (P <0.01 vs healthy controls for
both groups). In patients with RF and no apparent SHPT, BSP
concentrations were on average 23.0 ± 14.7 µg/L, whereas in
those patients with both RF and SHPT, mean concentrations were
increased to 30.6 ± 18.9 µg/L (P <0.01 vs healthy
controls for both groups). Patients with alcoholic liver cirrhosis had
mean serum BSP values of 12.6 ± 7.6 µg/L, which were not
significantly different from healthy controls (Fig. 6
).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We present here a homologous RIA specific for human BSP in serum. First, the purification procedure of Fisher et al. (7) developed for fetal bone had to be adapted to our source of the immunogen, bone obtained from hip replacements performed in elderly patients. The crucial step in our modified procedure was wide-pore C4 reversed-phase HPLC, which yielded a highly purified protein, as confirmed by N-terminal amino acid sequencing.
Because the yield of BSP from elderly bone was rather low, the BSP-containing fractions obtained from anion-exchange chromatography were used for immunization. This was also feasible, as at the time of immunization we did not know whether the HPLC fractionation would in any way affect the antigenicity of BSP. Our results, however, show no change in BSP antigenicity, thus making reversed-phase HPLC a new strategy to purify BSP from human bone.
Chickens were selected for immunization because the minimum quantity of antigen required is about half of that used for conventional immunization of rabbits or mice. This advantage is attributed to the large phylogenetic distance between birds and mammals, which results in increased sensitivity to antigen exposure and decreased background noise in immunoassay (18). High antibody concentrations in egg yolk and easy techniques for IgY separation allow a cost-effective production on a routine basis. By using these BSP antibodies and a highly specific 125I-labeled human BSP tracer, a reliable assay for the measurement of BSP in human serum was developed.
Our immunoassay showed no cross-reactivity with other noncollageneous bone proteins such as osteocalcin, osteonectin, or alkaline phosphatase. Furthermore, osteopontin was undetectable by Western blotting. Therefore, the measurement of BSP appears not to be influenced by these proteins.
Interestingly, BSP could not be measured in EDTA-plasma. In our hands, the recovery of EDTA-plasma samples supplemented with purified BSP ranged between 18% and 25%. We propose that this behavior may be attributed to the calcium-binding properties of BSP. Thus, the addition of EDTA to blood samples may displace calcium ions from the BSP molecule, which in turn is likely to lead to changes in the tertiary structure of the molecule, promoting aggregation and precipitation of BSP. This process results in higher count rates in the pellet, simulating low serum BSP concentrations. Aggregation may also explain the loss of BSP during purification without chaotropic agents. Between the chromatographic steps, we noted a near total loss of BSP when the sample was concentrated in the absence of urea or guanidine HCl. On the other hand, recalcification of EDTA-plasma samples also led to a dose-dependent increase of the count rates caused by precipitation of 125I-labeled BSP (data not shown). Thus, being a bone matrix protein with calcium-binding properties, BSP appears to be sensitive to changes in calcium concentrations, a fact that should be kept in mind when performing the assay.
In the 90 healthy controls studied in this investigation, serum BSP
concentrations ranged from 6.2 to 22.4 µg/L, with a mean
concentration of 12.1 ± 5.0 µg/L (± SD). This value is
slightly but not significantly higher than that reported earlier by us
(19), due to the fact that most of the individuals were
different from those studied in our previous publication
(19). Interestingly, in both publications, mean serum
values of BSP are comparable with those of serum osteocalcin, another
noncollagenous protein of the extracellular bone matrix synthesized by
osteoblasts and released into serum. As described elsewhere
(20), serum concentrations of BSP and osteocalcin do not
correlate with each other, suggesting that the two components reflect
different processes of bone turnover. Compared with the healthy
controls, serum BSP concentrations were significantly increased in
patients with metabolic bone disease. Highest concentrations were seen
in subjects with active Paget disease of bone and in individuals with
secondary renal hyperparathyroidism. These values are somewhat but not
significantly higher than those reported earlier for these disease
groups (19). Again, this is because patients in the
present investigation were largely different from those evaluated
earlier. As both disorders are usually associated with high bone
turnover, i.e., increased bone formation and accelerated bone
resorption, classification of serum BSP to either of these processes
can currently not be achieved. Saxne et al. have recently shown
increased serum and synovial BSP in patients with active rheumatoid
arthritis (21). These authors attributed these changes to
an increase in bone turnover often observed in patients with
inflammatory joint disease (22). However, we also observed
significantly increased concentrations of serum BSP in patients with
asymptomatic PHPT. We have previously shown that this frequent disorder
of calcium homeostasis is associated with increased bone resorption, as
indicated by increased concentrations of urinary pyridinium
cross-links. In contrast, markers of bone formation (such as alkaline
phosphatase or osteocalcin) were normal in these patients
(20) (see also Table 1
). Since BSP is primarily a product
of osteoblasts, our preliminary observations indicate that serum BSP
may reflect processes related to both bone resorption and bone
formation. However, further data from our own group, including studies
on the effect of intravenous bisphosphonates on BSP and other markers
of bone turnover, suggest that circulating BSP may be associated with
osteoclast rather than osteoblast activity (19).
In the present investigation, we also noted that serum BSP concentrations were strongly affected by renal function, and values were significantly increased in those patients with additional SHPT. In contrast, no significant changes were seen in subjects with advanced liver failure. Taken together, these observations indicate that circulating BSP is processed and (or) eliminated by the kidney, whereas the liver appears to play only a minor role in the metabolism of this glycoprotein.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  H. W. Woitge, H. Oberwittler, S. Heichel, A. Grauer, R. Ziegler, and M. J. Seibel Short- and Long-Term Effects of Ibandronate Treatment on Bone Turnover in Paget Disease of Bone Clin. Chem., May 1, 2000; 46(5): 684 - 690. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. J. Diel, E.-F. Solomayer, M. J. Seibel, J. Pfeilschifter, H. Maisenbacher, C. Gollan, M. Pecherstorfer, R. Conradi, G. Kehr, E. Boehm, et al. Serum Bone Sialoprotein in Patients with Primary Breast Cancer Is a Prognostic Marker for Subsequent Bone Metastasis Clin. Cancer Res., December 1, 1999; 5(12): 3914 - 3919. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Wuttke, S. Muller, D. P. Nitsche, M. Paulsson, F.-G. Hanisch, and P. Maurer Structural Characterization of Human Recombinant and Bone-derived Bone Sialoprotein. FUNCTIONAL IMPLICATIONS FOR CELL ATTACHMENT AND HYDROXYAPATITE BINDING J. Biol. Chem., September 21, 2001; 276(39): 36839 - 36848. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
). B, cpm; B0, cpm of zero calibrator.
