Clinical Chemistry 46: 252-257, 2000;
(Clinical Chemistry. 2000;46:252-257.)
© 2000 American Association for Clinical Chemistry, Inc.
Evaluation of a Bead-based Enzyme Immunoassay for the Rapid Detection of Osteocalcin in Human Serum
Alexandra M. Cr
ciun1,
Cees Vermeer1,
Hans-Georg Eisenwiener2,
Norbert Drees2 and
Marjo H.J. Knapen1,a
1
Department of Biochemistry and Cardiovascular Research Institute, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands.
2
Hoffmann-La Roche Diagnostics, Basel 4070,
Switzerland.
a Address corresponding to this author at: Department of Biochemistry, University of Maastricht, P.O. Box 616, 6200 MD Maastricht, The Netherlands. Fax 31-43-367-0992; e-mail m.knapen{at}bioch.unimaas.nl
 |
Abstract
|
|---|
Background: Circulating osteocalcin is a well-known marker for
bone formation, but none of the commercial kits currently available can
be used in automated systems. Here we present the first semiautomated
assay for human serum osteocalcin.
Methods: Polystyrene beads were coated with antibodies against
the COOH terminus of osteocalcin and used in the COBAS®
EIA System. Osteocalcin was detected with peroxidase-conjugated
antibodies against the osteocalcin NH2 terminus.
Results: The time required to analyze an unknown sample was
60 min, with a lower detection limit of 4.5 µg/L and a linear
dose–response curve between 4.5 and 100 µg/L. The intraassay
imprecision (CV) was 5–8% (n = 21); the interassay variation was
6–9% (n = 14). In samples from human volunteers and patients,
data generated with the newly developed assay were comparable to those
obtained with standard microtiter plate-based assays.
Conclusions: The coated beads assay may be implemented on
fully automated analyzers, which not only may further reduce
imprecision but may also substantially increase the applicability of
osteocalcin as a marker for bone metabolism in the routine clinical
setting.
|
Introduction
|
|---|
Osteocalcin (OC),1
also known as bone Gla
protein, is the most abundant noncollagenous protein in mature
bone (1)(2). Osteoblasts are the exclusive site
of OC biosynthesis, which is regulated at the transcription stage by
vitamin D (3)(4), whereas vitamin K is required
for the posttranslational formation of its three 
-carboxyglutamate
(Gla) residues. Although its function on a molecular level has remained
unclear to date, increased bone formation, including higher bone mass
and improved bone strength, was observed in OC-deficient (knock-out)
mice (5). These experiments have demonstrated that OC has an
important role in the regulation of bone growth and in the correct
deposition of the mineral matrix in bone. Because 20–30% of the de
novo synthesized OC is not accumulated in the bone tissue, but is
secreted into the blood stream, circulating OC is widely used as a
biomarker for bone formation
(6)(7)(8).
During episodes of vitamin K deficiency or during treatment with
vitamin K antagonists, Gla-containing proteins are synthesized in an
undercarboxylated (Gla-deficient) form. Fully carboxylated and
undercarboxylated OC may be quantified separately on the basis of their
different affinities for hydroxyapatite (9). It has been
reported by various groups that in the general population, a
substantial fraction of circulating OC occurs in its undercarboxylated
form (9)(10)(11), and that the concentration of
undercarboxylated OC may have an independent diagnostic value for the
assessment of bone mass and bone fracture risk (12)(13)(14).
During recent years, an increasing number of test kits for serum OC
have become commercially available, all of which are either
radioimmunoassays or enzyme-based immunoassays for microtiter plates.
Some of the drawbacks of such tests are that they are laborious and
prone to errors by the persons performing the tests, and that they
generally require long incubation steps (either all day or overnight)
before the data become available. In addition, the various OC kits may
give widely different values when the same serum sample is tested
(6)(15). This may be related to different
specificities of the antibodies used for fully carboxylated and
undercarboxylated OC and whether OC degradation products are
recognized.
To improve the accuracy and reproducibility of OC quantification, we
used commercial antibodies that were coated onto polystyrene beads and
used in the Roche COBAS® EIA System. We compared
the performance of this semiautomated enzyme immunoassay with
commercial radiometric and microtiter plate-based assays for various
markers of bone metabolism.
|
Materials and Methods
|
|---|
subjects
The reference values for the concentrations of OC and
bone-specific alkaline phosphatase (BAP) were established in serum
samples obtained from 151 apparently healthy subjects (60 men and 91
women) between 19 and 86 years of age who were recruited via a local
newspaper. The within-day and day-to-day variations of OC and several
other markers were examined in 12 apparently healthy volunteers (6 men
and 6 women) between 19 and 27 years of age recruited from the students
of the Maastricht University. During the first 24 h of the
experiment, urine and blood samples were collected at 0900
(start) and at 1100, 1400, 1700, 2100, 2300, 0300, and 0700. Subsequent
samples were collected at 0900 on days 2, 8, 15, 22, 29, and 57. All
samples taken at 0900 were obtained after an overnight fast (only water
allowed after 1900 of the preceding night). Changes in bone markers
during the follow-up of treatment were recorded in serum and urine
samples from 30 osteoporotic women (>65 years of age), before and
after treatment with either bisphosphonate (Alendronate, 5–10 mg/day
for 15 months; Merck), estrogen (Livial, 2.5 mg/day for 6
months; Organon), or calcitonin (Miacalcic nasal application, 100
IU/day for 12 months; Sandoz). All studies were approved by the
University Hospital Medical Ethics Committee.
sample collection and storage
Blood (10 mL) was taken by venipuncture to prepare serum, 0.5-mL
aliquots of which were frozen at -80 °C within 2 h after
sample collection until use. All urine samples collected at the 0900
time points were obtained after an overnight fast (14 h) and represent
the 2-h second morning void. Other urine samples collected during the
first day were nonfasting samples. All urine samples were stored in
0.5-mL aliquots at -30 °C until use.
procedure for the semiautomated oc immunoassay
The assay (hereafter called the Coated Beads Assay) is based on
commercial monoclonal antibodies (mAbs) against synthetic peptides
homologous to OC residues 1–19 and 20–43 (Osteometer), and will be
designated here as mAb1–19 and
mAb20–43. Polystyrene beads (2 mm
diameter) were coated with mAb20–43 by
Osteometer, other components for the test were those in the
microtiter-based N-midTM Osteocalcin ELISA
(Osteometer), and are described by Rosenquist et al. (16).
The test procedure, for which we used the COBAS EIA System from
Hoffmann-La Roche (Basel, Switzerland), was as follows: 25 µL of
sample (calibrator, control, or serum) was pipetted into polycarbonate
tubes, together with 225 µL of buffer A (0.14 mol/L NaCl, 0.01 mol/L
sodium phosphate, pH 7.4) and 25 µL of peroxidase-conjugated
mAb1–19. Subsequently, one coated bead was added
to each tube and incubated for 30 min at 37 °C while shaking in the
COBAS EIA Incubator. The beads were then washed with distilled water in
the COBAS EIA Washer, and 250 µL of substrate solution
(tetramethylbenzidine) was added, after which the tubes were incubated
for another 15 min at 37 °C with constant shaking. The reaction was
stopped by the addition of 1 mL of 0.1 mol/L
H2SO4; within 1 h
after termination of the reaction, the absorbance at 450 nm was
recorded with a 25-channel COBAS EIA Photometer. Serum OC
concentrations were calculated using a four-parameter logistic curve
fit based on the calibrators of the assay (0, 6.25, 12.5, 25, 50, and
100 µg/L).
other tests used
The data obtained with the experimental OC assay were compared
with commercial kits for OC (ELSA-Osteo; CIS Bio-international) and BAP
(Tandem-R Ostase; Hybritech). Both tests are two-site IRMAs.
Intact parathyroid hormone (PTH) was quantified in serum with the
N-tact PTH radioimmunoassay from Incstar. Markers tested in urine were
hydroxyproline (OHPro; hypronosticon; Organon Teknika),
deoxypyridinoline (DPD; Pyrilinks-D; Metra Biosystems), and type I
collagen C-terminal telopeptide (CTX; CrossLaps; Osteometer
BioTech). Creatinine was assessed in urine by standard enzymatic
techniques (Boehringer Mannheim) on a Beckman Synchron CX7-2 automated
analyzer. Urinary calcium was determined by atomic absorption
spectrophotometry (Perkin-Elmer). Urinary markers are expressed as the
ratio between these markers and creatinine throughout this report.
data analysis
Statistical analysis was performed with the software package
SPSSWin, Ver. 7.5 (SPSS). All results are given as the mean value
± SD. Differences between the groups were investigated with the
unpaired Student t-test. The Wilcoxon test was used
for the evaluation of differences within groups. Differences were
considered significant at P <0.05. Correlations between the
data obtained with different test procedures were evaluated using
linear regression.
|
Results
|
|---|
calibration curve and test characteristics
Calibration curves were constructed on 12 different days using six
calibrator solutions with OC concentrations in the range of 0–100
µg/L. Each calibrator was measured in triplicate, and the mean
absorbance at 450 nm (± SD) was expressed as a function of the OC
concentration (Fig. 1
). The lower limit to detection, defined as the mean absorbance
+ 3 SD for calibrator A (0 µg/L), was 4.45 µg/L [absorbance =
0.05 + 3 x 0.02 = 0.11]. The intra- and interassay
variation of the test was determined using three serum pools of known
OC concentrations. The intraassay variation was calculated by
expressing the SD as a percentage of the mean concentration as
calculated from 21 replicates of each serum pool, which was repeated on
3 different days. Mean values obtained were 7.6%, 6.0%, and 4.8% for
serum pools containing 10, 22, and 83 µg/L of OC, respectively. The
same serum pools were measured in duplicate on 14 consecutive days, and
interassay variations of the means of duplicates were calculated by
expressing the SDs as percentages of the means: 6.0%, 9.1%, and
5.9%, respectively.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 1. Calibration curve for OC by Coated Beads Assay.
Calibrators used were from the commercial N-mid Osteocalcin
assay (Osteometer). Bars, SD.
|
|
To determine the linearity on dilution, we used three human serum
samples containing pathologically high OC concentrations (from patients
with renal failure). Samples with OC concentrations well above the
highest calibrator (calibrator F; 100 µg/L) were prediluted with
calibrator A (0 µg/L) before use to give the following OC
concentrations: sample A (undiluted), 97.4 ± 2.2 µg/L; sample B
(prediluted twofold), 94.3 ± 2.6 µg/L; and sample C (prediluted
fivefold), 75.3 ± 0.7 µg/L. On 3 different days, each sample
was diluted serially and tested in duplicate. Recoveries were
calculated and expressed as percentages of the starting values (Table 1
), and it was apparent that sample A could be diluted twofold
without significant loss of recovery. As was the case with all (n
= 6) of the microtiter plate-based kits and radioimmunoassays we have
checked to date (data not shown), the recovery declined at higher
dilutions. Both prediluted samples (B and C) showed a strong decrease
of recovery after further dilutions, which indicates that more than
twofold dilution of serum samples may lead to substantial
underestimation of the OC concentration in these samples if the
dilutions are prepared with the calibrator A (0 µg/L), which does not
have a matrix sufficiently similar to serum.
To investigate whether the antibodies used in our assay discriminate
between OCs containing different numbers of Gla residues, full-length
synthetic OCs (17) containing either 0 or 3 Gla residues
were dissolved in calibrator A and tested in various dilutions. Both
synthetic peptides were recognized well, but the antibodies did not
differentiate between fully carboxylated and noncarboxylated OC (data
not shown).
variations related to age and gender
In this experiment, we assembled serum samples from apparently
healthy men (n = 60) and women (n = 91) of different ages and
compared the new assay with commercially available test kits for OC and
for BAP. The results obtained with the two OC assays were very similar,
with all three bone formation markers slightly increased in men 19–40
years of age, possibly because in this group the peak bone mass was not
yet reached (Table 2
). Only in the case of the commercial OC kit was this difference
statistically significant, however. In addition, in women 59–86 years
of age, the bone formation markers had a tendency to increase, which
may be related to the increased bone turnover frequently seen during
postmenopausal bone loss. Only in the case of BAP was this increase
statistically significant, however.
day-to-day and within-day variations
Twelve subjects (6 men and 6 women) were enrolled in an experiment
in which blood and urine samples were collected at various time points
during the first 24 h and at weekly intervals during the first
month, with a final sample collection after 2 months. We measured serum
concentrations of OC (by two assays), BAP, and PTH, and urine
concentrations of DPD, CTX, OHPro, total Ca2+,
and creatinine. For each subject and each variable separately, the
individual mean values of the seven fasting morning samples (taken at
0900) were calculated and used to calculate the group mean values
(Table 3
, columns 2, 4, and 6). Individual day-to-day CVs were expressed
as percentages of the corresponding individual mean values and were
used to calculate the mean intraindividual variation in the group
(Table 3
, columns 3, 5, and 7). It turned out that the serum
markers (except PTH) were relatively stable with time and that both the
inter- and intraindividual variations of the Coated Bead Assay were
comparable with commercial kits for bone markers. The urinary markers,
notably OHPro and CTX, had large intra- and interindividual CVs.
The within-day variation of the various markers was assessed using the
samples obtained during the first 24 h of the experiment (Fig. 2
). The individual variation of the bone formation markers OC
(determined with both assays) and BAP are given in the different plots.
No distinct diurnal variation was found for any of the three assays. A
more pronounced diurnal pattern was observed for the bone
resorption markers, but because of the large interindividual variation,
the difference between zenith and nadir did not reach statistical
significance (data not shown).
View larger version (27K):
[in this window]
[in a new window]
|
Figure 2. Within-day variation of bone formation markers.
(Top), individual plots of OC determined with the Coated
Beads Assay during consecutive time points within 24 h;
(middle), individual plots of OC determined with the
Elsa-Osteo assay; (bottom), individual plots of BAP. All
curves are presented as the concentration of each marker (µg/L). The
mean values per time point are given by the thick line;
bars, SD for each time point.
|
|
follow-up during osteoporosis treatment
To test the ability of the Coated Beads Assay to detect changes in
serum OC concentration during therapy, three groups of 10
postmenopausal women were followed during their treatment with
bisphosphonates, estrogen, or calcitonin (Table 4
). Changes during therapy were more pronounced for OC than for
BAP, with similar relative changes for the two OC assays.
Therapy-induced changes were also large in the bone resorption
markers, but because of the large standard deviation of urinary
markers, the changes were not statistically significant in all cases.
correlation between bone formation markers
To investigate the correlation between the various tests for bone
formation, the data from the former experiments were pooled. In total,
the data of 181 subjects (60 men, 91 women, and the baseline
measurements of 30 postmenopausal women) were used to perform linear
regression. The correlation between both OC assays was
r = 0.879 (P <0.0001), and the standard
deviation of the residuals (Sy|x) was 4.66 µg/L.
Regression analysis between BAP and both OC assays produced rather poor
regression coefficients and higher Sy|x values
(Table 5
).
|
Discussion
|
|---|
OC is the most abundant noncollagenous protein in bone. It is
synthesized by the osteoblasts, and after its cellular secretion,
~80% is bound to the hydroxyapatite matrix. The remainder is
released into the blood stream, where it is available for detection
(2). Circulating OC is used as a marker for bone formation,
and high concentrations have been observed in children (2)
and in patients with high bone turnover, as is found in Paget
disease and postmenopausal osteoporosis
(18)(19). Commercial test kits for OC detection
are based on one of two principles: radioimmunoassays or enzyme-linked
immunoassays. The use of OC as a marker for bone metabolism in routine
clinical chemistry is mainly restricted because of (a) the
large differences between the various kits and (b) the fact
that no automated test is available to date. The former problem may be
related to the fact that serum OC occurs in at least two conformations:
fully carboxylated OC, which contains three residues of the unusual
amino acid Gla, and undercarboxylated OC, which contains 0–2 Gla
residues (12)(20). The antibodies
(either polyclonal or monoclonal) used in different kits differ from
each other in their relative affinity for fully carboxylated and
undercarboxylated OC (15), and thus in the amounts that are
detected in the same serum sample. An additional problem is that some
kits detect only intact OC, whereas others recognize both intact
OC and OC fragments. Hence, standardization of reference samples and
antibodies used is an absolute requirement before data obtained with
various kits may be compared.
From a technical point of view, a serious drawback of existing kits is
that they are test tube- or microtiter plate-based assays that must be
pipetted by hand. This is laborious and forms a potential source of
imprecision and mistakes. Therefore, we have used and evaluated a
semiautomated method for OC detection, using the antibodies and
calibrators from a commercial enzyme-linked immunosorbent assay. The
test is based on the sandwich principle, with antibodies directed
against epitopes at the NH2 and COOH termini of
OC outside the Gla domain, thus ensuring that only intact OC (both
fully carboxylated and undercarboxylated) is detected. It turned out
that the newly developed assay (provisionally designated as the
Coated Beads Assay) has a high precision and compares well with the kit
from CIS Biointernational in both within-day and day-to-day variation,
as well as in patient follow-up studies. Like other test for bone
formation, the standard deviation of the Coated Beads Assay was much
lower than that of bone resorption markers, and a good correlation was
observed with OC values determined with the CIS kit.
In conclusion, we have demonstrated that with the semiautomated
assay, OC may be determined within 1 h, and that the accuracy of
the assay is at least comparable to existing kits. The new test still
requires a few pipetting steps, but the coated beads may be used
equally well in fully automated analyzers.
|
Acknowledgments
|
|---|
This study was supported by Hoffmann-La Roche Diagnostics (Basel,
Switzerland). We thank Dr. P. Proost (University of Leuven, Belgium)
for kindly supplying us with synthetic OC.
|
Footnotes
|
|---|
1 Nonstandard abbreviations: OC, osteocalcin; Gla, -carboxyglutamate; BAP, bone-specific alkaline phosphatase; mAb, monoclonal antibody; PTH, parathyroid hormone; OHPro, hydroxyproline; DPD, deoxypyridinoline; and CTX, type I collagen C-terminal telopeptide.
|
References
|
|---|
-
Price PA. Role of vitamin K-dependent proteins in bone metabolism [Review]. Annu Rev Nutr 1988;8:565-583.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Hauschka PV, Lian JB, Cole DEC, Gundberg CM. Osteocalcin and matrix Gla protein: vitamin K dependent proteins in bone [Review]. Physiol Rev 1989;69:990-1047.
[Free Full Text]
-
Kesterson RA, Stanley L, DeMayo F, Finegold M, Pike JW. The human osteocalcin promotor directs bone-specific vitamin D-regulatable gene expression in transgenic mice. Mol Endocrinol 1993;7:462-467.
[Abstract/Free Full Text]
-
Ozono K, Liao J, Kerner SA, Scott RA, Pike JW. The vitamin D-responsive element in the human osteocalcin gene. Association with a nuclear proto-oncogene enhancer. J Biol Chem 1990;265:21881-21888.
[Abstract/Free Full Text]
-
Luo G, Ducy P, McKee MD, Pinero GJ, Loyer E, Behringer RR, Karsenty G. Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature 1997;386:78-81.
[Medline]
[Order article via Infotrieve]
-
Power MJ, Fottrell PF. Osteocalcin: diagnostic methods and clinical applications [Review]. Crit Rev Clin Lab Sci 1991;28:287-335.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Garnero P, Grimaux M, Demiaux B, Preaudat C, Seguin P, Delmas PD. Measurement of serum osteocalcin with a human-specific two-site immunoradiometric assay. J Bone Miner Res 1992;7:1389-1398.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Eastell R, Robins SP, Colwell T, Assiri AMA, Riggs BL, Russell RGG. Evaluation of bone turnover in type I osteoporosis using biochemical markers specific for both bone formation and bone resorption. Osteoporos Int 1993;3:255-260.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Knapen MHJ, Hamulyák K, Vermeer C. The effect of vitamin K supplementation on circulating osteocalcin (bone Gla-protein) and urinary calcium excretion. Ann Intern Med 1989;111:1001-1005.
-
Knapen MHJ, Jie K-SG, Hamulyák K, Vermeer C. Vitamin K-induced changes in markers for osteoblast activity and urinary calcium loss. Calcif Tissue Int 1993;53:81-85.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Plantalech L, Guillaumont M, Vergnaud P, Leclercq M, Delmas PD. Impairment of gamma carboxylation of circulating osteocalcin (bone Gla protein) in elderly women. J Bone Miner Res 1991;6:1211-1216.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Szulc P, Chapuy M-C, Meunier PJ, Delmas PD. Serum undercarboxylated osteocalcin is a marker of the risk of hip fracture in elderly women. J Clin Investig 1993;91:1769-1774.
-
Szulc P, Arlot M, Chapuy M-C, Duboeuf F, Meunier PJ, Delmas PD. Serum undercarboxylated osteocalcin correlates with hip bone mineral density in elderly women. J Bone Miner Res 1994;9:1591-1595.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Knapen MHJ, Nieuwenhuijzen Kruseman AC, Wouters RSME, Vermeer C. Correlation of serum osteocalcin with bone mineral density in women during the first 10 years after menopause. J Bone Miner Res 1998;63:375-379.
-
Knapen MHJ, Eisenwiener H-G, Vermeer C. Osteocalcin detection in aging serum and whole blood: stability of different osteocalcin fractions. Clin Chim Acta 1996;256:151-164.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Rosenquist C, Qvist P, Bjarnason N, Christiansen C. Measurement of a more stable region of osteocalcin in serum by ELISA with two monoclonal antibodies. Clin Chem 1995;41:1439-1445.
[Abstract/Free Full Text]
-
Houben RJTJ, Jin D, Stafford DW, Proost P, Ebberink RHM, Vermeer C, Soute BAM. Osteocalcin binds tightly to the -glutamyl carboxylase at a site distinct from that of the other known vitamin K-dependent proteins. Biochem J 1999;341:265-269.
-
Gundberg CM, Lian JB, Gallop PM, Steinberg JJ. Urinary gammacarboxyglutamic acid and serum osteocalcin as bone markers: studies in osteoporosis and Paget’s disease. J Clin Endocrinol Metab 1983;57:1221-1225.
[Abstract/Free Full Text]
-
Kaddam IMS, Iqbal SJ, Holland S, Wong M, Manning D. Comparison of serum osteocalcin with total and bone specific alkaline phosphatase and urinary hydroxyproline:creatinine ratio in patients with Paget’s disease of bone. Ann Clin Biochem 1994;31:327-330.
-
Cairns JR, Price PA. Direct demonstration that the vitamin K-dependent bone Gla protein is incompletely -carboxylated in humans. J Bone Miner Res 1994;9:1989-1997.
[Web of Science][Medline]
[Order article via Infotrieve]
Top
Abstract
Introduction
