1. Digestion and absorption of carbohydrates—from molecules and mem-
branes to humans. Scand J Clin Lab Invest 1995;55:149–52.
19. Ledochowski M, Widner B, Fuchs D. Small intestinal bacterial overgrowth
syndrome (SIBOS) and neopterin. Pteridines 2000;11:9.
20. O’Dell BL, Emery M. Compromised zinc status in rats adversely affects
calcium metabolism in platelets. J Nutr 1991;121:1763–8.
ELISA Methodology for Detection of Modified Osteo-
protegerin in Clinical Studies, De Chen,*
Nihal A.
Sarikaya, Han Gunn, Steven W. Martin, and John D. Young
(Department of Pharmacokinetics and Drug Metabolism,
Amgen, Inc., One Amgen Center Dr., Thousand Oaks, CA
91320; * author for correspondence: fax 805-499-4953, e-
mail dchen@amgen.com)
Osteoprotegerin (OPG), also known as osteoclast inhibi-
tory factor, is a soluble receptor of the tumor necrosis
factor receptor superfamily. The protein is secreted as a
covalent, disulfide-linked homodimer, which is the pre-
dominant extracellular form (1), and is expressed in
multiple tissues (1–3). OPG-mediated pathways might
have a role in osteoporosis (3–6) because estrogen in-
creases OPG gene expression (4, 5). OPG maintains the
structure of healthy bone and inhibits osteoclast activa-
tion and differentiation (3, 7). In the vascular system,
OPG inhibits pathological calcification in the media in-
tima (3). OPG has been proposed for therapy of os-
teopenic disorders, such as postmenopausal osteoporosis,
Paget disease, rheumatoid arthritis, hypercalcemia, and
lytic bone metastases (8).
Initially, we developed an antibody-based ELISA
method with an anti-human OPG monoclonal antibody
for capture and an anti-human OPG polyclonal antibody
for detection. Yano et al. (9) raised the concern for us that
we may not detect the active dimeric OPG with antibody
capture because they reported that serum OPG increased
with age and that the monomer was the predominant
form of OPG in human serum. Although they used a
different antibody-dependent ELISA method (monoclo-
nal capture and detection), the results reported by Yano et
al. (9) do not correspond with the work performed at
Amgen (1, 3–5, 7, 8, 10–12). We reasoned that OPG ligand
(OPGL) (2, 7, 8, 10–16), also known as osteoclast differen-
tiation factor, is a potential alternative capture protein for
an OPG assay. In an attempt to develop an assay that
would measure all bioactive form(s) of OPG, we devel-
oped an ELISA assay that uses OPGL as the capture
protein. To avoid problems posed by batch-to-batch vari-
ability of human serum pools for use as assay diluent,
assay development was used to define a serum substitute.
Assay development was performed with AMGN-0007,
a modified OPG. Calibrators and quality-control (QC)
materials were prepared in human serum or serum sub-
stitute. Whereas calibrators were serially diluted, QC
materials were prepared individually. Calibration curves
were prepared using calibrators containing 0.020–500
␮g/L AMGN-0007. Each calibration curve contained at
least nine points, including the zero calibrator.
OPGL, AMGN-0007, and murine monoclonal antibody
were purified essentially as described previously (1, 7).
OPGL was coated onto 96-well microtiter plates (Costar).
Plates were blocked with 2 mL/L I-Block (Tropix) and 5
mL/L Tween 20 (Pierce) in phosphate-buffered saline
(PBS). Assay buffer, calibrators, and QC materials were
added to the wells. After all unbound substances were
removed by washing, murine anti-human OPG monoclo-
nal antibody was added to the wells. After another wash,
goat anti-mouse IgG conjugated with horseradish perox-
idase (IgG-HRP; Zymed) was added to the wells. After the
final wash, KPL TMB Microwell Peroxidase Substrate
(Kirkegaard & Perry Laboratories) was added to the
wells. The colorimetric reaction was stopped with 0.812
mol/L phosphoric acid. The color intensity was measured
at 450–650 nm with a ThermoMax Microtiter Plate Reader
(Molecular Devices).
The full-length OPG homodimer (OPG-FLD) and the
full-length OPG monomer (OPG-FLM) were purified
from conditioned medium with Sepharose columns and
concentrated into PBS. Sodium dodecyl sulfate-polyacryl-
amide gel electrophoresis was performed to confirm size
and purity (95%) of the monomer and dimer. Calibrators
and QC materials were prepared for each OPG analog:
AMGN-0007, OPG-FLD, and OPG-FLM. The assay was
performed as described above, except that we used HRP-
conjugated murine anti-human OPG monoclonal anti-
body.
Prepared in PBS, human serum substitute buffers con-
tained 30 mL/L human serum albumin (HSA; Biorecla-
mation, Inc.) and various concentrations (0–500 mL/L) of
fetal bovine serum (FBS; Sigma Chemical Co.). Calibrators
and QC materials were prepared and assayed as de-
scribed above.
OPGL was immobilized on the surface of the microtiter
plate. AMGN-0007 was then added to the plate at con-
centrations of 0.244–31.25 ␮g/L, the calibration curve
range. The analyte was detected with murine anti-human
OPG monoclonal antibody plus goat anti-mouse IgG-
HRP. Defined as two times the zero calibrator signal, the
detection limit was 0.244 ␮g/L. Other assay configura-
tions, such as monoclonal capture with polyclonal detec-
tion and ligand capture with polyclonal detection, were
studied and demonstrated low signal-to-noise ratios
throughout the calibration curve. For the dynamic range
of interest, OPGL capture followed by monoclonal detec-
tion produced the best signal-to-noise ratio throughout
the calibration curve; most likely, the result was attribut-
able to the specificity of both the ligand and the mono-
clonal antibody for AMGN-0007.
The abilities of OPG analogs to bind to solid-phase-
bound OPGL were compared (Fig. 1). The signal at 50
␮g/L OPG analog for OPG-FLD (2.352 absorbance units)
was very similar to that of AMGN-0007 (2.411 absorbance
units), whereas OPG-FLM had signal strength of 1.693
absorbance units. Compared with AMGN-0007 and OPG-
FLD, OPG-FLM demonstrated a curve shift to the right
Clinical Chemistry 47, No. 4, 2001 747

2. and a lower signal strength throughout the calibration
curve. Although monomeric OPG did bind to OPGL, the
ligand appeared to have greater affinity for dimeric OPG.
Results from a previous study using pulse-chase labeling
and immunoprecipitation of extracts from Chinese ham-
ster ovary cells transfected with OPG suggested that the
primary extracellular form of OPG is the homodimer (1).
Tomoyasu et al. (17) also determined that dimeric OPG
was secreted in greater amounts than monomeric OPG in
mammalian cell systems and was biologically more active
than monomeric OPG. In addition, they suggested that
dimeric OPG might bind to heparin and be transported to
the target sites more rapid than monomeric OPG (17); this
may explain the observations of Yano et al. (9).
Plapp et al. (18) suggested that QC materials be com-
posed of the same matrix as the specimens. With human
serum as diluent, problems may arise for the following
reasons: (a) the presence of infectious agents that cannot
be screened; (b) batch-to-batch variability attributable to
endogenous factors (19); and (c) a finite shelf life that may
fall short of the time spans of clinical studies. To minimize
the variability for pharmacokinetic profile studies, a
buffer that could be substituted for human serum as an
assay diluent was sought. HSA makes up 30–55 g/L of
serum (20, 21). Therefore, development was focused on
the normal concentration of albumin to maintain a total
protein content similar to the low end of the HSA refer-
ence interval. Weber et al. (19) suggested that the addition
of bovine serum might eliminate interference of hetero-
philic antibodies. After comparing the absorbance and
slope obtained for bovine serum with those obtained for
human serum, we chose 30 mL/L HSA with 100 mL/L
FBS as the serum substitute and diluent for AMGN-0007
assays. Differences in the assay signals obtained for the
serum substitute and human serum were tested for sta-
tistical significance using one-way ANOVA (Table 1). No
significant differences (P Ͼ0.05) were observed among the
signals obtained at different concentrations of human
serum in substitute diluent, suggesting that the human
serum substitute was an adequate substitute for human
serum as assay diluent.
In conclusion, we developed an ELISA AMGN-0007,
with OPGL for analyte capture and a monoclonal anti-
body specific for detection, and found that an albumin-
based diluent can be used. The ELISA will be useful for
analyzing samples from clinical trials as well as for
monitoring therapeutic efforts for osteopenic diseases.
Finally, this tool may enable the development of an assay
to measure endogenous OPG concentrations.
We wish to thank D. Chang, P. Campbell, and D. Yanagi-
hara for providing antibodies. W.J. Boyle, M. Kelley, and
E. Davy were instrumental in providing OPGL and sup-
port. We also thank C.R. Dunstan for guidance.
References
1. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS, Luthy R, et al.
Osteoprotegerin: a novel secreted protein involved in the regulation of bone
density. Cell 1997;89:309–19.
2. Yasuda H, Shima N, Nakagawa N, Mochizuki SI, Yano K, Fujise N, et al.
Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin
(OPG): a mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro.
Endocrinology 1998;139:1329–37.
3. Bucay N, Sarosi I, Dunstan CR, Morony S, Tarpley J, Capparelli C, et al.
Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial
calcification. Genes Dev 1998;12:1260–8.
4. Hofbauer LC, Dunstan CR, Spelsberg TC, Riggs BL, Khosla S. Osteoprote-
gerin production by human osteoblast lineage cells is stimulated by vitamin
D, bone morphogenic protein-2, and cytokines. Biochem Biophys Res
Commun 1998;250:776–81.
Fig. 1. Comparison of OPG analogs.
Various OPG forms [ AMGN-0007 (f), OPG-FLD (E), and OPG-FLM (Œ)] were
prepared for analysis by ELISA. Each compound was serially diluted to 0.049–50
␮g/L in human serum.
Table 1. ANOVA results for QC materialsa
prepared in different diluents.
Diluent
QC 1
(20 ␮g/L)
QC 2
(5 ␮g/L)
QC 3
(1 ␮g/L)
QC 4
(0.7 ␮g/L)
QC 5
(0.5 ␮g/L)
30 mL/L HSA ϩ 100 mL/L FBS in PBS 21.930 5.035 0.891 0.600 0.426
1 mL/L human serum in substitute diluent 22.850 5.186 0.928 0.600 0.426
10 mL/L human serum in substitute diluent 21.660 4.862 0.880 0.649 0.401
100 mL/L human serum in substitute diluent 21.250 4.775 0.880 0.612 0.401
Human serum only 21.740 4.580 0.880 0.637 0.426
F 3.741 4.902 0.424 0.534 0.870
P 0.090 0.056 0.787 0.718 0.541
F4,5 critical 5.192 5.192 5.192 5.192 5.192
a
QC materials and calibrators were prepared with AMGN-0007 in substitute diluent, human serum, or human serum combined with substitute diluent. Assay range
was 0.484–62 ␮g/L.
748 Technical Briefs

3. 5. Hofbauer LC, Khosla S, Dunstan CR, Lacey DL, Spelsberg TC, Riggs BL.
Estrogen stimulates gene expression and protein production of osteoprote-
gerin by human osteoblastic cells. Bone 1998;23(Suppl):S172.
6. Mizuno A, Amizuka N, Irie K, Murakami A, Fujise N, Kanno T, et al. Severe
osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteopro-
tegerin. Biochem Biophys Res Commun 1998;247:610–5.
7. Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess T, et al.
Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation
and activation. Cell 1998;93:165–76.
8. Kong YY, Yoshida H, Sarosi I, Tan HL, Timms E, Capparelli C, et al. OPGL is
a key regulator of osteoclastogenesis, lymphocyte development and lymph-
node organogenesis. Nature 1999;397:315–23.
9. Yano K, Tsuda E, Washida N, Kobayashi F, Goto M, Harada A, et al.
Immunological characterization of circulating osteoprotegerin/osteoclasto-
genesis inhibitory factor: increased serum concentrations in postmeno-
pausal women with osteoporosis. J Bone Miner Res 1999;14:518–27.
10. Shalhoub V, Faust J, Boyle WJ, Dunstan CR, Kelley M, Kaufman S, et al.
Osteoprotegerin and osteoprotegerin ligand effects on osteoclast formation
from human peripheral blood mononuclear cell precursors. J Cell Biochem
1999;72:251–61.
11. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki SI,
et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/
oscteoclastogenesis inhibitory factor and is identical to TRANCE/RANKL.
Proc Natl Acad Sci U S A 1998;95:3597–602.
12. Hsu H, Lacey DL, Dunstan CR, Solovyev I, Colombero A, Timms E, et al.
Tumor necrosis factor receptor family member RANK mediates osteoclast
differentiation and activation induced by osteoprotegerin ligand. Proc Natl
Acad Sci U S A 1999;96:3540–5.
13. Burgess TL, Qian YX, Kaufman S, Ring BD, Van G, Capparelli C, et al. The
ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts.
J Cell Biol 1999;145:527–38.
14. Anderson DM, Maraskovsky W, Billingsley WL, Dougall WC, Tometsko ME,
Roux ER, et al. A homologue of the TNF receptor and its ligand enhance T-cell
growth and dendritic-cell function. Nature 1997;390:175–9.
15. Fuller K, Wong B, Fox S, Choi Y, Chambers TJ. TRANCE is necessary and
sufficient for osteoblast-mediated activation of bone resorption in oste-
oclasts. J Exp Med 1998;188:997–1001.
16. Nakagawa N, Kinosaki M, Yamaguchi K, Shima N, Yasuda H, Yano K, et al.
RANK is the essential signaling receptor for osteoclast differentiation factor
in osteoclastogenesis. Biochem Biophys Res Commun 1998;253:395–
400.
17. Tomoyasu A, Goto M, Fujise N, Mochizuki S, Yasuda H, Morinaga T, et al.
Characterization of monomeric and homodimeric forms of osteoclastogen-
esis inhibitory factor. Biochem Biophys Res Commun 1998;245:382–7.
18. Plapp FV, Essmyer CE, Zucker ML. ClinLab Online.com. http://www.clinla-
bonline.com/quality_control.shtrr (Accessed July 2000).
19. Weber TH, Kapyaho KI, Tanner P. Endogenous interference in immunoas-
says in clinical chemistry. A review. Scand J Clin Lab Invest 1990;50(Suppl
201):77–82.
20. Wyngaarden TB, Smith LH, eds. Cecil textbook of medicine. Philadelphia:
WB Saunders, 1988:1211pp.
21. Andreoli TE, Bennet JC, Carpenter CCJ, Plum F, eds. Cecil essentials of
medicine, 4th ed. Philadelphia: WB Saunders, 1997:321,323,954.
Umbilical Cord and Maternal Plasma Thiol Concentra-
tions in Normal Pregnancy, Maarten T.M. Raijmakers,1
Eva
Maria Roes,2
Eric A.P. Steegers,2
Bas van der Wildt,3
and
Wilbert H.M. Peters1*
(Departments of 1
Gastroenterology
and 2
Obstetrics and Gynecology, University Hospital St.
Radboud, PO Box 9101, 6500 HB Nijmegen, The Nether-
lands; 3
Department of Obstetrics and Gynecology, ‘Nij
Smellinghe’ Hospital, PO Box 20200, 9200 DA Drachten,
The Netherlands; * author for correspondence: fax 31-24-
3540103, e-mail w.peters@gastro.azn.nl)
Cysteine is the only sulfhydryl-containing amino acid in
proteins and is the thiol residue in glutathione (1). In
addition to its importance in the storage and transport of
cysteine, glutathione plays a pivotal role in detoxification
by the action of glutathione S-transferases and in scav-
enging free (oxygen) radicals by the action of glutathione
peroxidases (1).
Thiol metabolism may be altered during pregnancy. In
healthy pregnant women, plasma concentrations of cys-
teine, glutathione, and homocysteine are decreased,
whereas increased homocysteine and cysteine concentra-
tions are seen in pathologic conditions, such as pre-
eclampsia, in which oxidative stress may play an impor-
tant role (2–4).
Thiols may have important physiological functions in
fetal metabolism. Although protein and amino acid turn-
over in the human placenta has been studied extensively
(5–7), few data concerning fetal concentrations of thiols
and placental maternal-fetal thiol interactions are cur-
rently available (8). During normal pregnancy, fetal
growth depends on the supply of nutrients from the
mother, and a clear correlation between maternal and
fetal amino acid and homocysteine concentrations has
been shown (6–8). Decreased concentrations of amino
acids in the umbilical artery, as compared with the
umbilical vein, have been interpreted as an uptake of
amino acids into fetal tissues where they may be used in
protein biosynthesis or as sources of energy (5). We
studied fetal and maternal thiol plasma concentrations in
normal pregnancies to determine reference concentrations
for cysteine, homocysteine, and cysteinylglycine in arte-
rial and venous umbilical cord plasma to gain insight into
maternal-fetal thiol interactions.
Arterial and venous umbilical cord blood samples from
320 consecutive neonates were drawn in preheparinized
tubes immediately after birth (cat. no. 260545; Kemper
Medical) from March 1997 to January 1998 at the Depart-
ment of Obstetrics/Gynecology of the ‘Nij Smellinghe’
Hospital. Approval for this study was granted by the
Institutional Review Board of the hospital. Blood gas
values were assessed on an ABL-330 analyzer (Radiome-
ter Nederland). Samples with a pH difference between
arterial and venous umbilical cord blood Ͻ0.02 pH units
and samples from neonates with a gestational age Ͻ37
weeks, an umbilical artery pH Ͻ7.20, a birth weight below
the 10th percentile according to Kloosterman (9), or who
were the offspring of women with a diastolic blood
pressure Ͼ90 mmHg during gestation were excluded.
Antecubital maternal venous blood of 35 women was
collected in parallel with the umbilical cord samples after
informed consent. Samples were taken in heparinized
tubes (cat. no. 367684; Becton and Dickinson) Ͻ4 h before
elective caesarian delivery or Ͻ15 min after vaginal birth.
Blood was centrifuged within 10 min at 1200g for 10 min
at room temperature. Plasma samples were stored at
Ϫ30 °C for ϳ2 years until analysis. Plasma concentrations
of total cysteine (tCys), total homocysteine (tHcy), and
total cysteinylglycine (tCysGly) in 195 umbilical cord (102
males and 92 females; no gender was recorded for 1
neonate) and 35 maternal samples were determined by
HPLC as described previously (2).
After log transformation to approach normalization, we
analyzed the data by the paired t-test to assess statistical
differences between maternal venous blood and arterial
Clinical Chemistry 47, No. 4, 2001 749