17
Journal of Immunological Methods, 139 (1991) 17-23
© 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 0022175991001554 JIM 05900
A one step sandwich enzyme immunoassay for 3,-carboxylated osteocalcin using monoclonal antibodies N o b u t o K o y a m a 1, K a n a k o Ohara 1, Hiroko Yokota 1, Tohru K u r o m e 1, Masahiko K a t a y a m a 1, Fumitsugu Hino 1, Ikunoshin K a t o i and Toshihiro Akai 2 I Biotechnology Research Laboratory, Takara Shuzo Co. Ltd., Otsu, Shiga, 520-21, Japan, and 2 Yoda Hospital, Sennan, Osaka, Japan
(Received 8 October 1990, revised received 12 November 1990, accepted 14 January 1991)
A highly sensitive, simple and reliable one-step sandwich enzyme immunoassay (EIA) for the ~,carboxylated form of osteocalcin (Gla-OC) has been developed using a monoclonal antibody. The minimum amount of Gla-OC detected by this EIA was approximately 0.2 n g / m l when a 10/tl aliquot of the sample was used. The serum Gla-OC level in 30 healthy subjects was 3.6 + 2.19 ng/ml (mean + SD). A significant increase was seen in patients with chronic renal failure (20.3 + 4.60 ng/ml), atherosclerosis (8.3 + 4.94 ng/ml) and osteoporosis (10.1 + 4.60 ng/ml). The correlation between the values obtained by the sandwich EIA and competitive RIA methods was given by the linear regression equation, y = 2.896 + 0.759x, for which the correlation coefficient (r) was 0.815 (n = 58). This newly developed Gla-OC specific EIA may be useful for the diagnosis of metabolic bone disease and ectopic calcification. Key words: Monoclonal antibody; Sandwich enzyme immunoassay; Osteocalcin; "r-Carboxylation
Introduction
Osteocalcin (OC) is a vitamin K-dependent, Ca 2+-binding single chain polypeptide of molecular weight 5900 with three y-carboxyglutamic acid residues, at positions 17, 21, and 24 (Poser et al., 1980). Although its function is unknown, these three residues cause it to bind strongly to hy-
Correspondence to: F. Hino, Biotechnology Research Laboratory, Takara Shuzo Co. Ltd., Otsu, Shiga, 520-21, Japan. Abbreviations: EIA, enzyme immunoassay; MAb, monoclonal antibody; RIA, radioimmunoassay; POD, horseradish peroxidase; PBS, phosphate-buffered saline; BSA, bovine serum albumin; HPLC, high performance fiquid chromatography; FPLC, fast protein liquid chromatography; KLH, keyhole lympet hemocyanin; GMBS, N-(y-maleimidobutyrylozyl)-succinimide; ELISA, enzyme-linked immunosorbent assay
droxyapatite. It constitutes 15% of the non-collagenous bone matrix proteins and is synthesized by bone cells (Price et al., 1976a). Circulating OC has been shown to be a sensitive indicator of bone turnover in patients with various metabolic bone diseases (Price et al., 1980; Delmas et al., 1983; Brown et al., 1984; Malluche et al., 1984). The measurement of OC in serum is usually performed by a competitive radioimmunoassay (RIA), first described by Price and Nishimoto (1980). A competitive EIA for OC has been developed using enzyme-labeled OC and a polyclonal antibody (Tanaka et al., 1986). However, such conventional immunoassays cannot distinguish between the ~,-carboxylated form ((GIa-)OC) and the decarboxylated form ((GIu-)OC). We describe here the production of hybridoma cell lines secreting MAb to OC and their application in a rapid
18 and sensitive sandwich EIA and in determining the content of Gla-OC in patients with metabolic bone disease and ectopic calcification.
Materials and methods
Osteocalcin Bone OC was extracted from bovine bone powder and was purified by the method of Gundberg et al. (1984). Thermal decarboxylation of the y-carboxyglutamic acid residue was accomplished by the method of Price (1984). Trypsin and V8 protease digestion and acid cleavage were accomplished by the methods of Price et al. (1976b) and of Fraser et al. (1972), respectively. The generated peptide fragments were purified by reverse phase HPLC and identified by amino acid analysis. Recombinant DNA derived human Glu-OC and Gla-OC were kindly provided by Mr. H. Matsushita, Protein Engineering Section, Biotechnology Research Laboratory, Takara Shuzo Co. Ltd., Japan. Monoclonal antibodies A conjugate of bovine Gla-OC with keyhole limpet hemocyanin (KLH) was prepared by the reaction of 2-iminothiolane-modified KLH with N-('/-maleimidebutyryloxy)-succinimide (GMBS) and Gla-OC in a step-wise manner (Hillel and Wu, 1977). BALB/c mice were then immunized with the conjugate mixed with Freund's complete adjuvant (Gibco). Splenocytes of the immunized mice were taken 3 days after administering a booster injection of the conjugate, and cell fusions were performed using a mouse myeloma cell (P3X63Ag8-U1) according to standard procedures (Galfr~ and Milstein, 1981). Growing hybridomas were screened by the enzyme-linked immunosorbent assay (ELISA) method, and positive cultures were cloned by limiting dilution. Polystyrene microtiter plates were coated with 100/~1 of 5/~g/ml bovine Gla-OC solution in phosphate-buffered saline (PBS). Free sites on the wells were blocked by incubating with PBS containing 1% bovine serum albumin (BSA, RIA grade, Sigma), at 37°C for 1 h. Cultured supernatants were incubated in the antigen coated wells at 37°C for 1 h. Then
goat antibodies to mouse IgG, labeled with horseradish peroxidase (POD), were added and the plates incubated at 37°C for 1 h. Enzyme activity was measured using 100/~1 of 0.01% H202 in 0.1 M sodium citrate buffer, pH 5.0, containing 1 mg/ml o-phenylenediamine (OPD). Finally, four MAbs (OC4-30, OCG2, OCG3, OCG4) were selected and purified from ascites fluids by precipitation with ammonium sulfate, followed by Mono Q FPLC (Pharmacia). Determination of the epitope specificities of the MAbs The epitope specificities of the MAbs were determined by a competitive ELISA. POD-labeled MAbs were prepared by the periodate method (Nakane and Kawaoi, 1974). 100 /~1 of PODlabeled MAb (1 /~g/ml in PBS containing 1% BSA) were pre-incubated with 100/~1 of OC, or its fragments, (0.1/Lg/ml) overnight at 4°C. The mixture was added to a microplate (100 /zl/well) which was precoated with bovine Gla-OC, and the bound POD activity was measured for 15 min at room temperature using OPD and H202 as described above. Selection of the M A b combination for use in sandwich EIA 2/zg of capture MAb (MAbl) in 200/~1 of PBS were added to the wells of a microplate and incubated overnight at 4°C. The wells were blocked with 200 ktl of PBS containing 1% BSA for 1 h at 37°C and then washed four times with 250 ~tl of PBS. 1 ng of bovine GIa-OC in 10 /~1 of PBS containing 1% BSA was added to the wells and then 100/~1 of a 1/1000 dilution of POD-labeled MAb (MAb2), corresponding to about 200 ng of antibody, in dilution buffer (PBS containing 1% BSA and 0.05% Tween 20) were added. The plate was incubated for 1 h at room temperature and then washed four times with PBS. The POD activity bound to the plate was assayed for 15 min at room temperature with a solution (100/~l/well) of 0.1 M citrate-0.2 M phosphate buffer, pH 5.0, containing 1 g / 1 0 P D and 0.01% (v/v) H202. The reaction was stopped by the addition of 100 /~l/well of 1 N H2SO4. The absorbance value at 492 nm was measured using a microplate reader (Titertek Multiskan, Model MCC, Flow Lab.).
19 G l a - O C specific E I A was p e r f o r m e d in a similar m a n n e r as above, b u t OC4-30 a n d O C G 4 were used as M A b l a n d M A b 2 , respectively, a n d the i n c u b a t i o n time of the i m m u n o l o g i c a l r e a c t i o n was increased to 1.5 h.
Competitive RIA for OC Serum s a m p l e s were assayed with the R I A kit f r o m C o m p a g n i e Otis Industries, S.A. ( G r e e n Cross Ltd., J a p a n ) a c c o r d i n g to the m a n u f a c t u r e r ' s directions. This kit is a c o m p e t i t i v e i m m u n o a s s a y using a p o l y c l o n a l a n t i b o d y d i r e c t e d a g a i n s t b o v i n e OC.
Serum samples B l o o d s a m p l e s were collected f r o m 30 h e a l t h y volunteers (14 w o m e n a n d 16 men; m e a n age 45 years, range 2 0 - 6 5 years), 30 p a t i e n t s with osteop o r o s i s (20 w o m e n a n d 10 men; m e a n age 58 years, range 4 8 - 7 5 years), eight p a t i e n t s with c h r o n i c renal failure (three w o m e n a n d five m e n ; m e a n age 48 years, range 2 5 - 6 2 years), a n d 20 p a t i e n t s with atherosclerosis (seven w o m e n a n d 13 men; m e a n age 54 years, range 3 5 - 7 3 years). A l l d o n o r s were fasting w h e n the collections were m a d e . T h e serum s a m p l e s o b t a i n e d were s t o r e d at - 4 0 ° C until analysis.
Results
Monoclonal antibodies F o u r M A b s (OC4-30 ( I g G 2 a ) , O C G 2 (IgG1), O C G 3 ( I g G 3 ) a n d O C G 4 ( I g G 1 ) ) were o b t a i n e d . T o d e t e r m i n e the antigenic e p i t o p e s on the O C m o l e c u l e recognised b y each M A b , c o m p e t i t i v e i n h i b i t i o n b y OC, or its fragments, of M A b b i n d ing to b o v i n e G l a - O C , i m m o b i l i z e d o n the p o l y styrene m i c r o t i t e r plate, was tested using the c o m petitive E L I S A . A s shown in T a b l e I, the b i n d i n g of every M A b to b o v i n e G l a - O C was i n h i b i t e d b y h u m a n G l a - O C . T h e b i n d i n g of O C 4 - 3 0 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l a - O C f r a g m e n t c o r r e s p o n d i n g to residues 8 - 3 1 , b u t was n o t inhibited by bovine Glu-OC, human Glu-OC or t r y p t i c f r a g m e n t 1 - 1 9 . T h e s e results suggested that OC4-30 recognized the residues 8 - 3 1 , which cont a i n e d three 3,-carboxyglutamic a c i d residues a n d t h a t O C 4 - 3 0 was specific for G l a - O C . T h e b i n d -
TABLE I EPITOPE SPECIFICITY OF THE MAbs INVESTIGATED OC and its fragments
Inhibition of MAb binding OC4-30 OCG2 OCG3 OCG4
(oc) Bovine Glu-OC Human GIa-OC Human Glu-Oc
+++ -
+++ +++ +++
(Fragments of bovine Gla-OC) a Acid-cleaved Frl-14 ND Acid-cleaved Fr15-48 ND ++ Tryptic Frl-19 Tryptic Fr21-42 Tryptic Fr45-49 ++ V8-digested Fr8-31 ++ -
+++ +++ +++
+++ +++ +++
+++ ++ +++
+ ++ -
a Percentage of A280compared with bovine Gla-OC as competitor in the competitive ELISA. Key: ND, not done; - , 70%.
ing of O C G 2 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l u - O C , h u m a n G l u - O C , a n d the acidcleavage f r a g m e n t c o r r e s p o n d i n g to residues 1 5 - 4 9 or t r y p t i c f r a g m e n t 4 5 - 4 9 . T h o s e results i n d i c a t e d t h a t O C G 2 r e c o g n i z e d the residues 4 5 - 4 9 b u t could not distinguish between Gla-OC and GluOC. T h e b i n d i n g of O C G 3 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l u - O C , h u m a n G I u - O C , the acid cleavage f r a g m e n t c o r r e s p o n d i n g to residues 1 5 - 4 9 , the t r y p t i c f r a g m e n t 2 1 - 4 2 , a n d the V8 p r o t e a s e d i g e s t e d f r a g m e n t 8 - 3 1 . These results showed the O C G 3 r e c o g n i z e d the residues 2 1 - 3 1 , which c o n t a i n e d two v - c a r b o x y g l u t a m i c acid residues, b u t c o u l d n o t d i s t i n g u i s h b e t w e e n G l a - O C a n d G l u - O C . T h e b i n d i n g of O C G 4 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l u - O C , h u m a n G l u - O C , the a c i d cleavage f r a g m e n t c o r r e s p o n d ing to residues 1 - 1 4 , a n d the t r y p t i c f r a g m e n t 1 - 1 9 . T h e s e results s h o w e d that O C G 4 recognized residues 1 - 1 4 b u t c o u l d n o t distinguish b e t w e e n Gla-OC and Glu-OC.
Two-site binding assay I n o r d e r to find M A b p a i r s suitable for use in the o n e step s a n d w i c h E I A for OC, f o r m a t i o n of the s a n d w i c h c o m p l e x was d e t e c t e d in a two-site b i n d i n g assay. T h e f o u r P O D - l a b e l e d M A b s were p a i r e d in their i m m o b i l i z e d f o r m in all 16 p o s s i b l e
20 TABLE I1
TABLE III
TWO-SITE BINDING ASSAY FOR OC USING TWO DIFFERENT MONOCLONAL ANTIBODIES
RECOVERY TEST
MAbl
MAb2
0c4-30 OCG2 OCG3 OCG4
OC4-30
OCG2
OCG3
OCG4
++ -
+ -
++ +++
+++ -
Sample (ng/ml)
Added OC (ng/ml)
Theoretical value (ng/ml)
Measured Recovery value (%) (ng/ml)
A (7.4)
0 10 30
3.7 8.7 18.7 7.5 12.5 22.5
3.6 9.3 17.3 7.3 11.5 24.7
B (15.0)
Key: Mabl, capture MAb; MAb2, Tag MAb; - , 50 × background. c o m b i n a t i o n s a n d tested at a fixed c o n c e n t r a t i o n of bovine G l a - O C . As shown in T a b l e II, at least five pairs of M A b appeared to form a sandwich complex with b o v i n e G I a - O C a n d of these OC4-30 a n d O C G 4 (used as capture a n d detector M A b respectively) were selected for G l a - O C specificity a n d the highest a b s o r b a n c e r e a d i n g at a fixed c o n c e n t r a t i o n of b o v i n e G l a - O C . Moreover, this c o m b i n a t i o n of M A b showed the most sensitive detection limit (data n o t shown).
Assay validity of sandwich EIA Fig. 1 shows the s t a n d a r d curve generated b y this assay using b o v i n e G l a - O C as standard. The
1.5
coefficient of v a r i a t i o n (CV) at each dilution of the s t a n d a r d was less t h a n 10%. T h e sensitivity of the assay was calculated as the c o n c e n t r a t i o n which was two s t a n d a r d deviations (2SD) above the zero s t a n d a r d a n d for the s t a n d a r d b o v i n e G l a - O C was f o u n d the be 0.2 n g / m l . T h e precision of the assay was tested b y assaying three s e r u m samples ten times (within-assay) a n d in five consecutive assays (between-assay). T h e CV values were less t h a n 15%. I n experiments to d e t e r m i n e analytical recovery, two serum samples c o n t a i n i n g 7.4, a n d 15.0 n g / m l respectively were used. Each serum sample was assayed following the a d d i t i o n of b o v i n e G l a - O C s t a n d a r d (0, 10, 30 n g / m l ) a n d the calculated recoveries were 92-110% (Table III).
Specificity of the EIA system
q' ~'
/
16"01
0.5"
f
3.6)
i(( 8.8 ) (g.o) L917) 0.0
1=0
,
,
20
30
40
GLA-OC (NGIML)
Fig. 1. A standard curve for Gla-OC in the sandwich EIA. The coefficent of variation (CV%) at each concentration is shown in parentheses, and vertical bars indicate the range of withinassay variation (SD, n = 6).
T o establish that the materials detected in the s e r u m c o r r e s p o n d e d to O C a n d n o t to other serum proteins of different m o l e c u l a r mass, we fractionated a 1.0 ml aliquot of a serum sample c o n t a i n i n g a p p a r e n t l y high c o n c e n t r a t i o n s of O C b y passage through a Sephadex G - 1 0 0 c o l u m n . The resulting fractions were a n a l y z e d using the sandwich E I A (Fig. 2). I m m u n o r e a c t i v e materials were f o u n d in the fraction i n which a p r o t e i n of the same molecular weight as O C w o u l d be expected to be eluted. I n addition, a b o u t 20% of the total activity detected was f o u n d to be associated with fractions eluting at the void volume. Cross-reactivities in the present E I A system were e x a m i n e d using b o v i n e G I u - O C , h u m a n GlaOC, h u m a n G l u - O C a n d other "y-carboxyglutamic a c i d - c o n t a i n i n g s e r u m proteins (protein C, p r o t e i n S, p r o t h r o m b i n ) . N e a r l y 100% cross-reactivity was
21 10
VO
GLA-OC
SERUM GLA--OC 2O
2
(NG/ML)
4O
OSTEOPOROSlS N=30 1
.a
CHRONIC RENAL FAILURE
N----8
z v
?,
ATHEROSCLEROSIS N---- 20
I
.a ~9
NORMAL N : 30
....
10
20
30
40
_ Z60 2
50
....
o
70
FRACTION NO
Fig. 2. Location of fractions containing Gla-OC immunoreactivity after fractionation of a 1 ml serum sample on a Sephadex G-100 column. The sample was added to a 15 ×700 mm column equilibrated with 5 mM N H n C O 3 and eluted in 3 ml fractions with the same buffer, e, A280; ©, Gla-OC determined by EIA. Arrows indicate the void volume (Vo), and the elution position of purified Gla-OC.
observed for human Gla-OC, but no cross-reaction was observed with any of the other proteins tested. Correlation of EIA with RIA To evaluate the accuracy of the present sandwich EIA method, we used 58 serum samples to compare the values obtained by the EIA with those obtained by the commercially available RIA. As shown in Fig. 3, there was a good correlation between the two methods over the entire range
i
40
•
•
-.
o
I
I
20
I
40 RIA--OC
60 (NG/ML)
Fig. 3. Correlation between data from the new EIA system using two MAbs (EIA-OC), and those from the conventional RIA using a polyclonal antibody (RIA-OC). The first order regression line is illustrated, y = 2.896+0.759x, n = 58, r = 0.815 ( p < 0.001)
Fig. 4. Serum concentrations determined with the newly developed EIA system using two MAbs. Dashed lines show mean values.
studied. The correlation coefficient (r) was 0.815 ( p < 0.001). Serum concentration of Gla-OC The Gla-OC content of sera from healthy volunteers and patients are shown in Fig. 4. The serum Gla-OC values of the 30 healthy volunteers ranged from 0.5 to 9.7 n g / m l (3.6 + 2.19 (mean + SD)). Significantly elevated Gla-OC concentrations were observed in the sera of patients with osteoporosis (10.1 + 4.60, p < 0.001), chronic renal failure (20.3 + 16.56, p < 0.05), and atherosclerosis (8.3 _ 4.94, p < 0.01).
Discussion
We have been developing MAbs against bovine GIa-OC, and the four MAbs tested recognized four different epitopes on OC as revealed by the competitive ELISA (Table I). In particular OC4-30 recognized the Gla-form of OC but not the Gluform, and we developed the sandwich EIA using this MAb, which was specific for Gla-OC. Serum OC measurements are used in various clinical situations, including metabolic bone disease, renal disorders, and hyperthyroidism (Brown et al., 1984). RIAs for OC have been described based on the competitive RIA method using radio-iodinated OC and polyclonal antibody (Price et al., 1980). The inherent problems of RIAs include the relatively short half-life of the label, relatively long incubation times, and a requirement for specialized facilities and equipment. An EIA for OC using a MAb and a POD-labeled OC
22
was recently described (Power and Fottrell, 1989), but we believe that our newly developed one-step sandwich EIA system using a Gla-OC specific MAb has certain advantages. For example, the entire assay time is 2 h which is much shorter than conventional assay systems. Furthermore, sample volumes of 10/H were used without prior dilution, in contrast to most published assays in which larger volumes have been used. Several investigators have suggested that subforms of OC exist in human serum samples and that antibodies differ in their ability to recognize all of them (Gundberg and Weinstein, 1986; Power et al., 1989). It has been suggested that OC has only one major epitope near the C-terminus (Price and Nishimoto, 1980), where the amino acid sequence is identical in the bovine and human proteins but it has been suggested that there is another epitope located between positions 20 and 37 (Taylor et al., 1988). In our newly developed sandwich EIA, the binding site was located within the sequence 1-31, which contains three y-carboxyglutamic acid residues. It has been reported that the apparent concentration of OC in a serum sample, as measured by immunoassay, depends on the antibody used (Power et al., 1989). In this study, the EIA using monoclonal antibodies underestimated the OC concentration present by about 24% when compared with the conventional RIA with polyclonal antibody. However, there was a good correlation between the EIA and conventional RIA and the presence of a high-molecular-weight immunogen was confirmed by gel filtration of the serum on Sephadex G-100, as reported previously (Price et al., 1980; Power et al., 1989; Taylor et al., 1990). An explanation of this underestimation of OC concentration may be that the measured OC in the EIA is Gla-OC, whereas the polyclonal antibody based RIA detected not only Gla-OC but also a partly degraded or decarboxylated form of OC, i.e., total OC. The calcium-binding property of the vitamin K-dependent T-carboxyglutamic acid containing plasma proteins is dependent on the T-carboxyglutamic acid residue, which is necessary for biological activities such as activation of the blood coagulation cascade. The T-carboxyglutamic acid residue of the OC is necessary for the formation of a high affinity mineral-protein complex. Thus, the
Gla-OC must be an active form and it is suggested that the measurement of Gla-OC by our EIA may provide better clinical information than does the conventional assay with measures the total of active and inactive forms of OC. In this study, a significant increase in serum Gla-OC levels was observed in patients with chronic renal failure and bone disease, as reported previously (Price et al., 1980). Moreover, we have demonstrated for the first time a significant increase in serum Gla-OC levels in a patient with atherosclerosis. Levy et al. (1979, 1983) reported the accumulation of Glacontaining proteins, such as OC and atherocalcin, in calcified plaque. From these results, it was suggested that the circulating levels of GIa-OC might correlate with disease status, and could, therefore, be a valuable diagnostic parameter in atherosclerosis.
References Brown, J.P., Delmas, P.D. and Malaval, L. (1984) Serum bone Gla-protein: a specific marker for bone formation in postmenopausal osteoporosis. Lancet l, 1091-1093. Delmas, P.D., Stenner, D. and Wahner, H.W. (1983) Increase in serum bone T-carboxyglutamic acid protein with aging in women. J. Clin. Invest. 71, 1316-1321. Fraser, K.J., Poulsen, J.K. and Haber, E. (1972) Specific cleavage between variable and constant domains of rabbit antibody light chains by dilute acid hydrolysis. Biochemistry 11, 4974-4977. Galfr/~, C. and Milstein, C. (1981) Preparation of monoclonal antibodies: strategies and procedures. Methods Enzymol. 73, 2-46. Grundberg, C.M. and Weinstein, R.S. (1986) Multiple immunoreactive forms of osteocalcin in uremic serum. J. Clin. Invest. 77, 1762-1767. Gundberg, C.M., Hauschka, P.V., Lian~ J.B. and Gallop, P.M. (1984) Osteocalcin: isolation, characterization and detection. Methods Enzymol. 107, 516-544 Gundberg, C.M., Wilson, P.S., Gallop, P.M. and Parfitt, A.M. (1985) Determination of osteocalcin in human serum: results with two kits compared with those by a well characterized assay. Clin. Chem. 31, 1720-1723. Hillel, H. and Wu, C.W. (1977) Subunit topography of RNA polymerase from Escherichia coli. A cross-linking study with bifunctional reagents. Biochemistry 16, 3334-3342. Levy, R.J., Liam, J.B. and Gallop, P.M. (1979) Atherocalcin, a T-carboxyglutamic acid containing protein from atherosclerotic plaque. Biochem. Biophys. Res. Commun. 91, 41-49. Levy, R.J., Gundberg, C. and Scheinman. P. (1983) The identification of the vitamin K-dependent bone protein osteocalcin as one of the T-carboxyglutamic acid containing
23 protein present in calcified atherosclerotic plaque and mineralized heart valves. Atherosclerosis 46, 49-56. Malluche, H.M., Faugere, M.C. and Fanti, P. (1984) Plasma level of bone Gla-protein reflect bone formation in patients on chronic maintenance dialysis. Kidney Int. 26, 869-874. Nakane, P.K. and Kawaoi, A. (1974) Peroxidase-labelled antibody. A new method of conjugation. J. Histochem. Cytochem. 22, 1084-1088. Poser, J.W., Esch, F.S., Ling, N.C. and Price, P.A. (1980) Isolation and sequence of the vitamine K-dependent protein from human Bone. J. Biol. Chem. 255, 8685-8691. Power, M.J. and Fottrell, P.F. (1989) Solid-phase enzyme immunoassay for osteocalcin in human serum or plasma, with use of a monoclonal antibody. Clin. Chem. 35, 2087-2092. Power, M.J. Gosling, J.P. and Fottrell, P.F. (1989) Radoimmunoassay of osteocalcin with polyclonal and monoclonal antibodies. Clin. Chem. 35, 1408-1415. Price, P.A. (1984) Decalboxylation of T-carboxyglutamic acid residues in proteins. Methods Enzymol. 107, 548-551. Price, P.A. and Nishimoto, S.K. (1980) Radioimmunoassay for
the vitamin K-dependent protein of bone and its discovery in plasma. Proc. Natl. Acad. Sci. U.S.A. 77, 2234-2238. Price, P.A., Otsuka, A.S., Poser, J.W., Kristaponis, J. and Raman, N. (1976a) Characterization of -y-carboxyglutamic acid containing protein from bone. Proc. Natl. Acad. Sci. U.S.A. 73, 1147-1151 Price, P.A., Poser, J.W. and Raman, N. (1976b) Primary structure of the T-carboxyglutamic acid-containing protein from bovine bone. Proc. Natl. Acad. Sci. U.S.A. 73, 3174-3375. Price, P.A., Parthemore, J.G. and Deftos, L.T. (1980) New biochemical marker for bone metabolism. J. Clin. Invest. 66, 878-883. Tanaka, H., Kuwada, M., Shiraki, M. and Katayama, K. (1986) An enzyme immunoassay for osteocalcin. J. Immunol. Methods 94, 19-24. Taylor, A.K., Linkhadt, S.G., Mohan, S. and Baylink, D.J. (1988) Development of a new RIA for human osteocalcin: evidence for a midmolecule epitope. Metabolism 37, 872877.
Journal of Immunological Methods, 139 (1991) 17-23
© 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 0022175991001554 JIM 05900
A one step sandwich enzyme immunoassay for 3,-carboxylated osteocalcin using monoclonal antibodies N o b u t o K o y a m a 1, K a n a k o Ohara 1, Hiroko Yokota 1, Tohru K u r o m e 1, Masahiko K a t a y a m a 1, Fumitsugu Hino 1, Ikunoshin K a t o i and Toshihiro Akai 2 I Biotechnology Research Laboratory, Takara Shuzo Co. Ltd., Otsu, Shiga, 520-21, Japan, and 2 Yoda Hospital, Sennan, Osaka, Japan
(Received 8 October 1990, revised received 12 November 1990, accepted 14 January 1991)
A highly sensitive, simple and reliable one-step sandwich enzyme immunoassay (EIA) for the ~,carboxylated form of osteocalcin (Gla-OC) has been developed using a monoclonal antibody. The minimum amount of Gla-OC detected by this EIA was approximately 0.2 n g / m l when a 10/tl aliquot of the sample was used. The serum Gla-OC level in 30 healthy subjects was 3.6 + 2.19 ng/ml (mean + SD). A significant increase was seen in patients with chronic renal failure (20.3 + 4.60 ng/ml), atherosclerosis (8.3 + 4.94 ng/ml) and osteoporosis (10.1 + 4.60 ng/ml). The correlation between the values obtained by the sandwich EIA and competitive RIA methods was given by the linear regression equation, y = 2.896 + 0.759x, for which the correlation coefficient (r) was 0.815 (n = 58). This newly developed Gla-OC specific EIA may be useful for the diagnosis of metabolic bone disease and ectopic calcification. Key words: Monoclonal antibody; Sandwich enzyme immunoassay; Osteocalcin; "r-Carboxylation
Introduction
Osteocalcin (OC) is a vitamin K-dependent, Ca 2+-binding single chain polypeptide of molecular weight 5900 with three y-carboxyglutamic acid residues, at positions 17, 21, and 24 (Poser et al., 1980). Although its function is unknown, these three residues cause it to bind strongly to hy-
Correspondence to: F. Hino, Biotechnology Research Laboratory, Takara Shuzo Co. Ltd., Otsu, Shiga, 520-21, Japan. Abbreviations: EIA, enzyme immunoassay; MAb, monoclonal antibody; RIA, radioimmunoassay; POD, horseradish peroxidase; PBS, phosphate-buffered saline; BSA, bovine serum albumin; HPLC, high performance fiquid chromatography; FPLC, fast protein liquid chromatography; KLH, keyhole lympet hemocyanin; GMBS, N-(y-maleimidobutyrylozyl)-succinimide; ELISA, enzyme-linked immunosorbent assay
droxyapatite. It constitutes 15% of the non-collagenous bone matrix proteins and is synthesized by bone cells (Price et al., 1976a). Circulating OC has been shown to be a sensitive indicator of bone turnover in patients with various metabolic bone diseases (Price et al., 1980; Delmas et al., 1983; Brown et al., 1984; Malluche et al., 1984). The measurement of OC in serum is usually performed by a competitive radioimmunoassay (RIA), first described by Price and Nishimoto (1980). A competitive EIA for OC has been developed using enzyme-labeled OC and a polyclonal antibody (Tanaka et al., 1986). However, such conventional immunoassays cannot distinguish between the ~,-carboxylated form ((GIa-)OC) and the decarboxylated form ((GIu-)OC). We describe here the production of hybridoma cell lines secreting MAb to OC and their application in a rapid
18 and sensitive sandwich EIA and in determining the content of Gla-OC in patients with metabolic bone disease and ectopic calcification.
Materials and methods
Osteocalcin Bone OC was extracted from bovine bone powder and was purified by the method of Gundberg et al. (1984). Thermal decarboxylation of the y-carboxyglutamic acid residue was accomplished by the method of Price (1984). Trypsin and V8 protease digestion and acid cleavage were accomplished by the methods of Price et al. (1976b) and of Fraser et al. (1972), respectively. The generated peptide fragments were purified by reverse phase HPLC and identified by amino acid analysis. Recombinant DNA derived human Glu-OC and Gla-OC were kindly provided by Mr. H. Matsushita, Protein Engineering Section, Biotechnology Research Laboratory, Takara Shuzo Co. Ltd., Japan. Monoclonal antibodies A conjugate of bovine Gla-OC with keyhole limpet hemocyanin (KLH) was prepared by the reaction of 2-iminothiolane-modified KLH with N-('/-maleimidebutyryloxy)-succinimide (GMBS) and Gla-OC in a step-wise manner (Hillel and Wu, 1977). BALB/c mice were then immunized with the conjugate mixed with Freund's complete adjuvant (Gibco). Splenocytes of the immunized mice were taken 3 days after administering a booster injection of the conjugate, and cell fusions were performed using a mouse myeloma cell (P3X63Ag8-U1) according to standard procedures (Galfr~ and Milstein, 1981). Growing hybridomas were screened by the enzyme-linked immunosorbent assay (ELISA) method, and positive cultures were cloned by limiting dilution. Polystyrene microtiter plates were coated with 100/~1 of 5/~g/ml bovine Gla-OC solution in phosphate-buffered saline (PBS). Free sites on the wells were blocked by incubating with PBS containing 1% bovine serum albumin (BSA, RIA grade, Sigma), at 37°C for 1 h. Cultured supernatants were incubated in the antigen coated wells at 37°C for 1 h. Then
goat antibodies to mouse IgG, labeled with horseradish peroxidase (POD), were added and the plates incubated at 37°C for 1 h. Enzyme activity was measured using 100/~1 of 0.01% H202 in 0.1 M sodium citrate buffer, pH 5.0, containing 1 mg/ml o-phenylenediamine (OPD). Finally, four MAbs (OC4-30, OCG2, OCG3, OCG4) were selected and purified from ascites fluids by precipitation with ammonium sulfate, followed by Mono Q FPLC (Pharmacia). Determination of the epitope specificities of the MAbs The epitope specificities of the MAbs were determined by a competitive ELISA. POD-labeled MAbs were prepared by the periodate method (Nakane and Kawaoi, 1974). 100 /~1 of PODlabeled MAb (1 /~g/ml in PBS containing 1% BSA) were pre-incubated with 100/~1 of OC, or its fragments, (0.1/Lg/ml) overnight at 4°C. The mixture was added to a microplate (100 /zl/well) which was precoated with bovine Gla-OC, and the bound POD activity was measured for 15 min at room temperature using OPD and H202 as described above. Selection of the M A b combination for use in sandwich EIA 2/zg of capture MAb (MAbl) in 200/~1 of PBS were added to the wells of a microplate and incubated overnight at 4°C. The wells were blocked with 200 ktl of PBS containing 1% BSA for 1 h at 37°C and then washed four times with 250 ~tl of PBS. 1 ng of bovine GIa-OC in 10 /~1 of PBS containing 1% BSA was added to the wells and then 100/~1 of a 1/1000 dilution of POD-labeled MAb (MAb2), corresponding to about 200 ng of antibody, in dilution buffer (PBS containing 1% BSA and 0.05% Tween 20) were added. The plate was incubated for 1 h at room temperature and then washed four times with PBS. The POD activity bound to the plate was assayed for 15 min at room temperature with a solution (100/~l/well) of 0.1 M citrate-0.2 M phosphate buffer, pH 5.0, containing 1 g / 1 0 P D and 0.01% (v/v) H202. The reaction was stopped by the addition of 100 /~l/well of 1 N H2SO4. The absorbance value at 492 nm was measured using a microplate reader (Titertek Multiskan, Model MCC, Flow Lab.).
19 G l a - O C specific E I A was p e r f o r m e d in a similar m a n n e r as above, b u t OC4-30 a n d O C G 4 were used as M A b l a n d M A b 2 , respectively, a n d the i n c u b a t i o n time of the i m m u n o l o g i c a l r e a c t i o n was increased to 1.5 h.
Competitive RIA for OC Serum s a m p l e s were assayed with the R I A kit f r o m C o m p a g n i e Otis Industries, S.A. ( G r e e n Cross Ltd., J a p a n ) a c c o r d i n g to the m a n u f a c t u r e r ' s directions. This kit is a c o m p e t i t i v e i m m u n o a s s a y using a p o l y c l o n a l a n t i b o d y d i r e c t e d a g a i n s t b o v i n e OC.
Serum samples B l o o d s a m p l e s were collected f r o m 30 h e a l t h y volunteers (14 w o m e n a n d 16 men; m e a n age 45 years, range 2 0 - 6 5 years), 30 p a t i e n t s with osteop o r o s i s (20 w o m e n a n d 10 men; m e a n age 58 years, range 4 8 - 7 5 years), eight p a t i e n t s with c h r o n i c renal failure (three w o m e n a n d five m e n ; m e a n age 48 years, range 2 5 - 6 2 years), a n d 20 p a t i e n t s with atherosclerosis (seven w o m e n a n d 13 men; m e a n age 54 years, range 3 5 - 7 3 years). A l l d o n o r s were fasting w h e n the collections were m a d e . T h e serum s a m p l e s o b t a i n e d were s t o r e d at - 4 0 ° C until analysis.
Results
Monoclonal antibodies F o u r M A b s (OC4-30 ( I g G 2 a ) , O C G 2 (IgG1), O C G 3 ( I g G 3 ) a n d O C G 4 ( I g G 1 ) ) were o b t a i n e d . T o d e t e r m i n e the antigenic e p i t o p e s on the O C m o l e c u l e recognised b y each M A b , c o m p e t i t i v e i n h i b i t i o n b y OC, or its fragments, of M A b b i n d ing to b o v i n e G l a - O C , i m m o b i l i z e d o n the p o l y styrene m i c r o t i t e r plate, was tested using the c o m petitive E L I S A . A s shown in T a b l e I, the b i n d i n g of every M A b to b o v i n e G l a - O C was i n h i b i t e d b y h u m a n G l a - O C . T h e b i n d i n g of O C 4 - 3 0 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l a - O C f r a g m e n t c o r r e s p o n d i n g to residues 8 - 3 1 , b u t was n o t inhibited by bovine Glu-OC, human Glu-OC or t r y p t i c f r a g m e n t 1 - 1 9 . T h e s e results suggested that OC4-30 recognized the residues 8 - 3 1 , which cont a i n e d three 3,-carboxyglutamic a c i d residues a n d t h a t O C 4 - 3 0 was specific for G l a - O C . T h e b i n d -
TABLE I EPITOPE SPECIFICITY OF THE MAbs INVESTIGATED OC and its fragments
Inhibition of MAb binding OC4-30 OCG2 OCG3 OCG4
(oc) Bovine Glu-OC Human GIa-OC Human Glu-Oc
+++ -
+++ +++ +++
(Fragments of bovine Gla-OC) a Acid-cleaved Frl-14 ND Acid-cleaved Fr15-48 ND ++ Tryptic Frl-19 Tryptic Fr21-42 Tryptic Fr45-49 ++ V8-digested Fr8-31 ++ -
+++ +++ +++
+++ +++ +++
+++ ++ +++
+ ++ -
a Percentage of A280compared with bovine Gla-OC as competitor in the competitive ELISA. Key: ND, not done; - , 70%.
ing of O C G 2 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l u - O C , h u m a n G l u - O C , a n d the acidcleavage f r a g m e n t c o r r e s p o n d i n g to residues 1 5 - 4 9 or t r y p t i c f r a g m e n t 4 5 - 4 9 . T h o s e results i n d i c a t e d t h a t O C G 2 r e c o g n i z e d the residues 4 5 - 4 9 b u t could not distinguish between Gla-OC and GluOC. T h e b i n d i n g of O C G 3 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l u - O C , h u m a n G I u - O C , the acid cleavage f r a g m e n t c o r r e s p o n d i n g to residues 1 5 - 4 9 , the t r y p t i c f r a g m e n t 2 1 - 4 2 , a n d the V8 p r o t e a s e d i g e s t e d f r a g m e n t 8 - 3 1 . These results showed the O C G 3 r e c o g n i z e d the residues 2 1 - 3 1 , which c o n t a i n e d two v - c a r b o x y g l u t a m i c acid residues, b u t c o u l d n o t d i s t i n g u i s h b e t w e e n G l a - O C a n d G l u - O C . T h e b i n d i n g of O C G 4 to b o v i n e G l a - O C was i n h i b i t e d b y b o v i n e G l u - O C , h u m a n G l u - O C , the a c i d cleavage f r a g m e n t c o r r e s p o n d ing to residues 1 - 1 4 , a n d the t r y p t i c f r a g m e n t 1 - 1 9 . T h e s e results s h o w e d that O C G 4 recognized residues 1 - 1 4 b u t c o u l d n o t distinguish b e t w e e n Gla-OC and Glu-OC.
Two-site binding assay I n o r d e r to find M A b p a i r s suitable for use in the o n e step s a n d w i c h E I A for OC, f o r m a t i o n of the s a n d w i c h c o m p l e x was d e t e c t e d in a two-site b i n d i n g assay. T h e f o u r P O D - l a b e l e d M A b s were p a i r e d in their i m m o b i l i z e d f o r m in all 16 p o s s i b l e
20 TABLE I1
TABLE III
TWO-SITE BINDING ASSAY FOR OC USING TWO DIFFERENT MONOCLONAL ANTIBODIES
RECOVERY TEST
MAbl
MAb2
0c4-30 OCG2 OCG3 OCG4
OC4-30
OCG2
OCG3
OCG4
++ -
+ -
++ +++
+++ -
Sample (ng/ml)
Added OC (ng/ml)
Theoretical value (ng/ml)
Measured Recovery value (%) (ng/ml)
A (7.4)
0 10 30
3.7 8.7 18.7 7.5 12.5 22.5
3.6 9.3 17.3 7.3 11.5 24.7
B (15.0)
Key: Mabl, capture MAb; MAb2, Tag MAb; - , 50 × background. c o m b i n a t i o n s a n d tested at a fixed c o n c e n t r a t i o n of bovine G l a - O C . As shown in T a b l e II, at least five pairs of M A b appeared to form a sandwich complex with b o v i n e G I a - O C a n d of these OC4-30 a n d O C G 4 (used as capture a n d detector M A b respectively) were selected for G l a - O C specificity a n d the highest a b s o r b a n c e r e a d i n g at a fixed c o n c e n t r a t i o n of b o v i n e G l a - O C . Moreover, this c o m b i n a t i o n of M A b showed the most sensitive detection limit (data n o t shown).
Assay validity of sandwich EIA Fig. 1 shows the s t a n d a r d curve generated b y this assay using b o v i n e G l a - O C as standard. The
1.5
coefficient of v a r i a t i o n (CV) at each dilution of the s t a n d a r d was less t h a n 10%. T h e sensitivity of the assay was calculated as the c o n c e n t r a t i o n which was two s t a n d a r d deviations (2SD) above the zero s t a n d a r d a n d for the s t a n d a r d b o v i n e G l a - O C was f o u n d the be 0.2 n g / m l . T h e precision of the assay was tested b y assaying three s e r u m samples ten times (within-assay) a n d in five consecutive assays (between-assay). T h e CV values were less t h a n 15%. I n experiments to d e t e r m i n e analytical recovery, two serum samples c o n t a i n i n g 7.4, a n d 15.0 n g / m l respectively were used. Each serum sample was assayed following the a d d i t i o n of b o v i n e G l a - O C s t a n d a r d (0, 10, 30 n g / m l ) a n d the calculated recoveries were 92-110% (Table III).
Specificity of the EIA system
q' ~'
/
16"01
0.5"
f
3.6)
i(( 8.8 ) (g.o) L917) 0.0
1=0
,
,
20
30
40
GLA-OC (NGIML)
Fig. 1. A standard curve for Gla-OC in the sandwich EIA. The coefficent of variation (CV%) at each concentration is shown in parentheses, and vertical bars indicate the range of withinassay variation (SD, n = 6).
T o establish that the materials detected in the s e r u m c o r r e s p o n d e d to O C a n d n o t to other serum proteins of different m o l e c u l a r mass, we fractionated a 1.0 ml aliquot of a serum sample c o n t a i n i n g a p p a r e n t l y high c o n c e n t r a t i o n s of O C b y passage through a Sephadex G - 1 0 0 c o l u m n . The resulting fractions were a n a l y z e d using the sandwich E I A (Fig. 2). I m m u n o r e a c t i v e materials were f o u n d in the fraction i n which a p r o t e i n of the same molecular weight as O C w o u l d be expected to be eluted. I n addition, a b o u t 20% of the total activity detected was f o u n d to be associated with fractions eluting at the void volume. Cross-reactivities in the present E I A system were e x a m i n e d using b o v i n e G I u - O C , h u m a n GlaOC, h u m a n G l u - O C a n d other "y-carboxyglutamic a c i d - c o n t a i n i n g s e r u m proteins (protein C, p r o t e i n S, p r o t h r o m b i n ) . N e a r l y 100% cross-reactivity was
21 10
VO
GLA-OC
SERUM GLA--OC 2O
2
(NG/ML)
4O
OSTEOPOROSlS N=30 1
.a
CHRONIC RENAL FAILURE
N----8
z v
?,
ATHEROSCLEROSIS N---- 20
I
.a ~9
NORMAL N : 30
....
10
20
30
40
_ Z60 2
50
....
o
70
FRACTION NO
Fig. 2. Location of fractions containing Gla-OC immunoreactivity after fractionation of a 1 ml serum sample on a Sephadex G-100 column. The sample was added to a 15 ×700 mm column equilibrated with 5 mM N H n C O 3 and eluted in 3 ml fractions with the same buffer, e, A280; ©, Gla-OC determined by EIA. Arrows indicate the void volume (Vo), and the elution position of purified Gla-OC.
observed for human Gla-OC, but no cross-reaction was observed with any of the other proteins tested. Correlation of EIA with RIA To evaluate the accuracy of the present sandwich EIA method, we used 58 serum samples to compare the values obtained by the EIA with those obtained by the commercially available RIA. As shown in Fig. 3, there was a good correlation between the two methods over the entire range
i
40
•
•
-.
o
I
I
20
I
40 RIA--OC
60 (NG/ML)
Fig. 3. Correlation between data from the new EIA system using two MAbs (EIA-OC), and those from the conventional RIA using a polyclonal antibody (RIA-OC). The first order regression line is illustrated, y = 2.896+0.759x, n = 58, r = 0.815 ( p < 0.001)
Fig. 4. Serum concentrations determined with the newly developed EIA system using two MAbs. Dashed lines show mean values.
studied. The correlation coefficient (r) was 0.815 ( p < 0.001). Serum concentration of Gla-OC The Gla-OC content of sera from healthy volunteers and patients are shown in Fig. 4. The serum Gla-OC values of the 30 healthy volunteers ranged from 0.5 to 9.7 n g / m l (3.6 + 2.19 (mean + SD)). Significantly elevated Gla-OC concentrations were observed in the sera of patients with osteoporosis (10.1 + 4.60, p < 0.001), chronic renal failure (20.3 + 16.56, p < 0.05), and atherosclerosis (8.3 _ 4.94, p < 0.01).
Discussion
We have been developing MAbs against bovine GIa-OC, and the four MAbs tested recognized four different epitopes on OC as revealed by the competitive ELISA (Table I). In particular OC4-30 recognized the Gla-form of OC but not the Gluform, and we developed the sandwich EIA using this MAb, which was specific for Gla-OC. Serum OC measurements are used in various clinical situations, including metabolic bone disease, renal disorders, and hyperthyroidism (Brown et al., 1984). RIAs for OC have been described based on the competitive RIA method using radio-iodinated OC and polyclonal antibody (Price et al., 1980). The inherent problems of RIAs include the relatively short half-life of the label, relatively long incubation times, and a requirement for specialized facilities and equipment. An EIA for OC using a MAb and a POD-labeled OC
22
was recently described (Power and Fottrell, 1989), but we believe that our newly developed one-step sandwich EIA system using a Gla-OC specific MAb has certain advantages. For example, the entire assay time is 2 h which is much shorter than conventional assay systems. Furthermore, sample volumes of 10/H were used without prior dilution, in contrast to most published assays in which larger volumes have been used. Several investigators have suggested that subforms of OC exist in human serum samples and that antibodies differ in their ability to recognize all of them (Gundberg and Weinstein, 1986; Power et al., 1989). It has been suggested that OC has only one major epitope near the C-terminus (Price and Nishimoto, 1980), where the amino acid sequence is identical in the bovine and human proteins but it has been suggested that there is another epitope located between positions 20 and 37 (Taylor et al., 1988). In our newly developed sandwich EIA, the binding site was located within the sequence 1-31, which contains three y-carboxyglutamic acid residues. It has been reported that the apparent concentration of OC in a serum sample, as measured by immunoassay, depends on the antibody used (Power et al., 1989). In this study, the EIA using monoclonal antibodies underestimated the OC concentration present by about 24% when compared with the conventional RIA with polyclonal antibody. However, there was a good correlation between the EIA and conventional RIA and the presence of a high-molecular-weight immunogen was confirmed by gel filtration of the serum on Sephadex G-100, as reported previously (Price et al., 1980; Power et al., 1989; Taylor et al., 1990). An explanation of this underestimation of OC concentration may be that the measured OC in the EIA is Gla-OC, whereas the polyclonal antibody based RIA detected not only Gla-OC but also a partly degraded or decarboxylated form of OC, i.e., total OC. The calcium-binding property of the vitamin K-dependent T-carboxyglutamic acid containing plasma proteins is dependent on the T-carboxyglutamic acid residue, which is necessary for biological activities such as activation of the blood coagulation cascade. The T-carboxyglutamic acid residue of the OC is necessary for the formation of a high affinity mineral-protein complex. Thus, the
Gla-OC must be an active form and it is suggested that the measurement of Gla-OC by our EIA may provide better clinical information than does the conventional assay with measures the total of active and inactive forms of OC. In this study, a significant increase in serum Gla-OC levels was observed in patients with chronic renal failure and bone disease, as reported previously (Price et al., 1980). Moreover, we have demonstrated for the first time a significant increase in serum Gla-OC levels in a patient with atherosclerosis. Levy et al. (1979, 1983) reported the accumulation of Glacontaining proteins, such as OC and atherocalcin, in calcified plaque. From these results, it was suggested that the circulating levels of GIa-OC might correlate with disease status, and could, therefore, be a valuable diagnostic parameter in atherosclerosis.
References Brown, J.P., Delmas, P.D. and Malaval, L. (1984) Serum bone Gla-protein: a specific marker for bone formation in postmenopausal osteoporosis. Lancet l, 1091-1093. Delmas, P.D., Stenner, D. and Wahner, H.W. (1983) Increase in serum bone T-carboxyglutamic acid protein with aging in women. J. Clin. Invest. 71, 1316-1321. Fraser, K.J., Poulsen, J.K. and Haber, E. (1972) Specific cleavage between variable and constant domains of rabbit antibody light chains by dilute acid hydrolysis. Biochemistry 11, 4974-4977. Galfr/~, C. and Milstein, C. (1981) Preparation of monoclonal antibodies: strategies and procedures. Methods Enzymol. 73, 2-46. Grundberg, C.M. and Weinstein, R.S. (1986) Multiple immunoreactive forms of osteocalcin in uremic serum. J. Clin. Invest. 77, 1762-1767. Gundberg, C.M., Hauschka, P.V., Lian~ J.B. and Gallop, P.M. (1984) Osteocalcin: isolation, characterization and detection. Methods Enzymol. 107, 516-544 Gundberg, C.M., Wilson, P.S., Gallop, P.M. and Parfitt, A.M. (1985) Determination of osteocalcin in human serum: results with two kits compared with those by a well characterized assay. Clin. Chem. 31, 1720-1723. Hillel, H. and Wu, C.W. (1977) Subunit topography of RNA polymerase from Escherichia coli. A cross-linking study with bifunctional reagents. Biochemistry 16, 3334-3342. Levy, R.J., Liam, J.B. and Gallop, P.M. (1979) Atherocalcin, a T-carboxyglutamic acid containing protein from atherosclerotic plaque. Biochem. Biophys. Res. Commun. 91, 41-49. Levy, R.J., Gundberg, C. and Scheinman. P. (1983) The identification of the vitamin K-dependent bone protein osteocalcin as one of the T-carboxyglutamic acid containing
23 protein present in calcified atherosclerotic plaque and mineralized heart valves. Atherosclerosis 46, 49-56. Malluche, H.M., Faugere, M.C. and Fanti, P. (1984) Plasma level of bone Gla-protein reflect bone formation in patients on chronic maintenance dialysis. Kidney Int. 26, 869-874. Nakane, P.K. and Kawaoi, A. (1974) Peroxidase-labelled antibody. A new method of conjugation. J. Histochem. Cytochem. 22, 1084-1088. Poser, J.W., Esch, F.S., Ling, N.C. and Price, P.A. (1980) Isolation and sequence of the vitamine K-dependent protein from human Bone. J. Biol. Chem. 255, 8685-8691. Power, M.J. and Fottrell, P.F. (1989) Solid-phase enzyme immunoassay for osteocalcin in human serum or plasma, with use of a monoclonal antibody. Clin. Chem. 35, 2087-2092. Power, M.J. Gosling, J.P. and Fottrell, P.F. (1989) Radoimmunoassay of osteocalcin with polyclonal and monoclonal antibodies. Clin. Chem. 35, 1408-1415. Price, P.A. (1984) Decalboxylation of T-carboxyglutamic acid residues in proteins. Methods Enzymol. 107, 548-551. Price, P.A. and Nishimoto, S.K. (1980) Radioimmunoassay for
the vitamin K-dependent protein of bone and its discovery in plasma. Proc. Natl. Acad. Sci. U.S.A. 77, 2234-2238. Price, P.A., Otsuka, A.S., Poser, J.W., Kristaponis, J. and Raman, N. (1976a) Characterization of -y-carboxyglutamic acid containing protein from bone. Proc. Natl. Acad. Sci. U.S.A. 73, 1147-1151 Price, P.A., Poser, J.W. and Raman, N. (1976b) Primary structure of the T-carboxyglutamic acid-containing protein from bovine bone. Proc. Natl. Acad. Sci. U.S.A. 73, 3174-3375. Price, P.A., Parthemore, J.G. and Deftos, L.T. (1980) New biochemical marker for bone metabolism. J. Clin. Invest. 66, 878-883. Tanaka, H., Kuwada, M., Shiraki, M. and Katayama, K. (1986) An enzyme immunoassay for osteocalcin. J. Immunol. Methods 94, 19-24. Taylor, A.K., Linkhadt, S.G., Mohan, S. and Baylink, D.J. (1988) Development of a new RIA for human osteocalcin: evidence for a midmolecule epitope. Metabolism 37, 872877.
