development and characterization of a polyclonal antiserum-based radioimmunoassay for dog...

JOURNAL OF BONE AND MINERAL RFSEARCH Volume 8, Number 6, 1993 Mary Ann Liebert, Inc., Publishers

Development and Characterization of a Polyclonal Antiserum-Based Radioimmunoassay for Dog Osteocalcin

PAOLO FANTI,' GIOVANNA COLOMBO,' CHINGHUA YAO,I STEPHEN A. BROWN,' MICHAEL W. VERNON,' and HARTMUT H. MALLUCHEI

ABSTRACT

Determination of the serum concentration of the protein osteocalcin (OC) is useful for the noninvasive evaluation of bone metabolism. Because the dog is an excellent experimental model for the study of bone, we produced and characterized a polyclonal antiserum specific for dog OC and used it to develop a radioimmunoassay (RIA) for the measurement of the concentration of this protein in dog serum. The antiserum ex- presses higher affinity for Ca"-bound than for Ca'*-free OC (Bso at versus 2 x lo-' dilution). Also, in the presence of Ca'* affinity is higher for the carboxylated than for the decarboxylated form of the protein, and under Ca'*-free conditions the affinity is equal for the two forms. The study of peptide fragments of OC demonstrates competitive binding of the peptide comprising amino acids 20-44 but not of other fragments; this suggests that the antigenic epitope of dog OC is located in the midmolecular region of the protein. The RIA displays excellent sensitivity for the measurement of OC in blood (detection limit 0.31 ng/ml), with intraassay and interassay variations of 4.6 and 6.8%, respectively. Analysis of gel chromatography fractions of normal dog serum shows that greater than 90% of the antigenic material coelutes with purified radiolabeled dog OC. Test of parallelism reveals lack of interference of serum constituents with the binding assay. The antiserum displays limited species specificity since it cross-reacts with human OC, but not with the protein from rodents. Consistent with previous observations in other in vivo models, the serum concentration of OC in experimental dogs is decreased significantly 7-10 days after thyroparathyroidectomy and it is unchanged 1 month following ovariohysterectomy.

INTRODUCTION

STEOCALCIN (OC) also known as bone Gla protein, is a 0 bone-specific protein (molecular weight 5800) produced by the osteoblast, that is present in the circulation in nanomolar concentrations") and in the bone matrix in amounts equimolar to tropocollagen. 1 1 )

Over the last decade, assays have been developed for the measurement of circulating OC in several species. A posi- tive correlation has been described between the blood levels of this protein and histomorphometric parameters of bone turnover in such conditions as uremic osteodystro- phy,") osteoporosis,''.') and aging.I6)

We have established the dog model as a useful tool to study bone remodeling and diseases of bone remodeling.c'.8) For noninvasive evaluation of bone metabolism in dogs, the availability of a method for the determination of circulating OC in this species would be highly desirable.

For this reason, we recently purified and completely se- quenced OC from dog bone (see companion paper, p. 733), and we report here the development of a radioimmunoassay (RIA) for the measurement of OC in dog serum. Also, we report the application of this assay to monitor changes in mineral homeostasis following endocrine manipulations in dogs.

'Division of Nephrology, Bone and Mineral Metabolism, University of Kentucky and VA Medical Center, Lexington. 'Department of Biochemistry, University of Kentucky Medical Center, Lexington. 'Research Services, VA Medical Center, Lexington, Kentucky. 'Department of Physiology, University of Kentucky Medical Center, Lexington.

745

746

I

FANTI ET AL.

MATERIALS AND METHODS Materials

X-ray diffraction-defined hydroxyapatite (HAP) was the kind gift of Dr. E. Moreno (Forsyth Dental Center, Bos- ton, MA). la510dine was purchased from New England Nu- clear (Boston, MA), donkey antirabbit serum (lot 04810) from Pel-Freez Biologicals (Rogers, AZ), and polyethylene glycol (molecular weight 15,000-20,OOO) from Sigma Chemicals Co. (St. Louis, MO). Unless otherwise speci- fied, all other chemicals were from Sigma Chemical. Bea- gle dogs were purchased from Warren Hobble Beagles (Leesburg, OH), a USDA-licensed dealer. Dog feed (Purina Dog Chow) was from Ralston-Purina (St. Louis, MO), thiamylal sodium (Surital) from Parke-Davis (Morris Plains, NJ), and L-thyroxine sodium salt (Synthroid) from Boots Pharmaceuticals (Lincolnshire, IL).

Preparation of native and decarboxylated dog OC and peptide fragments

The native protein and peptide fragments were prepared as described in detail in the companion paper. Decarbox- ylated OC (protein with glutamic acid instead of y-carboxyglutamic acid residues) was prepared essentially as described by Poser and Price.(19) Purified dog OC (300 pg) was acid decarboxylated in 50 mM HCI at 110°C for 4 h under vacuum (HCI-treated OC). A second sample was subjected to the same procedure except for being handled under basic conditions (10 mM KOH) to avoid decarboxylation (KOH-treated OC). These OCs were purified by reversed-phase high-performance liquid chromatography and their integrity confirmed by reversed-phase HPLC and by amino acid composition.

Decarboxylation of the protein was verified by HAP binding studies, as described by Poser and Price,(9) with modifications. X-ray diffraction-defined HAP (400 pg/ml final concentration) was added in 50 pg/ml increments to solutions containing 35 pg/ml of HCI-treated and KOH- treated OC. After each addition, the solutions were incubated at room temperature on a rotating mixer, centri- fuged to pellet the HAP and the HAP-OC complex, and the supernatant sampled for OC concentration. Incremen- tal additions of HAP lead to a linear fall in OC in solution in the carboxylated protein sample (from 35 to 10 pg/ml), but they caused a modest change in OC in solution (from 33 to 29 pg/ml) in the acid-treated, decarboxylated protein sample.

Preparation of radiolabeled dog OC

Dog OC ( 5 pg) were radiolabeled with 1 mCi of T (12.9 mCi per pg iodine) by the lactoperoxidase method.(10’ The iodinated protein was purified by gel filtration on a Sepha- dex (3-50 column (0.8 x 15 cm) equilibrated with phosphate-buffered saline (PBS) EDTA buffer (10 mM phosphate, pH 7.4, 124 mM NaCI. 25 mM disodium EDTA, 0.1% bovine serum albumin (BSA), 0.1% Tween 20, and 0.02% NaN,) and by batch adsorption with HAP.‘L1)

I 100

80 - $ Y

60 F \

40 m

20

0

FIG. 1. Titration of antidog OC antiserum under different ionic conditions. Percentage binding (BIT) of radiolabeled OC to serially diluted antiserum 61-21690 in PBS- EDTA- (filled circles) or TBS-CaC1,-containing (open circles) reaction mixtures.

Preparation of the antisera The immunization protocol was essentially that of Woro-

bec et al.‘”) Purified dog OC (0.5-1 mg) was injected intra- dermally at monthly intervals into four rabbits. To this pur- pose, OC was dissolved in 120 mM phosphate buffer @H 7.4), mixed with four volumes of polyvinyl pyrrolidone (PVP-40), and emulsified in either complete (initial challenge) or incomplete Freund’s adjuvant. Serum samples were collected 1-3 weeks after each immunization and tested for antibody titer by equilibrium binding assays with 1251-OC. All four immunized rabbits produced antibody to OC, whereas antibodies were not detected in the serum of control rabbits. The effect of Ca’+ on the antiserum-ligand interaction was evaluated by titrating the antisera in tris-buffered- saline (TBS)-Ca’* buffer (25 mM Tris, pH 7.4, 124 mM NaCI, 2 mM CaCI,, 0.1% albumin, and 0.02% NaN,) and in PBS-EDTA. The serum with highest affinity and titer (61-21690) was used to develop the RIA described here.

Radioimmunoassay The assay was developed in PBS-EDTA buffer and per-

formed in duplicate, under equilibrium and nonequilibrium conditions, essentially as described by Price and Nishimoto.(l) The dog OC standards were prepared in RIA buffer from a stock of homogeneously pure protein, the concentration of which was determined by ultraviolet

RADIOIMMUNOASSAY FOR DOG OSTEOCALCIN 141

(UV) absorptiometry and amino acid analysis (see companion paper). In the equilibrium assay, borosilicate tubes received, in order of additions, 100 p1 standard (0-80 ng/ml) or serum sample (unknowns), 100 pl of "'I-OC (tracer, 20,000 cpm per tube), 100 pl of 1.5% normal rabbit serum, 100 pl antiserum 61-21690 (final dilution 1:15,ooO). and 200 pl assay buffer. To measure nonspecific binding, duplicate tubes received 100 pl of 1251-OC, 200 pl of 1.5% normal rabbit serum, and 300 pl assay buffer. After 24 h incubation at 4"C, rabbit antiserum was precipitated by the addition to each tube of 1 ml of 2070 donkey antiserum to rabbit IgG in precipitation buffer (0.1 M phosphate, pH 7.4, 2.5% polyethylene glycol, and 0.02% NaN,), followed by incubation for 2 h at 4°C and centrifugation at 1000 x g for 15 minutes. The pellets were washed twice with dis- tilled water and radioactivity counted on a LKB gamma- radiation counter (Wallac, Finland). In the nonequilibrium assay, the tubes containing OC standards or unknowns, normal rabbit serum, and antiserum were incubated for 24 h at 4OC before the addition of the tracer. Following an additional 24 h incubation at 4"C, the antiserum-bound OC was precipitated using the same procedure as just out- lined. The fraction of 1151-OC bound to antiserum B is defined as cpm in the precipitate minus cpm due to nonspecific binding divided by total cpm in the assay; Bo is the value of B when no unlabeled OC is present. To ascertain agreement between different dilutions, all unknown samples were tested undiluted and at 1:2, 1:4, and 1:8 dilutions.

Validation of the RIA

Cross-reactivity was tested by measuring the displacement of dog "'I-OC from antiserum 61-21690 with known

2o 1 Y

0.1 1 10 100

O s t e o c a l c i n ( n g / r n l )

amounts of standard-grade human and rat osteocalcin (Biomnedical Technologies, Stoughton, MA) and with serial dilutions of human, rat, mouse, rabbit, and dog serum. Potential nonspecific effects of dog serum on the RIA were tested by assessing the ability of the method to measure accurately OC in serially diluted sera and to detect the addition of known amounts of purified dog OC to these samples.

To confirm that the serum component recognized by antiserum 61-21690 is indeed OC and to exclude affinity of the antiserum for other dog serum constituents, we compared the gel filtration elution profile of purified dog OC and of the antigenic constituent(s) of dog serum. Dog serum (3-5 ml) and "'I-labeled dog OC (20,000 cpm) were chromatographed separately on a Sephadex G-100 column (1.6 x 80 cm) with 50 mM Tris, p H 8.0, as mobile phase. Fractions (4 ml) were collected and analyzed for content of radioactivity from exogenously added l*'I-OC, for dog OC and/or other proteins expressing affinity for the antiserum using our dog OC RIA, and for total protein content by spectrophotometry at 280 nm.

Determination of blood 1.25-dihydroxyvitamin D [I,ZS-(OH),D], 25-hydroxyvitamin D (25-OHD). ionized calcium, and creatinine

Levels of 1,25-(OH),D were determined in 2 ml serum samples by competitive receptor binding assay, as previ- ously de~cr ibed . ( '~ .~ ' ) 25-OHD was measured with a competitive protein binding assay that uses the vitamin D binding protein of diluted, unpurified rat serum as assay binder.(15) Ionized calcium was determined in heparinized samples of whole blood using a F2111Ca Selectrode ion- - ized calcium electrode (Radiometer, Copenhagen,

m"

m \

40

20

0 0.1 1 10 100

O s t e o c a l c i n ( n g / m l )

Den-

FIG. 2. Displacement of radiolabeled dog OC by carboxylated and decarboxylated ligand under different ionic conditions. Relative fraction (B/Bo) of radiolabeled OC bound to antiserum 61-21690 following incubation with increasing concentrations of KOH-treated (carboxylated, open circles), HCI-treated (decarboxylated, filled circles), and control (native carboxylated, triangles) ligand, in EDTA-containing (a) and Ca'+-containing (b) reaction mixtures.

748 FANTl ET AL.

mark). Phosphorus and creatinine were measured in serum by colorimetric techniques using an autoanalyzer (Techni- con, Tarrytown, NY).

Experimental protocol for the dog studies

Female beagle dogs were fed a balanced diet containing 1.3% calcium and 0.9% phosphorus and were enrolled in the studies only if serum parameters, including calcium, phosphorus, and creatinine, were in the normal range and a thorough veterinary checkup revealed no abnormalities.

A total of 27 dogs, mean age 4.6 f 0.4 years and mean weight 10.6 i 0.4 kg, were bled by venipuncture before and 7-10 days following thyroparathyroidectomy (TPTX, n = 17) or sham surgery (sham, n = lo), while fasting. The surgery was performed under general anesthesia induced with intravenous administration of thiamylal sodium at a dose of 200 mg/kg body weight. Success of the TPTX (i.e., absence of circulating parathyroid hormone, PTH) was ascertained by diminished levels of whole-blood ionized calcium (iCa) from baseline of at least 0.25 mmol/ liter during the 3 days following the surgery. After TPTX, life-threatening hypocalcemia was prevented by supple- menting the diet with calcium lactate powder (0.8-1.0 g daily of elemental calcium) to maintain serum ionized calcium between 0.75 and 1.0 mmol/liter. L-thyroxine was given orally at a dose of 20 &kg daily to the TPTX dogs only to maintain euthyroidism.'*.16)

A group of 12 dogs, mean age 2.2 f 0.3 years and mean weight 9.9 f 0.5 kg, were bled before and 30 days after ovariohysterectomy (OHX, n = 6) of sham operation (sham, n = 6). The conditions of the surgical procedures were as just described.

Statistics

Experiments for the characterization of the OC antiserum and for the development of the RIA were repeated at least twice. Results of the dog experiment are given as mean f standard error of the mean (SEM). Statistical analysis was performed using two-tailed paired statistics. All computations were performed using the SPSS-PC soft- ware package (SPSS, Inc., Chicago, IL) on an 1BM PC- A T computer.

RESULTS Characterization of rhe anriserum

The antiserum chosen for these studies (antiserum 61- 21690) displayed different binding affinities for dog OC depending on the presence of Ca'+ in the reaction mixture. In titration studies (Fig. l ) , half-maximal binding of "'1- OC was achieved with a 1:1OO,OOO dilution of the antiserum when the reaction was conducted in TBS-Ca'+ buffer and with a 1:2O,OOO dilution in PBS-EDTA buffer. Nonspecific binding was below 2% in both instances.

Displacement of tracer (**SI-radiolabeled, fully carboxylated OC) with standard amounts of native, KOH-treated (carboxylated) and HC1-treated (decarboxylated) protein led to different results depending on the reaction buffer used. In PBS-EDTA buffer there was minimal divergence of the isotherms and no difference in affinity to the antibody (Fig. 2a). In TBS-Ca'+ buffer the isotherm generated with decarboxylated standards was shifted to the right and divergent compared to the curves generated with KOH- treated or native protein (Fig. 2b).

Characterization of the RIA

Although absence of Ca2+ in the reaction mixture resulted in lower affinity of antiserum 61-21690 for OC, the RIA was developed in PBS-EDTA buffer because in this buffer the antiserum showed the same affinity for the carboxylated and decarboxylated forms of the protein.

The equilibrium and nonequilibrium assays yielded standard curves with excellent sensitivity for the measurement of OC levels in blood, the detection limit being 0.62 ng/ml with the equilibrium assay and 0.31 ng/ml with the nonequilibrium assay. The intraassay and interassay variations were 4.6 and 6.8%, respectively.

The assay allowed accurate quantitation of OC in serially diluted setum samples and total recovery of known amounts of purified OC that were added to these samples (Table 1).

Radiolabeled dog bone OC and immunoassayable dog serum OC coeluted in gel filtration chromatography, demonstrating comparable molecular size (Fig. 3). Further- more, analysis of all fractions from the gel filtration col-

TABLE 1 . MEASUREMENT OF OC IN SAMPLES PREPARED BY THE SERIAL DILUTION OF DOG SERUM AND BY THE ADDITION OF PURIFIED DOG

BONE OC TO THE DILUTED SERA

Added purrfed dog bone OC (nglrnl): measured OC (ng/ml)a

Serum dilutions 0 2.5 5.0

Whole serumb 16.4 f 0.5 18.6 i 0.2 21.1 f 0.4 1:2 serum 9.5 f 0.2 11.5 f 0.4 14.0 f 0.1 1:4 serum 5.0 f 0.3 7.3 f 0.1 10.1 f 0.2 1:8 serum 2.4 f 0.1 5.0 i 0.4 7.0 ? 0.1

aMean f SEM of triplicate measurements. bPooled serum from two mildly uremic dogs.


P Q I

C - 0)

c

0 L

a - m c

0 I-

7 / 6 > - - E .

3 0

C I

2 ‘ :

U - a

1 0 0

Q - u)

O O I . . . L . * . .

70 90 110 130 150 170 190210

E f f l u e n t ( m l )

FIG. 3. Gel filtration of dog OC from serum and bone. Undiluted dog serum or 1251-labeled dog bone OC were chromatographed on a Sephadex G-100 column (1.6 x 80 cm) equilibrated with 50 mM Tris, pH 8.0. The elution profile of the serum sample was monitored by UV spec- troscopy (open circles). The elution volumes of serum or bone O C were determined by nonequilibrium RIA (filled circles) and gamma counting (triangles), respectively.

umn with the OC RIA showed that serum proteins that did not coelute with OC did not bind significantly to O C antiserum (Fig. 3), thereby demonstrating a lack of nonspecific cross-reactivity of the OC assay with proteins of larger and smaller size than OC (Fig. 3).

The OC tryptic fragment comprising amino acids 20-44 displaced competitively radiolabeled, intact OC in the RIA (Fig. 4). Conversely, the C-terminal fragments containing residues 32-40, 41-49, and 45-49 did not displace the tracer but the N-terminal peptide 1-19 elicited noncom- petitive displacement of the tracer.

Purified human and rat OC displaced dog 12sI-OC from antiserum 61-21690 50 and 90% less efficiently than equiv- alent amounts of dog OC standard (Table 2). In the test of parallelism, dog OC standards and serial dilutions of human and dog sera (but not mouse and rabbit sera) displaced dog l’sI-OC from the antiserum in a parallel manner.

Effect of thyroparathyroidectomy and ovariohysterectomy on blood chemistry

As shown in Table 3, TPTX resulted in significant falls of circulating OC (6.2 f 0.5-45 f 0.8 ng/ml, p c 0.01), 1,25-(OH)1D (20.3 ? 3.0 to less than 2.0 pg/ml, p c 0.01), and iCa (1.23 f 0.02-0.91 f 0.04 mmol/liter, p c 0.01) and in the rise of phosphorus (3.5 f 0.4-5.0 f 0.4 mg/dl, p c 0.05). Serum creatinine and 25-OHD were un- affected by TPTX. Sham surgery resulted in no change of any of these parameters.

In the OHX experiment serum OC was not different before and 1 month after OHX (8.6 + 1.4 versus 9.9 f 0.8 ng/ml) or sham operation (9.5 f 2.1 versus 10.6 * 1.4 ng/ml).

1OC

90

h 80

70

60

OO” 50

v

\

m 40

30

20

1 10 100 lo00 0.1

O s t e o c a l c i n ( n g l r n l )

FIG. 4. Displacement of radiolabeled O C by pure intact protein and proteolytic fragments. Relative fraction of ”’I- labeled dog OC bound to antibody (B /Bo) at increasing concentrations of intact protein (filled circles), tryptic fragments 1-19 (triangles), 20-44 (open circles), and fragments 32-40, 41-49, and 45-49 (diamonds).

DISCUSSION

We report the development and characterization of a polyclonal antiserum (61 -21690) that has high affinity and specificity for canine O C and that interacts with its ligand in a unique manner, depending on the ionic conditions.

OC structure and/or overall charge are affected by Ca2+.(17-19) Binding of OC to various antibodies was pro- moted,‘18,20) not affected,(z’) and inhibited(”) by Ca”. Antiserum 61-21690 displays Ca’+-dependent affinity for OC, but in contrast to other antisera,11*~20) its affinity for the ligand is only partially dependent on the presence of calcium, since titration of the antiserum in the presence of excess EDTA reveals half-maximal binding at approxi- mately lo-‘ dilution.

Also, inclusion of Ca’+ in the reaction mixture often af- f e c t ~ ( ~ ~ . ’ ~ . ~ ’ ) but not universally(z1) the selective binding of antisera to the carboxylated and decarboxylated forms of the protein. The calcium-dependent variable affinity of antiserum 61-21690 for different forms of OC is qualita- tively unique in that presence of Caz+ is associated with higher affinity for the carboxylated than decarboxylated protein, but presence of excess amounts of a Ca’+ chelator, such as EDTA, eliminates differences in affinity.

Therefore, antiserum 61-21690 appears to express its highest affinity for the presumed biologically relevant form of OC, that is, the carboxylated and Ca’+-bound form. This feature, although trivial to developing an in

FANTI ET AL.

vitro RIA for the determination of the protein concentration in biologic fluid, may be useful for other in vivo and in vitro applications of the antiserum.

Antisera that bind to different regions of OC have been described. (12124) This suggests that OC may have more than one immunogenic epitope, possibly due to species- specific differences in protein structure and/or to technical differences in the development of the antisera. Since antiserum 61-21690 binds competitively to the tryptic fragment comprising residues 20-44 od dog OC but not to other fragments, it is likely that the antigenic epitope of dog OC includes the 7-carboxyglutamic acid-rich midmolecular region of the protein. Analysis of the dog O C sequence with the PC-GENE computer program (IntelliGenetics, Inc., Mountain View, CA) for the prediction of antigenic deter- minants supports this conclusion. This analysis indicated that the points of highest hydrophilicity (i.e., presumed highest antigenicity) of dog OC are in the midmolecular region of the protein and comprise residues 17-21 and 28- 31, with the 17-21 segment having the highest score (data not shown). Since the y-carboxyglutamic acid residues are likely t o undergo a change in charge upon binding to Ca2+, it is unclear whether the Ca'+-dependent variable affinity of the antiserum for O C should be attributed to changes in tertiary structure of the protein or rather to a change in its charge.

The RIA described here was developed in PBS-EDTA buffer because the antiserum 61-21690 recognizes equally well carboxylated and decarboxylated ligand under Ca2*-

TABLE 2. MEASUREMENT OF KNOWN AMOUNTS OF PURIFIED HUMAN AND RAT O C USING THE DOG RADIOIMMUNOASSAY

Measured OC (ng/ml)

Added OC (ng/ml) Human Rat

20 10 5

7.3 I .2 4.5 0.9 2.7 0.5

free conditions. Under these conditions, antiserum 61- 21690 is effective at dilutions similar to those of other antisera for human and murine OC.(21,24-16)

Antiserum 61-21690 appears to bind OC selectively since analysis of the gel chromatography fractions of normal dog serum showed that more than 90% of the antigenic material coeluted with purified radiolabeled dog bone OC. Consistent with the observation by others,(1,20) in our experiment a small amount of high-molecular-weight antigenic material elutes at the void volme. It is at present unclear whether this high-molecular-weight form of immuno- reactive O C represents aggregation of the protein or its complexation to a higher molecular weight protein(s) of serum.

The binding assay is not appreciably affected by serum constituents since 1:2 to 1:s dilutions of dog serum result in progressive linear decrease of measured OC, and purified bone OC added to serum samples is completely recov- ered by the assay. Also, antiserum 61-21690 is species specific, since it cross-reacts rather strongly with human OC but poorly or not a t all with rat, mouse, or rabbit OC.

OC synthesis is affected by manipulations of the calcio- tropic hormones both in vitro and in vivo. Challenge with a wide range of 1,25-(OH)1D doses results consistently in increased OC synthesis, but the response to PTH depends on the experimental model and protocol.

Acute administration of exogenous PTH to adult fe- m a l e ~ , ( ~ " to subjects with familial rickets,(a*) and to prep- arations of rat h i n d q ~ a r t e r s ' ~ ~ ) results in decreased serum levels of OC. Also, PTH suppresses the 1,25-(OH)2D-mediated stimulation of OC synthesis in cultures of human bone cells(3o) and fetal rat calvariae.(3') However, PTH increases the steady-state levels of OC mRNA and the abun- dance of the protein in the culture medium of rat osteoblast-like cells. 02) In humans, hyper- and hypoparathyroidism are associated with high and low serum levels of OC, respectively. (3.15.3J)

Here we report decreased serum levels of OC in dogs 7-10 days following TPTX. These results agree with data in humans with normal renal and confirm our previous observation of decreased serum O C in patients with end-stage renal disease who underwent parathyroidectomy for correction of severe secondary hyperparathyroid- ism.(35) Whether the fall in serum OC is directly or indi-

TABLE 3 . EFFECT OF TPTX OR SHAM OPERATION ON CIRCULATING OC AND OTHER SERUM CHEMISTRIES IN ADULT FEMALE BEAGLE DOGS

TPTX dogs Sham dogs

Baseline Postsurgery Baseline Postsurgery

OC, ng/ml 6.2 f 0.5 4.5 f 0.8a 6.3 f 0.4 5.9 f 0.5 1,25-(OH),D, pg/ml 20.3 f 3.0 < 2a 21.1 f 4.2 22.0 f 2.0 25-OHD, ng/ml 19.1 f 4.1 22.2 f 4.4 20.0 f 2.8 24.5 f 3.3 iCa, mmol/liter 1.23 f 0.02 0.91 f 0.04a 1.23 f 0.03 1.25 f 0.05

Creatinine, mg/dl 0.6 f 0.1 0.6 f 0.1 0.6 f 0.1 0.6 f 0.1 Phosphorus, mg/dl 3.5 f 0.1 5.0 f 0.4b 3.4 f 0.4 3.5 f 0.2

aDifferent from baseline (p < 0.01). bDifferent from baseline (p < 0.05).


rectly related to lack of PTH awaits further studies. The effect of hypoparathyroidism on OC could be mediated by alterations in serum 1,25-(OH),D, since 1,25-(OH),D has a direct stimulatory effect on O C ~ y n t h e s i s ' ~ ~ . ~ ' ) and fell to undetectable levels in our dogs after TPTX.

Also, the interesting observation of a fall in O C of only 27% after TPTX, while PTH and 1,25-(OH),D were virtu- ally absent in the circulation, deserves further studies. This may be due to production of OC that escapes regulation, to the presence of as yet unidentified factors that regulate O C synthesis, or to insufficient elapsed time after TPTX for O C to reach trough levels. In our previous study in patients with renal failure, 2 weeks elapsed after parathyroidectomy before O C started to fall.(35)

Serum OC was unchanged in our experimental dogs 1 month after OHX. This is consistent with previous observations in ovariohysterectomized women,(3*) pharmaco- logically castrated monkeys, u9) and ovariectomized beagle dogs.'") Also, during the first 4 weeks after OHX we found histologically only indirect signs of hyperresorption and no changes in osteoblastic parameters.'") Thus the absence of changes in osteocalcin is not surprising. The study by Dannucci et al,(a) which measured OC biweekly in ovariectomized beagles using a rat OC RIA with favorable c ro~s- reac t iv i ty ,~~~) found a significant rise of serum OC starting 8 weeks after castration. This is consistent with observations in other ~ p e c i e s ( ~ ' . ~ ~ . ~ ~ ) in which serum OC rose 2-6 months after cessation of ovarian function.

ACKNOWLEDGMENTS

We thank Mr. Ken Westberry and Mrs. Louise Tipton for invaluable technical and secretarial help. This work was supported by VA Medical Research Funds (RAG) to P. Fanti, NIH Grant No. AR 35837-05 to H.H. Malluche, and NSF Grant No. EHR-9108764 to G. Colombo.

REFERENCES

1. Price PA, Nishimoto SK 1980 Radioimmunoassay for the vitamin K-dependent protein of bone and its discovery in plasma. Proc Natl Acad Sci USA 77:2234-2238.

2. Hauschka P 1986 Osteocalcin: The vitamin K-dependent CAI'-binding protein of bone matrix. Haemostasis 16:258- 269.

3. Malluche HH, Faugere MC, Fanti P , Price P 1984 Plasma levels of bone Gla-protein reflect bone formation in patients on chronic maintenance dialysis. Kidney Int M:869-874.

4. Brown J P , Malaval L, Chapuy MC, Delmas PD, Edouard C, Meunier PJ 1984 Serum bone Gla-protein: A specific marker for bone formation in postmenopausal osteoporosis. Lancet May 19:1091-1093.

5 . Delmas PD. Malaval L. Arlot ME, Meunier PJ 1985 Serum bone Gla-protein compared to bone histomorphometry in endocrine diseases. Bone 6:339-341.

6. Delmas PD. Stenner D, Wahner HW, Mann KG. Riggs BL 1983 Increase in serum bone gamma-carboxyglutamic acid protein with aging in women. J Clin Invest 71:1316-1321.

7. Malluche HH, Sherman D, Meyer W, Ritz E, Norman AW, Massry SG 1982 Effects of long-term infusion of physiologic doses of 1-34 PTH on bone. Am J Phvsiol 242:F197-F201.

8. Malluche HH, Matthews C, Faugere MC, Fanti P , Endres DB, Friedler RM 1986 1,25-dihydroxyvitamin D maintains bone cell activity, and parathyroid hormone modulates bone cell number in dogs. Endocrinology 119:1298-1304.

9. Poser JW, Price PA 1979 A method for decarboxylation of gamma-carboxyglutamic acid in proteins. J Biol Chem 254:

10. Marchalonis J 1969 An enzymatic method for the trace iodi- nation of immunoglobulins and other proteins. Biochem J

11. Price PA, Epstein DJ, Lothringer JW. Mashimoto SK, Poser JW, Williamson MK 1979 Structure and function of the vitamin K-dependent protein of bone. In: Suttie J W (ed.) Vita- min K Metabolism and Vitamin K-dependent Proteins. Uni- versity Park Press, Baltimore, pp. 219-226.

2. Worobec RB. Wallace JH, Huggins CG 1972 Angiotensin- antibody interaction. I . Induction of the antibody response. Immunochemistry 9:229-238.

3 . Reinhardt TA, Horst RL, Orf JW, Hollis BW 1984 A microassay for 1,25-dihydroxyvitarnin D not requiring high performance liquid chromatography: Application to clinical studies. J Clin Endocrinol Metab 58:91-98.

4. Baker LRI, Abrams SML. Rose CJ , Faugere M-C. Fanti P, Subayti Y, Malluche H H 1989 I ,25(OH)1D, administration in moderate renal failure: A prospective double-blind trial. Kidney Int 35561-669.

IS. Arnaud SB, Meger J 1983 Microassay for 25-hydroxyvitamin D: Method and interpretation. In: Bikle DD (ed.) Assay of Calcium-Regulating Hormones. Springer-Verlag, New York,

16. Hightower D. Kyzar JR, Chester DK, Wright EM 1973 Re- placement therapy for induced hypothyroidism in dogs. J Am Vet Med Assoc 163:979-980.

17. Hauschka PV, Carr SA 1982 Calcium-dependent alpha-heli- cal structure in osteocalcin. Biochemistry 21:2538-2547.

18. Delmas PD. Stenner DD, Romberg RW, Riggs BL, Mann KG 1984 lmmunochemical studies of conformational alterations in bone gamma-carboxyglutamic acid containing protein. Biochemistry 23:4720-4725.

19. Wians FH Jr , Krech KE, Hauschka PV 1983 Effects of magnesium and calcium on osteocalcin adsorption to hydroxyapatite. Magnesium 2:83-92.

20. Tracy RP, Andrianorivo A. Riggs BL, Mann KG 1990 Com- parison of monoclonal and polyclonal antibody-based immu- noassays for osteocalcin: A study of sources of variation in assay results. J Bone Miner Res 5451-461.

21. Taylor AK, Linkhart SG, Mohan S, Baylink DJ 1988 Devel- opment of a new radioimmunoassay for human osteocalcin: Evidence for a midmolecule epitope. Metabolism 37:872- 877.

22. Gundberg CM, Hauschka PV, Lian JB, Gallop PM 1984 Osteocalcin: Isolation, characterization, and detection. Methods Enzymol 1075 16-544.

23. Price PA, Lothringer JW, Nishimoto SK 1980 Absence of the vitamin K-dependent bone protein in fetal rat mineral. J Biol Chem 255:2938-2942.

24. Jiippner H , Schettler T, Giebel G, Wenner S, Hesch RD 1986 Radioimmunoassay for human osteocalcin using an antibody raised against the synthetic human (h37-49) sequence. Calcif Tissue Int 39:310-315.

25. Price PA, Parthemore JG, Deftos LF, Nishimoto SK 1980 New biochemical marker for bone metabolism. Measurement by radioimmunoassay of bone Gla protein in the plasma of normal subjects and patients with bone disease. J Clin Invest

431-436.

113:299-305.

pp. 65-83.

66:878-883. 26. Gundberg CM, Wilson PS, Gallop PM, Parfitt AM 1985 De-

752 FANTl ET AL.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

termination of osteocalcin in human serum: Results with two kits compared with those by a well-characterized assay. Clin Chem 31: 1720- 1723. Riggs BL, Tasi K, Mann KG 1986 Effect of acute increases in bone matrix degradation on circulating levels of bone-gla protein. J Bone Miner Res 1:539-541. Cole DEC, Gundberg CM 1985 Changes in serum osteocalcin associated with parathyroid hormone infusion in X-linked hypophosphatemic rickets. Clin Chim Acta 151:l-7. Calvo MS, Fryer MJ, Laakso KJ, Nissensen RA, Price PA, Murray TM, Heath H 1985 Structural requirements for parathyroid hormone action in mature bone. J Clin Invest 87:

Beresford JN, Gallagher JA, Poser JW. Russell RGG 1984 Production of osteocalcin by human bone cells in vitro. Ef- fects of 1 ,25(OH),D3, 24,25(OH)1D,, parathyroid hormone, and glucocorticoids. Metab Bone Dis 5:229-234. Lian JB, Courts M, Canalis E 1985 Studies of hormonal regulation of osteocalcin synthesis in cultured fetal rate calvariae. J Biol Chem 260:8706-8710. Noda M, Yoon K. Rodan GA 1988 Cyclic AMP-mediated stabilization of osteocalcin mRNA in rat osteoblast-like cells treated with parathyroid hormone. J Biol Chem 263:18574- 18577. Delmas PD, Wahner HW, Mann KG, Riggs BL 1983 Assess- ment of bone turnover in postmenopausal osteoporosis by measurement of serum bone Gla-protein. J Lab Clin Med 102:470-479. Deftos LJ, Parthemore JG, Price PA 1982 Changes in plasma bone Gla protein during treatment of bone disease. Calcif Tissue Int 34:121-124. Fanti P , Smith AJ, Price PA, Reitz RE, Malluche HH 1986 Effects of parathyroidectomy on circulating levels of la.25- dihydroxyvitamin D and bone Gla protein in dialyzed patients. J Clin Endocrinol Metab 62:869-873. Price PA, Baukol SA 1980 1,25-Dihydroxyvitamin D, in-

2348-2354.

creases synthesis of the vitamin K-dependent bone protein by osteosarcoma cells. J Biol Chem 255:11660-11663.

37. Yoon K, Rutledge SJC, Buenaga RF, Rodan G A 1988 Char- acterization of the rat osteocalcin gene: Stimulation of pro- moter activity of 1,25-dihydroxyvitamin D,. Biochemistry

38. Pacifici R, Brown C, Puscheck E, Friedrich E, Slatopolsky E, Maggio D, McCracken R , Avioli LV 1991 Effect of surgical menopause and estrogen replacement on cytokine release from human blood mononuclear cells. Proc Natl Acad Sci

39. Mann DR, Could KG, Collins DC 1990 A potential primate model for bone loss resulting from medical oophorectomy or menopause. J Clin Endocrinol Metab 71:105-110.

40. Dannucci GA, Martin RB, Patterson-Buckendahl P 1987 Ovariectomy and trabecular bone remodeling in the dog. Calcif Tissue Int 40:194-199.

41. Monier-Faugere MC. Friedler RM, Bauss HF, Malluche HH 1992 A new biphosphonate, BM21-9055, prevents bone loss occurring after cessation of ovarian function in experimental dogs (abstract). J Bone Miner Res 7(Suppl. l):S277.

42. Patterson-Allen P. Brautigam CE, Grindeland RE, Asling CW. Callahan PX 1982 A specific radioimmunoassay for osteocalcin with advantageous species cross-reactivity. Anal Biochem 12O:l-7.

27~8521-8526.

USA 88:5134-5138.

Address repr int requests to: Paolo Fanti. M . D .

Division of Nephrology, Bone and Mineral Metabolism

University of Kentucky Medical Center Lexington, K Y 40536

Received in original form July 13, 1992; in revised form Novem- ber 9, 1992; accepted December 5 , 1992.

development and characterization of a polyclonal antiserum-based radioimmunoassay for dog...

Documents