Sensitive enzyme immunoassay for human 28 kDa calbindin-D

Clinica Chimica Acta, 201 (1991) 183-192 .G 1991 Elsevier

ADONIS

CCA

Science Publishers

183

B.V.

All rights reserved OOW-8981/91/$03.50

(NH9898 19 IOf,1900

05061

Sensitive enzyme immunoassay for human 28 kDa calbindin-D Yuanyuan

Zhu ‘, Munehisa

Takashi

I, Koji Miyake

’ and Kanefusa

Kato 2

’ Department of UroloR): Nagoya UniL,ersitySchool of Medicine, Nagoya, and ’ Department of Biochemistry Institute for Dwelopmental Research, Aichi Prefectural Colony. Kasugai, Aichi (Japan) (Received

2 February

Key words: Calcium

1991; revision received

I June 1991;

binding protein; Vitamin

accepted 5 June 1991)

D-dependent:

Immunoassay

Summary

A sandwich-type enzyme immunoassay for human 28 kDa vitamin D-dependent calcium binding protein (calbindin-D) was established with a sensitivity of 1 pg/tube. Antisera were generated in rabbits injected with highly purified human kidney calbindin-D, and specific antibodies to calbindin-D were purified by the use of a column of calbindin-D-coupled Sepharose. The purified antibodies showed a single band at the position corresponding to calbindin-D on an immunoblotting test with a crude extract of human kidney. The assay system consisted of polystyrene balls with immobilized F(ab’), antibodies and the same antibodies labeled with /3-D-galactosidase from Escherichiu coli. The assay was specific to 28 kDa calbindin-D, showing no cross-reactivity with other calcium binding proteins such as S-lOOa,, ((Y(U), S-1OOb (@PI, parvalbumin and calmodulin. The assay was also reproducible (coefficients of variation between assays were < 10%). With the present method, immunoreactive calbindin-D could be detected in various human tissues, with major concentrations in kidney and brain. The values for immunoreactive calbindin-D in various body fluids of healthy subjects varied from undetectable in serum and semen to 3.8 A 2.0 (SD) pg/g creatinine in urine and 2.9 k 0.8 (SD) pg/l in cerebrospinal fluid. Immunohistochemically, the calbindin-D in human kidney was localized in epithelial cells of distal tubules.

Correspondence

and request for reprints

sity School of Medicine,

65 Tsuruma-cho,

to: Yuanyuan Showa-ku,

Zhu,

Department

Nagoya 466. Japan.

of Urology,

Nagoya Univer-

184

Introduction Human vitamin D-dependent calcium binding protein (calbindin-D) which is a member of the large family of intracellular calcium-binding proteins has two general types, differing predominantly in molecular size (28 kDa and 10 kDa) and in specific distribution [1,2]. In human kidney, however, calbindin-D has been purified and identified as the 28 kDa form [3-S], and studied immunohistochemitally [6,7]. Because existing assays for calbindin-D are insensitive [6,8-10,261, there is little information on its precise distribution in various human tissues, or its levels in body fluids. We now describe here a sensitive and specific immunoassay for human calbindin-D with a sensitivity of 1 pg/tube, which enables us to quantitate the calbindin-D in various tissues, and body fluids. Materials and methods Purification of human 28 kDa calbindin-D

The calbindin-D was purified from histologically normal human kidney tissues obtained at nephrectomy, according to the method described by Hitchman et al. [3] with modifications. After being cooled and recentrifuged, the heat-precipitated supernatant solution was dialyzed at 4’ C overnight against 0.02 mol/l Tris-HCl buffer, pH 7.4, containing 1 mmol/l EDTA, 0.07 mol/l NaCl and 0.01% 2mercaptoethanol, and then applied on a column (3.5 x 40 cm) of DEAE-Sephadex A50 equilibrated with the same buffer. The column was eluted with a linear gradient of NaCl (0.07 mol/l to 0.45 mol/l> in the buffer. The calbindin-D activity was eluted in the last minor peak. Fractions containing the activity peak were collected, dialyzed and reseparated on the column as before except that CaCl, was substituted for EDTA in the buffers. Fractions containing the activity peak were then further purified on the same column in the presence of EDTA as in step one. Fractions containing the activity peak were collected, concentrated and subjected to gel filtration on Sephadex G-100 (2.5 X 85 cm). As shown in Fig. lA, the final preparation revealed a single band on sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Antibodies

The antisera were generated in Japanese white rabbits by an injection of the purified calbindin-D (0.5 mg/rabbit) emulsified with an equal volume of Freund’s complete adjuvant. Similar immunizations were repeated 4 times every 2 wk. The rabbits were bled 1 wk after the last booster immunization. Monospecific IgG antibody was purified by the use of calbindin-D-coupled Sepharose 4B (Pharmacia Fine Chemical, Tokyo, Japan) as described previously [ill. About 20 mg of IgG antibody were obtained per 100 ml of the antisera. The specificity of the antibody preparation was confirmed by the immunoblotting test with a crude extract of human kidney. As shown in Fig. lB, the crude extract revealed an immunoreactive

185

Fig. 1. Sodium dodecyl-sulfate polyacrylamide gel electrophoresis of purified 28 kDa calbindin-D and a crude human kidney extract (A) and immunoblots with the purified anti-calbindin-D antibodies (B). Lane 1, standard proteins with molecular mass in kDa; Lane 2, 0.7 fig of purified calbindin-D; Lane 3, human kidney extract containing 0.35 pg of immunoreactive calbindin-D. A: polyacrylamide gel stained with Coomassie blue. B: nitrocellulose sheet stained with anti-calbindin-D.

single band at a position corresponding to calbindin-D. After digestion with pepsin (porcine stomach mucosa, from Sigma, St. Louis, MO, USA), the resultant F(ab’), fragments obtained from the purified antibody were used for preparing the immunoassay system [121. Immunoassay preparation and reagents

A sandwich enzyme immunoassay system was used with F(ab’), antibodies immobilized on to solid phase (polystyrene balls, 3.2 mm in diameter, Immuno Chemical, Okayama, Japan) noncovalently [13]. The same antibodies were labeled with P-D-galactosidase (from E. coli) by the use of o-phenylenedimaleimide [14,15]. Labeled antibody was quantitated by its galactosidase activity (1 U = 1 mol product under the conditions described below). The polystyrene balls with immobilized antibody were kept at 4°C at overnight before use in buffer A [O.Ol mol/l sodium phosphate buffers-, pH 7.0, containing 0.1 mol/l NaCI, 1 mmol/l MgCI,, 0.1% bovine serum albumin (demineralized, Organon Teknika, The Netherlands), and 0.1% NaN,]. The F(ab’), antibody was also labeled with horseradish peroxidase (HRP) by the use of N-succinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate [16], and the conjugate was used for immunoblotting test. Immunoassay conditions and procedures

In each test tube (10 x 75 mm>, a single polystyrene ball with immobilized antibodies was incubated at 30 ’ C with 10 ~1 of the sample or standard calbindin-D in a final volume of 0.5 ml with buffer B (0.01 mol/l sodium phosphate buffer, pH 7.0, containing 0.3 mol/l NaCl, 1 mmol/l MgCI,, 0.1% bovine serum albumin, 0.5% protease-treated gelatin [17], and 0.1% NaN,). After 4 h of incubation with

186

shaking, the reaction medium was discarded by aspiration, each ball was washed twice with 1 ml of chilled buffer A in each test tube, and then transferred to a fresh test tube (7.5 X 7.5 mm) containing 0.2 ml buffer A with 1 mU of the galactosidase-labeled F(ab’), antibody, and left standing at 5 o C overnight with shaking. Then the ball was washed as before and transferred to a fresh test tube (7.5 x 7.5 mm) containing 100 ~1 of buffer A. The galactosidase activity bound on the ball was assayed by incubation of the mixture with 50 ~1 of 0.3 mmol/l 4-methylumbelliferyl-/3-D-galactoside (Sigma) at 30 ’ C for 20 min with shaking. The reaction was terminated by adding 2.5 ml of 0.1 mol/l glycine-NaOH buffer (pH 10.3), and the fluorescence intensity of the 4-methylumbelliferone released was measured against a freshly prepared standard solution of 1 pmol/l 4-methylumbelliferone in the glycine-NaOH buffer. The wave-lengths used for excitation and emission analysis were 360 and 450 nm, respectively. Unless otherwise specified, all measurements were carried out in duplicate. Electrophoresis and immunoblot

Sodium dodecyl-sulfate (SDS)-polyacrylamide gel electrophoresis was performed according to Laemmli’s method [18] using 15% polyacrylamide slab gels, and proteins were visualized with Coomassie blue. Immunoblots were performed by a method similar to that of Towbin and associates [19]. In brief, the nitrocellulose sheet, on which proteins were electrophoretically transferred from polyacrylamide gels, was incubated in buffer B without NaN, for 30 min at room temperature, and then the HRP-labeled F(ab’), fragments of anti-calbindin-D antibody (0.06 pg/ml) were added for a further 60-min incubation. The sheet was washed with PBS buffer (50 mmol/l Tris-HCl, pH 7.5 containing 0.01% Nonodet P-40 and 0.9% NaCl), and the peroxidase activity visualized 3,3’-diaminobenzidine and hydrogen peroxide as described previously [20]. Tissue and body fluid samples

Various normal adult tissues were obtained at autopsy or surgical operation. Each tissue (0.3-0.5 g> was homogenized at 0’ C in 10 vol (v/w) of 50 mmol/l Tris-HCl buffer (pH 7.5) containing 5 mmol/l EDTA, using a Teflon-glass homogenizer. The homogenate was centrifuged at 100,000 X g at 4 o C for 1 h, and the soluble fraction was used for the analysis. Urine and serum samples of healthy subjects (60 adults and 30 children) were obtained at regular physical examinations from a local Institute for Public Health. Cerebrospinal fluid samples (6 adults and 4 children) were provided by the department of neurology in our hospital. The samples were taken for diagnostic purpose and found to be normal. Seminal samples of healthy subjects (5 adults) were obtained from healthy volunteers. All above samples were stored at -80” C until analysis. Immunoassays were performed in duplicate on 50 ~1 aliquots of body fluid samples and 10 ~1 portions

187

of tissue extracts, with kidney and brain extracts diluted 1: 100 with buffer B. The results were expressed as calbindin-D pg/g protein for tissues, calbindin-D pg/g creatinine for urine and calbindin-D pg/l for cerebrospinal fluid or for the precision of calbindin-D assay using urine. Other methods

Protein concentrations in extracts were determined by the Bio-Rad Protein Assay Kit which is based on dye-binding method of Bradford [21]. Concentrations of the purified 28 kDa calbindin-D were determined using the same method. The immunohistochemical staining of calbindin-D was performed by the indirect peroxidase-labeled antibody method with formalin-fixed, paraffin-embedded sections as described previously [22]. Determination of urinary concentrations of creatinine was carried out using an automated high performance liquid chromatography (HPLC) as described by Ogata and Taguchi [23]. After dilution lOO-fold with mobile phase [20 mmol/l NaH*PO,, pH 3.3, containing 3 mmol/l sodium 1-decanesulfonate]/CH,CN(BO/ 20), the urine samples were centrifuged at 4 ’ C at 2,000 X g for 5 min, and 10 ~1 of supernatant were used for HPLC. Human S-lOOa, (aa) and S-100b (p/3), used for the present study were those described previously [24]. Bovine calmodulin was obtained from Amano Pharmaceutical Co., Nagoya, Japan. Rat parvalbumin was purified from skeletal muscle as described by Endo et al. [25]. Results Specificity, sensitivity and precision of the immunoassay for human 28 kDa calbindinD

A standard curve of the assay of calbindin-D and the cross-reactivity with other calcium binding proteins are shown in Fig 2. The assay was specific to calbindin-D, showing no cross-reactivity with human S-lOOa, (CM>,S-100b (pp) rat parvalbumin and bovine calmodulin. The assay has a minimum detection limit of 1 pg/assay tube (the lowest amount giving a galactosidase activity significantly greater than that of the zero standard, P = 0.99). Within assay precision (n = 20) was tested with four urine samples and between assay precision with same samples in 20 consecutive assays (Table I>. Distribution of immunoreactive 28 kDa calbindin-D in various normal human tissues

Table II summarizes the data for concentrations of calbindin-D in soluble extracts of normal human tissues. Immunoreactive calbindin-D was present in large amounts in the kidney and brain, and in much smaller amounts in other tissues examined. In the kidney, concentrations of calbindin-D in the cortex were significantly higher than those of the medulla (P < 0.01).

Calcium-binding

proteins (P@

Fig. 2. Standard curve for the assay of calbindin-D (0) and its cross-reactivity with other calcium-binding proteins (0: human SlO&,, (aa). 0: human SlOOb (/.3p), A: rat parvalbumin. n : bovine calmodulin). A single polystyrene ball with immobilized antibodies was incubated in duplicate with the indicated amounts of the proteins assayed, and the polystyrene ball was then incubated with the same antibody Fab’ labeled with /3-D-galactosidase as described in the text.

Human 28 kDa calbindin-D levels in various normal body fluids

In the serum of healthy subjects (n = 20), no immunoreactive be detected.

calbindin-D could

TABLE I Precision of calbindin-D assay Assay

Urine

Within-

Sample Sample Sample Sample Sample Sample Sample Sample

Between-

I 2 3 4

1 2 3 4

No. of assays

Calbindin-D mean * 1 SD (pg/l)

CV (%)

20 20 20 20 20 20 20 20

0.5 f 0.04 3.3+0.14 4.9 + 0.40 10.0+0.97 0.5 * 0.04 3.OkO.20 4.5 rt 0.25 9.8tO.52

7.5 4.1 8.3 9.7 H.7 7.9 5.5 5.6

189 TABLE

II

Distribution

of 28 kDa

calbindin-D

listed, or actual observations

in various

normal

human

tissues. Median

values and ranges are

if n < 3

Tissue

Calbindin-D

n

(~g/g

protein)

median

range

Urogenital: renal cortex

10

854

(558

renal medulla

10

136

(75

testis other tissues *

-1348) -

288)

3

4.7

(2.3-

8.7)

I6

< I.0

(0.2-

0.8)

Non-urogenital: cerebral

I

cortex

124

small intestine

5

3.4

(1.7-

4.2)

adrenal

5

2.8

(2.7-

5.8)

stomach

4

2.1

(0.8-

4.3)

colon

4

2.1

(2.0-

2.8)

esophagus

7

2.0

6.2)

gall bladder

2

1.6,2.5

(0.3_

gland

appendix

I

2.0

adipose

2

0.5, 3.1

aorta

4

I.4

(0.4-

2.7)

pancreas

3

1.3

(0.4-

4.2)

bone (tibia)

1

I.3 WI-

1.0)

other tissues * * * prostate **

< I.0

28

3, urethra

3, urinary

heart 3. muscle 5. rectum

bladder

4, ureter

3, renal pevis 3.

1, lymph node 3, thyroid

3, skin I. lung 3, mammary

gland 1, spleen 6,

bronchus 2.

The levels of immunoreactive calbindin-D in normal cerebrospinal fluids were 2.9 f 0.8 pg/I for adults (mean f SD n = 6) and 2.9 k 0.9 pg/l for children (rz = 41, no age differences were found (P > 0.05). Table III summarizes data for the levels of calbindin-D in the urine of normal adults and children. No age and sex differences were found (P > 0.05). The average value of calbindin-D was 3.8 f 2.0 pg/g creatinine ( f SD, n = 90).

TABLE Urine

111 calbindin-D

concentrations

Age

No. of

W

samples

in healthy subjects Urine calbindin-D

(kg/g

Cr *

1(Mean

Male

Female 3.5 f 1.4

I-10

I5

3.9 -f 2.2

20-49

I5

4.2 + 2.7

3.8+2.0

50-79

15

4.0*

3.6*

a Urinary

creatinine.

2.1

I.8

*SD)

Fig. 3. Localization of calbindin-D in normal renal tissues. (A) In the cortex, calbindin-D was localized in the cytoplasm of epithelial cells of distal tubules only. (B) In the medulla, no positive cells for calbindin-D were detectable. (Indirect immunoperoxidase method, methyl green counterstain, x 115.)

In the semen of normal male adults (n = 51, no immunoreactive could be detected. Immunohistochemical

calbindin-D

localization of 28 kDa calbindin-D in human kidney

In the cortex, calbindin-D was immunostained in the cytoplasm of epithelial cells of distal tubules only (Fig. 3A). In the medulla, no calbindin-D-positive cells were detected (Fig. 3B). Discussion We report here the development of a sensitive sandwich-type enzyme immunoassay for human 28 kDa calbindin-D, with a detection limit of 1 pg/assay tube, l,OOO-fold more sensitive than the previously reported radioimmunoassay (RIA) [10,26]. The immunoassay system was monospecific to calbindin-D, showing no reactivity with other calcium-binding proteins such as S-lOOa, ((Y(W),S-1OOb (p/3), parvalbumin and calmodulin. The distribution of calbindin-D in normal human tissues was determined. Immunoreactive calbindin-D was present predominantly in human kidney and brain tissues as reported previously [4,9,10], but was widely distributed in all tissues examined in much smaller amounts. Parkes et al. [lo] reported a systematic survey of this protein in various human tissues. Their failure to detect it in many tissues examined was probably due to the insensitivity of their RIA method. The correlation between the widespread distribution of calbindin-D and receptors for 1,25-dihydroxyvitamin D remains to be determined. The concentration of immunoreactive calbindin-D in normal body fluids ranged from undetectable for serum and semen to mean value of 3.8 pg/g creatinine for urine and 2.9 Fg/l for cerebrospinal fluid. There were no age-linked differences were found in urine or cerebrospinal fluid, and no sex differences in urine levels.

191

Immunohistochemical staining of the human kidney showed calbindin-D in the epithelial cells of distal tubules (Fig. 3) as reported previously [6,7]. Approximately 84% of immunoreactive calbindin-D was present in the cortex, only 16% of that in the medulla, and very low concentrations the renal pelvis, ureter, urinary bladder, urethra and prostate (0.3-0.8 Fg/g protein). Although the testis gave a median value of 4.7 pg/g protein, no immunoreactive calbindin-D was detected in semen. These results support that urinary calbindin-D is derived mainly from distal renal tubules. References 1 Bruns ME, Kleeman E, Burns DE. Vitamin D-dependent calcium-binding protein of mouse yolk sac. Biochemical and immunochemical properties and responses to 1,25dihydroxycholecalciferol. J Bio Chem 1986;261:7485-7490. 2 Delorme AC, Danan JL, Mathieu H. Biochemical evidence for the presence of two vitamin D-dependent calcium-binding proteins in mouse kidney. J Bio Chem 1983;258:1878-1884. 3 Hitchman AJW, Kerr MK, Harrison JE. The purification of pig vitamin D-induced intestinal calcium binding protein. Arch Biochem Biophys 1973;155:221-222. 4 Morrissey RL, Bucci TJ, Empson RN, Lufkin EG. Calcium-binding protein: its cellular localization in jejunum, kidney and pancreas. Proc Sot Exp Biol Med 1975;149:56-60. 5 Morrissey RL, Rath DF. Purification of human renal calcium binding protein from necropsy specimens. Proc Sot Exp Biol Med 1974;145:699-703. 6 Roth J, Brown D, Norman AW, Orci L. Localization of the vitamin D-dependent calcium-binding protein in mammalian kidney. Am J Physiol 1982;243:243-252. 7 Orci L, Brown D, Roth J, Norman AW. Vitamin-D-dependent calcium binding protein content is reduced in kidneys of juvenile-onset diabetics. Lancet 1982;2:102-103. 8 Schneider LE, Wilson HD, Schedl HP. Effects of alloxan diabetes on duodenal calcium-binding protein in the rat. Am J Physiol 1974; 227:832-838. 9 Bruhlmann PM, Garcia SLM, Pizzolato GP, Norman AW, Orci L. Immunohistochemically detectable vitamin D-dependent calcium-binding protein is reduced in cerebellum of diabetic subjects. Diabetes 1984;33;917-922. 10 Parkes CO, Thomasset M, Baimbridge KG, Henin E. Tissue distribution of human calcium-binding protein (28,000 g mall’). Eur J Clin Invest 1984;14:181-183. 11 Kato K, Suzuki F, Semba R. Determination of brain enolase isozymes with an enzyme immunoassay at the level of single neurons. J Neurochem 1981;37:998-1005. 12 Kato K, Umeda Y, Suzuki F, Kosaka A. Interference in a solid-phase enzyme immunoassay system by serum factors. J Appl Biochem 1979;1:479-488. 13 Kato K, Hamaguchi Y, Okawa S, Ishikawa E, Kobayashi K, Katunuma N. Use of rabbit antibody IgG-loaded silicone pieces for the sandwich enzymoimmunoassay of macromolecular antigens. J Biochem 1977;81:1557-1566. 14 Kato K. Use of activated thiol-Sepharose in a separation method for enzyme immunoassay. Meth Enzymol 1983;92:345-359. 15 Kato K, Fukui H, Hamaguchi Y, Ishikawa E. Enzyme-linked immunoassay: conjugation of the Fab’ fragment of rabbit IgG with p-D-galactosidase from E. coli and its use for immunoassay. J Immunol 1976;116:1554-1560. 16 Yoshitake S, Imagawa M, Ishikawa E, Niitsu Y, Urushizaki I, Nishiura M, Kanazawa R, Kurosaki H, Tachibana S, Nakazawa N, Ogawa H. Mild and efficient conjugation of rabbit Fab’ and horseradish peroxidase using a maleimide compound and its use for enzyme immunoassay. J Biochem 1982;92:1413-1424. 17 Kato K, Umeda Y, Suzuki F, Kosaka A. Improved reaction buffers for solidphase enzyme immunoassay without interference by serum factors. Clin Chim Acta 1980;120:261-265.

192 18 Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680-685. 19 Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Nat1 Acad Sci USA 1979;76:4350-4354. 20 Haimoto H, Kurobe N, Hosoda S, Kato K. Sensitive enzyme immunoassay for human aldolase B. Clin Chim Acta 1989;181:27-36. 21 Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-254. 22 Takashi M, Haimoto H, Murase T, Mitsuya H, Kato K. An immunochemical and immunohistochemical study of SlOO protein in renal cell carcinoma. Cancer 1988;61:889-895. 23 Ogata M, Taguchi T. Simultaneous determination of urinary creatinine and metabolites of toluene, xylene, styrene, ethylbenzene and phenol by automated high performance liquid chromatography. Int Arch Occup Environ Health 1988;61:131-140. 24 Kato K, Suzuki F, Kurobe N, Okajima K. Enhancement of S-100 protein in blood of patients with Down’s syndrome. J Mol Neurosci 1990;2:109-113. 25 Endo T, Takazawa K, Kobayashi S, Onaya T. Immunochemical and immunohistochemical localization of parvalbumin in rat nervous tissues. J Neurochem 1986;46:892-898. 26 Christakos S, Friedlander EJ, Frandsen BR, Norman AW. Studies on the mode of action of calciferol. XIII. Development of a radioimmunoassay for vitamin D-dependent chick intestinal calcium-binding protein and tissue distribution. Endocrinology 1979;104:1495-1503.