Radioimmunoassay of Aspartate ... - Clinical Chemistry

4 downloads 544 Views 1017KB Size Report
We describe the development of a sensitive, specific radioim- munoassay ... serum. Over the last two decades, radioimmunoassay. (RIA) has been widely used.
CLIN.

CHEM.

30/8,

1361-1365

(1984)

Radioimmunoassay Serum Fred V. Leung,

of Aspartate Aminotransferase

Anne E. Niblock,

lsoenzymes

in Human

and A. Ralph Henderson

We describe the development of a sensitive, specific radioimmunoassay for the cytoplasmic and mitochondrial isoenzymes of human aspartate aminotransferase (L-aspartate:2oxoglutarate aminotransferase; EC 2.6.1.1). Isoerizymes from human heart tissue were purified to homogeneity and used to raise high-titer antisera in rabbits. We partly purified the antisera by selective column chromatography. The Bolton-Hunter reagent was used to radioiodinate the isoenzymes. The assay requires 100 L of serum, includes a

ford, IL 61105. We coated on cellulose antibody (Wellmark 7C7). Bovine serum St. Louis, MO 63178. Scientific Co., Don higher purity grade.

solid-phase

Total AST activity was measured at 37#{176}Caccording to the modified IFCC method (11), with a kit (Boehringer Mannheim Canada, St. Laurent, Quebec H4R 1V8), in the absence of exogenous pyridoxal phosphate, with an Model 8600 Reaction Rate Analyzer (LKB, Bromma, Sweden). Pyridoxal phosphate activation was not used in our study because it reportedly results in wide variations in activation of samples from different patients with heart disease (12). Isoenzyme activity was determined for each specimen by a modified immunoprecipitation technique (8). We added 100 /LL of specific antisera to c-AST or m-AST isoenzymes, as described below, to 100 L of the serum sample. The mixture was briefly vortex-mixed, then incubated at 37#{176}C for 1 h, centrifuged (1500 x g, 30 miii, 4#{176}C), and 50 L of the supernatant fluid was assayed for AST activity as described above for total AST activity. This remaining activity represented the isoenzyrne fraction that was not inhibited by the added antisera.

second-antibody

separation,

and

can

be com-

pleted in less than 3 h. There was no cross reactivity between the two isoenzymes. As little as 5 .tg (50 pmol) of each aspartate

aminotransferase

can

be measured

per

liter of

serum. Over the last two decades, radioimmunoassay (RIA) has been widely used in clinical laboratories to measure low (picogram) concentrations of hormones and drugs, to assist diagnosis, or to monitor patient therapy (1). Application of RIA to clinical enzymology has been restricted to a few enzymes such as prostatic acid phosphatase (EC 3.1.3.2) (2), amylase (EC 3.2.1.1) (3), creatine kinase (EC 2.7.3.2) isoenzymes (4,5), trypsin (EC 3.4.2 1.4) (6), and chymotrypsin (EC 3.4.21.1) (7). Immunological methods have been developed for measuring aspartate aminotransferase (L-aspartate:2oxoglutarate aminotransferase; EC 2.6.1.1; AST) isoenzymes in human serum and tissues (8,9), but many of them lack the sensitivity and precision of RIA, qualities illustrated by an RIA technique developed for quantifying AST isoenzymes in chicken tissue (10). We report here the development of two specific, sensitive RIA techniques for the independent determination of the mass concentrations of cytoplasmic (c-AST) and mitochondrial (m-AST) aspartate aminotransferase isoenzymes from human serum. The direct quantification of the mass of these isoenzymes is compared with an immunoinhibition assay (8), which measures the catalytic activity of the two.

Materials

and Methods

Protein-A-Sepharose CL-4B, Sephadex G-50, DEAE-Sephacel, and CM-Sephadex C-50 were obtained from Pharmacia (Canada) Inc., Dorval, Quebec H9P 9Z9, and DEAE Afil-Gel Blue was from Bio-Rad Laboratories (Canada) Ltd., Mississauga, Ontario TAX 2C8. The Bolton and Hunter reagent for protein iodination (--200O kCi/mol) was from Amersham, Oakville, Ontario L6H 2R3); iodine-125, 2 mCi in 100 L of 0.1 jimol/L NaOH, was from New England Nuclear, Lachine, Quebec H8T 3C9; and 1,3,4,6-tetrachloro-3a,6a-diphenylglycoluril (“lodogen”) and N-chlorobenzenesulfonamide on polystyrene beads (“lodo-beads”) were from Pierce Chemical Co., Rock-

Department of Clinical Biochemistry, University Hospital (University of Western Ontario), P.O. Box 5339, Postal Stn. A, London, Ontario, Canada N6A 5A5. Received April 16, 1984; accepted May 22, 1984.

Assay

for AST

Preparation

used donkey antibody to rabbit IgG, (“Sac-Cel”), as the solid-phase second Diagnostics Ltd., Guelph, Ontario NiH albumin was from Sigma Chemical Co., All other chemicals were from Fisher Mills, Ontario M3A 1A9, in reagent or

Catalytic

of Purified

Activity

Heart

AST

isoenzymes

Human heart tissue, obtained at autopsy, was used to prepare c-AST and m-AST isoenzymes as described previously (13). The isoenzymes were purified to homogeneity by affinity chromatography and isoelectric focusing as described elsewhere (14). The specific activities of each of these isoenzymes was 150 kU per gram of protein.

Preparation

of Antisera

Antibodies to c-AST and m-AST isoenzymes were separately prepared in rabbits. We made four subcutaneous injections of 1 mg/mL solutions of each isoenzyme in Freund’s complete adjuvant, at weekly intervals. After the fifth week, the rabbits were bled weekly and the antisera tested for antibody specificity. We had two successes out of three rabbits for each isoenzyme.

Purification

of Antisera

“Protein A” from Staphylococcus aureus acts as a broadrange IgO binding reagent, and when attached to Sepharose CL-4B it forms an affinity-chromatography medium (15). Rabbit antisera specific for c-AST or m-AST was equilibrated against phosphate buffer (100 mmol/L, pH 8.0) and added to a 9 mm x 150 mm column of protein A-Sepharose with the same buffer medium. The specific antibodies were eluted with a step-wise gradient of 100 mmol/L citrate buffer between pH 3-7, in steps of 0.5 pH unit. The fractions were monitored by their absorbance at 280 nm and by the immunoprecipitation technique. CLINICAL

CHEMISTRY,

Vol. 30, No. 8, 1984

1361

In a second procedure for IgG purification we used DEAEBlue, which binds most serum proteins except for IgG and transfer-in (16). A 13 mm x 230 mm column of this gel was equilibrated with Tris HC1 buffer (20 mmol/L, pH 8.0) containing 28 mmol of NaC1 and 20 mg of NaN3 per liter. The antiserum was loaded onto the column and eluted with a 50-500 mmol/L NaC1 gradient. Another procedure used to isolate and purify the IgG from rabbit antisera involved precipitation with caprylic acid (17): serum components other than IgG, ceruloplasmin, and some IgA are precipitated with caprylic acid and can then be removed by centrifugation at 4000 x g, 4#{176}C(17). The supernate containing the IgG components was extensively dialyzed against isotonic saline. The concentrated solution was then passed through a 15 mm x 70 mm column of DEAE-Sephacel, equilibrated and eluted with 15 mmol/L acetate buffer, pH 5.7. The IgG was not adsorbed onto the DEAE-Sephacel and was recovered, purified, in the void volume of the eluate. Affi-Gel

lodination

of c-AST

and m-AST

Isoenzymes

lodo-gen, pre-coated on polypropylene tubes as described (18), was reacted with about 20 j.g of isoenzyme in 30 L of sodium borate buffer (0.5 mol/L pH 8.5) and an aboutequivalent amount of Na1251. The mixture was agitated gently during a 10-mm incubation at 4#{176}C. The reaction was stopped by removing the sample from the reaction tube. We used lodo-beads to iodinate the enzymes in 100 minol/L Ti-is buffer, pH 7.0(21 #{176}C). The sample was removed from the beads to stop the reaction. Free unreacted Na’I was separated from enzyme-bound iodine by passing the mixture through Sephadex G-50 (Pharmacia). Radioiodination of the isoenzymes with the Bolton and Hunter reagent was as described elsewhere (19). The reagent supplied (1 mCi) in benzene was evaporated under a stream of nitrogen. Approximately 5 g of c-AST or m-AST in 10 L of 0.1 moIJL borate buffer, pH 8.0 (at 4 #{176}C), was added to the residue of dried iodinated ester, and the reaction mixture was stirred for 15 mm at 4 #{176}C. Then 200 tL of 400 mmol/L glycine in 100 mmol/L borate, pH 8.0 (at 4 #{176}C), was added and the incubation was continued for 5 mm, to trap unreacted ester. The ‘251-labeled isoenzyme was separated from low-molecular-mass labeled compounds by passing the mixture through a DEAE-Sephacel or CM-Sephadex C-50 column, to recover the m-AST or cAST labeled antigen, respectively (14). The labeled antigens were diluted with Tris HC1 buffer (10 mmoIJL, pH 7.8, containing 500 mg of bovine serum albumin and 20 mg of sodium aside per liter) to give between 35 000 and 40 000 counts/mm per 100 L. This solution, stored at 4#{176}C, was used for up to four weeks.

Determination

of Antisera

Avidity

Antisera to either c-AST or m-AST were titrated in dilutions up to 1/256 000 against 100 p.L of the radiolabeled isoenzymes. The free and bound ‘I-labeled antigens were separated by precipitating the antigen-antibody complex with the Sac-Cel donkey anti-rabbit IgG antiserum. The antiserum dilution used for c-AST was 5000-fold and for mAST was 18000-fold to obtain -50% binding of the total radioactivity. We used 100 L of this antiserum dilution in the RIA procedure.

Procedure

for Radioimmunoassay

The incubation mixture contained, per liter, 10 mmol of Ti-is HCI buffer (pH 7.8), 500 mg of bovine serum albumin, and 20 mg of sodium azide. Table 1 gives the protocol for

1362

CLINICAL CHEMISTRY, Vol. 30, No. 8, 1984

Table

1. Protocol Step

AlA

Volume,

1. Reaction mixture 2. Mix; room 3. Add 4. Mix; room 5. Add

for

of AST lsoenzymes mL

0.1 0.1 0.1

Serum or antisera a ‘21-labeled

standard

0.1

Sac-CeIl#{176}

1.0

Purified

isoenzymeb

incubate 1 h temperature incubate 30 mm, temperature water

6. Centrifuge (20 mm, 1500 x g,4#{176}C) 7. Decant supemate, count precipitate aSifjc antisera to c-AST or m-AST. b Isoenzyme c5acClI = donkey, anti-rabbit lgG coated on cellulose.

RIA of the isoenzymes. decreased nonspecific

=

c-A5T or m-AST.

The added bovine serum albumin binding to 3-5% of the total counts

added.

Active-Site

Titration

The

method used for the active-site titration of AST isoenzymes from human heart is based on fluorimetry of the half-cycle transamination of the phosphopyridoxal form of the isoenzyme (20). Human heart isoenzyme was saturated with pyridoxal phosphate and completely converted into the phosphopyridoxal form with a-ketoglutarate. Gel ifitration on Sephadex G-25 was used to remove excess keto acid, and the resulting enzyme eluate was divided into two parts. Enzymatic activity was determined in one part with the modified IFCC method as described above; the second was used for active-site titration. On addition of excess cysteine sulfinic acid, the enzyme is converted into the pyridoxamine form with formation of an equimolar amount of pyruvate by half-cycle transamination. Pyruvate was determined with a kit (cat. no. 124982; Boehringer Mannheim Canada, Dorval, Quebec H9P 9Z9), which measures the decrease in NADH absorbance at 25#{176}C in the presence of lactate dehydrogenase (EC 1.1.1.27). Pyruvate production and protein content data were used to calculate the catalytic constant (moles of product formed per second per mole of pure enzyme) based on relative molecular masses of 93000 and 90400 for c-AST and mAST, respectively (21). The turnover number (catalytic constant/number of active sites) is half the value of the catalytic constant, because aspartate aminotransferase contains two active sites per molecule (22).

Results Antisera

Preparation

were extracted from human heart tissue to homogeneity as indicated by single bands on analytical isoelectric focusing gels. These isoenzymes were used to induce polyclonal antibodies in rabbits, and these antibodies gave specific single precipitin lines when tested against the corresponding human AST isoenzyme. The specificity of each antisera for its own isoenzyme was exploited to inhibit its catalytic activity and allow the corresponding non-complexed isoenzyme to be determinedthe basis of the mimunoinhibition technique for measuring AST isoenzyme by catalytic means (23). With this technique the unpurified antisera (up to a 80-fold dilution) fully inhibited each of the isoenzymes. AST

isoenzymes

and purified

Antisera We

Purification

purified the antisera by precipitation with sulfate at a final 45% saturation, followed, more selectively, by column chromatography on protein A-Sepharose CL-4B, which binds IgG-type antibodies. Elution with a decreasing pH gradient buffer allowed us to recover the specific antibodies to c-AST or m-AST. A representative elution proffle is illustrated (Figure 1), showing six fractions of c-AST antibodies between pH 4 and 7. Different binding affinities for protein A represented by these fractions may be due to various subclasses of IgG. We used a single chromatographic method involving DEAE Affi-Gel-Blue in place of ammomum sulfate precipitation and a second column procedure. The separation of antibodies to c-AST isoenzyme from other protein components was as shown in Figure 2. Most of the specific antiserum was eluted at the void volume. Some carryover partly

ammonium

8

ANTI-C-AST

ACTIVITY

was noted, and this was recovered with 100 mmollL NaCl in 20 mmol/L Ti-is buffer, pH 8.0. This latter procedure for IgG purification proved to be more efficient and was used for both isoenzyme-specific antisera preparations. We have, however, encountered some loss of antibody to isoenzyme inhibitory activity in some of the batches of antisera treated with this affinity resin, for reasons unknown. An alternative and established IgG-purification procedure involving caprylic acid and chromatography on DEAESephacel proved to be as effective for isolating the active immunoglobulins. Antisera to c-AST and m-AST were immunologically active, both when tested by the immunoprecipitation

assay

and

by RIA.

lodination The iodination of c-AST or m-AST by Na’251 with lodogen or lodo-beads resulted in isoenzymes containing labeled iodine activity as recovered from gel-ifitration columns, but less than 5% of these isoenzymes was bound to their respective specific antisera. Attempts to label the isoenzyme by the Chloramine-T method (24) were also unsuccessful. Evidently, either the iodine attachment of the tyrosyl residues of the isoenzyme impedes its binding with the antibody or the enzyme structure was degraded. Radioiodination of the isoenzymes by use of the BoltonHunter reagent produced good incorporation of the tracer with retention of its immunological activity. Dose-response standard curves for c-AST and m-AST produced against each specific antiserum showed good differentiation of concentrations between 5 and 1000 ng/L (Figure 3).

E C

0 c’J

w 0 z

Specificity

(I)

FRACTION

NUMBER

Fig. 1. Elution profile of rabbit antisera to human c-AST from protein ASepharose CL-4B column Antisera activity determined by immunoprecipitation with human heart c-AST isoenzyme

0.028

0.I

1,4

NoCI(MI

ANTI-C-AST

AC1vITY

E

of the Antisera

We evaluated the specificity of the antisera, i.e., the cross reaction between c-AST and m-AST isoenzymes, by testing the capacity of the isoenzymw counterpart to interfere with the binding of the labeled antigen by the homologous antiserum. As shown in Figure 4, there was no detectable immunological crossreactivity between c-AST and m-AST isoenzymes from heart tissue and either isoenzyme-specific antiserum. AST isoenzymes purified from human liver can be substituted for heart isoenzyme on an equivalent basis. When we used the corresponding liver m-AST in place of heart m-AST and generated a standard curve, the two proffles were superimposable. A similar equivalence was obtained with the c-AST isoenzyme from the two tissue sources for the antisera generated by heart c-AST. The liver isoenzyme also did not cross react with the opposite antisera, as was also shown for the heart isoenzymes by RIA. This non-crossreactivity was also demonstrated by the Ouchterlony

technique

c’J

when

uJ 0

with

the corresponding the specific antisera,

C

0

z

of immunodiffusion.

Precipitin

lines

formed

antigen isoenzyme was reacted but not against the other isoen-

zyme.

0

AST Isoenzymes

(I,

m

in Human

Serum

We active

48

16

FRACTION

NUMBER

Fig. 2. Elution profile of rabbit antisera to human c-AST from DEAE Aff iI

Gel-Blue column Anti-C-AST activity determined

as described for Fig. 1

determined the concentration of immunologically isoenzymes of c-AST and m-AST in human serum by RIA, and compared it to the amount of catalytically active isoenzyme as measured by immunoprecipitation and kinetic conversion of NADH. We found a significant correlation (p