Acta Med Scand 204: 453465, 1978

Quantitation of J Chain in Human Biological Fluids by a Simple Immunochemical Procedure Anders 0. Grubb From the Depcrrtment of Clinicul Chemistry. Universitj?qf Lund, Mulmii Generul Hospital. Mulmo. Su3rden

ABSTRACT. The molecular form and immunochemical properties of the J chain populations released on reduction and carboxymethylation of normal human plasma, milk, saliva and of plasma containing IgA or IgM M-components were investigated. A procedure was devised to release the entire J chain population from these various sources and to produce immun* chemically identical J chain populations containing only J chain monomers. An identical standard J chain population was purified and quantitated by physiochemical means. A specific rabbit anti-J chain antiserum was raised against this pure J chain population. A simple and rapid immunochemical method for J chain quantitation in complex biological fluids as well as in solutions of pure polymeric immunoglobulins was constructed on these grounds. The J chain concentration was found to be (mean fS.D.) 1.74f0.65 p M in normal human plasma, 1.94f1.21 f l in human milk and 0.48f0.26 f l in human saliva. The J chain/IgA molar ratio was found to be (mean33.D.) 0.45k0.07 in human milk and 0.52k0.09 in human saliva when the IgA concentration was expressed as monomeric units per volume unit. The range of the J chain/IgA molar ratios in plasma samples with highly concentrated IgA M-components was 0-0.64. The J chainAgM mdar ratio in plasma samples with highly concentrated IgM M-components was between 1 and 2 when the IgM concentration was expressed as pentameric units per volume unit. Key words: J chain, immunochemistry, quantitation. Acta Med Scand 204: 453. 1978.

A small polypeptide chain demonstrated to occur in some polymeric immunoglobulins (7, 8, 28) was shown in 1970 by Halpern and Koshland ( 1 I ) to be a novel component of human and rabbit polymeric IgA. The authors named this new immunoglobulin part the J chain since they considered that its function was to assist in the joining of the monomeric

immunoglobulin units into polymeric molecules. Mestecky et al. (24) demonstrated in 1971 the J chain to be a covalently bound part of human pentameric IgM. Subsequent studies have showii it to be present in the macroimmunoglobulins of a number of vertebrate species (17). In addition to its role in the joining of monomeric IgA and IgM units, the J chain seems to be required for the combination of the polymers with secretory component ( 5 ) . J chain synthesis has been demonstrated not only in plasma cells secreting polymeric IgA and IgM but also in IgG-containing lymphoid cells (2, 15). It has been speculated that J chain plays an important role in the regulation of the Ig synthesis by various antigen-stimulated lymphoid cells (4, 17, 23). Progress in the elucidation of the part played by the J chain in the antibody response has, however, been delayed by lack of a simple and rapid method for quantitation of J chain in complex biological fluids. Presently available physicochemical methods (9, 12, 25) all require a purification step and are therefore troublesome and sensitive to variations in recovery. No values for the J chain concentration in human plasma, milk or saliva have been reported. The present work describes a simple, rapid and sensitive immunochemical procedure for J chain quantitation in biological fluids. The method is based upon a complete release of J chains in monomeric form by reduction and alkylation of the biological fluid studied. A subsequent electroimmunoassay of the J chain population is performed using an antiserum against reduced and alkylated Abbreviations: SDS=sodium dcdecyl sulfate; barbitalEDTA buffer=0.075 M barbital buffer, pH 8.6, containing 2 m M ethylenediaminetetraacetic acid; DTT=dithiothrei-

tol; Tris buffer=2.0 M tris-(hydroxymethy1)-amino-methane buffer, pH 8.6. A r m Med Scond 204

A . 0. Gruhb

454

as described by Weber and Osborn (33)and analytical and

1.oJ

0.5-

Fig. 1 . Sephadex (3-200 Superfine chromatography in 0.1 M ammonium bicarbonate at +6"C of a reduced and alkylated IgM M-component. 3-ml fractions were collected at a flow rate of 3 ml/h in a 77- x 5-cm column. The J chain-rich fractions denoted by the horizontal line were pooled and lyophilized.

J chain and a J chain standard population of monomeric, reduced and alkylated molecules. The work also reports the J chain concentration as obtained by this method in normal human plasma, milk and saliva as well as in plasma samples with highly concentrated IgM or IgA M-components. MATERIALS Agarose with low electroendosmosis was purchased from L'lndustrie Biologique Franqaise S. A., Gennevilliers, France and from Marine Colloids Inc., Rockland, Maine, USA; Sephadex G-200 and Dextran T 10 from Pharmacia Fine Chemicals, Uppsala, Sweden; Ultrogel AcA 54 from LKB-produkter AB, Bromma, Sweden; DEAE-Cellulose from Whatman Ltd, Springfield, Mill, England; guanidinium chloride and Coomassie Brilliant Blue R 250 from Schwarz/Mann, Orangeburg, N.Y ., USA; and complete Freund's adjuvant from Difco Laboratories, Detroit, Mich., USA. All other chemicals were of reagent grade and obtained from British Drug Houses Ltd, Poole, England. Monospecific rabbit antisera against human polyclonal IgM, IgA and against IgG Fc-fragments were available at our laboratory. Fresh samples of human plasma and saliva were obtained from healthy laboratory personnel and fresh samples of milk from lactating healthy females. Plasma samples with highly concentrated IgM, IgG, IgD or IgA M-components were selected from samples sent to the Clinical Labotatory for Ig measurements. These samples were stored frozen at -20°C for up to a year before their J chain concentrations were determined.

METHODS Electrophoretic procedures. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was performed

preparative agarose gel electrophoresis as reported by Johansson (14). Iinriiunoclierfiictrl procedurrs. Single radial immunodiffusion was performed according to Mancini et al. (22). double radial immunodiffusion according to Ouchterlony (27). microimmunoelectrophoresis as described by Scheidegger (30) and crossed immunoelectrophoresis as described by Laurel1 (19) and Ganrot (10); 10% Dextran T 10 was incorporated in the antiserum-containing gel to increase the precipitation of antigen-antibody complexes ( 13). Laurell's procedure for electroimmunoassay (20) was slightly altered as described in Results. Aiiiino w i d cindysis. Quantitative amino acid analyses were carried out essentially according to Spackman et al. (31). Protein samples of about I mg were hydrolyzed in 6 M HCI at 110°C for 24 and 72 hours in sealed, evacuated Pyrex tubes. The hydrolyzates were analysed with the two-column system on a Joel model JLC-5 AH automatic amino acid analyzer with the Jeol resin RC 1 . Purifictrtion of 19s and 8s IgM M-coinponeitis. One 8s and two 19s IgM M-components were purified in large amounts from plasma samples from two patients with macroglobulinemia. One of the samples contained a highly concentrated 19s IgM(K) M-component in addition to an 8s IgM(K) M-component of lower concentration. The other sample contained a single concentrated 19s IgM(K) M-component. The 19s M-components were purified by ammonium sulfate precipitation followed by repeated euglobulin precipitation and gel chromatography on Sephadex G-200. The 8s IgM M-component was purified by stepwise ammonium sulfate precipitation, dialysis of the appropriate fraction against phosphate buffer and ion exchange chromatography of the fraction on a DEAEcellulose column in the same buffer with elution of the M-component by a linear salt gradient. The M-component solution was then passed through a glutaraldehyde-

ABSORBANCE

0.1 0

FRACTION NUMBER

Fig. 2 . Ultrogel AcA 54 chromatography in 0.05 M ammonium bicarbonate at +6"C of a J chain fraction partly purified by Sephadex (3-200 chromatography and preparative agarose gel electrophoresis. Ten-ml fractions were collected at a flow rate of 30 ml/h in a 98- x 2.5-cm column. The fractions denoted by a horizontal line consisted of pure monomeric J chains, while the preceding fractions contained J chain dimers and polymers. Lightabsorbing buffer salts from the preparative gel electrophoresis were eluted in fractions 41-46.

Quiintitcrtion of J chrrin in human biological fluids

Fig. 3. Analytical agarose gel electrophoresis in 0.075 M barbital buffer, pH 8.6, with 2 mM EDTA o f (from left

to right) normal human serum; a fraction containing monomeric J chains from the Ultrogel chromatography described in Fig. 2; a fraction containing polymeric J chains from the Ultrogel chromatography described in Fig. 2; the monomeric J chain population used as standard in the quantitative immunochemical procedure; a reduced and carboxymethylated 19s IgM M-component.

insolubilized ( I ) immunosorbent against IgG and IgA and finally gel chromatographed on Sephadex (3-200. Analytical agarose gel electrophoresis. SDS-polyacrylamide gel electrophoresis and microimmunoelectrophoresis with use of an antiserum against human plasma failed to reveal any contamination of the purified M-components. These were extensively dialyzed against distilled water and then lyophilized.

RESULTS Production c$ pure reduced rind ciirboxytnethyliite~~ J chain f o r immunization

Of a pure IgM(K) M-component. 8.0 g were dissolved/suspended in 200 mlO.075 M barbital buffer, pH 8.6. with 2 mM EDTA (barbital-EDTA buffer). To the turbid solution was added 8 ml of the barbital-EDTA buffer containing 250 mg dithiothreitol (DTT). After 2 hours in room temperature,

455

during which period the solution became clear, 16 ml of a 2.0 M Tris-HCI-buffer, pH 8.6 (Tris-buffer) containing 630 mg iodoacetic acid was added. After another 15 min in room temperature the mixture was chromatographed on a column of Sephadex (3-200 Superfine (Fig. I). The eluted fractions were monitored for their J chain content by analytical agarose gel electrophoresis (Fig. 3) and J chain-rich fractions were pooled and lyophilized. The lyophilized substance was then dissolved in a small amount of barbital-EDTA buffer and subjected to preparative agarose gel electrophoresis. Gel strips containing J chain were then cut out, frozen and thawed three times. Thereafter they were centnfuged to elute the J chain solution from the gel residue. This J chain solution was lyophilized, the substance dissolved in 0.05 M ammonium bicarbonate and chromatographed on a column of Ultrogel AcA 54 (Fig. 2). Two protein-containing peaks were obtained, one consisting of monomers and the other of dimers and higher polymers of J chain (Fig. 3). The monomeric J chain fractions were pooled and lyophilized. SDS-polyacrylamide gel electrophoresis, analytical agarose gel electrophoresis and microimmunoelectrophoresis were used to demonstrate the absence of contaminating proteins in this J chain preparation. Production of (it1 antiseruni against reduced iind carboxvmethylnted J chiiin und testing of its specificity

Of the lyophilized, reduced and carboxymethylated monomeric J chain preparation, 2 mg were dissolved in 1 ml 0.15 M NaCl. This solution was emulsified with 2 ml complete Freund’s adjuvant and injected in the back feet and at some subcutaneous sites on the backs of three rabbits. Similar booster injections were given 2, 7 and 1 1 weeks after the primary immunization. The rabbits were then bled every two weeks. The specificity of the antisera was tested by crossed immunoelectrophoresis with up to 7 % antiserum in the gel and with native human normal plasma or with reduced and carboxymethylated normal or agammaglobulinemic human plasma in the first electrophoretic step. Two of the three rabbits produced antisera which were considered monospecific since they only formed one precipitation peak with reduced and carboxymethylated normal human plasma but no precipitation at all with native normal plasma or with reduced and carboxymethylated A r m Med Srund 204

456

A. 0. Grubb

Results of Mole et al. (26)

amide gel electrophoresis demonstrated that this monomeric standard J chain population. although not as pure as the J chain population used for immunization, consisted to at least 90% of reduced and carboxymethylated J chain molecules.

21

Molur concentration o f J chtrin

Table I. Amino ticid coinposition oj thr niotioineric strrndard J chuin population

Constituent

FoundU

Aspartic acid Threonineb Serine" Glutamic acid Proline Glycine Alanine Carbox ymethylcysteine Cystine Valine' Methionine Isoleucine Leucine Tyrosine Phenylalanine Lysine Histidine Argi ni ne

17.3 13.6 10.3 13.9 8.1 3.7 6.5 6.6 0 9.7 1.8 8. I 8 5.2 I .6 5.5 1.4 8.3

II 8 15.5 8 2 6 8 11

I 8 8 5 I 5.5 I 9

"Calculated on the basis of 8 residues of leucine per molecule. "Values obtained by extrapolation to zero hours hydrolysis. 'Seventy-two-hour hydrolysis value only.

agammaglobulinemic human plasma. After 3 bleedings without booster injections a significant decrease in the antiserum titer was noted.

in the .stondurd solution

The J chain concentration in the J chain standard solution was measured by three different procedures. Firstly, a measured volume of the solution was carefully lyophilized at high vacuum and its protein content was weighed thereafter. Secondly, the light absorption of the solution at 278 nm was used to calculate its J chain concentration from the extinction coefficient for J chain (E1:=7.0) given by Wilde and Koshland (34). Thirdly, measured volumes of the standard solution were lyophilized. their protein contents hydrolyzed in 6 M HCI for 24 and 72 hours and their total amino acid content determined. The molar concentration of the various amino acids (excluding amino acids of low concentrations) were then used to calculate the molar J chain concentration in the standard solution by comparison with the total amino acid composition of J chain given by Mole et al. (26). The results obtained by the three procedures were 207, 191 and 167 p M , respectively, when a J chain molecular weight of 16422 daltons was assumed (26). The

Production qf'rr purified reduced carboxymethylated and monomeric J chain population for use as standard

A pure 19s IgM(K) M-component was reduced, carboxymethylated and fractionated on Sephadex G-200 exactly as described above for the production of J chain for immunization. J chain-rich fractions were identified by analytical agarose gel electrophoresis, pooled and concentrated by pressure ultrafiltration at +4C. The concentrated J chain solution was then chromatographed on a column of Ultrogel AcA 54 under the conditions described in Fig. 2 and the eluate monitored by agarose gel electrophoresis. The fractions containing J chain were pooled, concentrated by pressure ultrafiltration and rechromatographed on the Ultrogel column. A single protein peak containing monomeric highly purified J chain was obtained. Appropriate fractions were pooled, concentrated by pressure ultrafiltration and divided into small aliquots, which were frozen and stored at -24°C until used. Analytical agarose (Fig. 3) and SDS-polyacrylActri Mrd Scritid 204

CENT OF MAXIMAL >PERCHAIN RELEASE

la

so a

DlTHlOTHRElTOL CONCENTRATION 10

20

30

Fig. 4. Influence of increasing dithiothreitol concentration on the J chain release from normal human plasma (0), milk (A), saliva (O), a plasma sample with a J chain-rich IgA M-component ( 0 )and from a plasma sample with a 19s IgM M-component (+).

Qirtrntitation of J chain in human biologicalfluids PER CENT OF MAXIMAL J CHAIN RELEASE

t

457

Table 11. Release of J chain on reduction and carboxymethylntion of different biological fluids in 0.075 M barbital buffer, p H 8.6, with 2 m M EDTA compared to the release on reduction and carboxymethylation in 2 M Tris-HCI-buffer, p H 8.6, rontciining 6 M guanidinium chloride Release

L

2

6

Biological fluid

(%)

Normal human serum Human milk Human saliva Plasma sample with a concentrated J chain-rich IgA M-component Plasma sample with a concentrated 19s IgM M-component

92 99 117 100

109

Fig. 5 . Influence of increasing reduction periods on the J chain release from normal human plasma ( O ) , milk (A), saliva (0). a plasma sample, with a J chain-rich IgA M-component ( 0 )and from a plasma sample with a 19s

IgM M-component tion of 7 mM.

(+)

at a final dithiothreitol concentra-

amino acid composition of the standard J chain population is given in Table I. No trace of cystine could be seen in the chromatograms. Release of J chuin from various biological jluicls on reduction mtl carbox?,tnethylNtic,n

Normal human plasma, serum, saliva and milk were used in these experiments as well as a plasma sample with a highly concentrated 19s IgM Mcomponent, a plasma sample with a highly concentrated J chain-rich IgA M-component and a solution in barbital-EDTA buffer of a pure 8s IgM M-component (10 mglml). Four parts of barbitalEDTA buffer were added to one part of these solutions and thereafter increasing amounts of DTT in a constant small volume of barbital-EDTA buffer. After 2 hours in room temperature the solutions were mixed with Tris buffer, containing a molar amount of iodoacetic acid 2.1 times the amount of DTT used for reduction. After another I5 min in room temperature the J chain release in the samples was determined by the electroimmunoassay described further on. The results are shown in Fig. 4. As can be seen, no further J chain release from the different biological fluids was observed by increasing the final DTT concentration beyond 7 mM. No J chain release from the pure 8s IgM was observed at any DTT concentration.

Fig.6. Analytical agarose gel electrophoresis in 0.075 M barbital buffer, pH 8.6, with 2 mM EDTA of: (from left to right) a 19s IgM M-component reduced and carboxymethylated in 0.075 M barbital buffer, pH 8.6; the same 19s IgM M-component reduced and carboxymethylated in 2 M Tris-HCI-buffer. pH 8.6, containing 6 M guanidinium chloride; normal human plasma. Before the electrophoresis the guanidinium chloride-containing solution was dialyzed against barbital buffer and cleared by centrifugation. Acta Med Scand 204

458

A . 0. Gruhh

ABSORBANCE I CHAIN CONCENTRATION (280 nml per cent of maxlmrl) 0.3.

I

0.2.

I

1 0.1

. .. .

1.

:I

.+*

-..* : 10

.

\.

.

.:. . *

'. I

- * t..'.,.'

.*

*

.

20 FRACTION NUMBER

Fig. 7. Ulti gel AcA 54 chromatography in 0.05 M am-

monium bicarbonate at +6"C of reduced and carboxymethylated normal human plasma and of a marker protein solution. Two-ml fractions were collected at a flow rate of 6 ml/h in a 56- X I-cm column. The J chain concentration in the fractions was determined by electroimmunoassay. ---=280-nm absorption of reduced and carboxymethylated plasma, A-A=J chain concentration, . . .=280-nm absorption of the marker protein solution containing horse apoferritin, monomeric light Ig chains, cytochrome C and Kunitz's trypsin inhibitor.

Four parts of barbital-EDTA buffer were added to one part of the biological fluids described above (excluding the 8s IgM solution) and then a small volume of a DTT solution, giving a final DTT concentration of 7 mM. After increasing periods of time the reduction was terminated by addition of Tris buffer containing iodoacetic acid. After another 15 min at room temperature the J chain release was assessed by the electroimmunoassay . Fig. 5 gives the results. After a reduction period of 2 hours no further release of J chain from the samples occurred. To one part of each of the biological fluids described above was added five parts of a solution of 6 M guanidinium chloride in Tris buffer. Thereafter DTT in a small volume of Tris buffer was added to a final concentration of 7 mM and after 2 hours at room temperature the reduction was terminated by addition of Tris buffer containing iodoacetic acid. All solutions were then extensively dialyzed at +4"C in barbital-EDTA buffer. The volumes of the solutions after the dialysis were recorded. In parallel experiments the biological fluids were reduced and carboxymethylated as described above but with the guanidinium-containing Tris buffer replaced by barbital-EDTA buffer. The J chain release from the biological fluids on reduction and alkylation under denaturing and non-denaturing conditions was then

I ? 3

I **

'I

5

6

'

Fig. 8. Crossed immunoelectrophoresis with the use of anti4 chain antiserum of various reduced and carboxymethylated biological fluids. I =the standard J chain solution, 2=normal human plasma, 3=milk, I=saliva, 5=a plasma sample with a J chain-rich IgA M-component,6=a plasma sample with a 19s IgM M-component. The long vertical lines denote the migration of an albumin marker in the first electrophoretic runs and the short verticals the application slots of the first electrophoretic runs.

determined by electroimmunoassay . Table I1 demonstrates that the same amount of 3 chain seems to be released under the denaturing and the non-

Quantittrtion o j J chain in huniun hiologiccilfluids

‘ 6

5

4 *t* .2. J .

459

Fig. 9 . Double radial im-

1

munodiffusion of various reduced and carboxymethylated biological fluids. The central wells contain an anti-J chain antiserum. I =standard J chain solution, 2 =normal human plasma, 3 =milk, I=saliva, 5 = a plasma sample with a J chain-rich IgA M-component, 6 = a plasma sample with a 19s IgM M-component.

Physicochemical and immunochemical properties of the standard J chain population and the J chain populations released from different biologira1,fluids on reduction rind cirrhoxvrncth~liitiorl One part of normal human plasma, saliva and milk as well as of a plasma sample containing a concentrated 19s IgM M-component and of a plasma sample containing a concentrated J chain-rich IgA M-component was mixed with four parts of

barbital-EDTA buffer and reduced with DTT at a final concentration of 7 mM for 2 hours at room temperature. The reduction was stopped by addition of Tris buffer containing iodoacetic acid: 500 pI of these solutions and of the standard J chain solution were then gel chromatographed in sequential runs on a column of Ultrogel AcA 54 in 0.05 M ammonium bicarbonate. A small amount of Kunitz’s trypsin inhibitor was dissolved in the solutions before each run to mark the separation volume of the column. Two-ml fractions were collected and their J chain concentration determined by electroimmunoassay. The separation characteristics of the column were repeatedly monitored by running 500 pl of a solution containing horse

Fig. 10. Electroimmunoassay of J chain in different human plasma samples. The horizontal line on the cathodal

side of the application holes marks the precipitates of four dilutions of the J chain standard solution.

denaturing conditions. Agarose gel electrophoresis of a solution of a pure 19s IgM M-component reduced under both denaturing and non-denaturing conditions gave the patterns depicted in Fig. 6.

ACIU Mrd Srund204

460

A . 0.Grubb

Table 111. J chain concentrcition in various biologicd j1uicl.s Biological fluid Normal human plasma Plasma pool Normal human milk"

Normal human saliva Plasma samples with concentrated IgA M-components Plasma samples with concentrated IgM M-components

No. of samples 18

(8 d , 10 P Swedish Caucasians) From 1 000 registered blood donors 15

(from different Swedish Caucasians after varying postpartum periods) 16 (6 d , 10 P Swedish Caucasians) 92 33

Mean 1.74 Range 0.57-2.61 S.D. 0.65 1.88

Mean 11.76 Range 4.08-30.47 S.D.6.75 13.4

Mean 0.87 Range 0.38-1.5 I S.D. 0.37 0.80

Mean 1.94 Range 0.68-4.36 S.D. 1.21

Mean 4.30 Range 1.38-9.27 S.D.2.32

Range

Quantitation of J chain in human biological fluids by a simple immunochemical procedure.

Acta Med Scand 204: 453465, 1978 Quantitation of J Chain in Human Biological Fluids by a Simple Immunochemical Procedure Anders 0. Grubb From the Dep...
1MB Sizes 0 Downloads 0 Views