Characterization and use of monoclonal and polyclonal antibodies against the mouse interferon-y receptor Ross Paul C

D. LeClairet” J. Zavodny,”

Wilkinson

Laboratory

tPathology/Oncology, City,

Kansas,

Mitali Basu7’1’ Judith L. Pace7’1 of

and and

the

Cancer

Center

David M. Pinson7’1 Mark and Stephen W. Russell*,S and

SMicrobiology/Molecular

the #Schering

Corporation,

the Departments

Abstract: To facilitate investigation of its physical and functional properties, 11 monoclonal antibodies (mAbs) and a goat polyclonal IgG specific for the mouse interferon(IFN-’y) receptor were characterized and their potential uses studied. Eight of the mAbs interacted with epitopes on the extracellular domain of the receptor, two interacted with epitopes on the intracellular domain, and one interacted with an epitope that could not be localized definitively to either region. Of the 11 mAbs, the majority (8) were IgGs, 2 were IgMs, and 1 was an IgA. Relative avidities of the seven that could be determined ranged from 333 to 0.002 Both the polyclonal goat IgG and mAb GR-20 (the latter specific for an epitope in the binding site for IFN-y) blocked binding of the ligand and, as expected, prevented induction by IFN-y of priming of macrophages for tumor cell killing. None of the other mAbs had an effect despite the fact that GR-22 partially ( > 50%) blocked binding of IFN-y Neither the polyclonal IgG nor any of the mAbs had an agonist effect. The relative usefulness of the antibodies for immunoprecipitation, immunoblotting, immunoassay, and cell staining with and without prior fixation is described. The results of immunocytochemical staining directly confirmed that the majority of immunologically reactive receptor protein expressed by cells is intracellular. To facilitate use by other investigators, the hybridomas that produce these mAbs will be offered to the American Type Culture Collection. J. Leukoc. Biol. 51: 507-516; 1992. Key

Words:

immunology

. pharmacology

molecular

biology

New

Interferon-y (IFN-y) has a wide spectrum of biological activities, ranging from the ability to induce an antiviral state in cells to a variety ofeffects important in regulating immune responses [35]. Thus, it has an unquestioned place in the hierarchy of pluripotent cytokines. Interferon-7 acts through a receptor-binding protein expressed by normal cells other than mature erythrocytcs. Despite the fact that the cDNA that encodes this transmembranous receptor protein recently has been cloned for humans [1] and mice [8, 18, 20, 21, 29], little is known about how it transduces function-inducing signals after ligand has bound to it. The work to be reported here has resulted in a panel of receptor-specific immunologic reagents, both monoclonal antibodies (mAbs) and a polycional IgG, that should be of help in characterizing structural and functional properties of the receptor for IFN--y. The individual antibodies have been cataloged with respect to their subclasses, specificities, and relative binding avidities. Whether they interfere with the

S. Hunt,’1

University

of Kansas

School

of Medicine,

Kansas

Jersey

binding assessed. uses as metric identified.

of IFN-y and Finally, those immunoblotting, analysis, and

MATERIALS Cells

AND

and

block/induce function that have the greatest immunoprecipitation, immunocytochemistry

also utility

has been for such flow cytohave been

METHODS

Medium

Duibecco’s modified minimal essential medium (D-MEM) was prepared from a powdered mix (ICN Flow Laboratories, McLean, VA) and supplemented to contain 3.7 mg/mi sodium bicarbonate, 2 mM glutamine (both from ICN Flow), 100 U/mI penicillin G potassium, and 100 g/ml streptomycm (Pfizer, Inc., Atlanta, GA), 15 mM HEPES (Sigma Chemical Co., St. Louis, MO), sodium pyruvate, and nonessential amino acids (both from ICN/Flow; diluted 1:100 from box stocks). Modified Eagle’s minimal essential medium was prepared from a powdered mix (Auto-POW MEM, ICN Flow) and supplemented with glutamine, pcniciilin/strcptomycin, and HEPES (H-MEM) as described above. Reduced serum medium (Opti-MEM GIBCO, Life Technologies, Grand Island, NY) was prepared from a powdered mix and supplemented with glutamine (2 mM). Fetal bovine serum ( FBS) was obtained from Hyclone Laboratories (Logan, UT). dotoxin

All

for detectable enLimulus amebocyte (Associates ofCape Cod, Woods Hole, MA) at a sensilevel of 0.05 ng/ml. Hybridomas producing rat mAbs for the mouse IFN-y receptor (MuIFN-’yR) were as previously described [3]; they were maintained in

lysate tivity specific

INTRODUCTION

Joan

of Pharmacology/Toxicology/Therapeutics,

Genetics/Immunology, Bloomfield,

L. Redick7’

culture as determined

media

were by assay

negative [26] with

made D-MEM + 20% FBS. Culture supernates from outgrowths were screened initially for their reactivity with intact and acetone-fixed mouse bone marrow culture-derived macrophages. Those apparently reactive with the MuIFN-yR were cloned twice by limiting dilution and used for further study. Rat XC sarcoma cells (catalog # CCL 165, American Type Culture

[ATCC], Rockviile, MD) were cultured 10% FBS. XC cells stably transfected with the cDNA for MuIFN-yR (XCR) and others transfected to express a secreted form of the receptor that lacked the transin

Collection

H-MEM

+

Abbreviations: receptor; XCsR, main; tion

NS-lsR,

FBS, fetal bovine serum; MuIFN-yR, secreted mouse IFN-y receptor minus secreted

external

domain

of

the

mouse interferon-y transmembrane do-

MuIFN-yR;

K0,

dissocia-

constant. Reprint

Research, Rainbow Received

Journal

requests:

Stephen

W.

Russell,

Wilkinson

Laboratory

1008 Wahl Hall West, University of Kansas Medical Blvd., Kansas City, KS 66160-7184. October 3, 1991; accepted November 27, 1991.

of Leukocyte

Biology

Volume

51,

May

for

Center,

1992

Cancer

3901

507

membrane domain (XCsR) [12] were maintained in HMEM + 10% FBS containing 400 g/ml ofthe cytotoxic antibiotic G418 (Geneticin, GIBCO) in lieu of the penicillin and streptomycin. Survival was possible under these conditions because each had been cotransfected with pRSVnco, a plasmid containing the gene for resistance to G418 [5]. Mouse NS-1 myeloma cells, stably transfected with a cDNA encoding only the extraceilular portion of the binding protern for mouse IFN-y (NS-lsR cells), secreted the extracellular portion of the mouse IFN-’y receptor [10]. These cells were carried in H-MEM + 10% FBS containing the selec-

for 48 h in either H-MEM + 10% FBS or serum-free OptiMEM. Control supernates from XC and NS-1 cells were generated in a similar manner. Sample purity was confirmed for the affinity-purified receptor preparations by eiectrophoresis through duplicate sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels [22], one of which was stained with Coomassie blue dye and the other processed for western (immuno)blot analysis. Receptor preparations appropriate for specific experiments were added to either the round-bottomed wells (50 al/well) of polystyrene 96-well plates (Linbro/Titertek, ICN Flow Laboratories) or poly-

tion An

vinylchloride microtiter plates (Dynatech Chantilly, VA) and incubated overnight washed four times with phosphate-buffered

agent, embryonic

fected with generously

mycophenolic human

acid kidney

the cDNA for by Dr. Robert

(5 g/ml; cell line,

the MuIFN-yR D. Schreibcr

Sigma 293R,

Chemical). stably trans-

[18], was (Washington

provided Univer-

St.

Louis, MO). This cell line was maintained as described for the XCR cells. Mouse L-929 (CRL 1, ATCC) and mouse P815 mastocytoma cells (originally obtained from Dr. Theodore Brunner, Lausanne, Switzerland) were maintained in H-MEM + 10% FBS. Bone marrow culture-derived macrophages from endotoxin-hyporesponsive C3H/HcJ mice were grown as previously described [25]. sity,

Production

and

purification

goat anti-rat to Sepharose

tein ing

concentrations the BCA protein

IL)

with

then nonspecific binding was blocked with 200 4/well of PBS containing 2% normal rabbit serum. After washing, 50 i of purified test antibody was added to each well and incubated for 1 h at room temperature. Bound after washing using the appropriate stain ABC immunoperoxidase kit Burlingame, CA). Positive wells were velopment using O-phenylenediamine

antibody was detected species-specific Vecta(Vector Laboratories, identified by color dedihydrochioride as

the chromogen. natech MR700

at 490 nm using Laboratories,

Plates plate

were reader

scanned (Dynatech

a DyInc.).

of antibodies

Monocional antibodies were produced in terminal stationary cultures and isotyped by double immunodiffusion analysis (Ouchterlony) using polyclonal anti-rat heavy chain-specific antisera (ICN ImmunoBiologicals, Costa Mesa, CA). The light chain type of each mAb was determined by enzymelinked immunosorbent assay (ELISA) using biotinylated anti-rat light chain-specific mAb, detected by peroxidase conjugated to avidin (Zymed Laboratories, Inc., San Francisco, CA). Monoclonal antibodies of the IgG subclasses were affinity purified using Protein G Sepharose 4 Fast Flow (Pharmacia LKB Biotechnology Inc., Piscataway, NJ) according to the manufacturer’s instructions. Monoclonal antibodies of the IgA and 1gM subclasses were purified similarly using polyclonal manufacturer

Laboratories, Inc., at 4#{176}C.Wells were saline (PBS),

4B

IgG (H (Zymed

+

L) coupled Laboratories).

by

the Pro-

of purified antibodies were measured usassay (Pierce Chemical Co., Rockford,

bovine serum albumin (BSA) as the standard. A rat mAb, B54-2, which recognizes an antigen unrelated to MuIFN-yR on mouse chronic inflammatory macrophages and peritoneal mast cells [24], was purified similarly and used as a control. Polyclonal IgG was produced from immunc serum harvested from a goat injected with MuIFN-yR purified by immunoaffinity chromatography from solubilized membranes of the monomyelocytic ccli line WEHI-3 (TIB 68, ATCC). The receptor-specific mAb used was IG2a

GR-20 [4] covalently linked to cyanogen bromide (CNBr)activated Sepharose 4B (Sigma Chemical). The IgG fraction of the immune goat serum was obtained from a resolubilized 50% ammonium sulfate precipitate by affinity chromatography on Protein G Sepharose 4B Fast Flow. Similarly purified normal IgG was obtained from pre-immune serum of the same goat.

Flow

Cytometry

Fluorescence was analyzed using a Coulter Electronics EPICS series 752 fluorescence-activated cell sorter (Coulter Electronics, Inc., Hialeah, FL). Autoand background fluorescence ofcell populations were determined by omitting both primary and secondary antibodies or primary antibody, respectively. All incubations were conducted at 4#{176}C.Viable cells were washed twice with icc-cold D-MEM + 10% FBS (containing 50 mM sodium azide to minimize endocytosis) and resuspended washed twice

at with

paraformaldehyde cold PBS. They with

icc-cold

107/mi. icc-cold

in PBS, were then PBS

containing

Cells to be permeabilized PBS, fixed for 20 mm with and washed twice more incubated at 5 x 106/mi 25

g/ml

digitonin

were 0.2%

with icefor 5 mm [13].

The

concentrations of paraformaldehyde and digitonin were selected on the basis of pilot experiments designed to determine the least amount of each that would allow antibodies specific for actin to stain the cytoskeicton of treated cells. After washing once in ice-cold medium, permeabilized cells were resuspended at 2.5 x 107/ml. Aliquots (200 ii) of either viable or permeabilized cells were mixed with an equal volume of ice-cold test or control antibody (100 jg/ml) and incubated at 4#{176}Cwith constant agitation for 1 h. Cells then were washed thrice with Earl’s balanced salt solution (without phenol red) containing 50 mM sodium azide and 15 mM HEPES (H-EBSS), after which they were incubated with fluoresceinated antibodies, either goat anti-rat IgG (Zymed Laboratories) or rabbit anti-goat IgG (Sigma Chemical). A 1:20 dilution in ice-cold medium was used. After constant agitation for 1 h, cells were washed thrice with H-EBSS and resuspended for flow cytometric analysis in 500 l PBS contaming 2% paraformaldehyde.

Immunoprecipitation ELISA

for Secreted

MuIFN--yR

XCsR or NS-lsR to be used as antigens in ELISA either were culture supernates or were purified by immunoaffinity chromatography using GR-20 mAb immobilized on CNBractivated Sepharose 4B. To produce supernates, cells that secreted these forms of the receptor were rinsed and cultured

508

Journal

of Leukocyte

Biology

Volume

51, May

1992

To prepare solubilized receptor, monolayers of 293R cells were washed once with HBSS (without Ca or Mg2), detached using nonenzymatic Cell Dissociation Solution ( Sigma Chemical), and the cell suspension was washed twice with PBS. Cells were surface labeled with [t25ljlactoperoxidase as previously described [17]. Cells were lysed in hypo-

volume tamed

of 85 tl in 2 mM Tris-HC1 [pH 7.4] that also conio mM NaCl, 1.2 mM MgCl2, 0.1 mM dithiothreitoi, 0.5 mCi [y-32P]ATP (6,000 Ci/mmol; Dupont NEN Research Products, Boston, MA), and 37 units of the catalytic subunit of beef heart cAMP-dependent protein kinase

tonic buffer (10 mM Tris, 3 mM MgC12, 1 mM PMSF, and 1 mM leupeptin [pH 7.2]) at a concentration of 2 x 106/ml. The suspension was centrifuged (27,000g, 20 mm, 4#{176}C),and the resultant pellet was extracted for 1 h with ice-cold solubilization buffer (PBS containing 1% Triton X-100, 1 mM PMSF, and 0.25 U/mi aprotinin), 5 x 1O cell equivalents/mi. A final centrifugation (100,000g, 30 mm, 4#{176}C)removed remaining particulate matter. The resultant extract then was preabsorbed overnight at 4#{176}Cwith an equal volume of Sepharose 4B to which normal rat IgG had been bound. The immunosorbent beads had been equilibrated with solubiiization buffer prior to their use. Antibodies to be analyzed were bound to beads

(Sigma Chemical). The reaction was stopped by the of 0.4 ml of 10 mM sodium pyrophosphate (Sigma cal) [pH 6.7] containing 5 mg/mi BSA. Unbound tope then was removed by dialyzing against several

of Sepharose 4B using secondary linkers: for goat IgG, Protein G (Pharmacia LKB) was used; for rat 1gM, Protein G, to which rabbit anti-rat j chain (Zymed Laboratories) had been attached; for rat IgA, Protein G + goat anti-rat a chain (ICN); and for rat IgG, goat anti-rat IgG (H + L) (Zymed Laboratories) coupled directly to Sepharose. Beads bearing the appropriate linker combination were incubated with the antibody (2OOd, 5Og/ml) to be tested (100d packed beads, 2 h, 4#{176}C)and then washed twice by centrifugation (13,600g, I mm, 4#{176}C)through wash buffer (50 mM Tris-HC1 [pH 7.4] containing 0.15 M NaC1, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mM PMSF, and 0.25 U/mI aprotinin). A final wash in icc-cold solubilization buffer (PBS containing 1% Triton X-100, 1 mM PMSF, and 0.25 U/ml aprotinin) followed. The pelleted, washed beads then were resuspended in 150 l of solubilizcd membrane and incubated (2 h, 4#{176}C) with constant gentle agitation. After four rinses in wash

Western

(Immuno)

Solubilized with a half electrophoresed nitroccllulose NH) using rics,

samples were described [3] goat IgG as

analyzed by or by western the primary

autoradiblotting label as

or

and

dried

overnight

at

4#{176}C. After

re-

goat

IgG

at

30

j.g/ml

in

wash

buffer

that

contained 1% normal rabbit serum. Bound antibody then was detected with the appropriate biotinylated rabbit antiIgG (i.e., either rat or goat, 2 zg/ml; Vector Laboratories) and streptavidin-conjugated gold, followed by silver enhancement Uanssen Biotech NV., Olen, Belgium) per the manufacturer’s instructions.

Competitive

Binding that was

rMuIFN-y. cells were

After washed

2 h on three

a

ice, supcrnatcs times with ice-

cold medium. Cells then were lysed by the addition ofO.5 ml 1 N NaOH per well followed by trituration. After the addition of 3 ml of scintillation cocktail (Econo-Safe, Research Products International, Mt. Prospect, IL), radioactivity was quantified in a scintillation spectrometer. Killing of P815 mastocytoma cells by C3H/HeJ bone marrow culture-derived macrophages was assessed using a 16-h

x

106/mi).

For

functional

with

the

rMuIFN-y

was

added.

throughout

the

activation

studies,

test

macrophages

antibodies

for

Antibody phase

of

1 h,

was

kept

the

assay.

were

after

in

the

which system

and Assay

antibodies rMuIFN-y accomplished

of Macrophage-Mediated

had to compete was assayed. as previously

32]. Briefly, 5 ;g of rMuIFN--y provided by Schering Corp. through Society, Atlanta, GA) was incubated

(2

with the binding The labeling of described [23,

x 106 U/mg; kindly the American Cancer (30#{176}C, 1 h) in a total

Immunocytochemistry For light microscopy, XC and XCR cells were cultured in 75 cm2 flasks. Cell layers were trypsinized briefly, washed in serum-free medium, and centrifuged onto glass slides using a Shandon Cytospin (Shandon, Inc.; Pittsburgh, PA). The cells were air-dried for 1 h at room temperature and then were fixed for 7 mm in cold acetone or for 2 mm in freshly prepared 4% paraformaldehyde. After drying, acetone-fixed cells were stored at -80#{176}C, and paraformaldehyde-fixed cells were stored in 70% EtOH overnight. The cells were rehydrated in PBS (pH 7.2) and then were immunostained using procedures described previously [7]. Two control rcagents were used: normal rat IgG (Sigma irrelevant rat monoclonal antibody, B54-2 gents were used at 7.5 g/ml. Binding was otinylated rabbit anti-rat IgG from Vector an avidin-biotin immunoperoxidase staining Laboratories. Binding was indicated by The cells were lightly counterstained with

Cytotoxicity The ability of 32P-labcled rMuIFN-y

of unlabeled removed, and

and binding a similar

Blotting

CA),

a mAb

excess were

preincubated

wetting the membranes in wash buffer (PBS [pH 7.4] contaming 0.1% BSA), nonspecific binding sites were blocked for 0.5 h at 37#{176}Cwith PBS containing 5% BSA. Blots then were washed and incubated (room temperature, 2 h) with either

tamed 20 mM sodium azide, 1 nM [32P]rMuIFN-’y, 250 molar excess of the test competitor. Nonspecific was assessed by incubating [32P]rMuIFN-y with

2

membranes of culture supernates were mixed volume of 3X SDS sample buffer. Samples were as above, electrophoretically transferred to a membrane (Schleichcr and Schuell, Kenne, a Mini Trans-Blot apparatus (BioRad Laborato-

Richmond,

of the same liquid without BSA. Specific radioactivity of the labeled rMuIFN-y preparations ranged between 20 and 50 Ci/.tg protein. After radioiabeicd rMuIFN-y was prepared, XCR cells were seeded in H-MEM + 10% FBS into 16-mm-diameter wells of 24 well plates at 10 cells/well and incubated at 37#{176}Cfor 24 h. Before they were used, plates were chilled by incubation on ice for 20 mm. Supernatant medium then was replaced with icc-cold medium that con-

5tCr-reiease assay as previously described [31]. Macrophages from endotoxin-hyporesponsive C3H/HeJ mice were used to eliminate any effect that small amounts ofcontaminating endotoxin might have had. Cytolytic activity was triggered after the macrophagcs had been exposed to rMuIFN-y (3 U/ml) by adding heat-killed Listeria monocytogenes (HKLM;

buffer, the beads were resuspended in 150 l of 3X SDS sampie buffer (450 mM Tris-HC1 [pH 6.8], 3% SDS, 30% glycerol, 0.15% bromphenol blue, and 3% 2-ME) diluted 2:1 with distilled water and heated 5 mm in a 100#{176}Cwater bath. After SDS-PAGE [22], ography as previously using the polyclonal described below.

addition Chemiradioisochanges

Chemical) and an [24]. All these rcadetected using biLaboratories and kit from Zymed a red coloration. Gill’s hematoxylin.

Statistics Values expressed are mean noted). When it was necessary enccs for more than one group homogeneity of variances was

LeClaire

ci aL

Anti-mouse

interferon-7

1 SEM (unless otherwise to determine whether differwere statistically significant, tested. A one-way analysis of

±

receptor

antibodies

509

variance, means, ogeneous

followed by a Tukcy HSD was used with homogeneous data sets, a Kruskal-Wallis of variance was used [28]. statistically significant.

analysis

considered

post hoc comparison of data sets. For heternonparametric one-way Values of P < 0.05 were

pected, there often was a substantial reduction in binding after cells had been fixed and permeabilized. A notable exception was mAb GR-34, for which the extent of binding after fixation increased. Binding ofthc polyclonal IgG to fixed and permeabilized cells decreased less after fixation than was the case for the mAbs. Because of the negative effect of fixation on antibodyantigen reactivity, a second approach was used to confirm localizations of the epitopes. This involved comparison of binding to either a secreted form of the receptor that contamed both the intracellular and cxtracclluiar domains ( XCsR) or to another secreted form of the receptor that was

RESULTS Physical

Properties

Antibodies basis

of Anti-MuIFN-’yR

thought of

initial

ELISA present

for in

to be

specific

screening

for

were

their capacity the culture

Antibodies the

MuIFN-7R

examined

to bind supernates

on

by

the

solid-phase

to secreted of XCsR

composed only of the extracellular domain (NS-lsR). The right halfofTable 2 summarizes the data obtained by ELISA using immunoaffinity purified XCsR and NS-lsR proteins. On the basis of these results, GR-23 and GR-29 were confirmed to bind to the intracellular domain. Monoclonal antibody GR-65, which had bound unequivocally to intact

MuIFN-yR cells. Eleven

mAbs consistently reactive with the receptor were identified (Table 1). Of these, IgG subclasses predominated (8/11), with two IgMs and one IgA. Nine of the eleven contained kappa light chains, while the other two contained lambda chains. Relative avidities for the seven mAbs that could be evaluated ranged clonal average

IgG

cells using ELISA but lar domain. to epitopes

from

0.002 to 333 Mi. The value for the goat poiyfraction given in Table 1 (1.4 M) represents the avidity of receptor-specific antibodies in the enriched fraction.

IgG

Distinction Epitopes

Between

Extracellular

and

Effects on Ligand Function

Intracellular

As seen in the left and GR-29) of the mAbs to be on the extracellular

ing

I

Antibodies

.

(approximately

poiycional in interfering effect (data

half of Table 2, bound to epidomain. As cx-

TABLE

Re active

with

of Biological

55%)

of

rMuIFN-y

IgG

anti-receptor with binding, not shown).

the

while

to

IgG2,,

its

receptor.

The

was as effective as GR-20 pre-immune IgG had no

rMuIFN--yR

Relative

avidity’

XC

XCsR

M’

6 4 16 16 64

0.213 0.000 0.057 0.051 0.025

1 .280 0.416 0.596 0.499 0.471

1.4 333 77 77 12

64

0.027

0.271

0.7

64

0.030

0.155

0.009

g/ml

IgG fractiond IgG2a,

GR-22

Induction

Absorbance4j

Isotype

Polyclonal GR-20 GR-3

and

goat

Concentration4

Antibody

Binding

The capacity that antibodies had to interfere with the specific binding of rMuIFN-7 was measured (Fig. 2). In a 250-molar excess, mAb GR-20 completely inhibited binding of [32PJrMuIFN-y to its receptor as effectively as rMuIFN-y did. By contrast, the irrelevant mAb, B54-2, had no detectable effect on binding when used at the same concentration. The two mAbs specific for intracellular epitopes, GR-23 and GR-29, as expected, had no effect on binding. Of the remaining antibodies, only GR-22 significantly reduced bind-

Monoclonal antibodies first were evaluated for their capacity to bind to either permeabilized or intact cells (Fig. 1). Fixation, a necessary step for stabilization of permeabilized cells before flow cytometric analysis, was expected to reduce the capacity of most epitopes to bind antibody. To minimize that problem, 293R cells that expressed > 100,000 MuIFN-yR./ccll E18] were used. Monoclonal antibodies that were specific for intracellular cpitopes were expected to bind only to cells that had been pcrmeabiiized. all but two (GR-23 topes that appeared

flow cytometric analysis, bound to XCsR in the failed to bind to NS-lsR, the secreted extracelluAll ofthe other mAbs were confirmed as binding on the extracellular domain.

GR-27 GR-34 GR-23

1gM, 1gM, IgG2,. IgG2,

GR-29

IgG2a,,

128

0.038

0.189

0.002

GR-65

IgG2a,

GR-70

G2b.

128 128

0.035 0.056

GR-96 GR-150

IgA. IgG.

128 128

0.087 0.085

0.154 0.225 0.147 0.139

NCR’ NCR NCR NCR

‘The

concentration

bAbsorbance4o

for XC or XCsR

of primary observed

in an

antigen

sources

antibody ELISA

used using

to detect

the solid-phase

supernatants

from

antigen

XCsR

or

XC

(as described

(control)

cells

in as the

Materials source

and

Methods).

of solid-phase

antigen.

The

development

times

were equal for each antibody, required for development varied, maximal absorbance could ‘ Relative avidity is defined as the inverse of the dissociation

a double

reciprocal

concentrations analyzed

plot.

(abscissa) by

general

The

relative

against linear

avidity

the

absorbance

regression.

Coefficients

‘The IgG fraction from goat serum fraction was estimated to be 2 mg/15 V No consistent result (i.e. , coefficient

510

Journal

of Leukocyte

allowing for comparison of nonspecific and specific binding. However, because the time not be compared between antibodies. constant (KD), which is equivalent to the negative inverse of the abscissa intersect from for each antibody was estimated from concentrations in the linear range of a plot of geometrically doubling produced (ordinate) [30]. Values from a double reciprocal plot of concentration versus absorbance were

was

of correlation

isolated

mg total

using

IgG [19].

of correlation

Biology

Volume

>

Protein