- Email: [email protected]
A double determinant sandwich immunoassay for quantitation of serum monoclonal anti-I-A antibody
Journal oflmmunologicalMethods, 85 (1985) 279-294 Elsevier
279
JIM03731
A Double Determinant Sandwich Immunoassay for Quantitation of Serum Monoclonal Anti-I-A Antibody Ifor R. Williams and Linda L. Perry
1
Department of Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA 30322. U.S.A. (Received 16 August 1985~ accepted 4 September 1985)
A double-determinant sandwich radioimmunoassay (RIA) is described for the specific detection of anti-l-A k monoclonal antibody (mAb) in the sera of non-Ighb murine hosts undergoing anti-Ia immunotherapy. This RIA utilizes 2 previously undescribed mAb reagents generated against an Ia.17-specific m A b secreted by the 10-3.6 hybridoma. The first reagent, 7.34, is specific for Ighb-linked allotypic determinants on the Fc portion of IgG2a immunoglobulins as defined by the pattern of reactivity with normal sera from a panel of inbred and Igh recombinant inbred strains. The second reagent, 58.3, is an anti-idiotypic m A b recognizing unique determinants in the combining site of 10-3.6 immunoglobulins, as determined by the specificity of the 58.3 mAb in solid-phase RIA and the capacity of this reagent to inhibit the binding of labeled 10-3.6 m A b to l-Ak-expressing spleen cells. In an RIA procedure using purified 58.3 mAb as substrate and 12~l-labeled 7.34 as the detection reagent, serum concentrations of 10-3.6 as low as 1-5 n g / m l can be measured reproducibly after mathematical linearization of the sigmoid standard curve. In the present studies, the serum half-life of 10-3.6 m A b was calculated from assay data to be 3 - 5 h in I-A k homo- or heterozygotes and 72 h in non-l-A k mice. The serum level of 10-3.6 as a function of the mAb treatment protocol was also examined and results are considered with respect to the efficacy of different therapeutic regimens in prolonging transplant survival. Sandwich immunoassays of this type (RIA or ELISA) should provide a highly sensitive and specific means for monitoring serum mAb levels in individuals subjected to antibody immunotherapy for treatment of autoimmune disease, transplant rejection or tumor progression.
Key words: anti- la immunotherapy - monoclonal antibody - transplantation - solid-phase immunoassay
1 Please address correspondence to: Linda L. Perry, Ph.D., Department of Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA 30322, U.S.A. Abbreviations: H, histocompatibility; mAb, monoclonal antibody; MHC, major histocompatibility complex; PBS, phosphate-buffered saline; PBSA, phosphate-buffered saline containing 1% ( w / w ) bovine serum albumin; RIA, radioimmunoassay. 0022-1759/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
280 Introduction
Antibodies against the murine I-A and I-E glycoproteins have been shown by many investigators to inhibit in vitro immune responses requiring helper T cell recognition of antigen in the context of these class II major histocompatibility complex (MHC) molecules (Baxevanis et al., 1980; Dubreuil et al., 1982). Anti-Ia antibodies also manifest in vivo activity as evidenced by their ability to inhibit delayed-type hypersensitivity (Perry et al., 1980) and antibody production (Rosenbaum et al., 1981) in a haplotype-specific manner. These findings have prompted investigations into the immunosuppressive activity of host-specific monoclonal anticlass II antibodies in autoimmune disease and transplantation. Monoclonal anti-la immunotherapy has been demonstrated to reduce the symptomatology of autoimmune disease in murine experimental models of encephalomyelitis (Steinman et al., 1981), myasthenia gravis (Waldor et al., 1983), lupus (Adelman et al, 1983) and arthritis (Wooley et al., 1985). Antibody treatment is also effective in prolonging the survival of mouse skin allografts differing at multiple minor histocompatibility (H) antigens or class I MHC antigens (Perry and Williams, 1985). The immunoregulatory effects of anti-I-A immunotherapy in transplantation have been confirmed in a murine heart allograft model where a 10 day course of antibody treatment has permitted indefinite survival ( > 90 days) of multiple minor H antigen incompatible hearts in 50% of recipients (Shelby, Williams and Perry, manuscript in preparation). The studies described in this communication represent an initial characterization of the pharmacokinetics of in vivo anti-Ia-immunotherapy. A sensitive and specific sandwich immunoassay is described that permits measurement of serum concentrations of the 10-3.6 hybridoma-derived anti-I-A k antibody as low as 1-5 n g / m l and relies upon the use of monoclonal antibodies specific for 2 spatially distinct epitopes on the 10-3.6 antibody molecule. This technique h~as shown that the 10-3.6 monoclonal anti-I-A k antibody is initially cleared with approximate first-order kinetics and has a serum half-life of 3-5 h in mice expressing the relevant I-A glycoprotein. The implications of these results for the design of anti-Ia treatment protocols for autoimmune disease and transplantation are discussed. This general method may be of value in measuring the pharmacokinetics of other monoclonal antibodies infused in vivo for therapeutic purposes.
Materials and Methods Mice
A / J (H-2~), B10.A/SgSnJ (H-2a), ( C 3 H / H e J × D B A / 2 J ) F 1 (H-2 k/~l) and D B A / 2 J (H-2 d) mice 6-10 weeks of age were purchased from The Jackson Laboratory, Bar Harbor, ME. C 3 H / H e J (H-2 k) mice were the kind gift of Dr. H.K. Ziegler of this department. Mice used in in vivo experiments were generally between 2 and 4 months of age. Production o f monoclonal antibodies against an anti-I-A ~ monoclonal antibody
The 10-3.6 hybridoma line produces 72a, x antibodies reactive with the Ia.17
281 TABLE I HYBRIDOMA LINES USED IN THIS STUDY Hybridoma line
Reactivity
Specificity
Antibody class
Igh haplotype of spleen cell donor
References
10-3.6 10-2.16 11-4.1 11-5.2 14-4-4S 7.34 58.3
I - A k'r'f's
la.17 Ia.17 H-2m.93 Ia.19 la.7 Allotype Idiotype
y2a,~ y2b,K y2a, x 72a, K 3'2a,~ 3,1 y2b
b b a a j c c
Oi et al. (1978) Oi et al. (1978) Oi et al. (1978) Oi et al. (1978) Ozato et al. (1980) This paper This paper
I-Ak'r'r'~ H-2Kk'r i-Ak'r I-Ek'd'pr 10-3.6 10-3.6
epitope on the A~ chain (Oi et al., 1978) and was obtained from American Type Culture Collection, Rockville, MD. Monoclonal antibodies reactive with idiotypic and allotypic determinants on the 10-3.6 molecule were obtained by immunizing D B A / 2 J mice with protein A-Sepharose-purified (Ey et al., 1978) 10-3.6 antibody in phosphate-buffered saline (PBS) emulsified in an equal volume of Freund's adjuvant (Difco Laboratories, Detroit, MI). These mice received an initial subcutaneous immunization with 100 /~g of 10-3.6 in complete Freund's adjuvant and were boosted intraperitoneally with 100/~g of 10-3.6 in incomplete Freund's adjuvant at biweekly intervals thereafter. After a total of 5 immunizations, the immune mice received a final intravenous injection of 100 ~g of antibody and 3 days later their spleen cells were fused with hypoxanthine-aminopterin-thymidine (HAT)-sensitive P3X63-AG8.653 myeloma cells (originally obtained from Dr. T. Springer of Harvard Medical School) using the procedure of Galfr6 et al. (1977) with minor modifications. Supernatants from wells containing growing hybridoma clones were screened by their ability to inhibit the binding of 1251-10-3.6 to A / J splenocytes as described below. From this fusion 2 clones producing antibodies against 10-3.6 were identified and subcloned to yield the 7.34 and 58.3 hybridoma lines. Double-diffusion Ouchterlony analysis revealed that 7.34 produces a 3,1 antibody, while 58.3 is ~,2b immunoglobulin. Binding o f radiolabeled 7.34 and 58.3 to 10-3.6 and other a n t i - M H C monoclonal antibodies Wells of polyvinyl microtiter plates (Falcon, Becton-Dickinson and Co., Oxnard, CA) were coated with 2.5 #g/well of 10-3.6 or other anti-class II monoclonal reagents by overnight incubation at 4°C. After washing to remove unbound antibody, excess protein binding sites were saturated by a 2 h incubation with 1% bovine serum albumin in PBS (PBSA). Protein A-Sepharose purified 7.34 and 58.3 antibodies and protein A (Pharmacia, Piscataway, NJ) were labeled with 125I by the chloramine-T method to a specific activity of approximately 10 ~Ci/~g. Twenty-five ~1 of each of these 125I-labeled reagents (50,000-70,000 cpm) diluted in PBSA were added to separate sets of antibody-coated wells. After 1 h, the wells were washed, cut
282 with a heated wire, and counted in a gamma counter. Measurements were performed in triplicate and the results expressed as the mean cpm bound.
Expression of the epitope defined by 7. 34 in normal sera from mice of c,arious Igh haplotvpes To determine if 7.34 a n d / o r 58.3 recognized epitopes present on immunoglobulins present in normal mouse serum, an inhibition RIA technique was used. Microtiter wells were coated with 1 ~g of 7.34 or 58.3 by overnight incubation at 4°C and, after washing, excess protein binding sites were saturated by an additional incubation with PBSA. To these wells were added serial dilutions of normal mouse sera from a series of different inbred strains. Normal sera from A / J , BALB/cByJ, C 3 H e B / F e J , D B A / 2 J , A K R / J , SJL/J, and C57BL/10J mice were obtained from mice housed in the vivarium at Emory University. Fractions containing various immunoglobulin classes were prepared from C57BL/10 serum by protein A-Sepharose fractionation as described by Ey el al. (1978). Normal sera from the Igh recombinant and congenic mouse strains BAB/14, C.B-20, C . B R / 3 and C . B R / 8 were the generous gift of Dr. M. Cancro of the University of Pennsylvania School of Medicine. After a 1 h incubation with the serum dilutions, the wells were washed and 25/xl of 1251-10-3.6 (50,000-70,000 cpm) added to each well. After another 1 h incubation, the plates were washed and the individual wells counted in a gamma counter. Interaction of serum antibody with 7.34 or 58.3 in the initial incubation was revealed as a decrease in the binding of radiolabeled 10-3.6 in the second step. Results are expressed as the % inhibition of the binding of ~251-10-3.6. Inhibition of I-"~I-10-3.6 binding to I-A ~ bearing cells Microtiter wells were coated with monolayers of erythrocyte-free spleen cells using a modification of the method of Heusser et al. (1981). Briefly, individual wells were pretreated with 50 /xl of a 100 /~g/ml solution of poly-L-lysine hydrobromide (MW 4000) (Sigma Chemical Co., St. Louis, MO) in PBS for 1 h at 37°C. After washing, 50 /~1 of a suspension (2 × 107) of Tris-NH4Cl-treated splenocytes were added to each well and the entire plate centrifuged (500 x g, 5 min). The plates were then flicked into a sink to empty the wells and the cell monolayers fixed by addition of 200/~1 of freshly prepared 0.25% glutaraldehyde in PBS for 5 min. After fixation, the wells were washed with a 2.19% solution of glycine and then used immediately or stored at 4°C with all wells containing a buffer composed of 3.5% sucrose, 0.05 M sodium cacodylate and 0.1% sodium azide. The ability of the 7.34 and 58.3 antibodies to block binding of ~251-10-3.6 to I-Ak-expressing spleen cells was tested by mixing 25 p~l of ~251-10-3.6 (60,000 cpm) and 25 ktl of dilutions of 7.34 or 58.3 and after 1 h adding the mixture to PBSA-pretreated wells of fixed spleen cells. After another hour, the wells were washed and the bound radioactivity quantitated by g a m m a counting. The results are expressed as the % inhibition of 1251-10-3.6 binding relative to control wells in which only ~51-10-3.6 was present. Competition radioimmunoassay to determine spatial relationship of epitopes recognized by 7.34 and 58.3 To determine whether 7.34 and 58.3 bound to spatially related or separate
283 epitopes on the 10-3.6 antibody molecule, serial dilutions of 7.34 or 58.3 were added to PBSA-treated microtiter wells coated with 5 ~g of 10-3.6. After 1 h, the wells were washed and 25 ~1 of ~251-58.3 were added to each well. After a second 1 h incubation, the wells were washed and the bound radioactivity measured in a gamma counter.
Double determinant sandwich radioimmunoassay to quantitate serum levels of 10-3. 6 Microtiter wells were coated with 1 ~g of 58.3 antibody by overnight incubation at 4°C. The plates were then washed and treated with PBSA to block remaining non-specific protein binding sites. Next 25 ~1 of dilutions of a 10-3.6 standard or serum samples were added to the 58.3 coated wells in triplicate. After 1 h the plates were washed and 25 ~1 of 125I-7.34 diluted in PBSA (50,000 70,000 cpm) were added to each well. After another 1 h incubation, the wells were washed and the bound radioactivity quantitated by gamma counting. Mathematical analysis of sandwich immunoassay data The standard curve obtained by plotting cpm bound vs. concentration of 10-3.6 was sigmoid shaped. In order to represent the same data by a linear function, the technique of logit transformation was utilized (Wellington, 1980). Briefly, if R, = radioactivity bound for the ith standard; C, = concentration of the ith standard; R 0 = non-specific background binding of 125I-7.34 in the absence of any 10-3.6; and R m = the maximum radioactivity bound at an infinite dose of 10-3.6, then Y, is defined as Yi = ( R i - R o ) / ( R . , - Ro)
(1)
and the least squares solution to the equation ln(Yi/(1 - Yi)) = b + mlnC i
(2)
where b represents the y-intercept and m the slope, can be obtained by a linear regression method. These calculations were performed using the SuperCalc3 electronic spreadsheet ( S o r c i m / I U S Micro Software, San Jos6, CA) on an IBM-PC. Initially, R m was assigned the value of the maximum number of counts bound to a single well in the actual assay. Then the value of R m was optimized by an iterative procedure that determined what value of R,,, gave the maximum correlation coefficient for the least squares solution of equation (2). The concentration of 10-3.6 in serum samples was then calculated from equation (2). The half-life of 10-3.6 in individual strains was calculated from the slope (m~) of the initial linear portion of the semilogarithmic antibody clearance curve using the formula
t,/2
= - (log 2)/me
(3)
Skin grafting Split-thickness tail skin grafts were performed and monitored daily until rejection
284
as described previously (Perry and Williams, 1985). Anti-I-A-treated mice received intraperitoneal injections of protein A-Sepharose-purified 10-3.6 antibody. Results are reported as the mean survival time of groups of 5 - 6 mice. The significance of differences in graft survival time between groups was assessed using the nonparametric Wilcoxon rank sum test (Lehmann, 1975).
Results
Characterization of the 7.34 and 58.3 monoclonal antibodies Two hybridoma lines, 7.34 and 58.3, were selected on the basis of antibody binding to the 10-3.6 anti-Ia.17 mAb in solid-phase RIA. The specificity of these reagents was assessed by several methods including direct binding of radioiodinated 7.34 and 58.3 to microtiter welIs coated with one of a series of anti-H-2 mAbs directed against class I or class II M H C antigens (Table II). Out of the panel tested, both 7.34 and 58.3 reacted only with 10-3.6; neither 7.34 or 58.3 bound to 10-2.16, another anti-Ia.17 mAb derived from the same fusion, but with a different isotype (~,2b) and different idiotypic determinants (Sher et al., 1984). To further characterize the specificity of these anti-10-3.6 mAbs, ailoimmune anti-I-A k and normal sera were obtained from a panel of mouse strains representing several of the defined Igh haplotypes including Ighh (the haplotype of the CWB strain used as spleen cell donor in derivation of the 10-3.6 hybridoma) and tested for expression of the epitopes defined by 7.34 and 58.3. An inhibition assay was used in which the presence of reactive material in a serum sample was detected by the ability of serum to inhibit the interaction between 7.34 or 58.3 coated to microtiter wells and t251-10-3.6. Sera from Igh h mice or any other Igh haplotype tested was not found to inhibit the binding of ~251-10-3.6 and 58.3 (data not shown), suggesting that 58.3 is directed against a private determinant specific for the 10-3.6 antibody
TABLE II SPECIFICITY OF A N T I B O D I E S P R E P A R E D A G A I N S T 10-3.6 Monoclonal antibodies 7.34 and 58.3 were labeled with 125I and tested for their binding to 10-3.6 and other anti-MHC hybridoma antibodies that had been absorbed to the surface of microtiter wells. Radioiodinated protein A was used as a positive control to make certain that all antibodies tested as substrates had bound to the microtiter wells. Antibody absorbed to microtiter wells
cpm ~25I-labeled reagent bound 1251_7.34
~251_58.3
None 10-3.6 10-2.16 11-5.2 11-4.1 14-4-4S
36 29964 145 56 79 72
33 27619 28 96 82 64
~zSI_proteinA 48 41 805 16219 29 589 39753 8 898
285 TABLE III SPECIFICITY OF 7.34 MONOCLONAL ANTIBODY Serum or serum fractions of various inbred mouse strains were tested for their ability to react with the 7.34 mAb by measuring their ability to bind to microtiter wells coated with 7.34 (1 p,g per well) and inhibit the binding of 1251-10-3.6 in a subsequent binding step. The results are expressed as the % inhibition relative to the PBSA control wells (0% inhibition). Experiment
Inhibiting serum
Igh-C haplotype
% Inhibition of t251-10-3.6 binding at indicated dilution 1:10
1:100
Reactivity with 7.34
BALB/cByJ C57BL/10J (B10) DBA/2J AKR/J A/J C3HHeB/FeJ
a b c d e j
3.0 94.7 9.7 5.4 8.1 - 0.2
3.1 91.8 9.4 9.1 6.5 0.6
+ -
BALB/cByJ B10 SJL/J C.B-20 ~ BAB.14 " C.BR/3 ~' C.BR/8 a BI0 non-IgG b B10 IgG2a b
a b b b b a a b b
3.9 95.4 94.7 95.5 94.0 11.2 19.9 ND c ND
20.2 94.4 94.1 92.0 77.5 6.5 4.3 7.4 91.9
+ + + + +
a The alleles of the Igh congenic and recombinant mice at the Dex, lgh-C and Pre-1 loci on chromosome 12 are summarized below. Crossover regions for the recombinant strains are indicated by a / . These data were compiled from Weigert and Riblet (1978) and Forman et al. (1984). Strain
Allele/antigen Dex
C.B-20 BAB.14 C.BR/3 C.BR/8
+ +
/ /
Igh
Pre-1
b b a a
0 0 a 0
/
b C57BL/10 NMS was fractionated as described in Materials and Methods. The effluent from the column when the serum was applied to the column at pH 8.0 is termed the non-lgG fraction. The igG2a fraction is that material that initially bound to the column, was not eluted from the column by pH 5.5 buffer, but was subsequently eluted by application of pH 4.5 buffer. c Not done.
m o l e c u l e . H o w e v e r , t h e e p i t o p e d e f i n e d b y 7 . 3 4 w a s p r e s e n t in n o r m a l m o u s e s e r u m from 2 Ighb strains, C57BL10flJ and SJL/J,
but not serum from strains representing
t h e a, c, d, e o r j I g h h a p l o t y p e s ( T a b l e I I I ) . A l s o s h o w n in T a b l e I I I a r e t h e r e s u l t s of an experiment performed of the determinant
to more precisely map the genetic control of expression
r e c o g n i z e d b y 7.34. T h e a b i l i t y o f s e r u m f r o m t h e I g h c o n g e n i c
286 IOO
Ccn
8O
h5 6O"
FO
40 ¸
I
,~
20
o
fi ~,
20
-40 0,001
. . . . . . . .
j
0.01
. . . . . . .
,
O. 1
. . . . . . . .
,
l
. . . . . . . .
,
. . . . . . . .
10
i
100
. . . . . . . .
I000
Antibody Concentration (ug/ml)
Fig. 1. Inhibition of 1251-10-3.6binding to I-Ak by 7.34 and 58.3. Twenty-five/tl of a series of dilutions of purified 7.34 and 58.3 were mixed with 25/LI of 1251-10-3.6and added to microtiter wells containing fixed BI0.A spleen cells. After 1 h of incubation, the wells were washed and the bound radioactivity quantitated by radioimmunoassay. The results are expressed as the % inhibition of 1251-10-3.6binding compared to control wells in which only radiolabeled 10-3.6 and PBSA were present.
strain C-B.20 to inhibit the interaction of 7.34 and 58.3 demonstrated that the epitope defined by 7.34 was under the control of genes linked to the Igh locus. To ascertain whether these genes m a p p e d to the Igh-V or Igh-C, additional serum samples from strains recombinant within the Igh locus (BAB.14, C . B R / 3 and C . B R / 8 ) were tested. BAB.14 is an intra-V H recombinant strain that expresses Igh-V markers c o n t r i b u t e d . b y the B A L B / c background strain (Igh ") and Igh-C determinants controlled by the Igh b haplotype (Weigert and Riblet, 1978). C . B R / 3 expresses Igh-V markers of the Igh b haplotype and Igh-C determinants from the Igh ~ haplotype, while C . B R / 8 expresses Igh-V and Igh-C alleles of the Igh~ haplotype ( F o r m a n et al., 1984). Serum samples from these 3 Igh recombinant strains interacted with 7.34 in a pattern predicted by the Igh-C allele, suggesting that 7.34 was directed against an allotypic determinant specified by the Ig-1 locus within the Igh-C region. Results of protein A-Sepharose fractionation of serum from C 5 7 B L / 1 0 J mice (Igh b) are consistent with this interpretation; 7.34-reactive molecules were retained by a protein A-Sepharose i m m u n o a b s o r b e n t at p H 8.0 and eluted when the p H of the buffer was reduced to 4.5 (Table III). The relative ability of the 7.34 and 58.3 antibodies to inhibit the binding of ]251-10-3.6 to B10.A spleen cells was examined to determine whether either of these antibodies recognized binding site determinants on the 10-3.6 m A b molecule (Fig. 1). The 58.3 antibody inhibited the binding of radiolabeled 10-3.6 to I-A k over a wide range of concentrations, with 50% inhibition attained using a 58.3 concentration of approximately 0.03 /~g/ml. Significant inhibition of 1251-10-3.6 binding by 7.34 was only observed at very high concentrations ( > 10 /Lg/ml) and concentra-
287 100
8O
/
60
g I ~
4o-
,/
z g ~ )5
J
J J
~3 c
20-
/
/
/
C
-20 . • 0.01
.
.
. 0.1
Antibody
. I Concentration
10 (ug/ml)
Fig. 2. The 7.34 and 58.3 antibodies recognize spatially distinct epitopes. Microtiter wells were coated with 5 fig of 10-3.6 by overnight absorption at 4°C. After PBSA treatment to block residual protein binding sites, serial dilutions of 7.34 and 58.3 were added to the wells. After 1 h these antibodies were washed out and 25 p,l of 1251-58.3 added. After another hour, the wells were washed and the bound radiolabel quantitated by gamma counting. tions of 0.1 and 0.3 t t g / m l were reproducibly found to potentiate the binding of radiolabeled 10-3.6. These results viewed in light of those presented earlier indicate that 58.3 is directed against an idiotypic determinant present on the binding site of 10-3.6, while 7.34 recognizes an allotypic determinant elsewhere on the 10-3,6 molecule. To confirm that 7.34 and 58.3 recognized spatially distinct epitopes, an inhibition assay was performed in which 10-3.6 coated wells were allowed to interact with various dilutions of unlabeled 7.34 or 58.3 antibody prior to addition of radioiodinated 58.3. As shown in Fig. 2, 7.34 did not cause any inhibition of the binding of t251-58.3 to the 10-3.6 coated wells, while the homologous unlabeled reagent (58.3) was inhibitory in a dose-dependent fashion. Finally, other experiments have demonstrated that F(ab')2 fragments of 10-3.6 prepared by pepsin digestion of whole antibody retain the ability to bind to 58.3, but do not bind to 7.34 (data not shown). Thus the epitope recognized by 7.34 is an allotypic determinant controlled by Igh-C genes and located on the Fc portion of the immunoglobulin molecule.
Double determinant sandwich immunoassay for 10-3.6 using the 7.34 and 58.3 monoclonal antibodies Several different types of immunoassays for 10-3.6 using 7.34 a n d / o r 58.3 were evaluated with respect to sensitivity. The most sensitive assay was found to be a sandwich immunoassay employing 58.3 absorbed to plastic microtiter wells as a capture antibody and radiolabeled 7.34 as a detection antibody. Similar assays using monoclonal antibodies recognizing spatially distinct epitopes on an antigen of interest have been described for quantitating levels of alpha-fetoprotein (Uotila et
288
Untransformed
RIA Data
25000
/ ~0000
/
•
15000
~r~
/
10000
f
/ /
0_~--
,/
/
,/
-~
-3
0
LN C o n c e n t r a t . i o n
Logit A n a l y s i s of RIA Data 4
t3 Jd 2 //." o /-
2 -2
y
/,
/ i:I
/
/i
4
Ef~
-9
-6
-3
LN Concentration
0
--~
3
Fig. 3. Standard curve obtained rising double-determinant radioimmunoassay for 10-3.6. A series of standards ranging in concentration from 1 n g / m l to 4096 n g / m l was used. Part A demonstrates the sigmoid shape of the plot of cpm bound against the natural logarithm of 10-3.6 concentration. Part B shows the identical set of data after the logit function has been used to linearize the data.
al., 1981), human Ia antigens (Russo et al., 1983), and human gamma interferon (Tanaka et al., 1985). Fig. 3A represents a standard curve obtained using the sandwich immunoassay and samples of 10-3.6 mAb ranging in concentration from 1 n g / m l to 4096 n g / m l . A sigmoid-shaped curve is obtained that is typical for this type of RIA. In order to utilize this assay for determination of 10-3.6 concentrations in serum samples, it was necessary to develop a mathematical formulation of the relationship between the number of counts bound and the concentration of 10-3.6. The logit function was utilized to linearize the sigmoid standard curve so that a least-squares linear regression calculation could be performed. The linear relationship between the logit function and the natural logarithm of the antibody concentration is demonstrated in Fig. 3B for the same set of standards depicted in Fig. 3A.
289
1000 I~C DBA/2 ..... (I-A') (l--A'/d)II c~H {,-A,)
E
c
1
1oo
~o
c) c
8
'
cO I
0
O,I
I'0
'
2'0
'
3'0
~4'0
5'0
B/O
Hours after Antibody Injection Fig. 4. Clearance of 10-3.6 from 3 mouse strains. Groups of 3 C 3 H / H e J , C3D2F 1 and D B A / 2 mice were injected intraperitoneally with 200 t~g of purified 10-3.6 antibody. Pooled serum samples were collected from each group of mice every 6 h for 2 days. The serum was frozen and later assayed for 10-3.6 using the double determinant immunoassay.
Application of sandwich assay for 10-3.6 to determination of antibody half-life in serum of treated mice The sandwich assay described above was used to follow the clearance of a single dose of 200 /~g of 10-3,6 injected intraperitoneally into mice of several MHC haplotypes. Serum samples were obtained at 6 h intervals for the 2 days after antibody injection and several dilutions of each sample assayed for the level of 10-3.6. As seen in Fig. 4, the clearance pattern following antibody injection was determined by the I-A alleles possessed by each strain used. In both homozygous I-Ak mice (C3H) and heterozygous I-Ak mice (C3D2F 1), the level of 10-3.6 initially declined following approximately first-order kinetics. The half-lives calculated by linear regression analysis of the initial linear portions of these clearance curves and
TABLE IV HALF-LIFE OF 10-3.6 ANTIBODY IN AN I-A k H O M O Z Y G O U S STRAIN, AN I-Ak HETEROZYGOUS STRAIN, A N D A NON-I-A k STRAIN Mice from 3 inbred strains received an intraperitoneal injection of 200 ttg of 10-3.6 antibody. The serum half-life of 10-3,6 in these strains was determined by using the initial linear portion of the clearance curve. Mouse strain
I-A genotype
Range of time points used to determine half-life (h)
Calculated 10-3.6 half-life (h)
C3H/HeJ C3D2F~ DBA/2J
k k/d d
12-24 h 6-18 h 6-54 b
3.5 4.5 72.7
290 TABLE V RELATIONSHIP BETWEEN TIMING OF ANTI-Ia IMMUNOTHERAPY AND PROLONGATION OF SKIN GRAFT SURVIVAL Groups of 5 6 A/J mice received BI0.A tail skin grafts on day 0. Mice were left untreated, were treated with 1000 ~tg of 10-3.6 mAb on day - 1 and 500 ~tg on day 1 and 2. or were treated with 200/Lg of 10-3.6 on days 0-9. Host
Donor
n
Treatment schedule ~g 10-3.6 mAb (days treated)
Total 10-3.6 dose (mg)
MST_+SEM
A/J A/J A/J
B10.A B10.A B10.A
5 6 6
1000 ( - 1); 500 (1,2) 200 (0-9)
2.0 2.0
9.8 _+0.7 11.8 ± 0.5 ~' 16.0 _+1.0 b
a P < 0.05. b p < 0.005.
e q u a t i o n (3) are 3.5 h for C3H a n d 4.5 h for C 3 D 2 F 1 (Table IV). After 24 h the rate of clearance began to gradually decrease. In mice not expressing an I-A molecule reactive with 10-3.6 ( D B A / 2 ) , a significantly different p a t t e r n of clearance was observed. U s i n g the c o n c e n t r a t i o n s d e t e r m i n e d at all time points examined in D B A / 2 mice, a 10-3.6 serum half-life of 72.7 h was calculated.
Effect of anti-l-A treatment protocol on serum antibody levels and immunoregulatory effect In previous studies of a n t i - I - A i m m u n o t h e r a p y in t r a n s p l a n t a t i o n c o n d u c t e d in this laboratory, a treatment protocol of 200 ~g of a n t i b o d y daily on days 0 - 9 was routinely used (Perry a n d Williams, 1985). However, other investigators utilizing a n t i - I - A antibodies as a therapeutic approach in experimental a u t o i m m u n e disease have developed protocols in which all a n t i b o d y is administered within 1 or 2 days of a u t o a n t i g e n i m m u n i z a t i o n ( S t e i n m a n et al., 1981). To compare the i m m u n o r e g u l a tory effects of these 2 types of treatment protocols in t r a n s p l a n t a t i o n , groups of A / J mice were grafted with B10.A tail skin and treated with 200 #g of m A b on days 0 - 9 (Protocol A) or 1000 ~g o n day - 1 and 500 ~g on days 1 and 2 (Protocol B) (total doses of 2.0 mg of 10-3.6 in both cases). Protocol B used the same days of a d m i n i s t r a t i o n a n d the same a p p r o p r i a t i o n of total m A b dose as used by S t e i n m a n et al. (1981) in studies o n m u r i n e encephalomyelitis. The skin graft survival times listed in T a b l e V d e m o n s t r a t e that a n t i b o d y treatment over a 10 day interval was associated with a m e a n graft survival time significantly ( P < 0.005) greater than the same total m A b dose administered between day - 1 a n d day 2. Serum samples were also taken daily from the a n t i b o d y - t r e a t e d mice in this experiment. The serum samples were taken immediately prior to the daily a n t i b o d y injection, a n d thus represent trough rather than peak serum levels in the mice treated daily with antibody. The results of this analysis presented in Fig. 5 reveal that mice treated according to Protocol A m a i n t a i n e d trough a n t i b o d y levels between 5 a n d 25
291
10 3.6 Treotment Protocol I 0.2 m9 Days 0-9 1.0 mg Doy -1; 05 rnq Ooys ~,2 lOB
k~
?
-,~ io (J CD
? o
--
½
"
6. ; ; I '0 Days after Transplantation
I
'2
I '4
Fig. 5. Effect of treatment protocol on serum 10-3.6 levels. Groups of 6 A / J mice were given B10.A skin grafts on day 0. One treated group received 200 ~g of 10-3.6 on days 0-9. A separate treated group received 1000/~g of 10-3.6 on day 0 and 500 p.g on days 1 and 2. Serum samples were collected on a daily basis and assayed for 10-3.6 using the double determinant immunoassay.
/~g/ml through day 10. In contrast, mice treated by Protocol B had higher antibody levels on days 1 through 3 (20-120/~g/ml), similar levels on day 4, and less than 5 # g / m l by day 5 and thereafter. Based upon the graft survival data under these 2 treatment protocols, it appears that maintenance of a minimal serum concentration of mAb over a prolonged period is superior to administration of the same total dose near the time of transplantation. Discussion
Antibodies against class II MHC molecules have been demonstrated to have immunosuppressive activity in experimental autoimmune disease and in transplantation. Anti-Ia immunotherapy is associated with the development of antigen-specific T cell-mediated suppression (Perry and Greene, 1982) to antigens presented during the treatment interval, while secondary responses to antigens encountered outside the treatment interval remain intact (Perry and Williams, 1985). As with other therapeutic agents, the development of optimal treatment protocols using these agents must be predicated on an understanding of their pharmacokinetics. Many possible immunoassay procedures for measurement of serum levels of monoclonal anti-Ia antibodies exist. Other investigators have followed serum levels of monoclonal anti-I-A antibody by flow cytometric (Waldor et al., 1984) or microcytotoxicity (Kruisbeek et al., 1985) methods. The sandwich assay described in this manuscript is substantially more sensitive than either of these other methods while retaining a high degree of specificity. The assay utilizes 2 monoclonal antibodies recognizing spatially distinct epitopes
292 on the 10-3.6 antibody molecule. The 7.34 hybridoma produces antibodies reacting with an allotypic determinant found on IgG2a immunoglobulin molecules from mice of the Ighb haplotype. Monoclonal antibodies directed against allotypic determinants controlled by the Ig-1 h locus have been described previously by other laboratories (Oi and Herzenberg, 1979; DiPauli and Raschke, 1978), but the relationship between the epitope defined by 7.34 and epitopes defined previously has not been addressed. The 58.3 mAb is directed against an idiotypic determinant of the 10-3.6 molecule as evidenced by its ability to block the binding of 10-3.6 to I-A k. Further evidence of the anti-idiotypic activity of 58.3 has been obtained by immunizing syngeneic ( D B A / 2 ) mice with this antibody; some of these mice responded by producing anti-I-A antibodies with the same epitope specificity as 10-3.6-derived immunoglobulin (Williams and Perry, data not shown). Clearance data obtained after a single 200 #g dose of 10-3.6 antibody in 3 inbred mouse strains reveals that the antibody is rapidly cleared from the circulation in mice that express the relevant class II molecules. The half-life calculated on the basis of the initial rate of decay is slightly shorter in homozygous I-A k mice (3.5 h) than in heterozygous I-Ak/d mice (4.5 h) as might be expected given the greater total number of target class II molecules in the former group. The half-life of 10-3.6 antibody in D B A / 2 mice lacking the Ia.17 specificity was calculated to be approximately 3.0 days. Fahey and Sell (1965) found that 2 IgG2a myeloma proteins had half-lives of 4.8 and 5.4 days, respectively, in normal mice. Thus, the clearance of 10-3.6 from the serum of D B A / 2 mice follows kinetics similar to those described for other IgG2a immunoglobulins. The rapid clearance of 10-3.6 in C3H and C3D2F 1 mice cannot be attributed to the administration of denatured or damaged immunoglobulin molecules because of the extended half-life observed in the D B A / 2 controls. Extrapolation of the clearance curves (Fig. 4) to 0 h for all strains examined gives an initial serum level of approximately 100/~g/ml. Since 200/xg were injected into these animals, the volume of distribution can be estimated as 2.0 ml. Given that the mice used in these experiments weighed between 20 and 25 g and total extracellular body water is approximately 20% of total body weight (Ganong, 1979), these data suggest that the administered anti-I-A antibody is able to distribute itself throughout the plasma (5% of body weight) as well as some other components of the extracellular fluid compartment. Therapeutic agents with short in vivo half-lives are generally administered either as a continous infusion or intermittently at intervals that allow maintenance of therapeutic levels in the body (Goodman et al., 1980). The serum concentration of anti-class II antibody necessary to manifest maximal immunosuppression has not been defined in most systems, although dose dependence has been clearly demonstrated (Perry and Williams, 1985). The parameter correlating most closely with the efficacy of treatment may be the residual expression of class II antigens by host antigen-presenting cells following mAb treatment, as suggested by the studies of Kruisbeek et al. (1985). The rapid re-expression of class II antigens after mAb-mediated modulation probably represents one basis for the continuous requirement of anti-Ia antibody to maintain the blockade of antigen presentation in the context of class II restriction elements.
293
Evidence for a direct correlation between monoclonal anti-Ia antibody serum levels and the immunosuppressive effects of anti-Ia treatment were obtained in a murine skin graft model of transplantation. Administration of 2.0 mg of 10-3.6 in 10 equal daily doses beginning with the day of transplantation was associated with significantly greater graft prolongation than administration of the entire 2.0 mg between the day before transplantation and 2 days after. The first regimen was associated with significantly higher serum antibody levels from day 5 onward, demonstrating that the maintenance of a 5-25 ~ g / m l serum concentration of Ab for 10 days is more favorable to the prolongation of skin graft survival, presumably due to extended inhibition of antigen presentation under these conditions. Future studies using this radioimmunoassay will address more definitively the serum level of anti-class II antibody necessary to maintain maximal inhibition of Ia expression, antigen presentation and induction of antigen-reactive T cells. If anti-Ia immunotherapy continues to show promise as an immunotherapeutic approach in autoimmunity and transplantation, further characterization of the pharmacokinetics of these antibodies will be valuable in designing optimal treatment protocols. Radioimmunoassays patterned after the one described herein can be developed for the quantitation of serum and tissue levels of other monoclonal immunoglobulins being considered for in vivo therapeutic use. The sandwich immunoassay technique for measurement of in vivo mAb levels could also be performed as an enzyme immunoassay. Such assays will only require mAbs directed against separate epitopes on the antibody of interest that do not cross-react with the immunoglobulins normally present in the treated host.
Acknowledgements We thank Dr. Michael Cancro for providing serum samples from C.B-20, BAB.14, C . B R / 3 and C . B R / 8 mice in his colony and for his helpful suggestions. Ms. Brenda Hott and Ms. Nancy Martin are gratefully acknowledged for their technical assistance in preparation of the 7.34 and 58.3 hybridomas. These studies were supported by Grant CA-32018 from the National Cancer Institute.
References Adelman N.E., D.L. Watling and H.O. McDevitt, 1983, J. Exp. Med. 158, 1350. Baxevanis, C.N., D. Wernet, Z.A. Nagy, P.H. Maurer and J. Klein, 1980, Immunogenetics 11,617. DiPauli, R. and W.C. Raschke, 1978, Curr. Top. Microbiol. Immunol. 81, 37. Dubreuil, P.C., D.Z. Birnbaum, D.H. Caillol and F.A. Lemonnier, 1982, Immunogenetics 16, 407. Ey, P.L., S.J. Prowse and C.R. Jenkin, 1978, Immunochemistry 15, 429. Fahey, J.L. and S. Sell, 1965, J. Exp. Med. 122, 41. Forman, J., R. Riblet, K. Brooks, E.S. Vitetta and L.A. Henderson, 1984, J. Exp. Med. 159, 1724. Galfr6, G., S.C. Howe, C. Milstein, G.W. Butcher and J.C. Howard, 1977, Nature (London) 266, 550. Ganong, W.F., 1979, Review of Medical Physiology (Lange Medical Publications, Los Altos, CA) p. 6.
294 Goodman, A.G., L.S. Goodman and A. Gilman (eds.), 1980, The Pharmacological Basis of Therapeutics (MacMillan Publishing Co., New York) pp. 21-27. Heusser, C.H., J.W. Stocker and R.H. Gisler, 1981, Methods Enzymol. 73, 406. Kruisbeek, A.M., J.A. Titus, D.A. Stephany, B.L. Gause and D.L. Longo, 1985, J. Immunol. 134, 3605. Lehmann, E.L., 1975, Nonparametrics: Statistical Methods Based on Ranks (Holden-Day, San Francisco, CA). Oi, V.T. and L.A. Herzenberg, 1979, Mol. Immunol. 16, 1005. Oi, V.T., P.P. Jones, J.W. Goding, L.A. Herzenberg and L.A. Herzenberg, 1978, Curr. Top. Microbiol. Immunol. 81, 115. Ozato, K., N. Mayer and D.H. Sachs, 1980, J. Immunol. 124, 533. Perry, L.L. and M.I. Greene, 1982, J. Exp. Med. 156, 480. Perry, L.L. and I.R. Williams, 1985, J. Immunol. 134, 2935. Perry, L.L., M.E. Doff, B.A. Bach, B. Benacerraf and M.I. Greene, 1980, Clin. Immunol. lmmunopathol. 15, 279. Rosenbaum, J.T., N.E. Adelman and H.O. McDevitt, 1981, J. Exp. Med. 154, 1694. Russo, C., M.A. Pellegrino and S. Ferrone, 1983, Trans. Proc. 15, 57. Sher, M.R., T.P. Bender and J.E. Niederhuber, 1984, J. Immunol. 133, 338. Steinman, L., J.T. Rosenbaum, S. Sriram and H.O. McDevitt, 1981, Proc. Natl. Acad. Sci. U.S.A. 78, 7111. Tanaka, E., M. lmai, S. Usada, K. Tachibana, H. Okamoto, Y. Ohike, T. Nakamura, Y. Miyakawa and M. Mayumi, 1985, J. Immunol. Methods 77, 275. Uotila, M., E. Ruoslahti and E. Engvall, 1981, J. Immunol. Methods 42, 11. Waldor, M.K., S. Sriram, H.O. McDevitt and L. Steinman, 1983, Proc. Natl. Acad. Sci. U.S.A. 80, 2713. Waldor, M.K., R.R. Hardy, K. Hayakawa, L. Steinman, L.A. Herzenberg and L.A. Herzenberg, 1984, Proc. Natl. Acad. Sci. U.S.A. 81, 2855. Weigert, M. and R. Riblet, 1978, Springer Semin. immunopathol. 1, 133. Wellington, D., 1980, in: Enzyme-lmmunoassay, ed. E.T. Maggio (CRC Press, Boca Raton, FL) Ch. 12, p. 249. Wooley, P.H., H.S. Luthra, W.P. Lafuse, A. Huse, J.M. Stuart and C.S. David, 1985, J. lmmunol. 134, 2366.
279
JIM03731
A Double Determinant Sandwich Immunoassay for Quantitation of Serum Monoclonal Anti-I-A Antibody Ifor R. Williams and Linda L. Perry
1
Department of Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA 30322. U.S.A. (Received 16 August 1985~ accepted 4 September 1985)
A double-determinant sandwich radioimmunoassay (RIA) is described for the specific detection of anti-l-A k monoclonal antibody (mAb) in the sera of non-Ighb murine hosts undergoing anti-Ia immunotherapy. This RIA utilizes 2 previously undescribed mAb reagents generated against an Ia.17-specific m A b secreted by the 10-3.6 hybridoma. The first reagent, 7.34, is specific for Ighb-linked allotypic determinants on the Fc portion of IgG2a immunoglobulins as defined by the pattern of reactivity with normal sera from a panel of inbred and Igh recombinant inbred strains. The second reagent, 58.3, is an anti-idiotypic m A b recognizing unique determinants in the combining site of 10-3.6 immunoglobulins, as determined by the specificity of the 58.3 mAb in solid-phase RIA and the capacity of this reagent to inhibit the binding of labeled 10-3.6 m A b to l-Ak-expressing spleen cells. In an RIA procedure using purified 58.3 mAb as substrate and 12~l-labeled 7.34 as the detection reagent, serum concentrations of 10-3.6 as low as 1-5 n g / m l can be measured reproducibly after mathematical linearization of the sigmoid standard curve. In the present studies, the serum half-life of 10-3.6 m A b was calculated from assay data to be 3 - 5 h in I-A k homo- or heterozygotes and 72 h in non-l-A k mice. The serum level of 10-3.6 as a function of the mAb treatment protocol was also examined and results are considered with respect to the efficacy of different therapeutic regimens in prolonging transplant survival. Sandwich immunoassays of this type (RIA or ELISA) should provide a highly sensitive and specific means for monitoring serum mAb levels in individuals subjected to antibody immunotherapy for treatment of autoimmune disease, transplant rejection or tumor progression.
Key words: anti- la immunotherapy - monoclonal antibody - transplantation - solid-phase immunoassay
1 Please address correspondence to: Linda L. Perry, Ph.D., Department of Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA 30322, U.S.A. Abbreviations: H, histocompatibility; mAb, monoclonal antibody; MHC, major histocompatibility complex; PBS, phosphate-buffered saline; PBSA, phosphate-buffered saline containing 1% ( w / w ) bovine serum albumin; RIA, radioimmunoassay. 0022-1759/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
280 Introduction
Antibodies against the murine I-A and I-E glycoproteins have been shown by many investigators to inhibit in vitro immune responses requiring helper T cell recognition of antigen in the context of these class II major histocompatibility complex (MHC) molecules (Baxevanis et al., 1980; Dubreuil et al., 1982). Anti-Ia antibodies also manifest in vivo activity as evidenced by their ability to inhibit delayed-type hypersensitivity (Perry et al., 1980) and antibody production (Rosenbaum et al., 1981) in a haplotype-specific manner. These findings have prompted investigations into the immunosuppressive activity of host-specific monoclonal anticlass II antibodies in autoimmune disease and transplantation. Monoclonal anti-la immunotherapy has been demonstrated to reduce the symptomatology of autoimmune disease in murine experimental models of encephalomyelitis (Steinman et al., 1981), myasthenia gravis (Waldor et al., 1983), lupus (Adelman et al, 1983) and arthritis (Wooley et al., 1985). Antibody treatment is also effective in prolonging the survival of mouse skin allografts differing at multiple minor histocompatibility (H) antigens or class I MHC antigens (Perry and Williams, 1985). The immunoregulatory effects of anti-I-A immunotherapy in transplantation have been confirmed in a murine heart allograft model where a 10 day course of antibody treatment has permitted indefinite survival ( > 90 days) of multiple minor H antigen incompatible hearts in 50% of recipients (Shelby, Williams and Perry, manuscript in preparation). The studies described in this communication represent an initial characterization of the pharmacokinetics of in vivo anti-Ia-immunotherapy. A sensitive and specific sandwich immunoassay is described that permits measurement of serum concentrations of the 10-3.6 hybridoma-derived anti-I-A k antibody as low as 1-5 n g / m l and relies upon the use of monoclonal antibodies specific for 2 spatially distinct epitopes on the 10-3.6 antibody molecule. This technique h~as shown that the 10-3.6 monoclonal anti-I-A k antibody is initially cleared with approximate first-order kinetics and has a serum half-life of 3-5 h in mice expressing the relevant I-A glycoprotein. The implications of these results for the design of anti-Ia treatment protocols for autoimmune disease and transplantation are discussed. This general method may be of value in measuring the pharmacokinetics of other monoclonal antibodies infused in vivo for therapeutic purposes.
Materials and Methods Mice
A / J (H-2~), B10.A/SgSnJ (H-2a), ( C 3 H / H e J × D B A / 2 J ) F 1 (H-2 k/~l) and D B A / 2 J (H-2 d) mice 6-10 weeks of age were purchased from The Jackson Laboratory, Bar Harbor, ME. C 3 H / H e J (H-2 k) mice were the kind gift of Dr. H.K. Ziegler of this department. Mice used in in vivo experiments were generally between 2 and 4 months of age. Production o f monoclonal antibodies against an anti-I-A ~ monoclonal antibody
The 10-3.6 hybridoma line produces 72a, x antibodies reactive with the Ia.17
281 TABLE I HYBRIDOMA LINES USED IN THIS STUDY Hybridoma line
Reactivity
Specificity
Antibody class
Igh haplotype of spleen cell donor
References
10-3.6 10-2.16 11-4.1 11-5.2 14-4-4S 7.34 58.3
I - A k'r'f's
la.17 Ia.17 H-2m.93 Ia.19 la.7 Allotype Idiotype
y2a,~ y2b,K y2a, x 72a, K 3'2a,~ 3,1 y2b
b b a a j c c
Oi et al. (1978) Oi et al. (1978) Oi et al. (1978) Oi et al. (1978) Ozato et al. (1980) This paper This paper
I-Ak'r'r'~ H-2Kk'r i-Ak'r I-Ek'd'pr 10-3.6 10-3.6
epitope on the A~ chain (Oi et al., 1978) and was obtained from American Type Culture Collection, Rockville, MD. Monoclonal antibodies reactive with idiotypic and allotypic determinants on the 10-3.6 molecule were obtained by immunizing D B A / 2 J mice with protein A-Sepharose-purified (Ey et al., 1978) 10-3.6 antibody in phosphate-buffered saline (PBS) emulsified in an equal volume of Freund's adjuvant (Difco Laboratories, Detroit, MI). These mice received an initial subcutaneous immunization with 100 /~g of 10-3.6 in complete Freund's adjuvant and were boosted intraperitoneally with 100/~g of 10-3.6 in incomplete Freund's adjuvant at biweekly intervals thereafter. After a total of 5 immunizations, the immune mice received a final intravenous injection of 100 ~g of antibody and 3 days later their spleen cells were fused with hypoxanthine-aminopterin-thymidine (HAT)-sensitive P3X63-AG8.653 myeloma cells (originally obtained from Dr. T. Springer of Harvard Medical School) using the procedure of Galfr6 et al. (1977) with minor modifications. Supernatants from wells containing growing hybridoma clones were screened by their ability to inhibit the binding of 1251-10-3.6 to A / J splenocytes as described below. From this fusion 2 clones producing antibodies against 10-3.6 were identified and subcloned to yield the 7.34 and 58.3 hybridoma lines. Double-diffusion Ouchterlony analysis revealed that 7.34 produces a 3,1 antibody, while 58.3 is ~,2b immunoglobulin. Binding o f radiolabeled 7.34 and 58.3 to 10-3.6 and other a n t i - M H C monoclonal antibodies Wells of polyvinyl microtiter plates (Falcon, Becton-Dickinson and Co., Oxnard, CA) were coated with 2.5 #g/well of 10-3.6 or other anti-class II monoclonal reagents by overnight incubation at 4°C. After washing to remove unbound antibody, excess protein binding sites were saturated by a 2 h incubation with 1% bovine serum albumin in PBS (PBSA). Protein A-Sepharose purified 7.34 and 58.3 antibodies and protein A (Pharmacia, Piscataway, NJ) were labeled with 125I by the chloramine-T method to a specific activity of approximately 10 ~Ci/~g. Twenty-five ~1 of each of these 125I-labeled reagents (50,000-70,000 cpm) diluted in PBSA were added to separate sets of antibody-coated wells. After 1 h, the wells were washed, cut
282 with a heated wire, and counted in a gamma counter. Measurements were performed in triplicate and the results expressed as the mean cpm bound.
Expression of the epitope defined by 7. 34 in normal sera from mice of c,arious Igh haplotvpes To determine if 7.34 a n d / o r 58.3 recognized epitopes present on immunoglobulins present in normal mouse serum, an inhibition RIA technique was used. Microtiter wells were coated with 1 ~g of 7.34 or 58.3 by overnight incubation at 4°C and, after washing, excess protein binding sites were saturated by an additional incubation with PBSA. To these wells were added serial dilutions of normal mouse sera from a series of different inbred strains. Normal sera from A / J , BALB/cByJ, C 3 H e B / F e J , D B A / 2 J , A K R / J , SJL/J, and C57BL/10J mice were obtained from mice housed in the vivarium at Emory University. Fractions containing various immunoglobulin classes were prepared from C57BL/10 serum by protein A-Sepharose fractionation as described by Ey el al. (1978). Normal sera from the Igh recombinant and congenic mouse strains BAB/14, C.B-20, C . B R / 3 and C . B R / 8 were the generous gift of Dr. M. Cancro of the University of Pennsylvania School of Medicine. After a 1 h incubation with the serum dilutions, the wells were washed and 25/xl of 1251-10-3.6 (50,000-70,000 cpm) added to each well. After another 1 h incubation, the plates were washed and the individual wells counted in a gamma counter. Interaction of serum antibody with 7.34 or 58.3 in the initial incubation was revealed as a decrease in the binding of radiolabeled 10-3.6 in the second step. Results are expressed as the % inhibition of the binding of ~251-10-3.6. Inhibition of I-"~I-10-3.6 binding to I-A ~ bearing cells Microtiter wells were coated with monolayers of erythrocyte-free spleen cells using a modification of the method of Heusser et al. (1981). Briefly, individual wells were pretreated with 50 /xl of a 100 /~g/ml solution of poly-L-lysine hydrobromide (MW 4000) (Sigma Chemical Co., St. Louis, MO) in PBS for 1 h at 37°C. After washing, 50 /~1 of a suspension (2 × 107) of Tris-NH4Cl-treated splenocytes were added to each well and the entire plate centrifuged (500 x g, 5 min). The plates were then flicked into a sink to empty the wells and the cell monolayers fixed by addition of 200/~1 of freshly prepared 0.25% glutaraldehyde in PBS for 5 min. After fixation, the wells were washed with a 2.19% solution of glycine and then used immediately or stored at 4°C with all wells containing a buffer composed of 3.5% sucrose, 0.05 M sodium cacodylate and 0.1% sodium azide. The ability of the 7.34 and 58.3 antibodies to block binding of ~251-10-3.6 to I-Ak-expressing spleen cells was tested by mixing 25 p~l of ~251-10-3.6 (60,000 cpm) and 25 ktl of dilutions of 7.34 or 58.3 and after 1 h adding the mixture to PBSA-pretreated wells of fixed spleen cells. After another hour, the wells were washed and the bound radioactivity quantitated by g a m m a counting. The results are expressed as the % inhibition of 1251-10-3.6 binding relative to control wells in which only ~51-10-3.6 was present. Competition radioimmunoassay to determine spatial relationship of epitopes recognized by 7.34 and 58.3 To determine whether 7.34 and 58.3 bound to spatially related or separate
283 epitopes on the 10-3.6 antibody molecule, serial dilutions of 7.34 or 58.3 were added to PBSA-treated microtiter wells coated with 5 ~g of 10-3.6. After 1 h, the wells were washed and 25 ~1 of ~251-58.3 were added to each well. After a second 1 h incubation, the wells were washed and the bound radioactivity measured in a gamma counter.
Double determinant sandwich radioimmunoassay to quantitate serum levels of 10-3. 6 Microtiter wells were coated with 1 ~g of 58.3 antibody by overnight incubation at 4°C. The plates were then washed and treated with PBSA to block remaining non-specific protein binding sites. Next 25 ~1 of dilutions of a 10-3.6 standard or serum samples were added to the 58.3 coated wells in triplicate. After 1 h the plates were washed and 25 ~1 of 125I-7.34 diluted in PBSA (50,000 70,000 cpm) were added to each well. After another 1 h incubation, the wells were washed and the bound radioactivity quantitated by gamma counting. Mathematical analysis of sandwich immunoassay data The standard curve obtained by plotting cpm bound vs. concentration of 10-3.6 was sigmoid shaped. In order to represent the same data by a linear function, the technique of logit transformation was utilized (Wellington, 1980). Briefly, if R, = radioactivity bound for the ith standard; C, = concentration of the ith standard; R 0 = non-specific background binding of 125I-7.34 in the absence of any 10-3.6; and R m = the maximum radioactivity bound at an infinite dose of 10-3.6, then Y, is defined as Yi = ( R i - R o ) / ( R . , - Ro)
(1)
and the least squares solution to the equation ln(Yi/(1 - Yi)) = b + mlnC i
(2)
where b represents the y-intercept and m the slope, can be obtained by a linear regression method. These calculations were performed using the SuperCalc3 electronic spreadsheet ( S o r c i m / I U S Micro Software, San Jos6, CA) on an IBM-PC. Initially, R m was assigned the value of the maximum number of counts bound to a single well in the actual assay. Then the value of R m was optimized by an iterative procedure that determined what value of R,,, gave the maximum correlation coefficient for the least squares solution of equation (2). The concentration of 10-3.6 in serum samples was then calculated from equation (2). The half-life of 10-3.6 in individual strains was calculated from the slope (m~) of the initial linear portion of the semilogarithmic antibody clearance curve using the formula
t,/2
= - (log 2)/me
(3)
Skin grafting Split-thickness tail skin grafts were performed and monitored daily until rejection
284
as described previously (Perry and Williams, 1985). Anti-I-A-treated mice received intraperitoneal injections of protein A-Sepharose-purified 10-3.6 antibody. Results are reported as the mean survival time of groups of 5 - 6 mice. The significance of differences in graft survival time between groups was assessed using the nonparametric Wilcoxon rank sum test (Lehmann, 1975).
Results
Characterization of the 7.34 and 58.3 monoclonal antibodies Two hybridoma lines, 7.34 and 58.3, were selected on the basis of antibody binding to the 10-3.6 anti-Ia.17 mAb in solid-phase RIA. The specificity of these reagents was assessed by several methods including direct binding of radioiodinated 7.34 and 58.3 to microtiter welIs coated with one of a series of anti-H-2 mAbs directed against class I or class II M H C antigens (Table II). Out of the panel tested, both 7.34 and 58.3 reacted only with 10-3.6; neither 7.34 or 58.3 bound to 10-2.16, another anti-Ia.17 mAb derived from the same fusion, but with a different isotype (~,2b) and different idiotypic determinants (Sher et al., 1984). To further characterize the specificity of these anti-10-3.6 mAbs, ailoimmune anti-I-A k and normal sera were obtained from a panel of mouse strains representing several of the defined Igh haplotypes including Ighh (the haplotype of the CWB strain used as spleen cell donor in derivation of the 10-3.6 hybridoma) and tested for expression of the epitopes defined by 7.34 and 58.3. An inhibition assay was used in which the presence of reactive material in a serum sample was detected by the ability of serum to inhibit the interaction between 7.34 or 58.3 coated to microtiter wells and t251-10-3.6. Sera from Igh h mice or any other Igh haplotype tested was not found to inhibit the binding of ~251-10-3.6 and 58.3 (data not shown), suggesting that 58.3 is directed against a private determinant specific for the 10-3.6 antibody
TABLE II SPECIFICITY OF A N T I B O D I E S P R E P A R E D A G A I N S T 10-3.6 Monoclonal antibodies 7.34 and 58.3 were labeled with 125I and tested for their binding to 10-3.6 and other anti-MHC hybridoma antibodies that had been absorbed to the surface of microtiter wells. Radioiodinated protein A was used as a positive control to make certain that all antibodies tested as substrates had bound to the microtiter wells. Antibody absorbed to microtiter wells
cpm ~25I-labeled reagent bound 1251_7.34
~251_58.3
None 10-3.6 10-2.16 11-5.2 11-4.1 14-4-4S
36 29964 145 56 79 72
33 27619 28 96 82 64
~zSI_proteinA 48 41 805 16219 29 589 39753 8 898
285 TABLE III SPECIFICITY OF 7.34 MONOCLONAL ANTIBODY Serum or serum fractions of various inbred mouse strains were tested for their ability to react with the 7.34 mAb by measuring their ability to bind to microtiter wells coated with 7.34 (1 p,g per well) and inhibit the binding of 1251-10-3.6 in a subsequent binding step. The results are expressed as the % inhibition relative to the PBSA control wells (0% inhibition). Experiment
Inhibiting serum
Igh-C haplotype
% Inhibition of t251-10-3.6 binding at indicated dilution 1:10
1:100
Reactivity with 7.34
BALB/cByJ C57BL/10J (B10) DBA/2J AKR/J A/J C3HHeB/FeJ
a b c d e j
3.0 94.7 9.7 5.4 8.1 - 0.2
3.1 91.8 9.4 9.1 6.5 0.6
+ -
BALB/cByJ B10 SJL/J C.B-20 ~ BAB.14 " C.BR/3 ~' C.BR/8 a BI0 non-IgG b B10 IgG2a b
a b b b b a a b b
3.9 95.4 94.7 95.5 94.0 11.2 19.9 ND c ND
20.2 94.4 94.1 92.0 77.5 6.5 4.3 7.4 91.9
+ + + + +
a The alleles of the Igh congenic and recombinant mice at the Dex, lgh-C and Pre-1 loci on chromosome 12 are summarized below. Crossover regions for the recombinant strains are indicated by a / . These data were compiled from Weigert and Riblet (1978) and Forman et al. (1984). Strain
Allele/antigen Dex
C.B-20 BAB.14 C.BR/3 C.BR/8
+ +
/ /
Igh
Pre-1
b b a a
0 0 a 0
/
b C57BL/10 NMS was fractionated as described in Materials and Methods. The effluent from the column when the serum was applied to the column at pH 8.0 is termed the non-lgG fraction. The igG2a fraction is that material that initially bound to the column, was not eluted from the column by pH 5.5 buffer, but was subsequently eluted by application of pH 4.5 buffer. c Not done.
m o l e c u l e . H o w e v e r , t h e e p i t o p e d e f i n e d b y 7 . 3 4 w a s p r e s e n t in n o r m a l m o u s e s e r u m from 2 Ighb strains, C57BL10flJ and SJL/J,
but not serum from strains representing
t h e a, c, d, e o r j I g h h a p l o t y p e s ( T a b l e I I I ) . A l s o s h o w n in T a b l e I I I a r e t h e r e s u l t s of an experiment performed of the determinant
to more precisely map the genetic control of expression
r e c o g n i z e d b y 7.34. T h e a b i l i t y o f s e r u m f r o m t h e I g h c o n g e n i c
286 IOO
Ccn
8O
h5 6O"
FO
40 ¸
I
,~
20
o
fi ~,
20
-40 0,001
. . . . . . . .
j
0.01
. . . . . . .
,
O. 1
. . . . . . . .
,
l
. . . . . . . .
,
. . . . . . . .
10
i
100
. . . . . . . .
I000
Antibody Concentration (ug/ml)
Fig. 1. Inhibition of 1251-10-3.6binding to I-Ak by 7.34 and 58.3. Twenty-five/tl of a series of dilutions of purified 7.34 and 58.3 were mixed with 25/LI of 1251-10-3.6and added to microtiter wells containing fixed BI0.A spleen cells. After 1 h of incubation, the wells were washed and the bound radioactivity quantitated by radioimmunoassay. The results are expressed as the % inhibition of 1251-10-3.6binding compared to control wells in which only radiolabeled 10-3.6 and PBSA were present.
strain C-B.20 to inhibit the interaction of 7.34 and 58.3 demonstrated that the epitope defined by 7.34 was under the control of genes linked to the Igh locus. To ascertain whether these genes m a p p e d to the Igh-V or Igh-C, additional serum samples from strains recombinant within the Igh locus (BAB.14, C . B R / 3 and C . B R / 8 ) were tested. BAB.14 is an intra-V H recombinant strain that expresses Igh-V markers c o n t r i b u t e d . b y the B A L B / c background strain (Igh ") and Igh-C determinants controlled by the Igh b haplotype (Weigert and Riblet, 1978). C . B R / 3 expresses Igh-V markers of the Igh b haplotype and Igh-C determinants from the Igh ~ haplotype, while C . B R / 8 expresses Igh-V and Igh-C alleles of the Igh~ haplotype ( F o r m a n et al., 1984). Serum samples from these 3 Igh recombinant strains interacted with 7.34 in a pattern predicted by the Igh-C allele, suggesting that 7.34 was directed against an allotypic determinant specified by the Ig-1 locus within the Igh-C region. Results of protein A-Sepharose fractionation of serum from C 5 7 B L / 1 0 J mice (Igh b) are consistent with this interpretation; 7.34-reactive molecules were retained by a protein A-Sepharose i m m u n o a b s o r b e n t at p H 8.0 and eluted when the p H of the buffer was reduced to 4.5 (Table III). The relative ability of the 7.34 and 58.3 antibodies to inhibit the binding of ]251-10-3.6 to B10.A spleen cells was examined to determine whether either of these antibodies recognized binding site determinants on the 10-3.6 m A b molecule (Fig. 1). The 58.3 antibody inhibited the binding of radiolabeled 10-3.6 to I-A k over a wide range of concentrations, with 50% inhibition attained using a 58.3 concentration of approximately 0.03 /~g/ml. Significant inhibition of 1251-10-3.6 binding by 7.34 was only observed at very high concentrations ( > 10 /Lg/ml) and concentra-
287 100
8O
/
60
g I ~
4o-
,/
z g ~ )5
J
J J
~3 c
20-
/
/
/
C
-20 . • 0.01
.
.
. 0.1
Antibody
. I Concentration
10 (ug/ml)
Fig. 2. The 7.34 and 58.3 antibodies recognize spatially distinct epitopes. Microtiter wells were coated with 5 fig of 10-3.6 by overnight absorption at 4°C. After PBSA treatment to block residual protein binding sites, serial dilutions of 7.34 and 58.3 were added to the wells. After 1 h these antibodies were washed out and 25 p,l of 1251-58.3 added. After another hour, the wells were washed and the bound radiolabel quantitated by gamma counting. tions of 0.1 and 0.3 t t g / m l were reproducibly found to potentiate the binding of radiolabeled 10-3.6. These results viewed in light of those presented earlier indicate that 58.3 is directed against an idiotypic determinant present on the binding site of 10-3.6, while 7.34 recognizes an allotypic determinant elsewhere on the 10-3,6 molecule. To confirm that 7.34 and 58.3 recognized spatially distinct epitopes, an inhibition assay was performed in which 10-3.6 coated wells were allowed to interact with various dilutions of unlabeled 7.34 or 58.3 antibody prior to addition of radioiodinated 58.3. As shown in Fig. 2, 7.34 did not cause any inhibition of the binding of t251-58.3 to the 10-3.6 coated wells, while the homologous unlabeled reagent (58.3) was inhibitory in a dose-dependent fashion. Finally, other experiments have demonstrated that F(ab')2 fragments of 10-3.6 prepared by pepsin digestion of whole antibody retain the ability to bind to 58.3, but do not bind to 7.34 (data not shown). Thus the epitope recognized by 7.34 is an allotypic determinant controlled by Igh-C genes and located on the Fc portion of the immunoglobulin molecule.
Double determinant sandwich immunoassay for 10-3.6 using the 7.34 and 58.3 monoclonal antibodies Several different types of immunoassays for 10-3.6 using 7.34 a n d / o r 58.3 were evaluated with respect to sensitivity. The most sensitive assay was found to be a sandwich immunoassay employing 58.3 absorbed to plastic microtiter wells as a capture antibody and radiolabeled 7.34 as a detection antibody. Similar assays using monoclonal antibodies recognizing spatially distinct epitopes on an antigen of interest have been described for quantitating levels of alpha-fetoprotein (Uotila et
288
Untransformed
RIA Data
25000
/ ~0000
/
•
15000
~r~
/
10000
f
/ /
0_~--
,/
/
,/
-~
-3
0
LN C o n c e n t r a t . i o n
Logit A n a l y s i s of RIA Data 4
t3 Jd 2 //." o /-
2 -2
y
/,
/ i:I
/
/i
4
Ef~
-9
-6
-3
LN Concentration
0
--~
3
Fig. 3. Standard curve obtained rising double-determinant radioimmunoassay for 10-3.6. A series of standards ranging in concentration from 1 n g / m l to 4096 n g / m l was used. Part A demonstrates the sigmoid shape of the plot of cpm bound against the natural logarithm of 10-3.6 concentration. Part B shows the identical set of data after the logit function has been used to linearize the data.
al., 1981), human Ia antigens (Russo et al., 1983), and human gamma interferon (Tanaka et al., 1985). Fig. 3A represents a standard curve obtained using the sandwich immunoassay and samples of 10-3.6 mAb ranging in concentration from 1 n g / m l to 4096 n g / m l . A sigmoid-shaped curve is obtained that is typical for this type of RIA. In order to utilize this assay for determination of 10-3.6 concentrations in serum samples, it was necessary to develop a mathematical formulation of the relationship between the number of counts bound and the concentration of 10-3.6. The logit function was utilized to linearize the sigmoid standard curve so that a least-squares linear regression calculation could be performed. The linear relationship between the logit function and the natural logarithm of the antibody concentration is demonstrated in Fig. 3B for the same set of standards depicted in Fig. 3A.
289
1000 I~C DBA/2 ..... (I-A') (l--A'/d)II c~H {,-A,)
E
c
1
1oo
~o
c) c
8
'
cO I
0
O,I
I'0
'
2'0
'
3'0
~4'0
5'0
B/O
Hours after Antibody Injection Fig. 4. Clearance of 10-3.6 from 3 mouse strains. Groups of 3 C 3 H / H e J , C3D2F 1 and D B A / 2 mice were injected intraperitoneally with 200 t~g of purified 10-3.6 antibody. Pooled serum samples were collected from each group of mice every 6 h for 2 days. The serum was frozen and later assayed for 10-3.6 using the double determinant immunoassay.
Application of sandwich assay for 10-3.6 to determination of antibody half-life in serum of treated mice The sandwich assay described above was used to follow the clearance of a single dose of 200 /~g of 10-3,6 injected intraperitoneally into mice of several MHC haplotypes. Serum samples were obtained at 6 h intervals for the 2 days after antibody injection and several dilutions of each sample assayed for the level of 10-3.6. As seen in Fig. 4, the clearance pattern following antibody injection was determined by the I-A alleles possessed by each strain used. In both homozygous I-Ak mice (C3H) and heterozygous I-Ak mice (C3D2F 1), the level of 10-3.6 initially declined following approximately first-order kinetics. The half-lives calculated by linear regression analysis of the initial linear portions of these clearance curves and
TABLE IV HALF-LIFE OF 10-3.6 ANTIBODY IN AN I-A k H O M O Z Y G O U S STRAIN, AN I-Ak HETEROZYGOUS STRAIN, A N D A NON-I-A k STRAIN Mice from 3 inbred strains received an intraperitoneal injection of 200 ttg of 10-3.6 antibody. The serum half-life of 10-3,6 in these strains was determined by using the initial linear portion of the clearance curve. Mouse strain
I-A genotype
Range of time points used to determine half-life (h)
Calculated 10-3.6 half-life (h)
C3H/HeJ C3D2F~ DBA/2J
k k/d d
12-24 h 6-18 h 6-54 b
3.5 4.5 72.7
290 TABLE V RELATIONSHIP BETWEEN TIMING OF ANTI-Ia IMMUNOTHERAPY AND PROLONGATION OF SKIN GRAFT SURVIVAL Groups of 5 6 A/J mice received BI0.A tail skin grafts on day 0. Mice were left untreated, were treated with 1000 ~tg of 10-3.6 mAb on day - 1 and 500 ~tg on day 1 and 2. or were treated with 200/Lg of 10-3.6 on days 0-9. Host
Donor
n
Treatment schedule ~g 10-3.6 mAb (days treated)
Total 10-3.6 dose (mg)
MST_+SEM
A/J A/J A/J
B10.A B10.A B10.A
5 6 6
1000 ( - 1); 500 (1,2) 200 (0-9)
2.0 2.0
9.8 _+0.7 11.8 ± 0.5 ~' 16.0 _+1.0 b
a P < 0.05. b p < 0.005.
e q u a t i o n (3) are 3.5 h for C3H a n d 4.5 h for C 3 D 2 F 1 (Table IV). After 24 h the rate of clearance began to gradually decrease. In mice not expressing an I-A molecule reactive with 10-3.6 ( D B A / 2 ) , a significantly different p a t t e r n of clearance was observed. U s i n g the c o n c e n t r a t i o n s d e t e r m i n e d at all time points examined in D B A / 2 mice, a 10-3.6 serum half-life of 72.7 h was calculated.
Effect of anti-l-A treatment protocol on serum antibody levels and immunoregulatory effect In previous studies of a n t i - I - A i m m u n o t h e r a p y in t r a n s p l a n t a t i o n c o n d u c t e d in this laboratory, a treatment protocol of 200 ~g of a n t i b o d y daily on days 0 - 9 was routinely used (Perry a n d Williams, 1985). However, other investigators utilizing a n t i - I - A antibodies as a therapeutic approach in experimental a u t o i m m u n e disease have developed protocols in which all a n t i b o d y is administered within 1 or 2 days of a u t o a n t i g e n i m m u n i z a t i o n ( S t e i n m a n et al., 1981). To compare the i m m u n o r e g u l a tory effects of these 2 types of treatment protocols in t r a n s p l a n t a t i o n , groups of A / J mice were grafted with B10.A tail skin and treated with 200 #g of m A b on days 0 - 9 (Protocol A) or 1000 ~g o n day - 1 and 500 ~g on days 1 and 2 (Protocol B) (total doses of 2.0 mg of 10-3.6 in both cases). Protocol B used the same days of a d m i n i s t r a t i o n a n d the same a p p r o p r i a t i o n of total m A b dose as used by S t e i n m a n et al. (1981) in studies o n m u r i n e encephalomyelitis. The skin graft survival times listed in T a b l e V d e m o n s t r a t e that a n t i b o d y treatment over a 10 day interval was associated with a m e a n graft survival time significantly ( P < 0.005) greater than the same total m A b dose administered between day - 1 a n d day 2. Serum samples were also taken daily from the a n t i b o d y - t r e a t e d mice in this experiment. The serum samples were taken immediately prior to the daily a n t i b o d y injection, a n d thus represent trough rather than peak serum levels in the mice treated daily with antibody. The results of this analysis presented in Fig. 5 reveal that mice treated according to Protocol A m a i n t a i n e d trough a n t i b o d y levels between 5 a n d 25
291
10 3.6 Treotment Protocol I 0.2 m9 Days 0-9 1.0 mg Doy -1; 05 rnq Ooys ~,2 lOB
k~
?
-,~ io (J CD
? o
--
½
"
6. ; ; I '0 Days after Transplantation
I
'2
I '4
Fig. 5. Effect of treatment protocol on serum 10-3.6 levels. Groups of 6 A / J mice were given B10.A skin grafts on day 0. One treated group received 200 ~g of 10-3.6 on days 0-9. A separate treated group received 1000/~g of 10-3.6 on day 0 and 500 p.g on days 1 and 2. Serum samples were collected on a daily basis and assayed for 10-3.6 using the double determinant immunoassay.
/~g/ml through day 10. In contrast, mice treated by Protocol B had higher antibody levels on days 1 through 3 (20-120/~g/ml), similar levels on day 4, and less than 5 # g / m l by day 5 and thereafter. Based upon the graft survival data under these 2 treatment protocols, it appears that maintenance of a minimal serum concentration of mAb over a prolonged period is superior to administration of the same total dose near the time of transplantation. Discussion
Antibodies against class II MHC molecules have been demonstrated to have immunosuppressive activity in experimental autoimmune disease and in transplantation. Anti-Ia immunotherapy is associated with the development of antigen-specific T cell-mediated suppression (Perry and Greene, 1982) to antigens presented during the treatment interval, while secondary responses to antigens encountered outside the treatment interval remain intact (Perry and Williams, 1985). As with other therapeutic agents, the development of optimal treatment protocols using these agents must be predicated on an understanding of their pharmacokinetics. Many possible immunoassay procedures for measurement of serum levels of monoclonal anti-Ia antibodies exist. Other investigators have followed serum levels of monoclonal anti-I-A antibody by flow cytometric (Waldor et al., 1984) or microcytotoxicity (Kruisbeek et al., 1985) methods. The sandwich assay described in this manuscript is substantially more sensitive than either of these other methods while retaining a high degree of specificity. The assay utilizes 2 monoclonal antibodies recognizing spatially distinct epitopes
292 on the 10-3.6 antibody molecule. The 7.34 hybridoma produces antibodies reacting with an allotypic determinant found on IgG2a immunoglobulin molecules from mice of the Ighb haplotype. Monoclonal antibodies directed against allotypic determinants controlled by the Ig-1 h locus have been described previously by other laboratories (Oi and Herzenberg, 1979; DiPauli and Raschke, 1978), but the relationship between the epitope defined by 7.34 and epitopes defined previously has not been addressed. The 58.3 mAb is directed against an idiotypic determinant of the 10-3.6 molecule as evidenced by its ability to block the binding of 10-3.6 to I-A k. Further evidence of the anti-idiotypic activity of 58.3 has been obtained by immunizing syngeneic ( D B A / 2 ) mice with this antibody; some of these mice responded by producing anti-I-A antibodies with the same epitope specificity as 10-3.6-derived immunoglobulin (Williams and Perry, data not shown). Clearance data obtained after a single 200 #g dose of 10-3.6 antibody in 3 inbred mouse strains reveals that the antibody is rapidly cleared from the circulation in mice that express the relevant class II molecules. The half-life calculated on the basis of the initial rate of decay is slightly shorter in homozygous I-A k mice (3.5 h) than in heterozygous I-Ak/d mice (4.5 h) as might be expected given the greater total number of target class II molecules in the former group. The half-life of 10-3.6 antibody in D B A / 2 mice lacking the Ia.17 specificity was calculated to be approximately 3.0 days. Fahey and Sell (1965) found that 2 IgG2a myeloma proteins had half-lives of 4.8 and 5.4 days, respectively, in normal mice. Thus, the clearance of 10-3.6 from the serum of D B A / 2 mice follows kinetics similar to those described for other IgG2a immunoglobulins. The rapid clearance of 10-3.6 in C3H and C3D2F 1 mice cannot be attributed to the administration of denatured or damaged immunoglobulin molecules because of the extended half-life observed in the D B A / 2 controls. Extrapolation of the clearance curves (Fig. 4) to 0 h for all strains examined gives an initial serum level of approximately 100/~g/ml. Since 200/xg were injected into these animals, the volume of distribution can be estimated as 2.0 ml. Given that the mice used in these experiments weighed between 20 and 25 g and total extracellular body water is approximately 20% of total body weight (Ganong, 1979), these data suggest that the administered anti-I-A antibody is able to distribute itself throughout the plasma (5% of body weight) as well as some other components of the extracellular fluid compartment. Therapeutic agents with short in vivo half-lives are generally administered either as a continous infusion or intermittently at intervals that allow maintenance of therapeutic levels in the body (Goodman et al., 1980). The serum concentration of anti-class II antibody necessary to manifest maximal immunosuppression has not been defined in most systems, although dose dependence has been clearly demonstrated (Perry and Williams, 1985). The parameter correlating most closely with the efficacy of treatment may be the residual expression of class II antigens by host antigen-presenting cells following mAb treatment, as suggested by the studies of Kruisbeek et al. (1985). The rapid re-expression of class II antigens after mAb-mediated modulation probably represents one basis for the continuous requirement of anti-Ia antibody to maintain the blockade of antigen presentation in the context of class II restriction elements.
293
Evidence for a direct correlation between monoclonal anti-Ia antibody serum levels and the immunosuppressive effects of anti-Ia treatment were obtained in a murine skin graft model of transplantation. Administration of 2.0 mg of 10-3.6 in 10 equal daily doses beginning with the day of transplantation was associated with significantly greater graft prolongation than administration of the entire 2.0 mg between the day before transplantation and 2 days after. The first regimen was associated with significantly higher serum antibody levels from day 5 onward, demonstrating that the maintenance of a 5-25 ~ g / m l serum concentration of Ab for 10 days is more favorable to the prolongation of skin graft survival, presumably due to extended inhibition of antigen presentation under these conditions. Future studies using this radioimmunoassay will address more definitively the serum level of anti-class II antibody necessary to maintain maximal inhibition of Ia expression, antigen presentation and induction of antigen-reactive T cells. If anti-Ia immunotherapy continues to show promise as an immunotherapeutic approach in autoimmunity and transplantation, further characterization of the pharmacokinetics of these antibodies will be valuable in designing optimal treatment protocols. Radioimmunoassays patterned after the one described herein can be developed for the quantitation of serum and tissue levels of other monoclonal immunoglobulins being considered for in vivo therapeutic use. The sandwich immunoassay technique for measurement of in vivo mAb levels could also be performed as an enzyme immunoassay. Such assays will only require mAbs directed against separate epitopes on the antibody of interest that do not cross-react with the immunoglobulins normally present in the treated host.
Acknowledgements We thank Dr. Michael Cancro for providing serum samples from C.B-20, BAB.14, C . B R / 3 and C . B R / 8 mice in his colony and for his helpful suggestions. Ms. Brenda Hott and Ms. Nancy Martin are gratefully acknowledged for their technical assistance in preparation of the 7.34 and 58.3 hybridomas. These studies were supported by Grant CA-32018 from the National Cancer Institute.
References Adelman N.E., D.L. Watling and H.O. McDevitt, 1983, J. Exp. Med. 158, 1350. Baxevanis, C.N., D. Wernet, Z.A. Nagy, P.H. Maurer and J. Klein, 1980, Immunogenetics 11,617. DiPauli, R. and W.C. Raschke, 1978, Curr. Top. Microbiol. Immunol. 81, 37. Dubreuil, P.C., D.Z. Birnbaum, D.H. Caillol and F.A. Lemonnier, 1982, Immunogenetics 16, 407. Ey, P.L., S.J. Prowse and C.R. Jenkin, 1978, Immunochemistry 15, 429. Fahey, J.L. and S. Sell, 1965, J. Exp. Med. 122, 41. Forman, J., R. Riblet, K. Brooks, E.S. Vitetta and L.A. Henderson, 1984, J. Exp. Med. 159, 1724. Galfr6, G., S.C. Howe, C. Milstein, G.W. Butcher and J.C. Howard, 1977, Nature (London) 266, 550. Ganong, W.F., 1979, Review of Medical Physiology (Lange Medical Publications, Los Altos, CA) p. 6.
294 Goodman, A.G., L.S. Goodman and A. Gilman (eds.), 1980, The Pharmacological Basis of Therapeutics (MacMillan Publishing Co., New York) pp. 21-27. Heusser, C.H., J.W. Stocker and R.H. Gisler, 1981, Methods Enzymol. 73, 406. Kruisbeek, A.M., J.A. Titus, D.A. Stephany, B.L. Gause and D.L. Longo, 1985, J. Immunol. 134, 3605. Lehmann, E.L., 1975, Nonparametrics: Statistical Methods Based on Ranks (Holden-Day, San Francisco, CA). Oi, V.T. and L.A. Herzenberg, 1979, Mol. Immunol. 16, 1005. Oi, V.T., P.P. Jones, J.W. Goding, L.A. Herzenberg and L.A. Herzenberg, 1978, Curr. Top. Microbiol. Immunol. 81, 115. Ozato, K., N. Mayer and D.H. Sachs, 1980, J. Immunol. 124, 533. Perry, L.L. and M.I. Greene, 1982, J. Exp. Med. 156, 480. Perry, L.L. and I.R. Williams, 1985, J. Immunol. 134, 2935. Perry, L.L., M.E. Doff, B.A. Bach, B. Benacerraf and M.I. Greene, 1980, Clin. Immunol. lmmunopathol. 15, 279. Rosenbaum, J.T., N.E. Adelman and H.O. McDevitt, 1981, J. Exp. Med. 154, 1694. Russo, C., M.A. Pellegrino and S. Ferrone, 1983, Trans. Proc. 15, 57. Sher, M.R., T.P. Bender and J.E. Niederhuber, 1984, J. Immunol. 133, 338. Steinman, L., J.T. Rosenbaum, S. Sriram and H.O. McDevitt, 1981, Proc. Natl. Acad. Sci. U.S.A. 78, 7111. Tanaka, E., M. lmai, S. Usada, K. Tachibana, H. Okamoto, Y. Ohike, T. Nakamura, Y. Miyakawa and M. Mayumi, 1985, J. Immunol. Methods 77, 275. Uotila, M., E. Ruoslahti and E. Engvall, 1981, J. Immunol. Methods 42, 11. Waldor, M.K., S. Sriram, H.O. McDevitt and L. Steinman, 1983, Proc. Natl. Acad. Sci. U.S.A. 80, 2713. Waldor, M.K., R.R. Hardy, K. Hayakawa, L. Steinman, L.A. Herzenberg and L.A. Herzenberg, 1984, Proc. Natl. Acad. Sci. U.S.A. 81, 2855. Weigert, M. and R. Riblet, 1978, Springer Semin. immunopathol. 1, 133. Wellington, D., 1980, in: Enzyme-lmmunoassay, ed. E.T. Maggio (CRC Press, Boca Raton, FL) Ch. 12, p. 249. Wooley, P.H., H.S. Luthra, W.P. Lafuse, A. Huse, J.M. Stuart and C.S. David, 1985, J. lmmunol. 134, 2366.