CELLULAR
IMMUNOLOGY
14,
3 1l-323 ( 1992)
Behavior of the ldiotypic Network in Conventional Immune Responses I. Kinetics of ldiotypic and Anti-idiotypic Antibodies Following Immunization with T-Independent and T-Dependent Antigens ANNETTE
J. SCHLUETER, DIEGO SEGRE,MARK S. KUHLENSCHMIDT, AND MARIANCELA SEGRE’ Department of Veterinary Pathobiology, University of Illinois, 2001 South Lincoln Avenue, Urbana, Illinois 61801 Received April 13, 1992; acceptedJuly 6, 1992
A minimal requirement in investigations of the behavior of the idiotypic network during immunization is the ability to quantitate both the idiotypic (Abl) and anti-idiotypic (Ab2) responses. Quantitation of Ab2 in serum is complicated by the simultaneous presenceof Abl, so that AblAb2 immune complexes escapedetection. In contrast, immune complexes should not complicate the enumeration of Ab2-producing lymphocytes in a hemolytic plaque assay.This study utilizes a procedure that allows detection of Ab2-producing cells in such an assay.The procedure relies upon the insertion of the appropriate antibody (Abl) into the membrane of indicator SRBC through a covalently attached dipalmitoyl phosphatidylethanolamine (DPPE) tail. When the Ab2 response following murine immunization with DNP-Ficoll was analyzed using such an assay, peak plaque-forming cell (PFC) numbers were found to coincide with peak Abl PFC numbers in both the primary and secondary response. In addition, this Ab2 responsewas found to be T independent. The murine immune responseto DNP-HGG demonstrateda peak Ab2 PFC response which followed the peak Abl PFC response after both primary and secondary immunization. This Ab2 response appeared to be T dependent. The secondary responsesto both DNP-Ficoll and DNP-HGG showed increased levels of Ab2 PFC and decreasedlevels of Abl PFC in comparison to the primary responsesto the same antigens, suggestingthat immunoregulation may occur within these idiotypic networks. o 1992 Academic press, IIIC.
INTRODUCTION It has been over 15 years since Jerne (1) argued for the existence of an idiotypic network (“a web of interacting variable region domains”) involving free immunoglobulin (Ig) molecules and the B cells bearing membrane-bound Ig molecules as receptors. This network was theorized to internally regulate the immune system. The existence of an idiotypic network has been experimentally confirmed, but its regulatory functions during conventional immune responseshave been more difficult to verify. Most of the experiments designed to demonstrate the regulatory function of the idiotypic network have employed monoclonal antibodies or myeloma proteins as the ’ To whom correspondence and reprint requests should be addressed. 311 0008-8749192$5.00 Copyright 0 1992 by Academic Press, Inc. All rights of repreduction in any form reserved.
312
SCHLUETER ET AL.
idiotypic antibody. While this strategy has the obvious advantage of allowing the experimenter to focus on the limited number of idiotopes carried by the homogeneous idiotypic antibody, it does not adequately reflect the more complex situation that occurs in the immune response to most conventional antigens, which elicit a highly heterogeneous idiotypic antibody population. An additional reason for the failure to document a regulatory role of the idiotypic network may be a lack of appropriate technical tools. Most investigators have utilized serologic methods for the quantitation of network components. Evidence for immunoregulation from these studies is based on the cyclical appearance of idiotypic (Abl)’ and anti-idiotypic (Ab2) antibodies, and the fact that the peak titers of these antibodies are at times (2-4), but not always (5), out of phase with each other. Serologic methods are hampered by the formation of circulating Abl-Ab2 immune complexes that would be expected to allow the detection only of the antibody that is present in excess.Therefore, the measured cycles of Abl and Ab2 production may not be truly representative of the actual levels of Abl and Ab2 being produced in response to immune stimulation. A few investigators have studied the idiotypic network through the enumeration of Ab2 plaque-forming cells (PFC) (2-8), but were able to detect only low levels of Ab2 PFC even though the indicator sheep red blood cells (SRBC) were coated with a homogeneous, monoclonal Ab 1. We thought that the failure of previous investigations of cellular network behavior to yield insightful results may have been due to technical inadequacies rather than to inherent lack of power of the hemolytic plaque assay.By anchoring lipid-derivatized, affinity-purified polyclonal Abl in the membrane of indicator SRBC, we were able to detect large numbers of Ab2 in conventional, polyclonal immune responses.In this study, we describe the response of the murine network to immunization with DNP-Ficoll or DNP-human -y-globulin (HGG). Several unexpected results emerge, including simultaneous T-independent peak Abl and Ab2 responsesfollowing DNP-Ficoll immunization. MATERIALS AND METHODS Mice. Two- to three-month-old female BALB/c mice and 6-week-old female BALB/ c (nu/nu) mice were obtained from Harlan-Sprague-Dawley (Indianapolis, IN). Immunization schedule. Mice receiving primary immunization with DNPsO-Ficoll (Biosearch, San Rafael, CA) were injected with 100 pg of the antigen in saline intraperitoneally. Secondary immunizations were given in a similar manner, 2 1 days after primary immunization. Abl and Ab2 plaque-forming cell (PFC) responses of the spleen cells were measured in groups of three mice killed at 0, 3, 4, 5, 7, 10, 12, 15, 18, or 21 days after primary immunization, and at 3, 5, or 7 days after secondary immunization. Mice receiving primary immunizations with DNP,,-HGG were injected intraperitoneally with 100pg of the antigen adsorbedon bentonite (9). Secondary immunizations were given in a similar manner, 20 days after primary immunization. Ab 1 and Ab2 PFC responsesof the spleen cells were measured in groups of three mice killed at 0, 3, 5, 7, 10, 14, or 20 days after primary immunization and at 3, 5, or 7 days after secondary immunization. 2Abbreviations used Ab 1, idiotypic antibody; Ab2, anti-idiotypic antibody; Ab28, anti-idiotypic antibody with idiotope that is bound by the idiotypic antibody’s paratope (aka the internal image of the antigen); DNP, dinitrophenyl hapten; PFC, plaque-forming cell(s); DPPE, dipalmitoyl phosphatidylethanolamine; SPDP: N-succinimidyl-3-(2-pyridyldithio)propionate.
IMMUNOREGULATION
VIA THE IDIOTYPIC
NETWORK,
I
313
Rabbit anti-DNP antibodies. A rabbit was initially immunized SCwith 10 mg DNPBSA in complete Freund’s adjuvant. This was followed by monthly iv injections of 1 mg DNP-BSA. Serum was obtained 1 week after each iv immunization. Ig was precipitated from serum pooled from several bleedings with 50% saturated ammonium sulfate, was dialyzed against borate buffer, pH 8.2, and was passedthrough a column of DNP-chicken y-globulin covalently linked to cyanogen bromide-activated Sepharose 4-B (Pharmacia LKB, Piscataway, NJ). The column was thoroughly washed with borate buffer and the antibody was eluted with 4 X lop3 A4 DNP-lysine. Eluted antibodies were then dialyzed vs a 0.2 M solution of dinitrophenol in order to displace DNP-lysine. Finally, dinitrophenol was removed by passing the antibody through a Dowex 1 X 8 anion exchange resin. F(ab’)* fragments were obtained from the affinitypurified antibody by pepsin digestion (10). Undigested Ig molecules were removed by affinity chromatography on protein A-Sepharose 4-B (Pharmacia LKB, Piscataway, NJ). Hemolytic plaque assay. The hemolytic plaque assayfor the detection of Ab 1 PFC was performed as previously described ( 1l- 13). For the detection of Ab2 PFC, the indicator SRBC were coated with affinity-purified F(ab’)z fragments of rabbit antiDNP antibody (Abl) to which dipalmitoyl phosphatidylethanolamine (DPPE) was covalently attached. DPPE was first conjugated with N-succinimidyl-3-(2-pyridyldithio)propionate (SPDP). A 40 mM solution of DPPE in chloroform:methanol (9:2, v/v) was combined with triethylamine at an 83: 1 (v/v) ratio. This solution was mixed with a 0.12 M solution of SPDP in absolute ethanol at a 2.5: 1 (v/v) ratio and was incubated at 25°C for 2 hr. The organic phase was extracted three times with dH20, dried, and was redissolved in 100% dimethyl sulfoxide. The completeness of the reaction was verified by thin layer chromatography using chloroform:methanol:water (65:25:4) as the solvent, and the product was stored at -20°C. A solution of the antibody (1 to 2 mg/ml) in phosphate-buffered saline (PBS) pH 7.4, was combined with a solution of SPDP in absolute ethanol such that the final molar ratio of SPDP: antibody was 1O:l. The reaction was allowed to proceed for 30 min at 25°C with occasional stirring. The reaction was stopped by overnight dialysis vs 0.1 M acetate buffer, pH 4.5. The disulfide groups on the SPDP-conjugated antibody were reduced by addition of dithiothreitol to a final concentration of 50 mM. The reaction was allowed to proceed for 30 min at 25°C and was stopped by overnight dialysis vs. a 0.01 M NaCl solution containing 5 mM EDTA. The antibody solution was then concentrated to one-tenth its initial volume. A sufficient volume of SPDP-conjugated DPPE solution was added to the antibody to attain a calculated 620: 1 molar ratio of DPPE:antibody, and the reaction was allowed to proceed for 30 min at 25°C. The product was dialyzed vs three changes of PBS containing 0.002% deoxycholate and 0.02% sodium azide. The product was then sonicated for four 30-set cycles using the stepped microtip of a Branson model 185 sonifier cell disrupter (Heat SystemsUltrasonics, Inc., Plainview, NY) set at 30 W. The DPPE-conjugated antibody was then diluted appropriately, mixed with a SRBC suspension, and the mixture was agitated overnight in a rotating drum at 37°C. The integration of rabbit antibody into the SRBC membrane was confirmed by hemolysis upon addition of goat anti-rabbit Ig serum and complement. Optimal coating of SRBC was obtained with 30 to 60 pg of DPPE-derivatized rabbit F(ab’)*/ml of 2.5% erythrocyte suspension. The antibodycoated SRBC were washed, resuspended in saline, and were used in the hemolytic plaque assay. Spleen cells that were not used for the plaque assayswere routinely
314
SCHLUETER ET AL.
+Ab -Ab
g 80,000
sP 5
60,000
2 3
40,000
1 2
z a 20,000
0
5 10 15 Day after primary
20 25 immunization
30
FIG. 1. Primary and secondary Abl and Ab2 responsesafter DNP-Ficoll immunization. Two- to threemonth-old female BALB/c mice were immunized intraperitoneally with 100 pg DNP-Ficoll. Three weeks later, some of the mice received an identical secondary injection (indicated by the arrow). The Abl (m) and Ab2 (0) PFC responsesof the spleen cells were measured in groups of three mice killed at 0, 3, 4, 5, 7, 10, 12, 15, 18, and 21 days after priming and at 3, 5, and 7 days after booster immunization. Results are expressedas the number of direct PFC/spleen as a function of time after primary immunization. Error bars represent the standard error within each group of mice.
frozen with 10% DMSO and stored in liquid nitrogen. In some cases,frozen spleen cells were thawed and utilized for hemolytic plaque assays( 14). Frozen cells yielded 20 to 50% the number of PFC found with fresh cells. SRBC were also coated with other affinity-purified antibodies (rabbit anti-OVA) and protein antigens (OVA, HGG, F(ab’)* from nonspecific rabbit IgG). Successful insertion of the DPPE-conjugated proteins into the SRBC membrane was confirmed by hemagglutination or hemolysis with the appropriate antiserum. Becausedifferent preparations of DPPE-conjugated antibody yielded different numbers of plaques with the same spleen cell suspension, a single preparation was used for all experiments reported in this paper. Since the sensitivity of the indicator system affectsthe number of detectable PFC, direct comparisons between the number of Abl and Ab2 are not warranted. RESULTS Kinetics of the Primary and Secondary Abl and Ab2 Responsesto DNP-Ficoll In order to detect murine Ab2 PFC generated in response to immunization, SRBC were coated with affinity-purified F(ab’)2 fragments of polyclonal rabbit anti-DNP antibodies (Ab 1). This was accomplished by conjugating Ab 1 to DPPE. The conjugate was then spontaneously inserted into the SRBC membrane. We used this technique to examine the Ab2 PFC response of BALB/c mice following immunization with DNP-Ficoll, a T-independent type 2 antigen. Becausethe rabbit Ab 1 used to coat the indicator SRBC is unlikely to share idiotopes with the mouse Abl elicited by immunization, Ab2 detected by the plaque assaywas probably the internal image of the antigen (Ab2/3). The Ab 1 PFC responsewas assayedsimultaneously in the samespleen cells using SRBC coated with DNP-derivatized F(ab’)z fragments of rabbit anti-SRBC IgG (11, 12). Three weekslater, a secondary injection of the samedose of DNP-Ficoll
IMMUNOREGULATION
VIA THE IDIOTYPIC NETWORK, I
was administered and the Ab 1 and Ab2 PFC responseswere again assessed.The results, expressedas the number of direct PFC/spleen as a function of time after immunization, are shown in Fig. 1. Several observations are of interest. First, it is evident that, following both primary and secondary immunization, the peaks of the Abl and Ab2 responsescoincide in time, at least within the limitations imposed by the sampling frequency. Both the Ab 1 and Ab2 primary responseswere detectable at Day 3 and peaked on Day 5 (90,667 Abl and 29,750 Ab2 PFC/spleen). Similarly, both the Abl and Ab2 secondary responsespeaked on Day 5 after secondary immunization (39,833 Abl and 60,000 Ab2 PFC/spleen). Since the Ab2 response is presumably driven by Abl, and not by the immunogen, a rather unusual mechanism must exist by which the Ab2 lymphocyte population is stimulated. One possible explanation for the early rise in Ab2 PFC which is consistent with the network theory is that these PFC were induced by the clonal expansion of membrane Ab 1-bearing B cells, the earliest antigen-driven event in a Tindependent response, rather than by secretedAbl molecules. A process in which Ab2-producing cells are induced by direct contact with Ab lbearing B cells would seem to leave little room for T cell involvement. Thus, the Ab2 response might be expected to be T independent. The data in Fig. 1 refer to direct PFC for both the Abl and Ab2 responses.Indirect plaques, developed upon addition of rabbit anti-mouse Ig serum, were never more numerous than the direct plaques in either the primary or secondary response(data not shown). This was expected for Abl PFC, since the response to DNP-Ficoll is known to be of the IgM isotype, as is characteristic of most T-independent responses.The Ab2 response,however, is presumably induced by Abl which, as is the casefor protein antigens in general, should be expected to induce both IgG and IgM antibodies. The fact that the Ab2 response consisted of direct plaques only, therefore, was another indication of its possible T-independent nature. In order to confirm the T-independent nature of this Ab2 response, we measured the Abl and Ab2 PFC responsesof 3 BALB/c (nu/nu) mice 5 days after primary or secondary immunization with DNP-Ficoll (Table 1). The results provide conclusive evidence that the Ab2 response to DNP-Ficoll was indeed T independent. The second unexpected observation which may be made from the data in Fig. 1 is that the peak Abl response was significantly (P < 0.001) lower following secondary immunization than following primary immunization, whereas the peak Ab2 response was significantly (P < 0.00 1) higher following secondary than primary immunization. The secondary Abl response also appeared to rise more slowly, and the secondary Ab2 responsemore rapidly, than was the casefor the corresponding primary responses. These data suggesta regulatory role for the idiotypic network: downregulation of Ab 1 by Ab2 in the secondary response. The mechanism of downregulation is not evident from these experiments, but may result from the persistence of Ab2 antibody in circulation or complexed with Ab 1 receptors on the surface of B cells at the time of the secondary immunization. Kinetics of the Primary and Secondary Abl and Ab2 Responsesto DNP-HGG Using the same assaysystems,we have also investigated the Abl and Ab2 primary and secondary responses of BALB/c mice immunized with a T-dependent antigen: DNP-HGG adsorbed on bentonite. Results, expressed as the numbers of indirect
316
SCHLUETER ET AL. TABLE 1 Response of BALB/c (nu/nu) Mice to DNP-Ficoll or DNP-HGG Immunization” PFC/spleen f SE
Immunization
Abl
Primary DNP-Ficoll Day 5 Secondary DNP-Ficoll Day 5 Primary DNP-HGG Day 7
Ab2
Background
37,800 k 13,183
23,650 + 9,358
500 f 100
61,600 f 41,669
114,900 f 52,337
567 f 367
817 f 109
533 f 60
767 f 203
’ Groups of three 6-week-old female BALB/c (nu/nu) mice were immunized intraperitoneally according to one of the following protocols: 100 rg DNP-Ficoll on Day 0, 100 pg DNP-Ficoll on Days 0 and 2 1, or 100 rg DNP-HGG on Day 0. The splenic Abl and Ab2 PFC responsesof these mice were determined at Day 5 (DNP-Ficoll) or Day 7 (DNP-HGG) after final immunization. Background PFC were determined for the same mice by counting the number of plaques formed in a lawn of uncoated, native SRBC. Plaque numbers for mice immunized with DNP-Ficoll represent direct PFC counts, and those for mice immunized with DNP-HGG represent indirect PFC counts (developed upon addition of diluted rabbit anti-mouse IgG antiserum to the plates).
PFC/spleen, are shown in Fig. 2. Several features of these kinetic curves contrast with the curves obtained from mice immunized with DNP-Ficoll. First, the peaks of the Ab2 PFC responsesto this antigen occurred later than the Abl peaksafter both primary and secondary immunization. In the case of the primary response, Abl PFC levels peaked on Day 7 (277,833 PFC/spleen) while Ab2 PFC levels did not peak until Day 14 (5733 PFC/spleen). In the secondary response, Abl PFC levels essentially peaked
---c ---c 102
’ 0
I
I
I
5
10
15
Day after primary
1
Ab 1 Ab 2 I
20 25 immunization
I
30
FIG. 2. Primary and secondary Abl and Ab2 responsesafter DNP-HGG immunization. Two- to threemonth-old female BALB/c mice were immunized intraperitoneally with 100 rg DNP-HGG adsorbed on bentonite. Three weeks later, some of the mice received an identical secondary injection (indicated by the arrow). The Abl (m) and Ab2 (0) PFC responsesof the spleen cells were measured in groups of three mice killed at 0, 3, 5, 7, 10, 14, and 20 days after primary immunization and at 3, 5, and 7 days after secondary immunization. Results are expressed as the number of indirect PFC/spleen as a function of time after primary immunization. Error bars represent the standard error within each group of mice.
IMMUNOREGULATION
VIA THE IDIOTYPIC NETWORK, I
317
on Days 3 and 5 (96,325 and 12I, 167 PFC/spleen, respectively; there was no statistical difference between the two values), whereasAb2 PFC did not peak until Day 5 (40,3 17 PFCfspleen). Such a delay is compatible with the view that secreted Abl may be handled by the immune system like other conventional protein antigens. Second, both the peak Abl and Ab2 PFC responses occurred more quickly following secondary immunization than primary immunization. This suggeststhat both Abl and Ab2 memory B cells are generated following primary immunization with this antigen. Finally, both Abl and Ab2 PFC responsesdetected after DNP-HGG immunization were largely indirect (only the numbers of indirect PFC are shown), indicating that the mice were mounting principally IgG responsesto the immunogen. All of these observations lend support to the notion that both the Ab 1 and Ab2 responsesto DNPHGG immunization are T dependent. As expected, BALB/c (nu/nu) mice failed to develop Ab 1 or Ab2 plaque numbers which were significantly different from those of background plaques (Table I). The absence of Ab2 PFC in nude mice immunized with a T-dependent antigen may be simply a consequence of the lack of an Abl response, or it may point to a fundamental difference in network behavior following immunization with T-dependent and T-independent antigens. Further experimentation will be necessaryto resolve this question. As in the caseof the responsesto DNP-Ficoll, the peak numbers of Abl PFC were significantly (P < 0.001) lower in the secondary than in the primary response, while the peak numbers of Ab2 PFC were significantly (P = 0.035) higher in the secondary than in the primary response.The finding of lower Abl PFC numbers in the secondary response was certainly unexpected, since accepted immunological theory states that the secondary antibody response to a T-dependent antigen should be higher than the primary response. Once again, this low secondary Abl response to DNP-HGG may be due to downregulation by Ab2. Specificity of the Ab2 Detection System In view of the unexpected results reported above (e.g., the similarity of the kinetics of Ab 1 and Ab2 responsesto DNP-Ficoll and the T independence of the Ab2 response to DNP-Ficoll), it was important to ascertain the specificity of the Ab2 detection system.Specificity was confirmed by plaque inhibition with DNP-lysine. No inhibition was found when irrelevant antigens, such as BSA, and rabbit, mouse or human Ig, were incorporated into the agar layer (Table 2). To make sure that the observed inhibition of Ab2 PFC by DNP-lysine was not due to toxicity of this chemical for the lymphocytes, spleen cells from a mouse immunized with chicken y-globulin (CGG) were plated out in agar containing CGG-coated SRBC and various concentrations of DNP-lysine. In the absence of DNP-lysine, 580 PFC/106 cells were counted. With DNP-lysine at 10e3M there were 5 10 PFC, at 1O-4h4, 658 PFC, and at 10m5M, 675 PFC. In the presenceof 100 pg CGG there were only 53 PFC/ 1O6cells. Similar results were obtained with cells from a mouse immunized with SRBC. Thus, there was a slight ( 12%) nonspecific decreasein the number of PFC at the highest concentration of DNP-lysine ( lop3 A4) but not at concentrations of 10m4M or less, while the PFC were 9 1% inhibited by the specific antigen. No Ab2 PFC above background were found in the spleensof mice immunized with irrelevant antigens, such as ovalbumin or chicken Ig (Table 3). Similarly, no PFC were found in the spleen of DNP-immune mice with indicator SRBC coated with an irrelevant antibody or with nonspecific
318
SCHLUETER ET AL. TABLE 2 Inhibition of Ab2 PFC by DNP-Lysine and by Irrelevant Antigens
Immunogen
Inhibitor
PFC”
Percent inhibition
DNP-FicolP
None DNP-lysine, 10e4M BSA, 100 pg None Mouse Ig, 100 pg Rabbit Ig, 100 ng None DNP-lysine, lo-’ M HGG, 100 pg
96,375/spleen 1,50O/spleen 149,OOO/spleen 169/106cells 182/106cells 2 12/ lo6 cells 6,625Jspleen 850/spleen 5,52S/spleen
99 0 0 0 87 17
DNP-FicollC DNP-HGGd
’ Results are expressed as PFC/spleen when the assayswere done with fresh spleen cells or as PFC/106 cells when frozen cells were used for the assay. b Spleen cells from a mouse immunized 5 days prior to assay with 100 rg DNP-Ficoll. c Spleen cells from a mouse immunized 5 days previously with 100 pg DNP-Ficoll were frozen with 10% DMSO, stored in liquid nitrogen, and thawed on the day of assay. d Spleen cells from a mouse immunized 26 and 5 days prior to assaywith 100 pg DNP-HGG adsorbed on bentonite.
rabbit F(ab’)* (Table 4). In addition, we did not find any increase in background PFC (specific for native SRBC) in mice immunized with DNP, indicating that the Ab2 PFC had not been elicited by inadvertent polyclonal B cell activation. Further evidence of specificity derives from the failure to find substantial numbers of Ab2 PFC in nonimmune mice (Figs. 1 and 2) while large numbers of Ab2 PFC were always found at the time of peak primary and secondary responseto DNP-Ficoll and of secondary response to DNP-HGG. Becauseof the similarity of the cytokinetics of the Abl and Ab2 responsesto DNPFicoll (Fig. l), we were concerned lest the detection system, purported to detect Ab2 PFC, in effect might have detected Abl PFC. Even though it is difficult to see how Ab l-coated SRBC might react with Ab 1 secretedby mouse spleencells, it was important
TABLE 3 Failure of Anti-DNP-Coated SRBC to Reveal PFC in the Spleen of Mice Immunized with Irrelevant Antigens Immunogen
SRBC coat
PFC/ 1O6cells
Ovalbumin”
Rbt (u-DNP (Abl) Ovalbumin None Rbt a-DNP (Abl) None
9 584 2 6 4
Chicken I$
(1Spleen cells from a mouse immunized 26 and 5 days previously with 100pg OVA adsorbedon bentonite were frozen with 10% DMSO, stored in liquid nitrogen, and were thawed on the day of assay. b Spleen cells from a mouse immunized 8 days previously with 100 ng chicken Ig adsorbed on bentonite were frozen and thawed as described above.
IMMUNOREGULATION
319
VIA THE IDIOTYPIC NETWORK, I TABLE 4
Failure of SRBC Coated with an Irrelevant Antibody or with Normal Rabbit F(ab’), to Reveal PFC in the Spleen of DNP-Immunized Mice Immunization
SRBC coat
PFC”
DNP-Ficoll, 26 and 5 days prior to assay
DNP-Rbt a-SRBC Rbt a-DNP (Abl) Rbt ol-OVAb None DNP-Rbt a-SRBC Rbt (Y-DNP (Ab 1) Normal Rbt F(ab’), Rbt a-DNP (Abl)
20,50O/spleen 27,80O/spleen 150Jspleen 375/spleen 176,075Jspleen 6,625/spleen 425/spleen 130/1O6cells
Normal Rbt F(abhb None
l/106 cells 4/l O6cells
DNP-HGG, 26 and 5 days prior to assay DNP-HGGb, 26 and 5 days prior to spleen cell collection
n Results are expressed as PFC/spleen when the assayswere done with fresh spleen cells, or as PFC/106 cells when frozen cells were used for the assay. b Successfulcoating with flab’h from rabbit anti-OVA and from normal rabbit IgG was verified by agglutination with goat anti-rabbit Ig serum and by plaque formation with mouse anti-rabbit Ig lymphocytes. ’ Spleen cells were stored frozen and thawed on the day of assay.
to show experimentally that this is not the case.The results reported in Table 5 demonstrate that Abl-coated SRBC do not react with affinity-purified anti-DNP but do react with the hapten DNP. Having shown that Abl-coated SRBC react with DNP but not with anti-DNP, it was interesting to attempt to demonstrate that Ab 1-coated SRBC also react with Ab2/3, the internal image of the antigen. This was accomplished by subjecting the serum of a mouse immunized with DNP-Ficoll, whose spleen had both Ab 1 and Ab2 PFC, to appropriate absorptions prior to titrating it by hemagglutination (Table 6). DISCUSSION Investigations of the idiotypic network have suffered from the lack of a reliable method for the detection and enumeration of cells producing anti-idiotypic antibodies. TABLE 5 Failure of Anti-DNP Antibody to Agglutinate Abl-Coated SRBC SRBC coat DNP-Rbt (Y-SRBC Rabbit cu-DNP (Ab 1) Rabbit cu-DNP(Abl)
Solution being titrated Affinity purified Rbt ol-DNP (70 wg/ml) Affinity purified Rbt a-DNP (70 fig/ml) DNP-BSA (300 &ml)
Hemagglutination titer 1:500 1:500
’ SRBC in tubes containing serial dilutions of DNP-BSA were washed and resuspendedin rabbit antiBSA serum diluted 1:400.
320
SCHLUETER ET AL. TABLE 6
Agglutination of DNP-Coated SRBC by Abl and Abl-Coated SRBC by Ab2 Derived from the Serum of a Mouse immunized with DNP-Ficoll SRBC coat
Serum absorbed with
Hemagglutination titer
DNP-Rbt ol-SRBC
Unabsorbed DNP-coated SRBC oi-DNP-coated SRBC Unabsorbed DNP-coated SRBC ai-DNP-coated SRBC
I:640
IMMUNOLOGY
14,
3 1l-323 ( 1992)
Behavior of the ldiotypic Network in Conventional Immune Responses I. Kinetics of ldiotypic and Anti-idiotypic Antibodies Following Immunization with T-Independent and T-Dependent Antigens ANNETTE
J. SCHLUETER, DIEGO SEGRE,MARK S. KUHLENSCHMIDT, AND MARIANCELA SEGRE’ Department of Veterinary Pathobiology, University of Illinois, 2001 South Lincoln Avenue, Urbana, Illinois 61801 Received April 13, 1992; acceptedJuly 6, 1992
A minimal requirement in investigations of the behavior of the idiotypic network during immunization is the ability to quantitate both the idiotypic (Abl) and anti-idiotypic (Ab2) responses. Quantitation of Ab2 in serum is complicated by the simultaneous presenceof Abl, so that AblAb2 immune complexes escapedetection. In contrast, immune complexes should not complicate the enumeration of Ab2-producing lymphocytes in a hemolytic plaque assay.This study utilizes a procedure that allows detection of Ab2-producing cells in such an assay.The procedure relies upon the insertion of the appropriate antibody (Abl) into the membrane of indicator SRBC through a covalently attached dipalmitoyl phosphatidylethanolamine (DPPE) tail. When the Ab2 response following murine immunization with DNP-Ficoll was analyzed using such an assay, peak plaque-forming cell (PFC) numbers were found to coincide with peak Abl PFC numbers in both the primary and secondary response. In addition, this Ab2 responsewas found to be T independent. The murine immune responseto DNP-HGG demonstrateda peak Ab2 PFC response which followed the peak Abl PFC response after both primary and secondary immunization. This Ab2 response appeared to be T dependent. The secondary responsesto both DNP-Ficoll and DNP-HGG showed increased levels of Ab2 PFC and decreasedlevels of Abl PFC in comparison to the primary responsesto the same antigens, suggestingthat immunoregulation may occur within these idiotypic networks. o 1992 Academic press, IIIC.
INTRODUCTION It has been over 15 years since Jerne (1) argued for the existence of an idiotypic network (“a web of interacting variable region domains”) involving free immunoglobulin (Ig) molecules and the B cells bearing membrane-bound Ig molecules as receptors. This network was theorized to internally regulate the immune system. The existence of an idiotypic network has been experimentally confirmed, but its regulatory functions during conventional immune responseshave been more difficult to verify. Most of the experiments designed to demonstrate the regulatory function of the idiotypic network have employed monoclonal antibodies or myeloma proteins as the ’ To whom correspondence and reprint requests should be addressed. 311 0008-8749192$5.00 Copyright 0 1992 by Academic Press, Inc. All rights of repreduction in any form reserved.
312
SCHLUETER ET AL.
idiotypic antibody. While this strategy has the obvious advantage of allowing the experimenter to focus on the limited number of idiotopes carried by the homogeneous idiotypic antibody, it does not adequately reflect the more complex situation that occurs in the immune response to most conventional antigens, which elicit a highly heterogeneous idiotypic antibody population. An additional reason for the failure to document a regulatory role of the idiotypic network may be a lack of appropriate technical tools. Most investigators have utilized serologic methods for the quantitation of network components. Evidence for immunoregulation from these studies is based on the cyclical appearance of idiotypic (Abl)’ and anti-idiotypic (Ab2) antibodies, and the fact that the peak titers of these antibodies are at times (2-4), but not always (5), out of phase with each other. Serologic methods are hampered by the formation of circulating Abl-Ab2 immune complexes that would be expected to allow the detection only of the antibody that is present in excess.Therefore, the measured cycles of Abl and Ab2 production may not be truly representative of the actual levels of Abl and Ab2 being produced in response to immune stimulation. A few investigators have studied the idiotypic network through the enumeration of Ab2 plaque-forming cells (PFC) (2-8), but were able to detect only low levels of Ab2 PFC even though the indicator sheep red blood cells (SRBC) were coated with a homogeneous, monoclonal Ab 1. We thought that the failure of previous investigations of cellular network behavior to yield insightful results may have been due to technical inadequacies rather than to inherent lack of power of the hemolytic plaque assay.By anchoring lipid-derivatized, affinity-purified polyclonal Abl in the membrane of indicator SRBC, we were able to detect large numbers of Ab2 in conventional, polyclonal immune responses.In this study, we describe the response of the murine network to immunization with DNP-Ficoll or DNP-human -y-globulin (HGG). Several unexpected results emerge, including simultaneous T-independent peak Abl and Ab2 responsesfollowing DNP-Ficoll immunization. MATERIALS AND METHODS Mice. Two- to three-month-old female BALB/c mice and 6-week-old female BALB/ c (nu/nu) mice were obtained from Harlan-Sprague-Dawley (Indianapolis, IN). Immunization schedule. Mice receiving primary immunization with DNPsO-Ficoll (Biosearch, San Rafael, CA) were injected with 100 pg of the antigen in saline intraperitoneally. Secondary immunizations were given in a similar manner, 2 1 days after primary immunization. Abl and Ab2 plaque-forming cell (PFC) responses of the spleen cells were measured in groups of three mice killed at 0, 3, 4, 5, 7, 10, 12, 15, 18, or 21 days after primary immunization, and at 3, 5, or 7 days after secondary immunization. Mice receiving primary immunizations with DNP,,-HGG were injected intraperitoneally with 100pg of the antigen adsorbedon bentonite (9). Secondary immunizations were given in a similar manner, 20 days after primary immunization. Ab 1 and Ab2 PFC responsesof the spleen cells were measured in groups of three mice killed at 0, 3, 5, 7, 10, 14, or 20 days after primary immunization and at 3, 5, or 7 days after secondary immunization. 2Abbreviations used Ab 1, idiotypic antibody; Ab2, anti-idiotypic antibody; Ab28, anti-idiotypic antibody with idiotope that is bound by the idiotypic antibody’s paratope (aka the internal image of the antigen); DNP, dinitrophenyl hapten; PFC, plaque-forming cell(s); DPPE, dipalmitoyl phosphatidylethanolamine; SPDP: N-succinimidyl-3-(2-pyridyldithio)propionate.
IMMUNOREGULATION
VIA THE IDIOTYPIC
NETWORK,
I
313
Rabbit anti-DNP antibodies. A rabbit was initially immunized SCwith 10 mg DNPBSA in complete Freund’s adjuvant. This was followed by monthly iv injections of 1 mg DNP-BSA. Serum was obtained 1 week after each iv immunization. Ig was precipitated from serum pooled from several bleedings with 50% saturated ammonium sulfate, was dialyzed against borate buffer, pH 8.2, and was passedthrough a column of DNP-chicken y-globulin covalently linked to cyanogen bromide-activated Sepharose 4-B (Pharmacia LKB, Piscataway, NJ). The column was thoroughly washed with borate buffer and the antibody was eluted with 4 X lop3 A4 DNP-lysine. Eluted antibodies were then dialyzed vs a 0.2 M solution of dinitrophenol in order to displace DNP-lysine. Finally, dinitrophenol was removed by passing the antibody through a Dowex 1 X 8 anion exchange resin. F(ab’)* fragments were obtained from the affinitypurified antibody by pepsin digestion (10). Undigested Ig molecules were removed by affinity chromatography on protein A-Sepharose 4-B (Pharmacia LKB, Piscataway, NJ). Hemolytic plaque assay. The hemolytic plaque assayfor the detection of Ab 1 PFC was performed as previously described ( 1l- 13). For the detection of Ab2 PFC, the indicator SRBC were coated with affinity-purified F(ab’)z fragments of rabbit antiDNP antibody (Abl) to which dipalmitoyl phosphatidylethanolamine (DPPE) was covalently attached. DPPE was first conjugated with N-succinimidyl-3-(2-pyridyldithio)propionate (SPDP). A 40 mM solution of DPPE in chloroform:methanol (9:2, v/v) was combined with triethylamine at an 83: 1 (v/v) ratio. This solution was mixed with a 0.12 M solution of SPDP in absolute ethanol at a 2.5: 1 (v/v) ratio and was incubated at 25°C for 2 hr. The organic phase was extracted three times with dH20, dried, and was redissolved in 100% dimethyl sulfoxide. The completeness of the reaction was verified by thin layer chromatography using chloroform:methanol:water (65:25:4) as the solvent, and the product was stored at -20°C. A solution of the antibody (1 to 2 mg/ml) in phosphate-buffered saline (PBS) pH 7.4, was combined with a solution of SPDP in absolute ethanol such that the final molar ratio of SPDP: antibody was 1O:l. The reaction was allowed to proceed for 30 min at 25°C with occasional stirring. The reaction was stopped by overnight dialysis vs 0.1 M acetate buffer, pH 4.5. The disulfide groups on the SPDP-conjugated antibody were reduced by addition of dithiothreitol to a final concentration of 50 mM. The reaction was allowed to proceed for 30 min at 25°C and was stopped by overnight dialysis vs. a 0.01 M NaCl solution containing 5 mM EDTA. The antibody solution was then concentrated to one-tenth its initial volume. A sufficient volume of SPDP-conjugated DPPE solution was added to the antibody to attain a calculated 620: 1 molar ratio of DPPE:antibody, and the reaction was allowed to proceed for 30 min at 25°C. The product was dialyzed vs three changes of PBS containing 0.002% deoxycholate and 0.02% sodium azide. The product was then sonicated for four 30-set cycles using the stepped microtip of a Branson model 185 sonifier cell disrupter (Heat SystemsUltrasonics, Inc., Plainview, NY) set at 30 W. The DPPE-conjugated antibody was then diluted appropriately, mixed with a SRBC suspension, and the mixture was agitated overnight in a rotating drum at 37°C. The integration of rabbit antibody into the SRBC membrane was confirmed by hemolysis upon addition of goat anti-rabbit Ig serum and complement. Optimal coating of SRBC was obtained with 30 to 60 pg of DPPE-derivatized rabbit F(ab’)*/ml of 2.5% erythrocyte suspension. The antibodycoated SRBC were washed, resuspended in saline, and were used in the hemolytic plaque assay. Spleen cells that were not used for the plaque assayswere routinely
314
SCHLUETER ET AL.
+Ab -Ab
g 80,000
sP 5
60,000
2 3
40,000
1 2
z a 20,000
0
5 10 15 Day after primary
20 25 immunization
30
FIG. 1. Primary and secondary Abl and Ab2 responsesafter DNP-Ficoll immunization. Two- to threemonth-old female BALB/c mice were immunized intraperitoneally with 100 pg DNP-Ficoll. Three weeks later, some of the mice received an identical secondary injection (indicated by the arrow). The Abl (m) and Ab2 (0) PFC responsesof the spleen cells were measured in groups of three mice killed at 0, 3, 4, 5, 7, 10, 12, 15, 18, and 21 days after priming and at 3, 5, and 7 days after booster immunization. Results are expressedas the number of direct PFC/spleen as a function of time after primary immunization. Error bars represent the standard error within each group of mice.
frozen with 10% DMSO and stored in liquid nitrogen. In some cases,frozen spleen cells were thawed and utilized for hemolytic plaque assays( 14). Frozen cells yielded 20 to 50% the number of PFC found with fresh cells. SRBC were also coated with other affinity-purified antibodies (rabbit anti-OVA) and protein antigens (OVA, HGG, F(ab’)* from nonspecific rabbit IgG). Successful insertion of the DPPE-conjugated proteins into the SRBC membrane was confirmed by hemagglutination or hemolysis with the appropriate antiserum. Becausedifferent preparations of DPPE-conjugated antibody yielded different numbers of plaques with the same spleen cell suspension, a single preparation was used for all experiments reported in this paper. Since the sensitivity of the indicator system affectsthe number of detectable PFC, direct comparisons between the number of Abl and Ab2 are not warranted. RESULTS Kinetics of the Primary and Secondary Abl and Ab2 Responsesto DNP-Ficoll In order to detect murine Ab2 PFC generated in response to immunization, SRBC were coated with affinity-purified F(ab’)2 fragments of polyclonal rabbit anti-DNP antibodies (Ab 1). This was accomplished by conjugating Ab 1 to DPPE. The conjugate was then spontaneously inserted into the SRBC membrane. We used this technique to examine the Ab2 PFC response of BALB/c mice following immunization with DNP-Ficoll, a T-independent type 2 antigen. Becausethe rabbit Ab 1 used to coat the indicator SRBC is unlikely to share idiotopes with the mouse Abl elicited by immunization, Ab2 detected by the plaque assaywas probably the internal image of the antigen (Ab2/3). The Ab 1 PFC responsewas assayedsimultaneously in the samespleen cells using SRBC coated with DNP-derivatized F(ab’)z fragments of rabbit anti-SRBC IgG (11, 12). Three weekslater, a secondary injection of the samedose of DNP-Ficoll
IMMUNOREGULATION
VIA THE IDIOTYPIC NETWORK, I
was administered and the Ab 1 and Ab2 PFC responseswere again assessed.The results, expressedas the number of direct PFC/spleen as a function of time after immunization, are shown in Fig. 1. Several observations are of interest. First, it is evident that, following both primary and secondary immunization, the peaks of the Abl and Ab2 responsescoincide in time, at least within the limitations imposed by the sampling frequency. Both the Ab 1 and Ab2 primary responseswere detectable at Day 3 and peaked on Day 5 (90,667 Abl and 29,750 Ab2 PFC/spleen). Similarly, both the Abl and Ab2 secondary responsespeaked on Day 5 after secondary immunization (39,833 Abl and 60,000 Ab2 PFC/spleen). Since the Ab2 response is presumably driven by Abl, and not by the immunogen, a rather unusual mechanism must exist by which the Ab2 lymphocyte population is stimulated. One possible explanation for the early rise in Ab2 PFC which is consistent with the network theory is that these PFC were induced by the clonal expansion of membrane Ab 1-bearing B cells, the earliest antigen-driven event in a Tindependent response, rather than by secretedAbl molecules. A process in which Ab2-producing cells are induced by direct contact with Ab lbearing B cells would seem to leave little room for T cell involvement. Thus, the Ab2 response might be expected to be T independent. The data in Fig. 1 refer to direct PFC for both the Abl and Ab2 responses.Indirect plaques, developed upon addition of rabbit anti-mouse Ig serum, were never more numerous than the direct plaques in either the primary or secondary response(data not shown). This was expected for Abl PFC, since the response to DNP-Ficoll is known to be of the IgM isotype, as is characteristic of most T-independent responses.The Ab2 response,however, is presumably induced by Abl which, as is the casefor protein antigens in general, should be expected to induce both IgG and IgM antibodies. The fact that the Ab2 response consisted of direct plaques only, therefore, was another indication of its possible T-independent nature. In order to confirm the T-independent nature of this Ab2 response, we measured the Abl and Ab2 PFC responsesof 3 BALB/c (nu/nu) mice 5 days after primary or secondary immunization with DNP-Ficoll (Table 1). The results provide conclusive evidence that the Ab2 response to DNP-Ficoll was indeed T independent. The second unexpected observation which may be made from the data in Fig. 1 is that the peak Abl response was significantly (P < 0.001) lower following secondary immunization than following primary immunization, whereas the peak Ab2 response was significantly (P < 0.00 1) higher following secondary than primary immunization. The secondary Abl response also appeared to rise more slowly, and the secondary Ab2 responsemore rapidly, than was the casefor the corresponding primary responses. These data suggesta regulatory role for the idiotypic network: downregulation of Ab 1 by Ab2 in the secondary response. The mechanism of downregulation is not evident from these experiments, but may result from the persistence of Ab2 antibody in circulation or complexed with Ab 1 receptors on the surface of B cells at the time of the secondary immunization. Kinetics of the Primary and Secondary Abl and Ab2 Responsesto DNP-HGG Using the same assaysystems,we have also investigated the Abl and Ab2 primary and secondary responses of BALB/c mice immunized with a T-dependent antigen: DNP-HGG adsorbed on bentonite. Results, expressed as the numbers of indirect
316
SCHLUETER ET AL. TABLE 1 Response of BALB/c (nu/nu) Mice to DNP-Ficoll or DNP-HGG Immunization” PFC/spleen f SE
Immunization
Abl
Primary DNP-Ficoll Day 5 Secondary DNP-Ficoll Day 5 Primary DNP-HGG Day 7
Ab2
Background
37,800 k 13,183
23,650 + 9,358
500 f 100
61,600 f 41,669
114,900 f 52,337
567 f 367
817 f 109
533 f 60
767 f 203
’ Groups of three 6-week-old female BALB/c (nu/nu) mice were immunized intraperitoneally according to one of the following protocols: 100 rg DNP-Ficoll on Day 0, 100 pg DNP-Ficoll on Days 0 and 2 1, or 100 rg DNP-HGG on Day 0. The splenic Abl and Ab2 PFC responsesof these mice were determined at Day 5 (DNP-Ficoll) or Day 7 (DNP-HGG) after final immunization. Background PFC were determined for the same mice by counting the number of plaques formed in a lawn of uncoated, native SRBC. Plaque numbers for mice immunized with DNP-Ficoll represent direct PFC counts, and those for mice immunized with DNP-HGG represent indirect PFC counts (developed upon addition of diluted rabbit anti-mouse IgG antiserum to the plates).
PFC/spleen, are shown in Fig. 2. Several features of these kinetic curves contrast with the curves obtained from mice immunized with DNP-Ficoll. First, the peaks of the Ab2 PFC responsesto this antigen occurred later than the Abl peaksafter both primary and secondary immunization. In the case of the primary response, Abl PFC levels peaked on Day 7 (277,833 PFC/spleen) while Ab2 PFC levels did not peak until Day 14 (5733 PFC/spleen). In the secondary response, Abl PFC levels essentially peaked
---c ---c 102
’ 0
I
I
I
5
10
15
Day after primary
1
Ab 1 Ab 2 I
20 25 immunization
I
30
FIG. 2. Primary and secondary Abl and Ab2 responsesafter DNP-HGG immunization. Two- to threemonth-old female BALB/c mice were immunized intraperitoneally with 100 rg DNP-HGG adsorbed on bentonite. Three weeks later, some of the mice received an identical secondary injection (indicated by the arrow). The Abl (m) and Ab2 (0) PFC responsesof the spleen cells were measured in groups of three mice killed at 0, 3, 5, 7, 10, 14, and 20 days after primary immunization and at 3, 5, and 7 days after secondary immunization. Results are expressed as the number of indirect PFC/spleen as a function of time after primary immunization. Error bars represent the standard error within each group of mice.
IMMUNOREGULATION
VIA THE IDIOTYPIC NETWORK, I
317
on Days 3 and 5 (96,325 and 12I, 167 PFC/spleen, respectively; there was no statistical difference between the two values), whereasAb2 PFC did not peak until Day 5 (40,3 17 PFCfspleen). Such a delay is compatible with the view that secreted Abl may be handled by the immune system like other conventional protein antigens. Second, both the peak Abl and Ab2 PFC responses occurred more quickly following secondary immunization than primary immunization. This suggeststhat both Abl and Ab2 memory B cells are generated following primary immunization with this antigen. Finally, both Abl and Ab2 PFC responsesdetected after DNP-HGG immunization were largely indirect (only the numbers of indirect PFC are shown), indicating that the mice were mounting principally IgG responsesto the immunogen. All of these observations lend support to the notion that both the Ab 1 and Ab2 responsesto DNPHGG immunization are T dependent. As expected, BALB/c (nu/nu) mice failed to develop Ab 1 or Ab2 plaque numbers which were significantly different from those of background plaques (Table I). The absence of Ab2 PFC in nude mice immunized with a T-dependent antigen may be simply a consequence of the lack of an Abl response, or it may point to a fundamental difference in network behavior following immunization with T-dependent and T-independent antigens. Further experimentation will be necessaryto resolve this question. As in the caseof the responsesto DNP-Ficoll, the peak numbers of Abl PFC were significantly (P < 0.001) lower in the secondary than in the primary response, while the peak numbers of Ab2 PFC were significantly (P = 0.035) higher in the secondary than in the primary response.The finding of lower Abl PFC numbers in the secondary response was certainly unexpected, since accepted immunological theory states that the secondary antibody response to a T-dependent antigen should be higher than the primary response. Once again, this low secondary Abl response to DNP-HGG may be due to downregulation by Ab2. Specificity of the Ab2 Detection System In view of the unexpected results reported above (e.g., the similarity of the kinetics of Ab 1 and Ab2 responsesto DNP-Ficoll and the T independence of the Ab2 response to DNP-Ficoll), it was important to ascertain the specificity of the Ab2 detection system.Specificity was confirmed by plaque inhibition with DNP-lysine. No inhibition was found when irrelevant antigens, such as BSA, and rabbit, mouse or human Ig, were incorporated into the agar layer (Table 2). To make sure that the observed inhibition of Ab2 PFC by DNP-lysine was not due to toxicity of this chemical for the lymphocytes, spleen cells from a mouse immunized with chicken y-globulin (CGG) were plated out in agar containing CGG-coated SRBC and various concentrations of DNP-lysine. In the absence of DNP-lysine, 580 PFC/106 cells were counted. With DNP-lysine at 10e3M there were 5 10 PFC, at 1O-4h4, 658 PFC, and at 10m5M, 675 PFC. In the presenceof 100 pg CGG there were only 53 PFC/ 1O6cells. Similar results were obtained with cells from a mouse immunized with SRBC. Thus, there was a slight ( 12%) nonspecific decreasein the number of PFC at the highest concentration of DNP-lysine ( lop3 A4) but not at concentrations of 10m4M or less, while the PFC were 9 1% inhibited by the specific antigen. No Ab2 PFC above background were found in the spleensof mice immunized with irrelevant antigens, such as ovalbumin or chicken Ig (Table 3). Similarly, no PFC were found in the spleen of DNP-immune mice with indicator SRBC coated with an irrelevant antibody or with nonspecific
318
SCHLUETER ET AL. TABLE 2 Inhibition of Ab2 PFC by DNP-Lysine and by Irrelevant Antigens
Immunogen
Inhibitor
PFC”
Percent inhibition
DNP-FicolP
None DNP-lysine, 10e4M BSA, 100 pg None Mouse Ig, 100 pg Rabbit Ig, 100 ng None DNP-lysine, lo-’ M HGG, 100 pg
96,375/spleen 1,50O/spleen 149,OOO/spleen 169/106cells 182/106cells 2 12/ lo6 cells 6,625Jspleen 850/spleen 5,52S/spleen
99 0 0 0 87 17
DNP-FicollC DNP-HGGd
’ Results are expressed as PFC/spleen when the assayswere done with fresh spleen cells or as PFC/106 cells when frozen cells were used for the assay. b Spleen cells from a mouse immunized 5 days prior to assay with 100 rg DNP-Ficoll. c Spleen cells from a mouse immunized 5 days previously with 100 pg DNP-Ficoll were frozen with 10% DMSO, stored in liquid nitrogen, and thawed on the day of assay. d Spleen cells from a mouse immunized 26 and 5 days prior to assaywith 100 pg DNP-HGG adsorbed on bentonite.
rabbit F(ab’)* (Table 4). In addition, we did not find any increase in background PFC (specific for native SRBC) in mice immunized with DNP, indicating that the Ab2 PFC had not been elicited by inadvertent polyclonal B cell activation. Further evidence of specificity derives from the failure to find substantial numbers of Ab2 PFC in nonimmune mice (Figs. 1 and 2) while large numbers of Ab2 PFC were always found at the time of peak primary and secondary responseto DNP-Ficoll and of secondary response to DNP-HGG. Becauseof the similarity of the cytokinetics of the Abl and Ab2 responsesto DNPFicoll (Fig. l), we were concerned lest the detection system, purported to detect Ab2 PFC, in effect might have detected Abl PFC. Even though it is difficult to see how Ab l-coated SRBC might react with Ab 1 secretedby mouse spleencells, it was important
TABLE 3 Failure of Anti-DNP-Coated SRBC to Reveal PFC in the Spleen of Mice Immunized with Irrelevant Antigens Immunogen
SRBC coat
PFC/ 1O6cells
Ovalbumin”
Rbt (u-DNP (Abl) Ovalbumin None Rbt a-DNP (Abl) None
9 584 2 6 4
Chicken I$
(1Spleen cells from a mouse immunized 26 and 5 days previously with 100pg OVA adsorbedon bentonite were frozen with 10% DMSO, stored in liquid nitrogen, and were thawed on the day of assay. b Spleen cells from a mouse immunized 8 days previously with 100 ng chicken Ig adsorbed on bentonite were frozen and thawed as described above.
IMMUNOREGULATION
319
VIA THE IDIOTYPIC NETWORK, I TABLE 4
Failure of SRBC Coated with an Irrelevant Antibody or with Normal Rabbit F(ab’), to Reveal PFC in the Spleen of DNP-Immunized Mice Immunization
SRBC coat
PFC”
DNP-Ficoll, 26 and 5 days prior to assay
DNP-Rbt a-SRBC Rbt a-DNP (Abl) Rbt ol-OVAb None DNP-Rbt a-SRBC Rbt (Y-DNP (Ab 1) Normal Rbt F(ab’), Rbt a-DNP (Abl)
20,50O/spleen 27,80O/spleen 150Jspleen 375/spleen 176,075Jspleen 6,625/spleen 425/spleen 130/1O6cells
Normal Rbt F(abhb None
l/106 cells 4/l O6cells
DNP-HGG, 26 and 5 days prior to assay DNP-HGGb, 26 and 5 days prior to spleen cell collection
n Results are expressed as PFC/spleen when the assayswere done with fresh spleen cells, or as PFC/106 cells when frozen cells were used for the assay. b Successfulcoating with flab’h from rabbit anti-OVA and from normal rabbit IgG was verified by agglutination with goat anti-rabbit Ig serum and by plaque formation with mouse anti-rabbit Ig lymphocytes. ’ Spleen cells were stored frozen and thawed on the day of assay.
to show experimentally that this is not the case.The results reported in Table 5 demonstrate that Abl-coated SRBC do not react with affinity-purified anti-DNP but do react with the hapten DNP. Having shown that Abl-coated SRBC react with DNP but not with anti-DNP, it was interesting to attempt to demonstrate that Ab 1-coated SRBC also react with Ab2/3, the internal image of the antigen. This was accomplished by subjecting the serum of a mouse immunized with DNP-Ficoll, whose spleen had both Ab 1 and Ab2 PFC, to appropriate absorptions prior to titrating it by hemagglutination (Table 6). DISCUSSION Investigations of the idiotypic network have suffered from the lack of a reliable method for the detection and enumeration of cells producing anti-idiotypic antibodies. TABLE 5 Failure of Anti-DNP Antibody to Agglutinate Abl-Coated SRBC SRBC coat DNP-Rbt (Y-SRBC Rabbit cu-DNP (Ab 1) Rabbit cu-DNP(Abl)
Solution being titrated Affinity purified Rbt ol-DNP (70 wg/ml) Affinity purified Rbt a-DNP (70 fig/ml) DNP-BSA (300 &ml)
Hemagglutination titer 1:500 1:500
’ SRBC in tubes containing serial dilutions of DNP-BSA were washed and resuspendedin rabbit antiBSA serum diluted 1:400.
320
SCHLUETER ET AL. TABLE 6
Agglutination of DNP-Coated SRBC by Abl and Abl-Coated SRBC by Ab2 Derived from the Serum of a Mouse immunized with DNP-Ficoll SRBC coat
Serum absorbed with
Hemagglutination titer
DNP-Rbt ol-SRBC
Unabsorbed DNP-coated SRBC oi-DNP-coated SRBC Unabsorbed DNP-coated SRBC ai-DNP-coated SRBC
I:640
