Functional Relationship Between T15 and J558 Idiotypes in BALB/c Mice

In inbred strains of mice, antiphosphorylcholine (PC) and anti-α1,3 dextran (DEX). antibodies are structurally distinct from each other and have been shown to exhibit noncrossreactive antigen binding and idiotypic specificities. However, the prototype anti-PC and anti-DEX antibodies, TEPC15 and J558, respectively, were shown to be connected via a common autoantiidiotypic monoclonal antibody isolated from newborn BALB/c mice. The capacity of various monoclonal anti-PC and anti-DEX antibodies as well as the antigens PC and DEX to modulate T15 and J558 idiotypes in BALB/c mice was tested by their administration to newborn mice. Anti-PC antibodies of the .T15 idiotype injected into 2-4-day-old mice, at a time when T15 anti-PC precursors develop in BALB/c mice, suppressed the anti- PC response of these mice at 6 weeks of age. Similarly, J558 antibodies injected into 8-12-day-old mice, at a time when J558 precursors normally develop, suppressed the response to DEX. As a further demonstration of this connectivity, the injection of J558 into 4-day-old mice led to a down modulation of T15 idiotype, whereas both T15 and a minor idiotypeexpressing antibody M167 when injected into 8-12-day-old mice caused a reduction in expression of the J558 idiotype. As predicted from in vitro analysis, injection of anti-PC antibodies of the M167 idiotype 2 to 4 days after birth enhanced the subsequent response to PC. However, anti-PC antibodies expressing another minor M603 idiotype did not affect the PC. response. The results parallel the in vitro enhancement of M167 antibodies but not M603 on T15 binding to antiidiotype in vitro. Similarly, anti-DEX antibodies expressing the M104E idiotype had no detectable effects on the capacity to respond to PC or DEX or on the expression of T15 and J558 idiotypes as adults. Exposure of newborn mice to PC led to a dramatic reduction in the response to DEX as adults, whereas exposure to DEX at this stage of development had no effect on response to PC as adults. Collectively, these observations provide evidence for a complex functional connectivity between T15 and J558 idiotype-bearing B cells during ontogeny and extend our previous observations that development of these idiotypes is regulated by idiotype-directed interactions between B cells or their immunoglobulin products.


INTRODUCTION
In mammals, antibody responses to certain defined antigens including bacterial polysaccharides are restricted to the expansion of one or a few clones of B cells resulting in idiotypic dominance of antibodies produced by these clones in the serum (Blomberg et al., 1972;Claflin et al., 1974;  1973; Lieberman et al., 1976;-Hansberg et ai., 1977;Mtikel/i and Karjalainen, 1977;Capra et al., 1977). It was shown that production of antibodies expressing these dominant idiotypes could be modulated by passive transfer of antiidiotypic antibodies (Cosenza and K6hler, 1972;Eichman, 1974). In addition to the numerous demonstrations of direct effects of polyclonal, as well as monoclonal antiidiotypic antibodies on antibody responses in vivo, evidence has also accumulated for the existence of regulatory B-cell networks (Cosenza and K6hler,214 M. VAKIL AND J. F. KEARNEY 1972; Cazenave, 1977; Kelsoe and Cerny, 1979;Bona et al., 1981;Rajewsky and Takemori, 1983;. During attempts to determine whether such idiotype-directed interactions occur under physiologically relevant situations, we isolated B-cell hybridoma-derived antibodies from perinatal mice that react with structurally and genetically defined immunoglobulin idiotypes Kearney et al., 1987). In addition, extensive connectivity, which is idiotype-associated, has been demonstrated between these perinatally derived antibodies (Holmberg et al., 1984;Kearney et al., 1987).
Based on the binding characteristics of these monoclonal antibodies in vitro and the demonstration that administration of these antibodies has potent longlived effects on the expression of target idiotypes in vivo , we proposed that B cells interact through immunoglobulin idiotypic determinants and that these interactions are involved in the establishment of the adult B-cell repertoire.
A monoclonal autoantiidiotypic antibody BD2 was shown to regulate the capacity of BALB/c mice to respond to both phosphorylcholine (PC) and c1,3 dextran (DEX). Furthermore, the nature and extent of these modulating effects depended on the age at which BD2 was injected into mice. Additionally, we also demonstrated that inactivation of similar early appearing antiidiotypic B cells in vivo led to a lack of development of T15-and. J558-1ike B cells and a diminished capacity to respond to PC and DEX . Based on these criteria, we proposed that idiotypic interactions among complemen-tary sets ot B cells during ontogeny were required for the development of T15 and J558 idiotypes.
In this report, we have extended these studies by analyzing the effects of administration of soluble idiotype-bearing antibodies and antigen on the development of these germline-encoded responses to PC and DEX. We demonstrate that development of the dominant idiotypes T15 and J558 can be altered by neonatal exposure to the antibodies expressing either T15 or J558 idiotypes as well as antigens PC and DEXo These observations establish a functional relationship between idiotypically connected B cells and suggest that such idiotypic connections may regulate the program of repertoire development.

Characteristics and Reactivity Patterns of Idiotype-Bearing Antibodies
Antigen-responsive T15-1ike and J558-1ike B cells in BALB/c mice first appear between 5 to 7 days and 12 to 15 days after birth, respectively (Sigal et al., 1977;Stohrer and Kearney, 1984). Since our hypothesis proposes that idiotypic interactions at a receptor level are essential for the selective expansion of T15 / and J558 / B cells, we wished to test whether passive administration of soluble idiotypebearing antibodies during this period would interfere with the normal development of T15-and J558like B cells. Table 1 summarizes the antigen-binding and idiotypic characteristics of the antibodies used in these studies.  Fig. 1A is a diagrammatic summary of the reactivities of several of these well-characterized monoclonal antibodies that bind to either PC or DEX. Also included are the monoclonal antiidiotypic antibodies BD2, DB3, and 4F-1:BD2 derived from 2-day-old mice is an autoantiidiotypic antibody that reacts with T15, M167, and J558 (Fig. 1B). DB3, derived from fetal mice, in turn, reacts with BD2 . 4F-1 was raised by immunizat'ion of adult BALB/c mice with a preparation of purified M167 and recognizes a determinant unique to M167 not expressed by T15 or J558 (Kang and K6hler, 1990 Fig. 2A). Partial reduction of antibody levels expressing the T15 idiotype was induced when treatment was delayed until 8 days after birth (results not shown). In mice injected with the same antibodies on day 10, the response to PC was not significantly different as compared to that in saline-treated controls (Fig. 2B). Injection of 10/zg of the non-PC binding IgA antibody J558 on day 2 led to partial suppression of T15 idiotype but only 15 to 20% reduction in the total antibody response to PC. In contrast, the response to PC in mice treated with J558 10 days after birth was no different than in saline-treated control mice in magnitude or idiotype profile (Fig. 2B). To test the possibility that the injected antibody forms antigenantibody complexes with PC that may be responsible for suppression of endogenous T15 idiotypebearing B cells, U10, a non-PC-binding somatic variant of $107, an IgA T15 / antibody (Behar et al., 1989) was injected into 2-day-old mice. This treatment also resulted in a reduction of the response to PC ( Fig. 2A)  These experiments showed that administration of saline-treated controls (Fig. 3A). The anti-DEX exogenous soluble T15 or J558 into neonatal mice response was also reduced by treatment with T15 resulted in the suppression of these idiotypes when antibodies between days 2 and 4. However, injection they were challenged as adults with the appropriate of J558 on day 10 resulted in a reduced response to antigen. Furthermore, this idiotypic suppression was DEX and completely abolished the production of the induced most efficiently at critical periods during J558 idiotype (Fig. 3B). Partial reduction of the development. In accord with our previous studies, response to DEX and J558 idiotype was still evident these periods coincided with the developmental in mice treated with 10/g of J558 antibody as late windows when B-cell clones expressing T15 or J558 as 15 days after birth (results not shown). Treatment idiotypes are normally generated in BALB/c mice with T15 idiotype on day 10 also brought about a (Stohrer and Kearney, 1984;Vakil et al., 1991 (Claflin, 1976). Similarly, antibodies expressing the M104E idiotype represent a variable A minor but significant portion of the response to DEX (Stohrer and Kearney, 1984). We have previously reported that a monoclonal antiidiotype antibody BD2 derived from 2-day-old BALB/c mice also reacts with antibodies expressing the M167 idiotype but not the M603 or M104E idiotypes  shown in Fig. 4A, injection of 10/g of the IgM antibody HPCM27, which expresses the M167 idiotype on day 4, led to a threefold increase in the total PC-specific IgM in the serum when these mice were challenged with R36A as adults, with T15 as the dominant idiotype. This treatment also led to a loss of J558 dominance (Fig. 4B) in response to DEX.
However, treatment with HPCM27 on day 8 or 12 did not induce significant alteration of the antibody responses to PC (Fig. 4A), but led to a diminished response to DEX and reduction of J558 idiotype (Fig.  4B) treatment In order to investigate the possibility that M167like B cells play a role in the development of T15and J558-1ike B cells in neonatal mice, the following experiment was conducted. Fifty /g of 4F-1, a monoclonal antibody specifically reactive with M167 but not the T15 idiotype, was injected into neonatal mice on days 2 and 4 in an attempt to alter the functional development of endogenous B cells expressing the M167 idiotype. When mice treated in this fashion were challenged with R36A 6 to 8 weeks later, their antibody response to PC was only 30% of that in saline-treated control mice with almost total loss of T15 idiotype although 4F1 does not react with T15 / antibodies (Fig. 5A)  treatment failed to alter the antibody response to effect on the anti-PC response. We have previously DEX (Fig. 5B). These observations lead us to con-shown that the first DEX-specific precursors. clude that the presence of functional B cells proappearing in the neonatal period express the ducing M167-1ike antibodies may assist in the normally minor M104E idiotype and that J558-1ike B development and expansion of B cells producing cells are not detected in the splenic focus assay until T15-1ike antibodies, but do not affect the develop-12 to 15 days after birth (Stohrer and Kearney, ment of J558-1ike B cells. 1984). This is reflected also by the priming effect of DEX treatment on days 2 and 7 (Fig. 6B), when Effects of Neonatal Exposure of Mice to PC and these non-J558anti-DEX precursors can be primed.

DEX
Thus, although the injection of J558 antibodies into The most striking feature of the observations 2or 8-day-old mice results in the modulation of described before was that soluble T15 and J558 could T15 idiotype, B cells expressing the J558 idiotype reciprocally inhibit the development of the J558 or may not be induced in mice before 12 to 15 days T15 idiotypes in vivo. This added further support to after birth, and, therefore, injection of dextran into our previous hypothesis that the development of mice before development of T15 / B cells does not T15-and J558-1ike B cells is regulated by a set of affect T15 expression later in life.
common autoantiidiotypic B cells. We previously demonstrated that exposure of 2-4-day-old mice to R36A resulted in the priming of T15 negative, DISCUSSION largely M167-1ike antibodies, whereas exposure to the same antigen between days 6 and 21 primed for Our previous studies indicated that B cells pro-T15-1ike antibodies (Vakil et al., 1991). The next ducing immunoglobulins that cross-react in vitro experiment was, therefore, designed to study con-with purified antibodies expressing either the T15 or nectivity between the T15 and J558 idiotypes in the J558 idiotype exist in neonatal mice and can be response to perinatal administration of PC and DEX isolated in the form of hybridomas. The dual reantigens, activity or cross-reactivity of one such autoanti-Groups of mice were injected with 2 xl07 R36A on idiotypic antibody BD2 has been confirmed by days 2, 9, or 15 after birth and they were challenged appropriate binding-inhibition assays (Vakil and with dextran at 7 weeks of age. As shown in Fig. Kearney, 1986). Administration of purified BD2 to 6A, in mice exposed to PC 2 days after birth, the neonatal mice was shown to enhance the developantibody response to DEX when challenged as ment of both the T15 and the J558 idiotypes. As an adults was-greatly diminished, whereas in mice extension of these previous findings, experiments exposed to PC on day 9, a moderate response to were designed to address whether the idiotypic DEX was observed although the J558 idiotype was connectivity between T15 and J558 antibodies was diminished in proportion. In contrast, treatment of functionally relevant with respect to the observed mice with PC at 15 days after birth did not affect program of development of the B-cell repertoire. the quantity or J558 portion of the anti-DEX We, therefore, sought to test the idea that adminiresponse. In a similar experiment, mice were stration of soluble idiotype-bearing immunoinjected with 5/g of DEX on days 2, 7, 14, or 21 globulins would block signaling through B-cellafter birth and challenged with R36A and DEX at 8 associated idiotypes on the precursors of endoweeks of age (Fig. 6B). All DEX-treated mice genous T15-and J558-1ike B cells. mounted an enhanced response to DEX as compared The first set of experiments demonstrated that to saline-treated controls. However, in mice treated administration of soluble antibodies expressing the with DEX on days 2 or 7 after birth, the J558 idio-T15 or the J558 idiotype specifically inhibits the type failed to dominate the immune response. All development of both T15-and J558-idiotype-bearing mice in this set also mounted normal anti-PC B cells, supporting the existence of a common reguresponses with dominant expression of the T15 idio-latory cell for both idiotypes. Next, we showed type (results not shown), antibodies expressing the M167 idiotype that Thus, exposure to PC prior to the development of normally constitutes a minor component of PC-T15and J558-1ike B cells resulted in a distortion of binding antibodies enhanced the response to PC anti-DEX responses, however, exposure to DEX while inhibiting the response to DEX. We have preduring these same windows of development had no viously noted that antibody M167 enhances binding DEVELOPMENTAL RELATIONSHIP BETWEEN T15 AND J558 IDIOTYPES 221 of BD2 to T15 in vitro, whereas it inhibits binding of the same antibody to J558   It is probable that M167-1ike B cells, which are suggests that there is fine-tuning of these interacamongst the earliest PC-responsive B cells to tions among newly formed B cells in the neonate. become functional (Vakil et al., 1991)

may in fact
It is also possible that clones of B cells generated serve as autoantiidiotypic B cells, as proposed in in the primary lymphoid organs during development Fig. 1B, and aid in the idiotype-directed selection of may engage in idiotypic interactions prior to surface T15-1ike B cells involving intermediate BD2-1ike B expression of conventional surface immunoglobulin. cells. The loss of T15 idiotype in mice treated with The expression of V H structures in conjunction with anti-M167 antibody 4F-1 supports this notion. The an invariant surrogate light-chain complex conidiotype-directed signals generated by BD2-1ike B sisting of X5 and Vpre-B molecules would also procells may induce a selective differentiation and/or vide a possibility for idiotypic selection to occur at expansion of the target T15-1ike B cells, the pre-B-cell level.
The reasons for inhibition of J558 idiotype by Our results show that interference with the M167 are unclear. However, normal expansion of developmental program of clonal selection can lead J558 idiotype occurs during the latter part of the to serious deficiencies in the immune response to second week of life in BALB/c mice By this time, exogenous antigens later in life. We have recently PC-specific B cells expressing the T15 idiotype are at obtained evidence that neonatal administration of adult levels. The initially predominant M167-PC also leads to a modulation of the immune idiotype-bearing B cells are now diluted and may response to group A Streptococcus (Vakfl and not affect the development of the J558 idiotype. It is Kearney, unpublished observations). It is likely that also possible that other anti-J558 B cells besides mice neonatally exposed to these antigens suffer those represented by hybridoma BD2 are responsible from several other deficits in their B-cell repertoires for development of J558-1ike B cells during the that are as yet unknown to us. It is known that second and third week of life in BALB/c mice.
human infants do not respond to bacterial polysac-One of the most striking findings in these, studies charide antigens until about 2 years of age and was that exposure of newborn mice to R36A also infections with S. pneumoniae, H. influenzae, and resulted in the loss of the capacity to respond to other Gram-positive bacteria are of common occur-DEX. This effect could also be mimicked by injection rence during infancy. The implications of our of antibodies expressing the M167 idiotype between studies, therefore, go beyond the purpose, which is days 4 and 12. These experiments added further an attempt to understand the cellular events that evidence that functional idiotypic connections exist lead to the selection and establishment of a normal between sets of B cells specific for PC and DEX and adult B-cell repertoire in mammals. They emphasize that perturbation of the developing PC-responsive B the importance of vigorous immunological monicells can adversely affect development of those that toring in human infants receiving synthetic conbind to a structurally distinct antigen such as DEX, jugate vaccines containing bacterial polysaccharides which appears later during development, that are currently being developed for the induction Based on our previous findings, the mininetwork of immunity in human infants. described in Fig. 1 must represent only a small portion of the overall potential for interactions within the early B-cell repertoire. From the reactivity pat-METHODS terns of other hybridoma antibodies isolated from fetal and neonatal BALB/c mice, it is evident that the perinatal B-cell repertoire is highly connected Mice through imnmunoglobulin idiotypes (Kearney et al., Male and female BALB/c mice were obtained from 1987). It is clear that during development, some Jackson laboratories, Bar Harbour, ME, and bred in clones increase in frequency while others decline our facility. (Sigal et al., 1977;Stohrer and Kearney, 1984;Vakil et al., 1991). The mechanisms involved in this Neonatal Treatments and Immunizations waxing and waning of clones are not understood, however, they may involve competition for space, Neonatal mice were injected with monoclonal anticytokines, or idiotype-directed signals. The fact that bodies in saline or 2x107 R36a or 5/zg dextran the pattern of development of the B-cell-specificity B1355S in saline by intraperitoneal route at ages repertoire repeats itself generation after generation indicated in the Results section. Control mice were DEVELOPMENTAL RELATIONSHIP BETWEEN T15 AND J558 IDIOTYPES 223 injected with saline. Adult mice were immunized with 2x108 R36A in saline or dextran B1355S in saline intraperitoneally, and were bled 7 days later.
Quantitation of Serum Antibodies PCand DEX-specific antibodies in sera of immunized mice were analyzed in a quantitative ELISA, as previously described (Stohrer and Kearney, 1984). Briefly, polyvinyl microtiter plates were coated with PC-BSA, DEX-BSA, or monoclonal antiidiotypic antibodies, blocked and incubated with serial dilutions of serum samples from individual mice, or known amounts of monoclonal standards.
The assays were developed with phosphataselabeled goat antimouse IgM antibodies and phosphatase substrate. The absorbance values were read in a Titertek multiscan (Flow Laboratories, McCleany, VA) and the standard equivalents were calculated based on reference O.D. values of monoclonal antibody standards. Data are represented as geometric means of groups of six to eight mice with standard deviations calculated using a rank test.