The role of receptor IgM and IgD in determining triggering and induction of tolerance in murine B cells

DNA vaccination and the immune responsiveness of neonates

Neonates often respond poorly to conventional vaccines or microbial infections. Immaturity of the immune system has been considered to play a role in this regard. However, accumulating evidence shows that in certain conditions, neonatal inoculation of antigens leads to protective immunity. In the particular case of DNA vaccines administered to neonates, the rule is immunity rather than tolerance. Exceptions to the rule give opportunities to further understand the neonatal responsiveness and the mechanism of DNA vaccination. Due to the very nature of the vaccine vector, inhibition of neonatal DNA vaccination by maternal antibodies may be limited to the humoral immunity.

Nature versus nurture: contributions of developmental programming and the microenvironment to B cell tolerance

The original Burnet Lederberg and Bretscher Cohn models of immunological tolerance are essentially incompatible, one considering tolerance to be the obligatory outcome of antigen recognition by an immature lymphocyte and the other considering it as one of two possible responses to antigen, the crucial determinant being interaction with a second antigen-reactive cell. The early experimental evidence was confusing, in that it appeared to support both theories. In response to this situation, a hybrid model retaining some of the features of the original models was proposed. In particular, immature B cells were regarded as 'hypersensitive to tolerance induction', but could also make a positive response to antigen under some circumstances. More recent data from B cell transgenic mice have challenged even these hybrid models, stimulating renewed interest in the question of how B cell tolerance is regulated in vivo. This article presents a new interpretation of the data, in which the increased resistance of mature B cells to tolerance induction is postulated to result from partial receptor desensitization in response to environmental antigen, rather than from a developmentally programmed change in B cell signalling. Thus, it is suggested that Burnet's 'window of tolerance induction' is determined by the environment rather than developmental pre-programming. If this postulate is accepted, induction of B cell self-tolerance in both the bone marrow and periphery follows the simple and elegant rules originally laid down by Bretscher and Cohn.

The role of T cells in the regulation of B cell tolerance

The study of conventional models of B cell tolerance has suggested that self-tolerance is imposed on B cells at an early stage in their development due to a peculiar sensitivity of immature B cells to tolerance induction. While this concept accounts for some aspects of central B cell tolerance, it is inconsistent with recent reports of tolerance induction in mature splenic B cells from immunoglobulin transgenic mice. We present an alternative model, the hierarchical model (Aust. N. Z. J. Med. 25, 761-767, 1995), in which regulation of naive B cell reactivity is a function of antigen signal strength and availability of T cell help, but is independent of B cell maturation stage. In turn, the development of tolerance or memory in the T cell compartment is dependent on a combination of antigen-MHC recognition by T cells and antigen-nonspecific signalling by antigen-presenting cells. Using a transgenic model of T-B collaboration, we have shown that both immature and mature self-reactive B cells can be rescued and induced to secrete auto-antibody if the B cell determinant is linked to a carrier protein bearing a foreign T cell determinant.

Cross-linking of surface IgM, but not surface IgD receptors, by soluble monoclonal antibodies primes murine B cells to secrete immunoglobulin in response to lymphokines

We have previously reported the development of a two-step culture system in which soluble anti-mu monoclonal antibodies prime small resting murine B cells to secrete immunoglobulin (Ig) in response to restimulation with a mixture of interleukin-4 (IL-4) and IL-5. Here we have extended these studies to investigate the effects of engaging surface IgD (sIgD). We find that, unlike anti-mu, three different anti-delta monoclonal antibodies did not prime B cells to secrete Ig. In addition, these anti-delta antibodies inhibited anti-mu-stimulated priming for Ig secretion, while enhancing DNA synthesis in response to anti-mu. Furthermore, anti-delta antibodies still inhibited anti-mu-induced priming when added 24-48 h after anti-mu. These results therefore suggest that triggering of sIgD on B cells induces a dominant inhibitory signal which is not necessarily dependent upon co-ligation of sIgM and sIgD receptors. In addition, these findings raise the possibility that ligating sIgM or sIgD receptors on mature B cells in the absence of T cell help, may produce different downstream effects.

A role for immunoglobulin D: interference with tolerance induction

We have studied the induction and maintenance of tolerance in the B lymphocyte compartment. Neonatal and adult transgenic mice which expressed either surface IgM (sIgM) or sIgM and sIgD anti-2,4,6-trinitrophenyl (TNP) were treated with soluble mono- and multivalent forms of TNP-modified carriers. We compared the B cell compartment of mice treated with antigen and of littermates injected with phosphate-buffered saline. Antigen-mediated cross-linking of membrane-bound IgM (sIgM) caused deletion of B cells both in neonatal and adult mice with mu and kappa transgenes. Deletion was the result of apoptosis. In mice that carried an additional delta transgene sIgD interfered with tolerance induction. The stage in which the cells were sensitive to deletion was characterized as a transitional stage between immature (sIgMdull, heat-stable antigenbright, B220dull, sIgD-) and more mature (IgMbright, heat-stable antigendull, B220bright, sIgD-) B cells. Surviving cells were functional as measured by receptor-mediated changes in the intracellular free Ca2+ concentration. We propose that when the immature B cells have reached the final stages of maturation IgM always transmits negative signals in the absence of T cell help. When B cells need to be screened against self reactivity IgM is the only antigen receptor expressed. The presence of sIgD protects resting B cells from deletion and allows them to initiate an effective immune response.

Transient down-regulation of PKC-zeta RNA following crosslinking of membrane IgM on WEHI-231 B lymphoma cells

Regulation of protein kinase C (PKC) isoform mRNAs has been studied in the immature, murine B lymphoma WEHI-231 by the MAPPing protocol and by slot blot analysis of unamplified mRNA. This membrane IgM (mIgM)-positive cell line has been previously used as a model to study signal transduction by mIgM in immature B lymphocytes and the role of those signals in the induction of immune tolerance in the B cell compartment. Stimulation of the cells by anti-mu antibodies, phorbol ester, or Ca2+ ionophore caused growth arrest and death of the cells. IL 4 and IL 5 slowed the growth of the cells. Of these stimuli, only anti-mu stimulation affected PKC mRNA levels. Anti-mu treatment caused a transient decrease in the amount of PKC-zeta isoform mRNA within 3 hr. Within 24 hr levels returned toward normal. Anti-mu had little or no effect on the expression of mRNA for the alpha, beta, delta, or epsilon isoforms of PKC. WEHI-231 cells do not express PKC-gamma. Although anti-mu treatment blocked progression of the cells from the G0/G1 stage into the S phase of cell cycle, viable sort selected cells in either the G0/G1 or the S/G2/M phases showed no clear difference in the expression of PKC-zeta message. Thus, there is not preferential regulation of expression of PKC-zeta during stages of the cell cycle. The results show that mIgM on WEHI-231 cells can transduce a signal that is not mediated by PKC or Ca2+ mobilization alone. The signal causes transient, selective down-regulation of mRNA encoding the zeta PKC isoform.

Differential regulation of in vitro humoral and cellular immune responsiveness in Brugia pahangi-infected jirds

Patent filarial infection is associated with the downregulation of parasite-specific immune reactivity. In the present study, the relationship between in vitro parasite antigen-specific humoral and cellular immune responsiveness was investigated in Brugia pahangi-infected jirds and in jirds immunized with soluble antigens. Spleen cells from B. pahangi-immunized jirds or from jirds with prepatent infections mounted significant in vitro proliferative and antibody responses to B. pahangi. The antigen concentration which elicited optimal antibody production was 10- to 10,000-fold lower than that required to stimulate optimal blastogenesis. Lymph node cells from both immunized and infected jirds consistently produced lower levels of parasite-specific antibody than spleen cells, yet generated higher proliferative responses to filarial antigen. A dissociation between in vitro antibody production and proliferation was also observed in experiments with spleen cells from microfilaremic jirds; spleen cells from patent animals did not proliferate when stimulated with B. pahangi antigen, but did produce significant levels of parasite-specific antibody. Depletion of adherent or histamine receptor-bearing cells restored the proliferative reactivity of spleen cells from microfilaremic jirds, but had limited effects on antibody production. In admixture experiments, spleen cells from microfilaremic animals suppressed the proliferative responsiveness of cells from keyhole limpet hemocyanin (KLH)-immunized jirds to KLH by 38%, but had no effect on KLH-specific antibody production. The present results support the hypothesis that parasite-specific cellular and humoral reactivity are differentially regulated in experimental filariasis.

Altered immunoglobulin expression and functional silencing of self-reactive B lymphocytes in transgenic mice

Immunological tolerance has been demonstrated in double-transgenic mice expressing the genes for a neo-self antigen, hen egg lysozyme, and a high affinity anti-lysozyme antibody. The majority of anti-lysozyme B-cells did not undergo clonal deletion, but were no longer able to secrete anti-lysozyme antibody and displayed markedly reduced levels of surface IgM while continuing to express high levels of surface IgD. These findings indicate that self tolerance may result from mechanisms other than clonal deletion, and are consistent with the hypothesis that IgD may have a unique role in B-cell tolerance.

Lymphoma models for B cell activation and tolerance. IV. Growth inhibition by anti-Ig of CH31 and CH33 B lymphoma cells

CH31 and CH33 are B cell lymphomas whose growth in vitro is inhibited by anti-Ig reagents, including both polyclonal and monoclonal anti-mu antibodies, and an anti-idiotype antiserum. Antibodies against class I or class II major histocompatibility complex antigens do not affect the growth of these cells. Inhibition is dependent on surface Ig cross-linking and does not require ligand binding to Fc receptors. Interestingly, the inhibition of growth by anti-mu is reversed in CH31 (but not CH33) by E. coli lipopolysaccharide. These lymphomas should provide excellent models to study the mechanisms of growth inhibition mediated by surface Ig cross-linking and the pathways of its reversal.

Kinetics of specific IgA, IgD, IgE, IgG, and IgM antibody responses in rubella

Rubella-specific IgD and IgE antibodies were determined with a solid-phase enzyme immunoassay using enzyme-labeled heavy-chain specific anti-immuno-globulins, and the antibody responses in rubella infection were compared to IgM, IgA, and IgG antibodies. IgD and IgE antibodies increased rapidly after the onset of infection, remained at a high level for at least 2 months, and declined slightly by 6 months. In comparison, the IgM antibodies decreased more rapidly, whereas the IgG antibodies persisted longer at a steady level. By 6 months the mean levels of the different antibodies had declined from their maximal mean levels as follows: IgM, 52%; IgA, 42%; IgE, 35%; IgD, 29%; and IgG, 8%. Thus IgD and IgE antibodies, in spite of their known short half lives, persisted longer than IgM and IgA antibodies, which limits their diagnostic value. The IgA antibody responses were found too variable to substitute for IgM antibody determination in diagnosis of a recent rubella virus infection from a single serum specimen. Comparison of maternal and cord blood sera indicated that, in addition to IgG antibodies, rubella-specific IgD antibodies were found to cross the placenta.

Macrophage-derived soluble factors mediate suppression induced by 2,4-dinitrophenyl-conjugated mouse IgG in hybridoma cells

Our laboratory has established that 2,4-dinitrophenyl-conjugated mouse IgG (DNP-MGG) can specifically suppress proliferation, antibody synthesis, and secretion in vivo in two anti-DNP secreting cell lines: hybridoma 35-12 and myeloma MOPC-315. In the present study an in vitro system was used to further analyze the mechanism of suppression of hybridoma 35-12 cells (HC) by DNP-MGG. It was found that DNP-MGG-induced suppression of HC requires macrophages (M phi) and occurs only in eclipsed HC which are mainly small, nonsecreting cells. The M phi-mediated suppression is DNP specific, requires no M phi-HC cell contact, and does not involve killing of eclipsed HC. M phi culture supernatant alone cannot mediate suppression, but supernatants obtained by culturing M phi with either HC or supernatant from HC culture can mediate suppression of eclipsed HC in the presence of DNP-MGG. DNP-MGG is not required for the generation of effective M phi factors, but it is required for suppression of HC in the presence of M phi factors. Indomethacin cannot reverse M phi-mediated suppression, suggesting prostaglandins may not be the M phi factors. These data suggest that M phi-derived factors which are not prostaglandins in nature may play a role in B-cell regulation and in B-cell suppression induced by tolerogenic forms of antigen.

Primed lymphoid cell tolerance. II. In vivo tolerization of highly tolerogen-sensitive hapten-primed, potentially IgG-producing B cells

The relative ease of tolerizing IgM-bearing versus IgG-bearing B cells was investigated. Previous work had shown that IgG-bearing trinitrophenyl (TNP)-specific B cells from mice primed and boosted with TNP-keyhole limpet hemocyanin (TNP-KLH) are highly susceptible to tolerization in vitro by TNP presented on an unrelated carrier. TNP-OVA was used as tolerogen, as it may represent a more general class of tolerogens than those which are nonmetabolizable or immunoglobulin containing. This study showed that highly primed B cells are tolerizable in vivo using TNP-OVA, with the IgG response to TNP-KLH easier to tolerize than the IgM response. To determine if the ease of tolerization of the IgG response in vivo was due to intrinsic differences in B-cell precursors of the IgM and IgG responses, tolerance was performed in vitro with B cells of defined surface isotypes. A T-independent antigen, TNP-endotoxin, was employed to minimize T-cell effects. At least 10 times as much TNP-OVA was required to tolerize B cells bearing the IgM surface isotype than those with the IgG surface isotype. Thus, the ease of inhibition of the IgG response as compared to the IgM response in vivo by preexposure to TNP-OVA may be at least partially explained by inherent differences in IgM and IgG B-cell precursors.

Sensitization of T and B cells by a normally tolerogenic macromolecule: the induction of unresponsiveness and of sensitization is polymorphic

Changes induced in B and T cells by tolerogen were experimentally defined in animals which resist down-regulation. Female A/J, C57BL/6J, MLR/MpJ-lpr/lpr, MLR/MpJ-+/+ and NZB/B1NJ mice were injected at various ages with a tolerogenic form of rabbit gamma globulin (sRGG) or were left uninjected, and all were then immunized with dinitrophenylated RGG (DNP-RGG) on alum. The degree of tolerance was estimated by measuring anti-DNP and anti-RGG spleen plaque-forming cell (PFC) numbers. In some cases, the state of T or B cells deduced from these experiments was further examined by cell transfer experiments. Four types of responsiveness to the tolerogenic form of RGG (sRGG) were distinguished: 1) persistent tolerance inducibility of T and B cells to tolerance induction (A/J); 2) T cells retaining tolerance-inducibility after initiation of age-dependent sensitization of B cells by the tolerogenic form of RGG (C57BL/6J, NZB/B1NJ); 3) age-dependent resistance of T cells to tolerance-induction and age-dependent sensitization of B cells by tolerogen, with the sensitization only manifesting itself in reactivity with T cells from immunized donors (SJL/J); and 4) sensitization of T cells by a usually tolerogenic form of RGG (MLR/MpJ-lpr-lpr). Thus, the development of resistance against tolerance induction is highly polymorphic, not only with respect to the age of onset, but also with respect to the cellular site of its first manifestation and the effect. Possible mechanisms of B-cell sensitization in sRGG-treated C57BL/6 and NZB mice are discussed in relation to the age-dependent emergence of a B-cell subpopulation and of non-specific helper T-cell factors.

Association of surface IgM with two membrane proteins on murine B lymphocytes detected by chemical crosslinking

Surface proteins of intact murine B lymphocytes were crosslinked by the bifunctional reagent, 4,4'-diphenyldiazoniumdisulphidfluoroborate and then radiolabelled with 125I. After solubilization of the cells, Ig-containing complexes (Ig and protein(s) covalently crosslinked to Ig) were isolated and analyzed by two-dimensional SDS polyacrylamide gel electrophoresis (2D-SDS-PAGE). Ig-containing complexes were separated in the first dimension by cylindrical SDS gels. Subsequently, the disulphide bridges of the isolated surface Ig molecules and of the crosslinking reagent were cleaved, and the products electrophoresed in the second dimension on SDS slab gels. In addition to the Ig polypeptide chains, two proteins with mol. wts of 46,000 and 56,000 could be identified. In order to answer the question whether these proteins are associated with IgM and/or IgD Ig-complexes were separated with regard to their isotype and analyzed separately by 2D-SDS-PAGE. It was found that both proteins are associated with subunits of IgM. In the case of IgD, no associated structures could be demonstrated by the method used.

Complete amino acid sequence of the delta heavy chain of human immunoglobulin D

We have determined the amino acid sequence of the variable (V) region of the delta heavy (H) chain of human IgD isolated from the plasma of myeloma patient WAH. This V region is unusual in its amino end group (arginine) and in its length (129 residues). The length is due to 10 insertions in the third complementarity-determining region (CDR3). A computer search showed that no reported CDR3-joining region (-JH) sequences are identical and that they appear to be unrelated to the constant (C) region sequences of immunoglobulins. The V region sequence together with our previous results for the C region give the complete sequence of the human delta chain WAH, which has 512 amino acid residues and a Mr congruent to 65,000. The human delta chain has four domains (V, C delta 1, C delta 2, and C delta 3) and a long hinge region; by comparison, the mouse delta chain lacks a continuous segment of 135 residues, including half the hinge region and the entire C delta 2 domain. The human and mouse delta chains also differ in the number, kind, and location of GlcN and GalN glycans and probably in conformation and quaternary structure. These and other considerations suggest that there may be multiple forms of both secreted and membrane-bound IgD that differ in size, structure, and function.

Ontogeny of acquired immunological tolerance to H-2 alloantigens

The ontogeny of mixed lymphocyte reaction and cell-mediated lymphocytotoxicity (CML) responsiveness was studied in mice that had been neonatally tolerized by the injection of 15 X 10(6) semiallogeneic bone marrow and spleen cells, and compared to that of unprimed littermates. Spleen cells from primed mice did not develop reactivity towards the tolerizing antigen but did develop reactivity towards third-party alloantigens. Similarily, thymus cells from primed mice also failed to develop CML reactivity towards the tolerizing antigen. Thymus cell mixed lymphocyte reactivity, which was demonstrable in neonates before the tolerizing injection, declined rapidly between 2 and 4 days of age, compared to that of controls. Thus neonatal tolerance was shown to operate at least at a thymic, and perhaps pre-thymic level; measurement of the levels of chimeric cells in the thymus and bone marrow favored the latter interpretation.

Amino acid sequence of the first constant region domain and the hinge region of the delta heavy chain of human IgD

We have determined the amino acid sequence of the first constant (C) region domain (C delta 1) and the hinge region of the delta heavy chain of human IgD WAH and also the sequence of the adjacent COOH-terminal portion of the variable (V) region, including the JH region. Together with the sequence of the Fc fragment already reported, this establishes the complete amino acid sequence of the C region of the human delta chain and confirms the presence of three C region domains in human IgD. Although the CH1 domains of the five classes of human heavy chains have the expected degree of homology (approximately 30%), the homology of the C delta 1 domains of the human and mouse chains is less than that exhibited by the CH1 domains of other pairs of human and mouse heavy chains. The hinge region of the human delta chain has an unusual structure; the NH2-terminal half has four (or five) GalN oligosaccharides attached, whereas the COOH-terminal half lacks carbohydrate, is dissimilar in sequence, and has a high charge. A computer search verified that the GalN-rich segment has a high degree of identity in sequence with the middle portion of the human C mu 2 domain and that the high-charge segment is related to the same sequence. We propose that the two segments of the human delta hinge have a common evolutionary origin and arose by duplication and independent mutation of a hinge exon derived from the ancestral gene for the C mu 2 domain.

In vitro tolerance induction of primed, IgD-negative murine spleen cells

The above observations demonstrated induction of immunological tolerance in vitro in primed IgD-, IgG+ B cells. In these studies, addition of trinitrophenylated (TNP) turkey gammaglobulin (TGG) or TNP ovalbumin conjugates suppressed the secondary in vitro response in mice primed with TNP keyhole limpet hemocyanin (TNP-KLH). Suppression was not a reflection of a shift in kinetics of the antibody response, was not dependent on suppressor T cells, and could only be eliciate when conjugate was added within 4 h of addition of TNP-KLH moreover, preincubation of the primed spleen cells with TNP-TGG for 20 h at 37 degrees C, followed by extensive washing, was as effective in inhibiting the response to TNP-KLH as when TNP-TGG was present throughout the 5 d of culture, reflecting induction of a tolerant state. Amounts of conjugate in the concentration range that have been shown by others to tolerize immature or neonatal B cells or mature B cells that have been stripped of surface IgD were sufficient to induce tolerance. The target cells being tolerized did not bear IgD, as determined by B cell depletion and blocking procedures with anti IgD. Whether the lack of surface IgD on the primed cells contributed to the relative ease of tolerance induction was not established by these studies, but the advantages of using primed B cells to examine further the role of surface IgD in tolerance susceptibility was discussed.

Primary structure of the Fc region of human immunoglobulin D: implications for evolutionary origin and biological function

We have determined the complete amino acid sequence of a tryptic Fc delta fragment generated from an intact human IgD (WAH); it is 226 residues long and includes the second (C delta 2) and the third (C delta 3) constant domains of the delta chain. Comparison of the homology of the Fc sequence of the five human immunoglobulin classes suggests that either the delta-chain gene evolved from the alpha-chain gene soon after the divergence of a mu-alpha common ancestor or it evolved from an ancestral gene distinct from both the mu-alpha and the gamma-epsilon common ancestors. Comparative study using a spatial model of the Fc region indicates that the structure of the C delta 3 domain differs extensively from that of the carboxy-terminal domains of other heavy chain classes; this, together with the unique hinge region structure, probably reflects the biological role of IgD as a receptor molecule on the B-lymphocyte surface.

Regulation of murine B cells through surface immunoglobulin. I. Monoclonal anti-delta antibody that induces allotype-specific proliferation

We describe the identification of a monoclonal antibody that recognizes a determinant on the delta chain of mice of the Iga, allotype groups. The monoclonal Ig in soluble form induces allotype-specific proliferation by splenic B lymphocytes from normal animals of these haplotypes. Spleen cells from mice bearing the X-linked defect of CBA/N mice fail to respond, although they bear the determinant. Proliferation is independent of T lymphocytes. The data indicate a direct triggering function for sIgD.

Activation of murine B lymphocytes by anti-immunoglobulin is an inductive signal leading to immunoglobulin secretion

Cultures of isolated mouse splenic B lymphocytes activated by the divalent F(ab')2 fragment of purified rabbit anti-mouse Fab or class-specific anti-mouse IgM antibodies can be driven on to high rate Ig secretion by the addition of the supernatant fluid of a 24-h culture of concanavalin A-activated spleen cells (SN). The polyclonal antibody response to anti-Ig pus SN is comparable in magnitude with the lipopolysaccharide response as measured in a reverse plaque assay. The addition of SN can be delayed for 24 h after addition of anti-Ig without changing the kinetics of the response. Addition at 48 h delays the response by 24 h. The response to F(ab')2 anti-Fab plus SN is sensitive to Fc-dependent inhibition because intact anti-Fab antibodies inhibit strongly at relatively low concentrations. The monovalent Fab' fragment fails to induce Ig secretion, indicating that cross-linkage of surface immunoglobulin is required. Although the production of active SN is T cell dependent, the response to anti-Ig plus SN is T independent. These findings are interpreted as a polyclonal model of a thymus-dependent antibody response. X

Memory B cells at successive stages of differentiation. Affinity maturation and the role of IgD receptors

The following evidence, mainly presented here, suggests that IgD receptors play a crucial role in determining the potential for affinity maturation in memory B cell populations. IgD receptors are present on the first memory B cells to appear after priming. These memory cells give rise to more-mature memory cells that have lost their IgD receptors. The proportions of early (IgD(+)) and mature (IgD(-)) memory cells found in individual donors vary with time, priming conditions, and the availability of T cell help, and both populations frequently coexist for long periods of time. IgD(+) and IgD(-) memory cells carry IgG receptors and give rise to IgG responses with identical isotype representation in adoptive recipients. IgD(+) memory cells, however, always give rise to predominantly low-affinity antibody responses, whereas IgD(-) memory cells consistently generate responses of substantially higher average affinity. This affinity differential is maintained between early and mature memory populations in the same donor and does not appear to be a result of selective differentiation of higher-affinity IgD(+) memory cells into the IgD(-) memory pool. Thus, the selective forces responsible for affinity maturation appear to operate mainly in mature memory cell populations that have already lost IgD receptors; or, stated conversely, little or no selection towards high-affinity memory appears to occur among memory cells that retain IgD receptors. In discussing these findings, we suggest that the IgD receptors themselves are responsible for maintaining early memory populations at a lower average affinity than IgD(-) populations in the same animal. The IgD receptors, we argue, serve to increase the antigen-binding capacity of lower-affinity memory cells so that these cells can survive, expand, and differentiate (to IgD(-)) at antigen concentrations that select against expansion of low- affinity memory cells no longer carrying IgD receptors. Thus, when antigen is limiting, IgD(-) memory populations will be selectively expanded to higher average affinities, whereas coexisting IgD(+) populations will retain their initial affinity profile. This hypothesis suggests that mechanisms that regulate expression and loss of IgD receptors are central to the adaptability of the immune system in its response to invading pathogens. Two related roles can be envisioned for the IgD receptors in this regard. First, they extend the lower boundary of the affinity range of early memory cell populations induced by a given antigenic stimulus and therefore broaden the diversity of responses obtainable from these populations. Secondly, they support the persistence of low-affinity memory populations under conditions where antigen becomes limiting and eventually disappears. These persisting populations then serve as a diversely reactive reservoir from which mature memory populations can be drawn with higher affinities either for the original antigen or, more importantly, for related antigens that the animal may subsequently encounter. Thus the existence of IgD receptors on early memory cells maintains the full range of response diversity despite ongoing selective expansion of (mature) memory populations to produce antibodies with high combining affinities for individual antigens. The flexibility inherent in such an organizational system, we believe, could be expected to account for the evolutionary development of IgD receptors and the regulatory capabilities that support operation of the system.