The kinetics of hemolytic plaque formation. II. Inhibition of plaques by hapten.

We present a mathematical theory of hapten inhibition of hemolytic plaque formation. The treatment is based upon the mathematical model for plaque growth presented by DeLisi &; Bell (1974). The lymphocyte under consideration is embedded in an infinite three-dimensional medium, and is secreting antibodies isotropically at a constant rate. As the antibodies diffuse from the source they can bind reversibly to hapten, and in the most general case reversibly to red blood cell (RBC) epitope. The model leads to a non-linear diffusion equation coupled to a set of first order differential equations. The system must, in general, be solved numerically. However, in many cases of experimental interest simplifications arise which permit closed form solutions to be obtained. In this paper we have developed solutions for three special cases.In the first example antibodies can bind only univalently to RBCs, as would be expected if the epitopes are sparsely distributed. In this case reaction between antibody site and RBC epitope is rapid ( ⪆ 1 sec) and reversible and local equilibrium is assumed. This leads to a “pure” diffusion equation in the free antibody concentration, but with a reduced diffusion coefficient.In another example univalent attachment of an antibody site to a RBC epitope is followed by a rapid irreversible intramolecular reaction. This might be expected for example if the epitope density is large. An exact solution to the resulting diffusion equation was also found in this case. In order to assess an intermediate situation, we also solved the equations for a model in which intramolecular reaction is slow and irreversible.The theory predicts that the type of information one can obtain from inhibition experiments depends critically upon the preparation of the RBC. If the cell is sufficiently haptenated so that rapid irreversible multivalent attachment is favorable, a differential plot of the inhibition curve will reflect the affinity distribution of antibody sites for free hapten. If only univalent attachment with RBCs is possible, so that antibody sites bind to RBC hapten in the same way they bind to free hapten, then a differential plot of the inhibition curve will reflect the secretion rate distribution.     

RNA conversion of lymphoid cells to synthesize allogeneic immunoglobulins in vivo

Ribonucleic acid extracts (“5 day immune” and “nonimmune”-RNA) obtained from lymph nodes and spleens of rabbits homozygous for the b⁴ or b⁵ allele of light chain immunoglobulin allotypes were injected iv into nonimmunized rabbits homozygous for the alternate allele. The recipient rabbits were then given multiple iv injections of sheep red blood cells (SRBC). The spleens were assayed 13, 21, and 37 days following the RNA injection for “direct” IgM and “indirect” IgG plaque forming cells (PFC) specific for SRBC. The b4 or b5 light chain allotype and the a1, a2, and a3 heavy chain allotype of the antibody in the plaques was identified by radioautography and by inhibition of plaque formation using anti-allotype antibodies. The b light chain allotype of the RNA donor was identified in 22–32% of the IgM plaques and in 25–42% of the IgG plaques. The allotype of the host rabbit b light chain allotype was identified in 56–67% of the IgM plaques and in 57–71% of the IgG plaques. Likewise the a heavy chain allotype of the RNA donor was identified in 10–19% of the IgM plaques and in 12–19% of the IgG plaques. The allotype of the host rabbit a heavy chain allotype was identified in 51–60% of the IgM plaques and in 55–63% of the IgG plaques. The concentrated lysates of spleen and lymph node cells were also analyzed for immunoglobulins of each light chain allotype by immunodiffusion with radiolabeled antibody. The allotype of both the RNA donor rabbit and host rabbit were found in most of the lysates of lymphoid tissues and in some of the IgG isolated from the serum and concentrated.     

The kinetics of hemolytic plaque formation. III. Inhibition of plaques by antigen.

Equations are presented which can be used to describe the inhibition of plaques by multifunctional antigen which binds γG antibody bivalently. The interaction is treated as a bimolecular reaction which is irreversible within the time of the experiment. It is shown that under these conditions the characteristics of the inhibition curve, and their relationship to kinetic and thermodynamic parameters are strongly dependent upon how antibody interacts with RBCs. When the epitope coating is dense, multivalent attachment of antibody is likely and the interaction is considered irreversible. When the epitope coating is sparse, only rapid, reversible univalent attachment is considered and local equilibrium is assumed to hold.The first case leads to an abrupt inhibition curve whose position is determined by the forward rate constant and RBC density. The second case leads to broad asymmetric curves. For this situation the relation between the extent of inhibition and the affinity of the population is generally complicated and reliable affinity information is difficult to obtain. This is contrasted to results obtained previously for unifunctional inhibitors from which reliable affinity information can, in principle, be obtained. The results emphasize the need for carefully designed experiments if affinity information is to be obtained from inhibition studies.     

Comparative analysis of amyloid-beta chemical structure and amyloid plaque morphology of transgenic mouse and Alzheimer's disease brains

We have undertaken an integrated chemical and morphological comparison of the amyloid-beta (Abeta) molecules and the amyloid plaques present in the brains of APP23 transgenic (tg) mice and human Alzheimer's disease (AD) patients. Despite an apparent overall structural resemblance to AD pathology, our detailed chemical analyses revealed that although the amyloid plaques characteristic of AD contain cores that are highly resistant to chemical and physical disruption, the tg mice produced amyloid cores that were completely soluble in buffers containing SDS. Abeta chemical alterations account for the extreme stability of AD plaque core amyloid. The corresponding lack of post-translational modifications such as N-terminal degradation, isomerization, racemization, pyroglutamyl formation, oxidation, and covalently linked dimers in tg mouse Abeta provides an explanation for the differences in solubility between human AD and the APP23 tg mouse plaques. We hypothesize either that insufficient time is available for Abeta structural modifications or that the complex species-specific environment of the human disease is not precisely replicated in the tg mice. The appraisal of therapeutic agents or protocols in these animal models must be judged in the context of the lack of complete equivalence between the transgenic mouse plaques and the human AD lesions.     

Defective primary and secondary IgG responses to thymic-dependent antigens in autoimmune PN mice

Palmerston North (PN) mice spontaneously develop autoimmune disease resembling SLE. Because immune responsiveness has not been defined in this strain, a study was designed to assay primary splenic plaque-forming cell (PFC) responses to thymus-dependent (TD) and thymus-independent (TI) Ag. Initial surveys of PN mice inoculated with the TD Ag SRBC showed adequate production of IgM PFC, but small numbers of IgG PFC were developed with polyspecific antiserum. In contrast, H-2-compatible DBA/1 control mice gave the expected responses to SRBC (IgG plaques elevated twofold compared with IgM plaques). PN mice had the usual responses to Ag that are largely TI; both PN and DBA/1 mice had active IgM and modest IgG responses to TNP-LPS and TNP-Ficoll. Additional experiments determined that PN mice had similar patterns of defective IgG responses to several different TD Ag (SRBC, horse RBC, and DNP-keyhole limpet hemocyanin). In each instance, the usual predominance of IgG1 plaques was absent, and total numbers of plaques developed with antisera specific for IgG isotypes were suppressed. Defective PN IgG production was evident as early as 3 wk of age, was not influenced by aging to 43 wk, and was not corrected by increasing the antigenic challenge 10-fold. PN spleen cells treated with monoclonal anti-Thy-1.2 and C were injected with pools of DBA/1 T cells into 850-rad irradiated (DBA/1 x PN)F1 hybrids. These recipients expressed low IgG1 responses to SRBC, suggesting that the B cell-containing fraction that was not lysed by anti-Thy-1.2 transferred the PN defect. PN mice, which do not respond to TD Ag with active IgG production, contradict the proposal that autoimmunity is associated with hyper-responsiveness to TD and TI Ag.     

Activation to IgG secretion by lipopolysaccharide requires several proliferation cycles.

We have investigated the relationship between IgG secretion and cell proliferation after polyclonal activation of murine spleen cells with lipopolysaccharide (LPS). It was found that IgG secretion was optimal when cells proliferated extensively. Under those conditions, DNA synthesis commenced 8 to 12 hr after exposure to LPS. Increased proliferative activity was observed up to day 3, when the majority of the lymphoblasts were mitotically active. Two inhibitors of DNA synthesis, thymidine (TdR) and hydroxyurea (HU) caused a reduction in both the IgM and the IgG response, but the latter was more severely reduced. The inhibition was strongest when TdR and HU were added to cultures early after exposure to LPS, indicating that the cells developing to Ig secretion were continuously proliferating. 5-bromo-2'-deoxyuridine (BrdU) caused a general inhibition of IgM and IgG secretion at high concentrations, and a selective inhibition at low concentrations. The selective inhibition of IgG secretion, when measured on day 4, was also observed after a pulse of BrdU on days 1 and 2. The data suggest that development to IgG secretion is a complex process, which requires several proliferation cycles.     

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.     

IgG autoantibody activity in normal mouse serum is controlled by IgM

In the serum of normal BALB/c mice, IgG antibody reactivity to mouse actin and tubulin, DNA, and TNP groups was very low compared to that of the IgM. This activity was considerably increased when IgG was separated, by affinity chromatography on protein A-Sepharose, whereas no difference in the IgM activity was observed. Addition of IgM to IgG isolated from the same serum resulted in the inhibition of IgG binding to these Ag. Isolation of IgG antibodies on actin, TNP, and tubulin immunoadsorbents has indicated that at least part of the IgG antibodies is polyreactive. In order to understand this inhibition better, experiments with F(ab')2 fragments of IgG were performed. IgM inhibited the binding of F(ab')2 to the antigens in a dose-dependent manner and reacted with immobilized F(ab')2. IgM isolated on F(ab')2 immunoadsorbent, as compared to the initial IgM preparation, were less active toward the Ag but more inhibitory for IgG binding to the Ag. In some pathologic situations, IgM failed to inhibit some IgG antibody activities. The anti-DNA IgG activity from (NZB x NZW)F1 mice was not affected by autologous IgM. Similarly the anti-tubulin IgG from mice infected with Trypanosoma cruzi were less inhibited by IgM from autologous serum than antitubulin IgG from normal mice. These results are compatible with the existence in normal mice of an idiotypic-like network, regulating via an IgM population in the serum, the binding of IgG autoantibodies to self Ag. Modifications of this idiotype-anti-idiotype system might lead to the expression and/or expansion of autoreactive IgG-producing clones.     

Effects of hydrophobic and hydrophilic bile salt mixtures on cholesterol crystallization in model biles

[Ru(2,2'-bipyridine)(2)(4,4'-dicarboxy-2,2'-bipyridine)](2+) (RuBDc) is a very photostable probe that possesses favorable photophysical properties including long lifetime, high quantum yield, large Stokes' shift, and highly polarized emission. In the present study, we demonstrated the usefulness of this probe for monitoring the rotational diffusion of high-molecular-weight (MW) proteins. Using frequency-domain fluorometry with a high-intensity, blue light-emitting diode (LED) as the modulated light source, we compared the intensity and anisotropy decays of RuBDc conjugated to immunoglobulin G (IgG) and immunoglobulin M (IgM), which show a six-fold difference in MW We obtained slightly longer lifetimes for IgM (=428 ns in buffer) than IgG (=422 ns in buffer) in the absence and presence of glycerol, suggesting somewhat more efficient shielding of RuBDc from water in IgM than in IgG. The anisotropy decay data showed longer rotational correlation times for IgM (1623 and 65.7 ns in buffer) as compared to IgG (264 and 42.5 ns in buffer). Importantly, the ratio of the long rotational correlation times of IgM to IgG in buffer was 6.2, which is very close to that of MW of IgM to IgG (6.0). The shorter correlation times are most likely to be associated with domain motions within the proteins. The anisotropy decays reflect both the molecular size and shape of the immunoglobulins, as well as the viscosity. These results show that RuBDc can have numerous applications in studies of high-MW protein hydrodynamics and in fluorescence polarization immunoassays (FPI) of high-MW analytes.     

Antigen-initiated B lymphocyte differentiation. IX. Characterization of memory AFC progenitors by buoyant density and sedimentation velocity separation.

The characteristics of memory B cell antibody-forming cell (AFC) progenitors from long-term hapten-primed CBA mice were investigated by using sedimentation velocity and buoyant density separation to isolate physically distinct B cell sub-sets. The isolated fractions were assayed by the adoptive immune response to NIP-POL antigen, under conditions where neither T cells nor other accessory cells were limiting the IgM or IgG AFC responses. The results were compared to previous studies on the IgM AFC-progenitors of unprimed adult mice. Splenic IgM and IgG memory AFC-progenitor activity was largely found among the typical B cells of slow to medium sedimentation rate, in contrast to the fastre sedimenting IgM AFC-progenitor activity of unprimed animals. Splenic IgM and IgG memory AFC-progenitor activity was found among the medium to light density cells, and so resembled by this parameter the IgM AFC-progenitor activity in unprimed animals. Thoracic duct lymphocytes from hapten-primed mice also exhibited memory IgM and IgG AFC-progenitor activity in the slow-medium sedimentation range. However, in contrast to spleen, the IgM and IgG memory AFC-progenitor activity in lymph was found among very dense B cells. Two physically distinct sub-populations of memory B cells have thus been identified, namely: i) small, medium-light density, presumably tissue-resident B lymphocytes found in spleen; and ii) small, dense, presumably recirculating B lymphocytes found in lymph. Both physical forms include IgM and IgG progenitors. Both forms are distinct from the larger, medium-light density "virgin" AFC-progenitors in the spleen of unprimed adult mice.     

Inhibition of cell growth by monoclonal anti-transferrin receptor antibodies.

Five anti-murine transferrin receptor monoclonal antibodies have been characterized with respect to immunoglobulin class, effects on binding of transferrin, and effects on AKR1 lymphoma cell growth in vitro. The immunoglobulin M (IgM) antibodies, but not the IgG antibodies, prevent cell growth. We suggest that the profound effects of the IgM antibodies on cell growth are probably due to extensive cross-linking of cell surface receptors. In support of this, we are able to mimic the growth-inhibiting effects of the IgM antibodies by adding antiimmunoglobulin to an IgG antibody. By flow microfluorimetry, we show that an IgG antibody by itself induces up to a 10-fold downward regulation in the cell surface transferrin receptor, which is accompanied by accelerated receptor degradation. A similar downward regulation is seen in mutant cells resistant to growth inhibition by an IgM antibody, when grown in the selecting antibody. Wild-type cells grown in the presence of IgM antibody do not show receptor downward regulation. Inhibitory effects of antibody plus antiimmuoglobulin on mutant cells are also consistent with extensive cross-linking causing inhibition of growth.     

The Fc portion of intact IgG blocks stimulation of human PBMC by anti-T3

The means by which normal human serum inhibits the activation of PBMC by OKT3 has been investigated. The Fc portion of intact IgG is shown to be the major inhibitor in human serum of this OKT3-induced stimulation. Furthermore, inhibition by IgG subclasses correlated with their ability to bind to the monocyte Fc receptor, i.e., IgG1 and IgG3 produced greater inhibition than IgG2 and IgG4, and this inhibition was competitive. In contrast, hypogammaglobulinemic serum, IgA, IgM, and F(ab')2 of IgG were not inhibitory relative to intact IgG or Fc of IgG. Because of the similarities between T3 and the idiotype-defined T cell receptor for antigen, these investigations suggest that IgG might modulate the interactions between the T cell recognition complex and anti-idiotype antibody, thus regulating the idiotype network.     

Vesicular stomatitis virus Indiana glycoprotein as a T-cell-dependent and -independent antigen.

The neutralizing immunoglobulin M (IgM) response to vesicular stomatitis virus (VSV) has been shown to be largely T-cell independent in several T-cell-deficient models of mice. By using different antigen froms of VSV, VSV antigen doses could be graded in vivo (infectious > > UV inactivated > formalin inactivated). The present study reveals a T-cell-dependent component of the neutralizing IgM response in nude mice given intravenous injections of low doses of noninfectious UV-inactivated VSV serotype Indiana (VSV-IND) only if the mice are transfused with VSV-IND-specific helper T cells. Instead, nude mice immunized with infectious VSV, which leads to greater antigen doses in vivo, were able to mount an IgM response in the absence of T cells. These results indicate that the IgM response to low doses of VSV-IND glycoprotein (G) is T-cell dependent. Nude mice immunized with infectious VSV also made a variable but low VSV-IND-neutralizing IgG response. A VSV-IND matrix (M)-specific helper T-cell line rendered this response more consistent, much higher, and longer lasting. Thus (i) VSV-G induces a mostly T-cell-independent but partially T-cell-dependent IgM (the latter can be visualized best at low doses of antigen) and (ii) the antibody response to VSV in nude mice proceeds through steps, i.e., IgM and IgG, that are dose dependent. The results suggest that the predominant role of helper T cells may be to expand and maintain the individual steps of differentiating B cells.     

The role of immunoglobulins in immunity to Trypanosoma brucei gambiense

The effects of IgM and IgG antibody molecules were compared on a weight basis by both agglutination and in vitro protection tests. It was shown that IgM is a better agglutinating antibody and in the presence of complement is also a better neutralizing antibody. However, when IgM and IgG were tested for their ability to passively protect infected animals, it was noted that the IgG fraction appeared to give greater protection. Based on differences in diffusion coefficients (and size) of the molecules, it is hypothesized that IgM is the antibody class responsible for the relapse phenomena observed in the blood, while, it is the IgG antibody molecule which is more active in extravascular locations.     

Nonionic block polymer surfactants enhance the avidity of antibodies in polyclonal antisera against Streptococcus pneumoniae type 3 in normal and Xid mice

Avidities of antibody (sub)classes in polyclonal antisera against Streptococcus pneumoniae type 3 (S3) can be (semi) quantitatively determined with a specific inhibition ELISA. A hexasaccharide was isolated from the hydrolyzed S3 capsular polysaccharide and coupled to a protein-carrier. Mixtures containing these conjugates and nonionic block polymer (NBP) surfactants were used for immunization. After various immunizations of these conjugates without NBP the anti-S3 specific antibodies of IgM and IgG2a isotype decreased in both antibody level and avidity. The adjuvants NBP 1501 and L121 not only enhanced the hexasaccharide-protein induced IgM and IgG antibody levels but also clearly increased the avidity of the two antibody (sub)classes IgM and IgG2a. This effect was observed in normal (data not shown) and X-linked immunodefective mice. A maturation of the IgG antibody response was realized by the second immunization with hexasaccharide-protein conjugate whereas the third immunization showed no further increase in antibody level and avidity.     

Duplicated variable region genes account for double isotype expression in a human leukemic B-cell line that gives rise to single isotype-expressing cells.

The SSK41 cell is an immunoglobulin IgM+ human neoplastic B-cell line that switches to IgG+ cells (SSK gamma) spontaneously. We isolated a derivative of SSK41 (SSKWB) that expressed both IgM and IgG. Studies on the Ig heavy chain gene organization have shown that the SSKWB cell had two identical VDJ genes on the same chromosome; one linked to the C mu gene and the other to the C gamma 1 gene, both of which are transcribed to produce mu- and gamma-mRNAs with the same VDJ sequence. We also isolated the two switch variants derived from the SSKWB cell by cell sorter: 1G (IgG+) and 11M (IgM+). The 1G cells contained two populations; one had a similar genetic organization as SSK gamma and expressed only IgG1, and the other carried the same genetic organization as the SSKWB cell but produced aberrantly spliced mu-chain mRNA, in which the hydrophobic signal sequence exon is directly joined to the C mu exon. Te 11M cell deleted the VDJ-C gamma 1 segment of the SSKWB cell and expressed IgM. These results raise the interesting possibility of another mechanism for class switching.     

Development of monoclonal IgA and an apparent IgG in a patient with macroglobulinemia: sharing of individually specific antigenic determinants among IgM, IgA, and IgG.

Patient CM, who initially was diagnosed as having macroglobulinemia (IgM, kappa) was subsequently found to develop a monoclonal IgA(kappa) protein. Rabbit antisera directed against the patient's IgAm and IgM were rendered specific for individual antigenic (ind) determinants. The anti-IgAm and IgM ind sera reacted with both 131I labeled monoclonal proteins in a double antibody radioimmunoassay (RIA). In addition, both monoclonal immunoglobulins inhibited the reaction between labeled immunoglobulin and both ind antisera, and statistical analysis of the data suggested that the shared ind determinants were identical. The IgG fraction of patient CM's serum also contained a component which competed with both monoclonal IgA (CM) and IgM (CM) in the RIA specific for ind determinants. Analysis of serum samples taken over a 2-year period revealed that, in addition to IgM, both the IgA and IgG components possessing the shared ind determinant(s) were present in low concentrations in the earliest sample, although not detected by conventional techniques. The monoclonal IgA and the IgG component were found to increase in concentration over this time interval with a concomitant decrease in IgM. The regulation of immunoglobulin expression with respect to the proposed models of gene organization in antibody-producing cells was discussed.     

Central role of T lymphocytes in specific recognition of tumor antigens in 51Cr-leukocyte adherence inhibition

The plaque-forming cell (PFC) response of C57BL/6 mice to NP-coupled sheep red blood cells can be inhibited by administration of microgram doses of a variety of monoclonal anti-NP antibodies of various isotypes prior to immunization. NP-specific PFC are most strongly affected but a drastic reduction of the SRC-specific response also occurs. The latter depends on physical association of the immunizing red cells with NP groups. Inhibition was obtained with antibodies of all IgG subclasses, but was marginal or absent when up to 100 μg IgM or IgD antibodies or F(ab′)₂ fragments were administered. Our experiments lead us to conclude that antibody feedback inhibition is not the function of a particular IgG subclass and that the role of the variable region in feedback inhibition is sufficiently explained by its affinity for the antigen. It was noted that small doses of a monoclonal NP-specific IgM antibody failed to enhance the response to NP-SRC.     

Regulation of T15 idiotype dominance. II. Genes unlinked to the Igh locus regulate T15 dominance of the secondary adoptive transfer response to phosphocholine

We have studied the idiotype and fine specificity of the secondary immune response to phosphocholine (PC) in C57BL (B10, B10.D2, and B.C8) and BALB (BALB/c, BAB-14, and C.B20) congenic strains of mice. In vivo IgM responses of mice from these two genetic backgrounds differed in their T15 idiotypic representation. BALB strains expressed the T15 idiotype on greater than 90% of their IgM, PC-specific plaque-forming cells (PFC), whereas C57BL strains expressed the T15 idiotype on approximately 50% of their IgM PFC. All strains examined expressed greater than 75% PC-inhibitable, VHPC idiotype-positive, IgM PFC. The IgG3 and IgA memory responses were similar to the IgM memory response; BALB strains produced a higher proportion of T15+ PFC than C57BL strains; however, the majority of IgG3 and IgA PFC in all strains were VHPC+, and PC-inhibitable. In contrast, the IgG1 memory response was not dominated by T15+, VHPC+, PC-inhibitable PFC in any of the strains tested. The IgG1 PFC required nitrophenylphosphocholine (NPPC) for efficient inhibition. The IgG2 memory response generally mimicked the IgG1 response with respect to idiotype and specificity. These data demonstrate that the representation of the T15 idiotype in the anti-PC immune response is determined by genes outside both the MHC and Igh genetic loci. Control of T15 expression in secondary IgM, IgG3, and IgA anti-PC responses was examined by using a cell-mixing protocol with primed T and B cells from BALB/c and B10.D2 mice. T15 representation in these responses was determined by the genotype of the B cell, not by the genotype of the helper T cell. Similarly, the B cell genotype was responsible for the idiotypic profile of a primary, in vitro, T-dependent, anti-PC response.     

Optimal strategies in immunology III. The IgM-IgG switch

Summary During a primary immune response generally two classes of antibody are produced, immunoglobulin M (IgM) and immunoglobulin G (IgG). It is currently thought that some lymphocytes which initially produce IgM switch to the production of IgG with the same specificity for antigen. During a secondary immune response IgG is the predominant antibody made throughout the response. In this paper we address the question of why such apparently complicated modes of response should have been adapted by evolution.We construct mathematical models of the immune response to growing antigens which incorporate complement dependent cell lysis. By comparing the times required to eliminate antigen we show that under certain conditions it is advantageous for an animal to switch some of its lymphocytes from IgM to IgG production during a primary response, but yet to secrete only IgG during a secondary response. The sensitivity of such a conclusion to parameter variations is studied and the biological basis and implications of our models are fully discussed.Portions of this work were performed under the auspices of the U.S. Department of Energy. A.S.P. was also supported by the National Science Foundation under Grant No. ENG-7904852 and BRSG grant S07 RR05664-11 awarded by the Biomedical Research Support Grant Program, Division of Research Resources, National Institute of Health. A.S.P. is the recepient of an NIH Research Career Development Award 1K04 AI 00357-01. S.R. was a recipient of NIH Fellowship 5 F32 AI05107-02