Activation and negative selection of functionally distinct subsets of antibody-secreting cells by influenza hemagglutinin as a viral and a neo-self antigen

We have compared transgenic mice that express the influenza virus PR8 hemagglutinin (PR8 HA) as a membrane-bound neo-self antigen (HA104 mice) with nontransgenic (non-Tg) mice for their ability to generate HA- specific B cell responses after primary immunization with PR8 virus. HA- specific, IgM-secreting B cells were induced with similar frequencies in HA104 and non-Tg mice. In addition, a B cell clonotype (C4) that is characteristic of anti-HA immune responses of BALB/c mice was identified among HA-specific IgM hybridomas from HA104 mice. A subset of HA-specific, IgG-secreting B cells that arises rapidly after primary virus immunization in non-Tg mice, however, was substantially reduced in HA104 mice. Likewise, a B cell clonotype (C12) that dominates HA- specific IgG hybridomas generated after primary immunization of non-Tg mice was present at greatly reduced frequencies among hybridomas from HA104 mice. Because HA-specific, IgG-secreting B cells were generated by HA104 mice in response to a mutant HA containing an amino acid interchange in a B cell antigenic site, we conclude that these PR8 HA- specific, IgG-secreting B cells are negatively selected in HA104 mice as a result of their specificity for the neo-self PR8 HA. The findings demonstrate that HA-specific B cells that display distinct phenotypic potentials in non-Tg mice also differ in their susceptibility to negative selection from the primary B cell repertoire of HA104 mice: a subset of B cells that undergo rapid differentiation to become HA- specific IgG antibody-secreting cells (ASC) after activation in non-Tg mice is negatively selected in HA104 mice. By contrast, a subset that gives rise to HA-specific, IgM-secreting ASC persists in the primary repertoire of HA104 mice and can be activated by virus immunization.     

The diversity of the influenza-specific primary B-cell repertoire in BALB/c mice

The primary immune response of BALB/c mice to influenza (PR8) hemagglutinin (HA), a complex protein antigen, has been examined by the splenic focus assay, and the resulting monoclonal anti-HA antibodies have been characterized by their reactivity with heterologous viruses. The analysis of the primary B-cell response to HA revealed marked differences from responses previously defined for haptenic determinants. There were following differences: (a) the frequency of HA-specific B cells in both conventional and germ-free BALB/c mice was 1 in 1.0-1.5 X 10(5) splenic B cells, which is substantially lower than the frequency of B cells responsive to various simple haptenic determinants; (b) monoclonal anti-HA antibodies were predominantly of the IgA or IgM isotypes instead of IgG, which dominates antihapten responses; and (c) after immunization, the frequency of anti-HA-specific B cells increases by 10- to 50-fold, which is much greater increase than that observed after immunization with haptenic determinants. Fine specificity analysis of primary monoclonal HA-specific antibodies revealed extensive diversity and a considerable overlap with the specificities obtained from immune mice. Given the low overall frequency of HA-specific B cells, it could be calculated that the representation of most HA-specific clonotypes within the B-cell repertoire could not exceed 1 in 10(7) B cells. These findings indicate that the primary B-cell clonotype repertoire is extremely diverse and largely antigen independent in its generation.     

Functional analysis of influenza-specific helper T cell clones in vivo. T cells specific for internal viral proteins provide cognate help for B cell responses to hemagglutinin

We compared the effects of adoptively transferred Th cell clones specific for the influenza hemagglutinin (HA), matrix (M), or nucleoprotein (NP) on the antibody response of nude mice infected with A/PR/8/34 influenza virus. We show that the production of antibodies to the HA absolutely requires the presence of virus-specific Th cells. Further, transfer of a Th clone specific for the internal proteins, M or NP, was as effective as was transfer of an HA-specific clone in supporting an antibody response to the HA. With each of the clones, the kinetics of the response were accelerated by approximately 3 d compared with the antibody response of normal BALB/c mice. The HA- and M-specific clones supported an isotype switch from IgM to IgG and IgA similar to that which occurs during a normal antibody response. Finally, as shown by coinfection experiments, the response required a cognate T-B interaction whether the determinants recognized by the Th and B cell are located on the same viral protein or on different viral proteins within the same virus particle. The implications of these findings for understanding the T-B interactions that occur during an effective antiviral antibody response are discussed.     

CD4+ T Cell Tolerance to Parenchymal Self-Antigens Requires Presentation by Bone Marrow–derived Antigen-presenting Cells

T cell tolerance to parenchymal self-antigens is thought to be induced by encounter of the T cell with its cognate peptide–major histocompatibility complex (MHC) ligand expressed on the parenchymal cell, which lacks appropriate costimulatory function. We have used a model system in which naive T cell receptor (TCR) transgenic hemagglutinin (HA)-specific CD4⁺ T cells are adoptively transferred into mice expressing HA as a self-antigen on parenchymal cells. After transfer, HA-specific T cells develop a phenotype indicative of TCR engagement and are rendered functionally tolerant. However, T cell tolerance is not induced by peptide–MHC complexes expressed on parenchymal cells. Rather, tolerance induction requires that HA is presented by bone marrow (BM)–derived cells. These results indicate that tolerance induction to parenchymal self-antigens requires transfer to a BM-derived antigen-presenting cell that presents it to T cells in a tolerogenic fashion.     

CD4+ follicular regulatory T cells optimize the influenza virus–specific B cell response

CD4⁺ follicular regulatory T (Tfr) cells control B cell responses through the modulation of follicular helper T (Tfh) cells and germinal center development while suppressing autoreactivity; however, their role in the regulation of productive germinal center B cell responses and humoral memory is incompletely defined. We show that Tfr cells promote antigen-specific germinal center B cell responses upon influenza virus infection. Following viral challenge, we found that Tfr cells are necessary for robust generation of virus-specific, long-lived plasma cells, antibody production against both hemagglutinin (HA) and neuraminidase (NA), the two major influenza virus glycoproteins, and appropriate regulation of the BCR repertoire. To further investigate the functional relevance of Tfr cells during viral challenge, we used a sequential immunization model with repeated exposure of antigenically partially conserved strains of influenza viruses, revealing that Tfr cells promote recall antibody responses against the conserved HA stalk region. Thus, Tfr cells promote antigen-specific B cell responses and are essential for the development of long-term humoral memory.     

Antibodies that are specific for a single amino acid interchange in a protein epitope use structurally distinct variable regions

We have analyzed how the immune system generates antibodies that are specific for analogues of an epitope on the influenza virus hemagglutinin (HA) that differ solely by the presence of Asp or Gly at amino acid 225. Most antibodies induced in response to HA(Asp225) use one of a few closely related variable (V) region structures that are encoded by characteristic VH/Vk gene segment combinations. Remarkably, none of these VH/Vk combinations was induced in response to HA(Gly225). Instead of modifying the HA(Asp225)-specific V regions by junctional variation or somatic mutation to recognize the altered epitope, new VH/Vk combinations were used. The expression of unique VH/Vk combinations appears to confer exquisite specificity to the selection of HA-specific B cells from the pre-immune repertoire.     

Extensive diversity in the recognition of influenza virus hemagglutinin by murine T helper clones

A panel of H-2k class II-restricted Th clones were established from individual CBA mice primed by infection with X31 influenza virus. 27 clones, which showed specific recognition of the HA surface glycoprotein, were all H3N2 subtype specific, in contrast to a T cell line which was crossreactive and which may have other specificities. 20 distinct HA-specific clones recognized a tryptic cleavage fragment of X31 consisting of residue 28-328 of HA1 (tops) which includes all the Ab-combining regions of the HA molecule. Seven other HA-specific clones failed to respond to either tops or to aggregate (the remainder of the virus after tryptic cleavage of tops). The specificity of these clones has been mapped, tentatively, to a conformational determinant in the interface antibody-binding region of the HA trimer. Analysis of the fine specificity of the HA-specific clones against a panel of H3N2 natural variant viruses isolated from major virus epidemics from 1968 to 1984 revealed a hitherto unrecognized diversity in T cell recognition of a HA. A total of 12 specificity groupings were evident, and varied from groups of clones that recognized all natural variants to one clone that responded only to isolates from 1968 to 1972. Six out of eight clones from a major specificity group, which failed to recognize variants TX/77, BK/79, or CN/84, responded to two overlapping peptides (48-68 and 53-87), corresponding to a region of HA1 that includes part of two antibody combining sites. An examination of the amino acid sequences of natural variant viruses from this region of HA revealed that residues Asn53 and Asn54 and/or Ile62 were critical for recognition by these clones. We conclude that recognition of HA by Th cells is not restricted to a limited number of epitopes in the conserved regions of the molecule, but is extremely diverse and includes specificities that map to variable antibody-combining regions of the molecule. In addition, the sensitivity of the T cell clones to the amino acid substitutions occurring in HA1 of natural variant viruses suggests that Th may play a role in the immune pressure for antigenic variation in the HA molecule.     

Differential antigen presentation regulates the changing patterns of CD8+ T cell immunodominance in primary and secondary influenza virus infections

The specificity of CD8+ T cell responses can vary dramatically between primary and secondary infections. For example, NP366-374/Db- and PA224-233/Db-specific CD8+ T cells respond in approximately equal numbers to a primary influenza virus infection in C57BL/6 mice, whereas NP366-374/Db-specific CD8+ T cells dominate the secondary response. To investigate the mechanisms underlying this changing pattern of immunodominance, we analyzed the role of antigen presentation in regulating the specificity of the T cell response. The data show that both dendritic and nondendritic cells are able to present the NP366-374/Db epitope, whereas only dendritic cells effectively present the PA224-233/Db epitope after influenza virus infection, both in vitro and in vivo. This difference in epitope expression favored the activation and expansion of NP366-374/Db-specific CD8+ memory T cells during secondary infection. The data also show that the immune response to influenza virus infection may involve T cells specific for epitopes, such as PA224-233/Db, that are poorly expressed at the site of infection. In this regard, vaccination with the PA224-233 peptide actually had a detrimental effect on the clearance of a subsequent influenza virus infection. Thus, differential antigen presentation impacts both the specificity of the T cell response and the efficacy of peptide-based vaccination strategies.     

Syngeneic monoclonal antiidiotype can induce cellular immunity to reovirus

A syngeneic monoclonal antiidiotypic antibody was generated in BALB/c mice after repeated immunization with a BALB/c monoclonal anti-reovirus hemagglutinin (HA) antibody. The resultant syngeneic monoclonal antiidiotypic antibody, in the absence of adjuvant, was found to be capable of priming both BALB/c (H-2d, Igh-1a) and C3H/Hej (H-2k, Igh-1j) mice for Lyt-1+- and Lyt-2+-dependent responses against the mammalian reovirus. By the use of intertypic reassortants and variant virus analysis, the specificity of the response was finely mapped to the neutralization domain of the viral hemagglutinin (HA). Using purified monoclonal antiidiotype, we were able to compare the potency of antiidiotype to virus in terms of induction of immunity. 8 X 10(8) protein molecules were able to prime for cellular responses to reovirus. These studies indicate that in the reovirus system, T cells and B cells share idiotypic configurations, and that antiidiotypic antibodies of the type described herein may be useful in the development of vaccines against certain viral infections.     

Active immunization against virus infections due to antigenic drift by induction of crossreactive cytotoxic T lymphocytes

We have examined whether active immunization with c13 protein, a hybrid protein of the first 81 amino acids of the viral NS1 nonstructural protein and the HA2 subunit of A/PR/8 (H1N1) hemagglutinin, could protect BALB/c mice from challenge with A/PR/8 H1 subtype virus. Mice immunized with the c13 protein had a significant reduction of pulmonary virus titers with A/PR/8 (H1) virus, but failed to limit the replication of A/PC (H3) virus, which reflects the in vitro CTL activity of c13 immune spleen cells. We observed that the epitope recognized by HA2 specific CTL, which are induced by a derivative of c13 protein, is highly conserved among H1 and H2 subtype virus strains. This led us to test whether active immunization with c13 protein would also limit pulmonary virus replication in mice infected with the A/TW virus, a virus of the H1 subtype, which was isolated in 1986, and with a virus of the H2 subtype, A/Japan/305/57. Immunized mice had significantly lower lung virus titers than did control mice, and did not possess any neutralizing antibodies to the challenger viruses. These results indicate that active immunization with a fusion protein containing the cross-reactive CTL epitope protects mice from influenza infection by inducing CTL against influenza A H1 and H2 subtype virus strains, which markedly vary in their antibody binding sites on the HA1. The ability to induce active cross-reactive immunization with a fusion protein which contains a highly conserved CTL epitope offers a model for vaccine approaches against viruses which undergo significant variations in their antibody binding sites.     

Differences in antigen presentation to MHC class I-and class II-restricted influenza virus-specific cytolytic T lymphocyte clones

We have examined requirements for antigen presentation to a panel of MHC class I-and class II-restricted, influenza virus-specific CTL clones by controlling the form of virus presented on the target cell surface. Both H-2K/D- and I region-restricted CTL recognize target cells exposed to infectious virus, but only the I region-restricted clones efficiently lysed histocompatible target cells pulsed with inactivated virus preparations. The isolated influenza hemagglutinin (HA) polypeptide also could sensitize target cells for recognition by class II-restricted, HA-specific CTL, but not by class I-restricted, HA-specific CTL. Inhibition of nascent viral protein synthesis abrogated the ability of target cells to present viral antigen relevant for class I-restricted CTL recognition. Significantly, presentation for class II-restricted recognition was unaffected in target cells exposed to preparations of either inactivated or infectious virus. This differential sensitivity suggested that these H-2I region-restricted CTL recognized viral polypeptides derived from the exogenously introduced virions, rather than viral polypeptides newly synthesized in the infected cell. In support of this contention, treatment of the target cells with the lysosomotropic agent chloroquine abolished recognition of infected target cells by class II-restricted CTL without diminishing class I-restricted recognition of infected target cells. Furthermore, when the influenza HA gene was introduced into target cells without exogenous HA polypeptide, the target cells that expressed the newly synthesized protein product of the HA gene were recognized only by H-2K/D-restricted CTL. These observations suggest that important differences may exist in requirements for antigen presentation between H-2K/D and H-2I region-restricted CTL. These differences may reflect the nature of the antigenic epitopes recognized by these two CTL subsets.     

CD27 promotes survival of activated T cells and complements CD28 in generation and establishment of the effector T cell pool

CD27, like CD28, acts in concert with the T cell receptor to support T cell expansion. Using CD27(-/-) mice, we have shown earlier that CD27 determines the magnitude of primary and memory T cell responses to influenza virus. Here, we have examined the relative contributions of CD27 and CD28 to generation of the virus-specific effector T cell pool and its establishment at the site of infection (the lung), using CD27(-/-), CD28(-/-), and CD27/CD28(-/-) mice. We find that primary and memory CD8+ T cell responses to influenza virus are dependent on the collective contribution of both receptors. In the primary response, CD27 and CD28 impact to a similar extent on expansion of virus-specific T cells in draining lymph nodes. CD27 is the principle determinant for accumulation of virus-specific T cells in the lung because it can sustain this response in CD28(-/-) mice. Unlike CD28, CD27 does not affect cell cycle activity, but promotes survival of activated T cells throughout successive rounds of division at the site of priming and may do so at the site of infection as well. CD27 was found to rescue CD28(-/-) T cells from death at the onset of division, explaining its capacity to support a T cell response in absence of CD28.     

Hyper immunoglobulin E response in mice with monoclonal populations of B and T lymphocytes.

A key event in the pathogenesis of allergies is the production of antibodies of the immunoglobulin (Ig)E class. In normal individuals the levels of IgE are tightly regulated, as illustrated by the low serum IgE concentration. In addition, multiple immunizations are usually required to generate detectable IgE responses in normal experimental animals. To define the parameters that regulate IgE production in vivo, we generated mice bearing monoclonal populations of B and T lymphocytes specific for influenza virus hemagglutinin (HA) and chicken ovalbumin (OVA), respectively. A single immunization of the monoclonal mice with the cross-linked OVA-HA antigen led to serum IgE levels that reached 30-200 microg/ml. This unusually high IgE response was prevented by the infusion of regulatory alpha/beta CD4(+) T cells belonging to both CD25(+) and CD25(-) subpopulations. The regulation by the infused T cells impeded the development of fully competent OVA-specific effector/memory Th2 lymphocytes without inhibiting the initial proliferative response of T cells or promoting activation-induced cell death. Our results indicate that hyper IgE responses do not occur in normal individuals due to the presence of regulatory T cells, and imply that the induction of regulatory CD4(+) T cells could be used for the prevention of atopy.     

The long-term maintenance of cytotoxic T cell memory does not require persistence of antigen

I have used the transfer of primed lymphocytes into syngeneic irradiated recipients to investigate whether the persistence of antigen is required in the long-term maintenance of cytolytic T cell memory to influenza virus. Animals were immunized with influenza virus (A/WSN) and used 17 wk later as either donors for T cells or as lethally irradiated recipients. Naive age-matched mice served as controls. At intervals of 4, 8, 16, and 25 wk after T cell transfer, experimental and control groups were immunized with a heterologous virus (A/JAP) and splenocytes tested for lytic activity to influenza virus 3 and 6 d after immunization. Lytic activity 3 d after infection (a property exclusive to a memory cytotoxic T cell response) (Effros, R. B., J. Bennink, and P. C. Doherty. 1978. Cell. Immunol. 36:345.; and Hill, A. B., R. V. Blanden, C. R. Parrish, and A. Mullbacher. 1992. Immunol. Cell Biol. 70:259), was only observed by primed and naive irradiated recipients reconstituted with memory T cells. No day 3 responses were observed when naive T cells were transferred into irradiated primed or unprimed recipients. These observations demonstrate that cytolytic T cell memory to influenza virus is long lived in the absence of antigen.     

CD40-independent pathways of T cell help for priming of CD8(+) cytotoxic T lymphocytes

In many cases, induction of CD8(+) CTL responses requires CD4(+) T cell help. Recently, it has been shown that a dominant pathway of CD4(+) help is via antigen-presenting cell (APC) activation through engagement of CD40 by CD40 ligand on CD4(+) T cells. To further study this three cell interaction, we established an in vitro system using dendritic cells (DCs) as APCs and influenza hemagglutinin (HA) class I and II peptide-specific T cell antigen receptor transgenic T cells as cytotoxic T lymphocyte precursors and CD4(+) T helper cells, respectively. We found that CD4(+) T cells can provide potent help for DCs to activate CD8(+) T cells when antigen is provided in the form of either cell lysate, recombinant protein, or synthetic peptides. Surprisingly, this help is completely independent of CD40. Moreover, CD40-independent CD4(+) help can be documented in vivo. Finally, we show that CD40-independent T cell help is delivered through both sensitization of DCs and direct CD4(+)-CD8(+) T cell communication via lymphokines. Therefore, we conclude that CD4(+) help comprises at least three components: CD40-dependent DC sensitization, CD40-independent DC sensitization, and direct lymphokine-dependent CD4(+)-CD8(+) T cell communication.     

Distinct Costimulatory Molecules Are Required for the Induction of Effector and Memory Cytotoxic T Lymphocytes

A successful T cell immune response has two major products: effector T cells which directly or indirectly remove the antigens, and memory T cells, which allow a faster and more efficient recall response when challenged by related antigens. An important issue is whether costimulatory molecules on the antigen-presenting cells are involved in determining whether T cells will differentiate into effector or memory cells after antigenic stimulation. To address this issue, we have produced mice with targeted mutations of either the heat-stable antigen (HSA), or both HSA and CD28. We show that CD28/B7 and HSA provide two alternative costimulatory pathways for induction of immunological memory to influenza virus. Furthermore, our results revealed that B7 is essential for the generation of effector T cells from either naive or memory T cells, while HSA is not necessary for the generation of effector T cells. Our results demonstrate that the induction of memory T cells and effector T cells can utilize distinct costimulatory molecules. These results have important implications on lineage relationship between effector and memory T cells.     

Generation of antibody diversity in the immune response of BALB/c mice to influenza virus hemagglutinin. I. Significant variation in repertoire expression between individual mice

The paratypic and idiotypic diversity of the BALB/c antibody response to the hemagglutinin (HA) of the influenza A/PR/8/34 virus (PR8) was investigated using a panel of 125 anti-HA hybridoma antibodies derived from 14 BALB/c mice. The paratypic diversity, as assessed by a fine specificity analysis using 51 related influenza viruses, was extensive: 104 distinct paratopes were observed. In three instances, antibodies with indistinguishable paratopes were isolated from two individual mice. A minimum estimate of the size of the adult BALB/c anti-HA paratypic repertoire, calculated from these data, is 1,500. The generation of this diverse repertoire was studied by screening the anti-HA hybridoma panel for the presence of idiotypes (Id) that are markers for variable (V) region sequences derived from related germ line V genes. Three cross-reactive Id (IdX) that are markers for the V(k)21C, V(k)21B, and V(k)21A, D, E, or F L chain subgroups were found, respectively on 16, 1, and 10 anti-HA hybridoma antibodies derived from seven individual BALB/c mice. Thus, the V(k)21 IdX(+) hybridomas constitute 22 percent of the anti-HA hybridoma panel. The V(k)21 IdX are also present on 8.6 percent of K-bearing immunoglobulin in normal BALB/c serum. This suggests that the V(k)21 group is used preferentially in the BALB/c anti-HA immune response. The generation of the anti-HA repertoire was further studied using large panels of anti-HA hybridomas derived from two individual adult BALB/c mice. Anti-idiotypic antisera were raised in rabbits against individual hybridomas from each mouse. One anti-Id serum defined a family of four idiotypically and paratypically related, but not identical, antibodies from mouse 36, which represented 31 percent of the hybridoma antibodies isolated from this mouse. None of the 112 anti-HA hybridoma antibodies derived from 13 other individual mice showed idiotypic cross-reactivity. Furthermore, this Id could not be detected in anti-PR8 antisera from 75 individual BALB/c mice. Another anti-Id serum defined a family of 27 idiotypically related antibodies from mouse 37, which represented 50 percent of the hybridoma antibodies isolated from this mouse. Only 1 of the 71 hybridoma antibodies isolated from 13 other individuals was idiotypically cross-reactive. These results demonstrate that individual adult BALB/c mice express paratypically and idiotypically distinct antibody repertoires to the HA of influenza virus PR8. Based on these observations, we suggest that somatic mutation plays an important role in the generation of the adult anti-HA repertoire. Mechanisms that could account for differences in repertoire expression among individual mice are discussed.     

Mechanism of protection against influenza A virus by DNA vaccine encoding the hemagglutinin gene

Influenza A virus with its two major antigenic surface proteins hemagglutinin (HA) and neuraminidase (NA) is a widely used model to study DNA immunizations in mice and other animals. Natural protection against influenza A virus infection is mediated by antibodies, which mostly are not protective against antigenic shift or drift variants of the original virus. Therefore, it would be a major task to induce a protective cellular immune response to more conserved proteins or epitopes. Injection of plasmid encoding a viral antigen is known to induce cellular as well as humoral immunity. In this study we investigate the mechanism of protection after intramuscular vaccination of C57Bl/6 mice with a DNA vaccine encoding HA of influenza A/PR/8/34. After a single injection, only a small percentage of mice survive the lethal challenge with homologous virus. The amount of protection can be doubled by applying a booster injection. Furthermore, by coinjection of plasmids encoding cytokines GM-CSF and IL-12, respectively, nearly all of the mice are protected. Mice with specific defects in the cellular immune response [perforin knockout (P-/-) mice] and in the humoral immune response [IgD/IgM knockout (muMT) mice], respectively, have been immunized with HA DNA with or without cytokine DNA. Protection could only be induced in P-/- mice, whereas muMT mice succumbed to the infection. Moreover, when muMT mice were infected with only 0.75 x50% lethal dose they died all the same, whereby mice that had been depleted of CD8+ T cells before infection showed an even greater progression of illness. Altogether these results demonstrate that antibodies mediate protection after immunization with plasmid coding for HA of influenza A virus, and that booster immunizations and coinjection of plasmids encoding GM-CSF or IL-12 can improve this protection.     

Low dose Leishmania major promotes a transient T helper cell type 2 response that is down-regulated by interferon gamma-producing CD8+ T cells

An unresolved issue in the field of T helper (Th) cell development relates to the findings that low doses of antigen promote Th2 cell development in vitro, whereas several classic in vivo studies suggest the opposite. Here we resolve this paradox by studying the early immune response in mice after infection with different doses of Leishmania major. We found that low parasite doses induced a Th2 response in C57BL/6 (B6) mice, whereas high doses induced a Th1 response. However, the Th2 response in low dose-infected mice was transient and the animals healed. The appearance of a Th1 response after low dose infection was dependent upon the concomitant activation of interferon gamma-producing CD8+ T cells. In the absence of CD8+ T cells, the Th2 response was maintained. However, either neutralization of interleukin (IL)-4 or administration of IL-12 promoted a Th1 response after low dose infection of CD8-deficient mice, indicating that the required role for CD8+ T cells was limited to modulation of CD4+ T cell responses. Thus, the discrepant results seen between in vivo and in vitro studies on the effects of antigen dose on Th cell differentiation may depend upon whether CD8+ T cells participate in the immune response.