T cell involvement in Ab responses to thymus-independent type 2 Ags is an immunologic enigma. The identity of these cells and the mechanisms of their TCR engagement to carbohydrate molecules remain unknown. We measured IgG Ab production after immunization with pneumococcal polysaccharides in mice with disruptions in selected genes of the T cell pathway. Nonclassical MHC class I-like CD1 molecules and MHC class I-dependent CD8+ cells were found to be essential. Our findings set forth a new paradigm for humoral responses in which CD1 expression as well as a subset of CD8+ cells are required to provide helper function for Ab production against thymus-independent type 2 polysaccharides, similar to MHC class II-restricted CD4+ cells for protein Ags.
Protective immunity against many bacteria, including Streptococcus pneumonia, Neisseria meningitidis, and Haemophilus influenzae, is dependent on the Ab responses to the capsular polysaccharides, which belong to a category of Ab currently classified as thymus-independent type 2 (TI-2)3 (1, 2). Nevertheless, mice depleted of all T cells were rendered incapable of mounting such responses (3, 4), suggesting that certain T cell interactions are essential. Unlike proteins, carbohydrates do not associate with classical MHC class I or class II molecules for TCR engagement. The type of T cell involved in TI-2 responses and their mechanisms of action remain an enigma.
Recently, a MHC class I-like molecule, CD1, has been shown to present a variety of glycolipids and nonpeptidic Ags to T cells (5, 6, 7). CD1 molecules are a family of transmembrane glycoproteins expressed in association with β2-microglobulin (β2m). They are mainly found on the surface of immature thymocytes and APC including dendritic cells, activated macrophages, and B lymphocytes (8). Mice express only one CD1 isotype, CD1d1, which appears to function similarly to the CD1a, b, and c in humans and is encoded by a homologue of the human CD1d gene. CD1 has been shown to present mycobacterial lipid and glycolipid Ags to T cells (5, 6). CD1 has also been shown to bind other types of molecules, including GPI (9) and peptides with a hydrophobic motif (10). Some of the responding T cells isolated have cytolytic activity. However, unlike mice deficient in perforin, CD1-deficient mice did not have a decreased survival after infection with Mycobacterium tuberculosis (11, 12), suggesting that CD1-restricted cytolysis is not a major protective factor in that model. The primary function of this nonclassical immune pathway remains uncertain. Various investigators have established CD1-restricted T cell clones over the past few years (13). Some of the molecules recognized by CD1-restricted clones appear to have the characteristics typical of TI-2 Ags (14, 15). We hypothesized that CD1 might be involved in the immune responses to other natural TI-2 Ags such as pneumococcal or meningococcal polysaccharides. By measuring specific IgG Ab production after immunization in mice with disruptions in selected genes affecting the T cell pathway, we identified the requirement for MHC class I-dependent CD8+ cells and for CD1+ APC. Here, we present a new paradigm of T cell involvement in Ab production against bacterial polysaccharide Ags.
Materials and Methods
Immunization of mice
BALB/cJ mice, 129S1/SvImJ mice, and CD1d knockout mice backcrossed to BALB/c mice over 10 times (C.129S2Cd1dtm1Gru/J) were obtained from Jackson ImmunoResearch Laboratories. Wild-type C57BL/6J mice and selected strains of genetic knockouts on C57BL/6 background, including β2m−/− (B6.129P2-β2mtm1Unc/J), CD4−/− (B6.129S6-Cd4tm1Knw/J), and CD8−/− (B6.129S2-Cd8atm1Mak/J) were also obtained from Jackson ImmunoResearch Laboratories. TAP-1−/− (B6.129S2-Tap1tm1Arp/J) and CD1−/− knockouts (B6.S129Sv-Cd1d) on C57BL/6 background were provided kindly by Dr. L. Van Kaer (Vanderbilt University, Nashville, TN) (16, 17). H2KbDb double knockout (DKO) mice on C57BL/6 background (B6. S129S-H2kdH2dd) were gifts from Dr. F. Lemonnier (Institut Pasteur, Paris, France) (18). For individual experiments, groups of 8- to 12-wk-old mice (n = 4–6) received i.p. injections of 0.1 ml of a 23-valent pneumococcal polysaccharide (PnPS) vaccine (Pneumovax; Merck) and 0.2 ml of tetanus toxoid (TT) (Pasteur Merieux Connaught), or 0.1 ml i.p. of a pneumococcal conjugate vaccine (Prevnar; Wyeth-Ayerst Laboratories), containing serotypes 4, 6B, 9V, 14, 18C, 19F, 23F conjugated to diphtheria CRM197 protein, or the meningococcal polysaccharide vaccine (Menomune; Aventis Pasteur). Serum levels of specific IgG Abs to TT, pneumococcal serotypes 4, 14, 19, and meningococcal group A (Men. A) were measured by ELISA, 3 wk after vaccination.
ELISA for PnPS specific IgG
ELISA for PnPS IgG was modified from a protocol established by Dr. G. Carlone at the U.S. Centers for Disease Control and Prevention (Atlanta, GA) (19). Briefly, 96-well Immunolon II plates (Dynex) were coated with PnPS Ags (American Type Culture Collection). Mouse serum samples, diluted 1/100 in a buffer containing the C polysaccharide, were incubated on the plates for 2 h at room temperature. After washing, bound Abs were probed with HRP-labeled goat anti-mouse IgG-Fc Abs (Jackson ImmunoResearch Laboratories) and TMB substrate (Kirkegaard & Perry Laboratories). The resulting change of OD was determined on an ELISA plate reader at 450 nm. ELISA for IgG against Men. A was performed following the method of Carlone et al. (20) as described previously.
ELISA for TT-specific IgG
ELISA for TT Ab assay was performed on Immunolon II plates coated with TT (Pasteur Merieux Connaught). Serum samples diluted 1/500 in PBS with 1% BSA and 3% goat serum were incubated on the plates for 2 h at 37°C. HRP-labeled goat anti-mouse IgG-Fc and TMB substrate were used to detect bound IgG.
Reconstitution of β2m−/− mice
Splenocytes from normal C57BL/6 mice were incubated in gelatin-coated plastic tissue culture flasks for 1 h at 37°C. Adherent cells were recovered with cell stripper (Mediatech) and then depleted of B cells using micromagnetic beads coated with anti-CD19 mAb (Miltenyi Biotec). These cells were >97% CD14+ monocytes. From nonadherent splenocytes, B cells were positively selected by CD19 micromagnetic beads (Miltenyi Biotec) and then depleted of contaminating T cells by using CD3 Ab-coated magnetic beads (Dynal) (21). The resulting preparation was >98% pure CD19+ B cells. The remaining nonadherent splenocytes, depleted of B cells, were then positively selected by CD8 Ab-coated micromagnetic beads (Miltenyi Biotec). The resulting preparation consisted of >95% CD3+CD8+ T cells. Each β2m−/− mouse received 107 CD8+ cells, 2.5 × 107 B cells, and 5 × 105 monocytes i.p. The mice were then immunized with 0.1 ml of Pneumovax and 0.2 ml of TT within 16 h. Serum Ab were measured by ELISA at 3 wk.
Results and Discussion
To investigate a possible role of CD1 in Ab responses to soluble bacterial polysaccharide, we utilized mice with homozygous disruption in the Cd1d genes (Cd1d−/−) for immunization. This strain, C.129S2Cd1dtm1Gru/J, originated on a C.129S background and was extensively backcrossed (N11) to BALB/cJ mice. CD1d−/− mice do not express CD1 molecules on their cell surface and lack NK+ T cells. They have normal numbers of CD4+ and CD8+ lymphocytes and normal expression of MHC class I and class II molecules (18). After immunization with polysaccharide vaccines, the CD1d−/− mice failed to produce any significant levels of IgG Abs to PnPS or Men. A polysaccharide, in contrast to the CD1d+/+ control mice (p < 0.02) (Fig. 1). Following immunization with TT, levels of specific IgG Abs were comparable between the two groups. These results suggested that the absence of CD1 rendered the mice incapable of producing IgG Ab specific for PnPS or Men. A polysaccharide, although retaining the ability to produce Ab to a MHC class II-restricted T-dependent Ag.
To confirm these findings and to characterize the regulatory pathway involved, we repeated the vaccination experiment in a second, larger, group of C57BL/6 mice with disruptions in selected gene loci that are critical in the classical and nonclassical MHC restricted T cell pathways. With the notable exception of the CD4−/− mice (22), all other groups were found to produce Ab to TT. The responses to the polysaccharide Ags were very different (Fig. 2). Similar to the results of the first experiment in the CD1d−/− mice with BALB/c genetic background, CD1d−/− mice with C57BL/6 background also had no detectable IgG Abs to PnPS, although they responded normally to TT. The β2m−/− mice responded to the PnPS and TT in the same pattern as the CD1 knockouts. The β2m chain is part of the heterodimer for the CD1, MHC class I (K/D allotypes), and the FcRn. Thus, gene disruption of β2m results in the absence of functional forms of all these molecules (23). Their TT Ab titers were lower than that of the C57BL/6 wild type, corresponding to the decreased total serum IgG due to a defect in the FcRn receptor (24). Pretreatment of the serum from these mice with DTT, to destroy serum IgM, before performing the ELISA, did not alter the negative findings with regard to specific PnPS IgG Abs (data not shown). Thus, the inability to detect specific IgG Abs by ELISA is unlikely to be due to Ab competition from IgM. The findings in these two groups of mice confirmed that CD1 is required for PnPS IgG Ab production, presumably as the Ag-presenting molecule. An indication for the phenotype of the T cells that responded to PnPS, in the context of CD1, was provided by two other groups of gene knockout mice. CD8−/− mice (25) were unable to produce PnPS type 4 and type 19-specific IgG, whereas CD4−/− mice produced these Abs at levels at least as high as the wild-type C57BL/6 controls. This pattern was exactly opposite to the responses against the MHC class II-restricted, T-dependent Ag TT. Thus, CD8+ cells seem to function as the Th cells in this CD1-dependent system, whereas CD4+ cells do not seem to enhance Ab production to the soluble PnPS, in contrast to what has been reported when intact bacteria were used (26).
An unexpected finding was the markedly diminished production of polysaccharide IgG Abs in H2KbDb DKO mice and TAP1−/− mice when compared with the wild-type control group (p < 0.04) (Fig. 2). The decrease was more severe in H2 KbDb DKO mice than in TAP1−/− mice. The effect cannot be attributed to a general deficiency in IgG production, since none of the H2KbDb DKO, TAP1−/−, CD1d−/−, or CD8−/− mice have decreased levels of serum total IgG, or of IgG3 and IgG1 (data not shown) that are the predominant isotypes for polysaccharide Abs (27). Both H2KbDb DKO and TAP1−/− strains express CD1 on the surface of their APC (12), therefore their inability to produce PnPS-specific IgG is unlikely to be due to a defect in Ag presentation by CD1. H2KbDb DKO mice do not express MHC class I molecules and are deficient in CD8+ T cells, due to a lack of positive selection during thymopoiesis (19). TAP1−/− mice lack the transporter that mediates the translocation of peptides from the cytoplasm into the endoplasmic reticulum. They express little or no MHC class I on the cell surface, because of the defect in loading MHC class I heavy chains in the cytosol (17, 28). The decrease in CD8+ T cells in TAP1−/− mice is not as marked as in H2KbDb DKO, and the CD8+ T cells that remain appear to be polyclonal with respect to TCR V β-chain expression (17). One possible explanation for the failure of these mice to produce PnPS IgG Ab may be that maturation of PnPS-specific CD8+ cells require MHC class I, similar to the classical CD8+ lymphocytes and a novel subset of CD8+NKT cells (29). This hypothesis is consistent with previous findings that CD1 does not contribute significantly to the development of CD8+ T cells (30). The relative reduction of PnPS Ab production in these two groups of mice (Fig. 2) also appears to match their relative depletion of CD8+ T cells. Alternatively, it may indicate the involvement of another, as yet unidentified, subset of MHC class I-dependent cells.
To exclude a possible role for elements other than those we have identified (i.e., CD1+ APC and CD8+ cells), we reconstituted β2m−/− mice, which were deficient in both MHC class I and CD1-dependent immune components, with selected subsets of splenic cells from normal donors. Four β2m−/− mice received adoptive transfer of purified CD8+ T cells, B cells, and macrophage/monocytes from unimmunized C57BL/6 wild-type mice, followed by vaccination with the PnPS vaccine and TT. Our results demonstrated that reconstitution with CD1+ APC and CD8+ T cells alone was sufficient to restore the ability of β2m−/− mice to produce specific IgG Abs to PnPS types 4, 14, and 19 (Fig. 3). The lower level of polysaccharide-specific IgG production in the reconstituted β2m−/− mice, compared with wild-type C57BL/6, is similar to their relative levels of TT Abs and consistent with the known FcR defect in nonlymphoid cells in these mice that cannot be corrected by splenic cell transfer. In addition, partial reconstitution of β2m−/− mice with CD1+ B cells and macrophages did not enable these mice to produce Abs to the PnPS (data not shown). These findings further confirm that PnPS Ab production does not require the MHC class II-restricted CD4+ Th cells but is critically dependent on both CD1+ APC and responsive CD8+ T cells.
In the absence of the MHC class I/CD1-restricted Th cells, such as in the β2m−/− mice, PnPS Ab production should still be possible if the required Th function is provided by an alternative pathway. This can be accomplished by conjugating the polysaccharides to a protein carrier so that they can be presented in the context of MHC class II to elicit CD4+ T cell help. As expected, after immunization with the pneumococcal conjugate vaccine, both controls and β2m−/− mice produced PnPS-specific IgG Abs to the three serotypes tested (Fig. 4). Thus, the B cells of β2m−/− mice are capable of producing PnPS IgG Abs when the polysaccharide Ags are presented by MHC class II with CD4+ T cell help. Data from these experiments does not preclude the involvement of a different subset of B cells involved in the responses to the conjugate vaccine.
In this report, we have presented evidence for a previously unrecognized function of the nonclassical MHC class I-like Ag-presenting molecule CD1. Our findings indicate that their primary functions may include providing T cell help in Ab production to some TI-2 Ags. Despite elucidation of the cellular basis for the thymus-dependent humoral immunity to protein/peptide Ags, no satisfactory model has been established for the mechanisms underlying the induction of the putatively thymus-independent immune responses to bacterial polysaccharides. The ability of peptides to associate with MHC class II molecules on dendritic cells/macrophages and B cells allows for specific engagement of TCRs. The resulting T cell priming and the interactions between the Ag-specific B cells and the T cells, through costimulatory receptor-ligand pairs and cytokines, lead to Ig secretion and Ig isotype switching. Polysaccharide Ags have been shown to bind CD21+ marginal zone B cells, and it has been suggested that these cells are involved in specific Ab production. However, the nature of the B cell population responsible for producing polysaccharide-specific IgG Abs is unknown. IgM production to these Ags can occur in vitro through the cross-linking of BCR molecules by the repetitive epitopes on polysaccharide Ags and is thought to involve immature B1 B cells (31). Both β2m−/− and CD1d−/− mice possess normal numbers of marginal zone and follicular B cells, which do not express CD1, yet they are incapable of producing PnPS IgG (32). Addition of exogenous cytokines may induce IgM secretion, but the B cells do not undergo isotype switching without T cell interaction. One of the stumbling blocks in any comprehensive model for TI-2 Ab responses has been the absence of a known mechanism for TCR engagement with polysaccharide Ags. Our evidence on the involvement of CD1 provides the missing link. Based on our current understanding of Ag presentation and Ab production, one may postulate that PnPS are presented by CD1 molecules expressed on Ag-presenting dendritic cells/macrophages and B cells. They can thus prime and elicit helper functions from CD1-responsive effector T cells. Although we have yet to fully identify the subsets of T cells involved, our results suggest that they have a CD8+ phenotype because mice lacking CD8 were unable to produce PnPS IgG Abs. Since TI-2 Ab responses can be found in athymic nude mice, the origin of the CD8+ cells involved poses another intriguing question that remains to be investigated.
We thank Dr. Luc Van Kaer for providing the B6.129sv-Cd1d knockout mice and for helpful discussions; and Dr. Francois Lemonnier for the gift of the H2KbDb DKO mice.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by the National Institutes of Health Grant HD01392 (to L.J.K).
Abbreviations used in this paper: TI-2, thymus-independent type 2; β2m, β2-microglobulin; DKO, double knockout; PnPS, pneumococcal polysaccharide; Men. A, meningococcal group A; TT, tetanus toxoid.