Understanding thymus-independent antigen-induced reduction of thymus-dependent immune responses

Abstract

Deficiencies in immune responses against polysaccharides can have direct consequences for patients, and therefore, a better understanding of these immune reactions is crucial. We have studied the immune response against the polysaccharide dextran B512 (Dx). Administration of immunogenic doses of thymus-independent (TI) Dx reduces the immunoglobulin G1 response to later challenges with a thymus-dependent (TD) form of Dx. We investigated if this suppression is a general phenomenon caused not only by Dx but also by other TI antigens, and examined possible mechanisms contributing to this unresponsiveness. We show that clonal exhaustion is not involved in modulating subsequent responses, nor is signalling via FcγRIIB or other antibody mediated pathways. The reduced TD response is not an exclusive Dx phenomenon; it is also induced by TI antigen oxazolone (Ox). However, responses against the hapten dinitrophenyl (DNP) are not affected, indicating that the TI priming negative effect is not a general process. This may be explained by the restricted immune response to both Dx and Ox, in contrast to the unrestricted DNP response. Our conclusion from these experiments is that the underlying mechanism for the TI-induced reduction of latter TD responses is a property of the TI activation itself.

Introduction

Thymus-independent type 2 (TI-2) antigens have the ability to stimulate B cells without the assistance of T lymphocytes. The classic example of a TI-2 antigen is a high molecular weight polysaccharide such as the dextran B512 (Dx). Dx derived from the bacteria Leuconostoc mesenteroides is used in this study as model antigen. Dx is formed by glucose units linked to 96% in a α1–6 position, creating a large near-linear structure.¹ Just as for many carbohydrates, the ability to respond to Dx is something that develops rather late in life; mice acquire full potential to respond against Dx as late as 3 months after birth.^2,3 The antibody response against native Dx is a typical thymus-independent (TI) response; it is primarily immunoglobulin M (IgM) and does not undergo significant affinity maturation or memory formation. Furthermore, the response against Dx is restricted in the use of a few variable (V) immunoglobulin genes and is in the mouse strain C57BL/6 dominated by the expression of the V_HB512 and the V_KOX-1 genes.^4,5

Unlike the responses against most TI antigens, immunizations with Dx induce the formation of germinal centres (GC) in the spleen.^6,7 During TD responses, the GCs provide a milieu where immunoglobulin class switch and somatic hypermutation occur, processes that lead to affinity maturation. However, despite the GC formation, the antibody response specific to Dx stays predominantly IgM and does not show signs of maturation even after successive immunizations. In fact, it is known that secondary responses against Dx are reduced compared to primary responses.^8–10 This unresponsiveness against Dx can be partially overcome by the use of an appropriate adjuvant, such as cholera toxin (CT).^7,11 When Dx is administered together with CT, the secondary response is at least as strong as the primary response, but it still consists mainly of IgM.¹¹

While the native form of Dx has a high molecular weight and is a strong TI-2 antigen, Dx with lower molecular weight are less immunogenic. Small Dx, with a molecular weight of 5 × 10⁵ and less, do not induce a response unless they are coupled to carrier molecules. When carbohydrates are coupled to thymus-dependent (TD) antigens, such as proteins, the conjugate primarily assumes the characteristics of a TD antigen¹² making it possible to create antigens that elicit a TD response against Dx epitopes. Repeated immunizations with a TD form of Dx gives a strong humoral response consisting of Dx specific IgG1 and development of immunological memory. An interesting feature of the Dx response is that priming with Dx as TI antigen, modulates following responses induced by the same antigen given in a TD form. The TD Dx response will consist mainly of IgM and very little Dx-specific IgG1, while the response against the protein carrier is unaffected.¹³ The reversed protocol, TI Dx challenge after priming with TD Dx, leads to the production of Dx-specific IgG.¹³ We have previously shown that this TI-induced reduction of the Dx specific IgG1, is long lived in mice and cannot be abrogated by the use of the adjuvant CT.^13,14

Revealing the mechanisms for this effect would not only contribute to a better understanding of the regulation of the immune responses, but may also have important implications in vaccination research, especially for vaccines carrying carbohydrates. In this work we have examined a number of aspects that may contribute to the unresponsiveness caused by TI antigens. Collectively, our results suggest that the reduction of IgG1 in TD Dx responses cannot be explained by clonal exhaustion nor by antibody mediated mechanisms such as regulation via Fc receptors. Furthermore, we show that this phenomenon is not unique for Dx. On the basis of our findings we discuss possible mechanisms and implications for this type of immunomodulating response.

Materials and methods

Mice

Founders for the FcγRIIB–/– were given to the department by Professor Heyman, Department of Genetics and Pathology, Uppsala University, Uppsala, Sweden, and were originally a gift from Professor J. V. Ravetch, Division of Molecular Biology, Sloan Kettering Cancer Center, NY, USA. Lk transgenic mice carrying immunoglobulin rat κ light chain were originally generated at MRC Laboratory of Molecular Biology, Cambridge, UK, and were bred into a C57BL/6 background and maintained at our animal facilities. The C57BL/6 mice were purchased from Charles River (Uppsala, Sweden) and M & B (Ry, Denmark). The animals were maintained in our animal facilities at the Department of Immunology (Stockholm University, Stockholm, Sweden). The mice were 8–14 weeks old at the initiation of the experiments.

Antigens and immunizations

The TI-2 antigen native Dx B512 was obtained from INC Pharmaceuticals Inc. (Cleveland, OH). A TD form of Dx was obtained by conjugating Dx with a MW of 10³ (3–5 glucose units) to the protein chicken serum albumin (CSA; Sigma Chemical Co., St Louis, MO). Dx was conjugated to hydrazide-CSA via its terminal aldehyde group using reductive amination as described before.¹⁵ TI forms of the haptens 2-phenyloxazolone (Ox), and dinitrophenyl (DNP) were obtained by conjugating 4-ethoxymethylene-2-phenyl-2-oxazolin-5-one (Sigma-Aldrich Corp., St. Louis, MO) and Nε-2,4-DNP-l-lysine (Sigma-Aldrich), respectively, to Dx with a MW of 2 × 10⁶ (Pharmacia Fine Chemicals, Uppsala, Sweden) as previously described.¹⁵ TD forms of Ox were obtained by conjugating Ox to bovine serum albumin (BSA; Sigma-Aldrich) and human gammaglobulin (Pharmacia) as described.¹⁶ The TD antigen BSA-DNP was obtained from Biosearch Technologies (Novato, CA). Dx-specific antibodies originated from hybridoma cell lines^5,17 were administered at the dose of 80 µg/immunization of IgM and 1·5 µg/immunization of IgA.

The mice were immunized intraperitoneally. Dx was given at doses of 10 µg/animal; BSA-Ox, BSA-DNP and CSA-Dx at doses of 100 µg/animal; and Dx-Ox and Dx-DNP at doses of 50 µg/animal. All antigens were administered together with the adjuvant CT (List Biological Laboratories Inc, Campbell, CA) 1 µg/animal/dose. Mice were bled by retro-orbital puncture under light anaesthesia. Serum was collected after centrifugation and kept at −20° until use.

After priming with the TI antigen (Dx, Dx-Ox or Dx-DNP), mice were immunized with the TD antigen (CSA-Dx, BSA-Ox or BSA-DNP, respectively) administered 4 and 7 weeks after priming. The animals were bled day 7 after final immunization. TD antigen-challenged mice not receiving TI antigen priming were treated in the same way except for the lack of TI priming. Dx-specific antibodies were given every other day for a total of three times. Four and 7 weeks after the administration of antibodies, mice were challenged with CSA-Dx and were bled day 7 after the last CSA-Dx immunization.

Detection of specific antibodies by enzyme-linked immunosorbent assay (ELISA)

Briefly, 96-well ELISA plates (Costar, Corning, NY) were coated with either 10 µg/ml Dx T250 (Pharmacia Fine Chemicals), 10 µg/ml CSA-Ox, 10 µg/ml OVA-DNP (Biosearch Technologies), 10 µg/ml OVA-TNP (Biosearch Technologies), 10 µg/ml human gamma globulin-Ox, 10 µg/ml BSA (Sigma-Aldrich Corp.) or 10 µg/ml CSA (Sigma-Aldrich Corp.) overnight at room temperature. Serial serum dilutions were applied to the plates and incubated overnight at room temperature. Bound antibodies were detected with alkaline-phosphatase labelled goat anti-mouse IgM or IgG1 (Southern Biotech. Assoc., Birmingham, AL) incubated for 2 hr at RT. When testing the hybridomas, an alkaline-phosphatase labelled mouse antibody against rat κ light chain (Sigma-Aldrich Corp.) was used in this last step, to verify the use of the transgenic light chain. After this, the plates were developed with the substrate p-nitrophenyl phosphate (Sigma-Aldrich Corp.). The absorbance was measured using an Anthos Reader 2001 (Anthos Labtech Instruments, Salzburg, Austria).

Hybridoma cell lines

Hybridoma cell lines used for determining the V_H usage originated from Lk mice bred onto C57BL/6 background. Mice were hyperimmunized with either Dx or CSA-Dx administered together with CT. Spleens were taken day 3 after the third immunization and fused with the non-secreting hybridoma cell line SP2/0 using polyethylene glycol solution Hybri-Max (Sigma-Aldrich Corp.) as fusing agent. The cultures were grown under the selection pressure of HAT-media Hybri-Max (Sigma-Aldrich Corp.). The antigen specificity of the clones was determined by ELISA as described above.

RNA isolation, reverse transcriptase–polymerase chain reaction (RT–PCR) amplification and sequencing

Total RNA was isolated from 1 to 5 × 10⁶ cells of each hybridoma using the SV Total RNA Isolation System (Promega, Madison, WI). cDNA was obtained using the CMFOR primer (5′GCT CCT GCA GAC GAG-3′), specific for the constant µ region, and the RT murine Moloney leukaemia virus (Promega). Five µl of cDNA reaction mixture were amplified in PCR reaction using the forward primer VH1FOR-2 (5′TGA GGA GAC GGT GAC CGT GGT CCC TTG GCC CC-3′), complementary to J_H segments, and the back primer VH1BACK (5′-AGG TSM ARC TGC AGS AGT CWGG-3′), designed to complement V_H genes from nucleotide position 2–23. PCR amplification was performed in 50 µl containing 375 µm dNTPs, 25 pmol of each primer and 5 units of BioTAQ^™ polymerase (Bioline UK Ltd, London, UK) in a Gene Cycler (Biometra, Göttingen, Germany). The PCR conditions were: 2 min at 94°, followed by 30 cycles of 45 s at 94°, 1 min at 55° and 2 min at 72°; the final cycle was completed by 10 min at 72°. The PCR products were cut out from agarose gels, purified with Nucleospin Extract kit (Macherey-Nagel GmbH & Co.KG, Germany), and their DNA concentration was measured.

The sequence reaction was performed with the primer VH1FOR-2 and 35–40 fmol of the PCR products using the Dye terminator Cycle Sequencing kit (Beckman Coulter, Fullerton, CA). The conditions were as follows: 30 cycles of 20 s at 96°, 20 s at 45° and 4 min at 60°. The products were washed, precipitated in ethanol, resuspended in deionized formamide and directly sequenced in an automated DNA sequencer (CEQTM 2000, Beckman Coulter). The identity of the V_H genes was determined using the program IgBlast (NCBI) (http://www.ncbi.nlm.nih.gov/igblast).

Results

The V_H repertoire of anti-Dx antibodies derived from secondary TI and TD responses are both restricted and use the same V_H repertoire

We and others have reported^4,5 that the immune response to Dx B512 is restricted because of the dominant expression of V_H and V_L genes. In the mouse strain C57BL/6 about 70% of the specific antibodies induced after primary immunizations with TI and TD forms of Dx carry the V_HB512 gene in combination with the V_ΚOX-1 gene.⁵ This restriction, combined with the relatively long time it takes to develop responses against Dx, suggests that clonal exhaustion could be a possible explanation for the reduced TD response after TI priming. Strong TI activation, in the absence of memory formation, could potentially cause the responding B-cell clones to mature into short-lived effector cells, thereby decreasing the capacity to mount secondary responses. However, in more recent studies we have shown that when Dx is given together with the adjuvant CT, secondary TI Dx responses are generated and at levels at least as high as in the primary responses.¹¹ This abrogation of the unresponsiveness after immunization with TI Dx using CT, could indicate that either Dx-specific clones are not clonally exhausted, or that in the presence of CT different clones are generated in the secondary response. To resolve this issue, we produced hybridoma cell lines from transgenic mice carrying the V_ΚOx1 gene linked to a κ rat light chain gene. The mice had been hyperimmunized with either TI or TD forms of Dx, native Dx and Dx coupled to CSA, respectively, in the presence of CT. We then studied the V_H Ig gene usage in 10 Dx-specific κ rat + hybridomas. The results show a restricted used of the V_HB512 gene in both TI and TD anti-Dx responses (Table 1), the same as previously reported for primary responses.⁵ Even if just a few clones were tested, it can be established that the V_H repertoire is unchanged, indicating that the mechanisms for shaping the restricted repertoire are the same in both TI and TD responses. Thus, we can conclude that neither the use of CT nor the primary immunizations influence the repertoire of subsequent Dx responses, and consequently that clonal exhaustion does not seem to be the mechanism contributing to the reduced TD responses induced by TI Dx priming.

Table 1.

V_H genes of IgM⁺ hybridomas obtained from mice immunized with native dextran B512 as TI antigen or CSA-Dx as TD antigen, and comparison with related known V_H genes

Immunization*	Hybridomas	Related	Homology‡	VH family
VH gene†
TI	A1D5/1	VHB512	100	J558
	A1D9/2	VHB512	100	J558
	A2B9/2	J558.k	94	J558
	A3B2/1	VHB512	100	J558
	A3H8/2	VHB512	100	J558
	A4H6/1	VHB512	100	J558
TD	M3E5	VHB512	100	J558
	M6E1	VHB512	100	J558
	M6E2	VHB512	100	J558
	M4H11	VHB512	96	J558

^*Transgenic C57BL/6 mice carrying a rat

^κlight chain were immunized with dextran in TI or TD forms and cholera toxin as adjuvant.

^†V_H genes were sequenced and compared with already published sequences of antidextran specific antibodies⁵ or with mouse immunoglobulin genes sequences using the program IgBlast (NCBI).

^‡Percentage of homology with related V_H genes. Nucleotide differences within the last four positions were ascribed to junctional events and they were not considered in the comparisons.

Passive administration of Dx-specific antibodies does not cause reduced IgG1 levels

Apart from clonal exhaustion, a number of immunoglobulin-dependent mechanisms have been proposed to be the cause for the reduction of the TD response to Dx mediated by TI priming. We can name the generation of anti-idiotype (Id) antibodies, and also remaining Dx-specific antibodies from the primary response that might deliver negative signals through Fc receptors. The immune response against Dx B512 is dominated by the production of antibodies of the IgM class with minor representation of other isotypes. In C57BL/6 mice, a high proportion of the anti-Dx antibodies express the major Id, 17-9 Id (Id⁺), that was first described in an IgA monoclonal antibodies (mAb).¹⁸

To elucidate if Dx specific antibodies remaining after TI Dx priming were responsible for the reduced TD IgG1 response, mice were given Dx-specific Id⁺ IgM or IgA mAb before challenge with antigen. Treated mice received three doses with 80 µg of IgM or 1·5 µg of IgA, at days 0, 2 and 4. The amount of IgM in one administration is comparable to serum levels of Dx-specific antibodies at day 10 of an anti-Dx immune response.¹³ The administered antibodies could be detected in sera and maintained their specificity (data not shown). One additional group, used as control, received antibodies with irrelevant specificity (DNP specific IgM). Four and 7 weeks after the administration of mAbs, the animals were challenged with CSA-Dx together with CT, and sera were collected 7 days after the last immunization. The serum levels of Dx specific IgM and IgG1 were compared to the serum levels in mice challenged with CSA-Dx as mentioned above, with or without previous TI Dx priming. The results show that the magnitude of the anti-Dx IgM response (Fig. 1a) induced by TD Dx was very similar in all groups, whether treated with soluble Dx-specific Id⁺ IgA or IgM, or irrelevant mAb (DNP-specific IgM). While Dx-primed mice showed the expected reduction of Dx-specific IgG1 antibodies (Fig. 1b), the treatment with anti-Dx antibodies did not result in a reduction of the IgG1 levels in the TD response. This suggests that down-regulation of Dx specific IgG1 is probably not caused by mechanisms induced by antigen-specific immunoglobulin, such as the generation of anti-Id antibodies or negative signals mediated by Fcµ or Fcα receptors.

Figure 1.

TD Dx responses in mice treated with Dx-specific antibodies. Different groups of 8 week old mice received, either Dx-specific IgM (IgM α-Dx) or IgA (IgA α-Dx) monoclonal antibodies or antibodies with irrelevant specificity (IgM α-DNP), prior to two immunizations with TD Dx (CSA-Dx). For comparison, TI Dx primed and non-primed groups are also shown. Values for individual mice (shown as absorbance levels) are shown as diamonds, and the line indicates the mean for each group. Statistical difference from the nonprimed control group is marked with * (P < 0·05 using Student's t-test). (a) Levels of Dx-specific IgM at serum dilution 1 : 1600. (b) Dx-specific IgG1 at serum dilution 1 : 3200.

Negative regulation by FcγRIIB is not responsible for TI antigen induced low IgG1 response against TD Dx

Even if IgM production is dominant, immunization with TI Dx also results in the generation of low levels of Dx-specific IgG2b and IgG3 antibodies.^19,20 Because B-cells express FcγRIIB, a receptor that binds IgG1, IgG2a and IgG2b subclasses²¹ and signalling through this receptor gives a negative signal that affects B-cell proliferation and antibody secretion,^22,23 we wanted to investigate if specific IgG could bind to the FcγRIIB and cause a down-regulation of later TD Dx responses. Wild type C57BL/6 and FcγRIIB knockout mice were either primed with TI Dx or not before challenge with the TD form, CSA-Dx. At day 7 after TD immunization, we analysed the humoral response of the different groups of mice by ELISA measuring Dx-specific IgG1 levels in sera. The results show no differences between the responses in the wild type and the FcγRIIB knockout mice (Fig. 2). After secondary immunizations with the TD form of Dx, non-TI-primed mice from both mouse strains elicited a strong IgG1 response, while priming with TI Dx induced a reduction of Dx specific IgG1. Because the effect of TI priming was as strong in FcγRIIB knockout mice as in wild type mice, we concluded that negative regulation by FcγRIIB mediated by soluble IgG antibodies did not cause the reduction of the TD Dx response.

Figure 2.

TD responses against Dx in TI Dx-primed FcγRIIB knock out mice. IgG1 levels in sera, taken from mice day 7 after secondary immunization with CSA-Dx, with or without previous priming with TI Dx, were measured by ELISA. Mean values for three mice are shown, error bars represent standard deviation. The levels of Dx-specific IgG1 differs significantly (P < 0·05 using Mann–Whitney U-test) between TI primed and non-primed groups for both mouse strains (FcγRIIB knock out and C57BL/6) while there is no difference in response between the knock out and wild type mice. Representative data are shown from one of two performed experiments.

Reduction of TD IgG1 responses induced by TI-priming is also observed in the restricted response against the hapten Ox

We wanted to investigate if the restriction observed in the immunoglobulin gene usage in the antibody response against Dx was a key element for the reduction of TD IgG1 responses after TI priming. To assess this, we chose two well-characterized haptens: namely Ox and DNP. The response to Ox in C57BL/6 mice is restricted and the V region of the anti-Ox antibodies light chains is coded by the V_ΚOX-1 gene^24,25 the same gene used in the response to Dx.⁴ In contrast, the response to DNP is more heterogeneous using both κ chains as well as different λ subtypes²⁶ and can induce antibodies of different idiotypes.²⁷ The haptens Ox and DNP were used coupled to either a high molecular weight Dx, to obtain TI forms of the antigens (Dx-Ox, Dx-DNP), or to BSA giving TD forms (BSA-Ox and BSA-DNP). Mice were primed with the TI form prior challenge with the TD form of the respective antigen and the specific IgG1 response was evaluated. Priming of mice with Dx-Ox previous to immunization with BSA-Ox caused a statistically significant reduction of the levels of Ox specific IgG1 compared to those mice that did not receive the TI priming (Fig. 3a). However, that reduction was lower than the observed when using Dx. In contrast, no significant differences were found in the levels of DNP specific IgG1 between mice primed with Dx-DNP (Fig. 3b) and those not primed.

Figure 3.

TD responses after TI priming. (a) Serum levels of Ox specific IgG1 in TI Ox primed and non-primed mice were measured by ELISA. The OD of the sample at 405 nm was adjusted against an internal standard to enable comparison between different ELISA plates. Values show the mean of the experimental groups (n = 7), error bars show standard deviation. Statistical differences (P < 0·05 using Mann–Whitney U-test) are marked with *. (b) IgG1 response against DNP in mice primed with TI DNP or nonprimed. Values show mean of experimental groups of three mice, and error bars represent standard deviation. No statistical differences between the groups could be detected using Mann–Whitney U-test. Data from one representative experiment of two performed.

Discussion

Though deficiencies in immune responses against polysaccharides can have important consequences for patients, the molecular basis for this type of immune reactions remains largely unknown. Mice and humans are competent to make antibodies against TD antigens more or less from birth, whereas the response to TI-2 antigens such as pneumococcal and Haemophilus influenzae type B polysaccharides is delayed in humans until the second year of life.²⁸ This is also the case with the polysaccharide Dx B512. In mice, the response to Dx is only fully developed by 3 months of age, which is well into adulthood. Another aspect of the response to Dx is the restriction to the dominant expression of few clones using the V_HB512 and V_ΚOX1 genes. In addition, and in common to other TI antigens, Dx is a poor inducer of immunological memory and IgG class switch occurs only to IgG2b and IgG3.^19,20

Experimentally induced unresponsiveness to Dx is long lasting, probably related to the characteristics of the response, which is clonally restricted and requires a long maturation time to recover from clonal deletion.⁹ Unresponsiveness can be caused by the administration of high doses of TI Dx, which induces tolerance²⁹ or by administration of immunogenic doses of TI Dx (in the presence of adjuvant), which reduces the IgG response to latter challenges with a TD form of Dx.⁷ Together with the poor immunogenicity of polysaccharides, this last named type of induced unresponsiveness may be a way used by encapsulated bacteria to evade the immune system. In the present work we investigate if this type of suppression is a general phenomenon caused not only by Dx but also by other TI antigens and examine possible mechanisms contributing to this unresponsiveness.

TI induced reduction of TD responses might be caused by clonal exhaustion of antigen-specific clones during the TI response. Such exhaustion could be due to signals causing differentiation of the responding B cells into short-lived effector cells, lack of signals needed for memory cell formation, or antibody-mediated negative signals inducing cell death. In this paper we show that after repeated TI or TD immunizations, Dx-specific IgM carry the same V_H repertoire as in primary responses, demonstrating not only that TI and TD challenges appear to evoke the same type of restricted B-cell response but also that repeated immunizations do not lead to a shift in the IgM repertoire. It thus appears unlikely that the TI response induces a total clonal exhaustion since the IgM repertoire stays constant. If the TI challenge exhausted the responding clones totally, one would expect a shift in V_H gene usage in subsequent responses. In addition, even if the responding cells were clonally deleted during TI priming and the animals were repopulated by new B cells carrying the same V_H genes in time before later challenges, the phenomenon of clonal exhaustion can not explain why these new cells do not switch to IgG1 after later TD challenges. Furthermore, the observation that the V_H gene usage in a TD response did not differ from a TI-induced response implies that both types of responses activate similar B-cell subsets. This suggests that TI and TD responses are not generated by different B cells but more likely that the TI activation pathway irreversibly shapes the latter responses.

We also show in this work that antibody-mediated mechanisms such as Id regulation or signalling through Fc receptors cannot explain the reduction of TD responses. Administration of Id⁺ antibodies, which could potentially lead to the formation of anti-Id antibodies or the down regulation of B cells through Fc receptors, did not influence later TD Dx responses. Furthermore, mice lacking the FcγRIIB reacted just as strong as wild type mice on TI Dx priming.

Our conclusion from these experiments is that the underlying mechanism for the TI-induced reduction of the TD response is a property of the TI activation. We favour the explanation that IgG1 reduction may be caused by signals directing differentiation of the responding B cells to a non-switch/non-memory pathway. In other words, that TI-2 antigens drive the responding B cells onto a path of differentiation that renders them unable of immunoglobulin class switch and memory formation during later responses. This explanation is supported by experiments showing that different types of activation drive responding B cells into acquiring different surface phenotypes.^30–32 Other possible explanations could be that the kinetics of the responses differ or that the reduced response is caused by the T-cell compartment. Studies of the kinetics in TD and TI Dx responses have shown that the timepoint used is optimal for both kind of responses⁷ suggesting that kinetic differences is not a major contributor to the variance in TD Dx responses in TI Dx primed versus not-primed mice. Because the response against the protein carrier is unaffected it appears as if the T-cell response is normal; however, further experiments would be needed to directly address this question. A possibility to determine the role of the T-cell compartment is by performing cell transfer experiments. The transfer of B cells from TI Dx-primed mice into B-cell deficient animals, which then could be challenged with TD Dx, could give important information on a possible role for T cells in this type of response.

The reduction of TD responses is probably a general phenomenon, not only related to the carbohydrate Dx B512 because TI Ox also induced a reduction in TD Ox responses. However, the suppressive effect is not universal because the response to the hapten DNP was not affected. This may be explained by the different characteristics of the immune response against the three antigens. The immune response to both Dx or to Ox displays a restricted repertoire while the response to DNP is unrestricted, suggesting that restriction could have a crucial role in the inhibition of TD responses. That the effect of priming with TI Ox (Fig. 3a) is not as strong as the effect of Dx priming (Fig. 2) could possibly be explained by lower Ox epitope density on the Dx-Ox conjugate. One important observation was that mice primed with Dx before challenge with CSA-Ox sometimes displayed a low, not statistically significant, reduction in the anti-Ox IgG1 response. Possibly this can be related to the preferential use of the V_KOx-1 chain in both Ox and Dx responses. This may indicate that the element of restriction is central for inducing a long-term effect on the immune response.

The finding that a reduction of TD responses can be induced by another restricted TI-2 epitope apart from Dx clearly shows the impact that TI priming can have for subsequent responses. It also suggests that the TI-induced reduction of the TD response is a result of the TI antigen-induced maturation of the responding B cells and not caused by suppressive signals in the B cell environment. To fully understand why TI and TD responses differ in such characteristics as immunoglobulin switch and memory formation, it is necessary to gain more knowledge of the differences in activation and maturation of the responses. A better comprehension of the basics of TI and TD responses will also give valuable insights in the effect of TI versus TD priming on subsequent immune responses. A knowledge that can be crucial for the development of effective vaccines, especially against encapsulated bacteria, pathogens that cause millions of deaths each year.

Abbreviations

bsa

bovine serum albumin

csa

chicken serum albumin

ct

cholera toxin

DNP

dinitrophenyl

dx

dextran

gc

germinal centre

Id

idiotype

Id⁺

17-9 Id

mAb

monoclonal antibodies

Ox

2-phenyloxazolone

td

thymus-dependent

ti

thymus-independent

ti-2

thymus-independent type 2