Idiotype-specific CD4+CD25+ T suppressor cells prevent, by limiting antibody diversity, the occurrence of anti-dextran antibodies crossreacting with histone H3



CD25+ suppressor T cells regulate the immune response against the type-2 "thymus independent" bacterial polysaccharide antigen α(1→3)dextran (Dex) in BALB/c mice. These T cells, represented by the clone 178-4 Ts, restrict the Dex-specific IgG antibody repertoire such that the J558 idiotype dominates. Antibodies with other structures in the heavy-chain variable region (VH region), predominantly within the CDR3 domain, occur when the T cell control fails. This increase of antibody diversity caused by a lack of CD25+ Ts cells, e.g. in nude mice, does not result in the appearance of antibodies with enhanced affinity to the antigen Dex, but often leads to a crossreactivity with autologous proteins. Twenty-two out of sixty Dex-specific hybridomas from nude mice, but no hybridomas from euthymic mice, crossreact with a nuclear protein, as tested by ELISA. This nuclear protein was identified as histone H3. Ten of the sixty hybridomas from nude mice were sequenced and show VH sequences that deviate from the original J558 sequence. Three of these ten hybridomas crossreact with the histone H3. Adoptive transfer of CD25+ Ts cells to nude mice leads to a marked increase of antibodies carrying the original J558 idiotype within the IgG pool after immunization with Dex. Our data demonstrate a CD25+ Ts cell-mediated restriction of VH usage, which prevents the appearance of crossreactive autoantibodies.


Antinuclear antibodies



VH region:

Heavy-chain variable region

1 Introduction

CD4+CD25+ suppressor T cells are known as potent immunoregulators (reviewed in 14). They play key roles in preventing the development of CD8+ T cell alloreactivity 5, in the inhibition of organ-specific autoimmunity 6, in the immunoregulation of alloresponses 7 and in the regulation of autoantibody production 8. It is presumed that CD4+CD25+ Ts cells require a TCR with high affinity for a self peptide 9 and require B cells for their expansion in the periphery 10.

The humoral response to the type-2 thymus-independent polysaccharide antigen α(1→3)dextran (Dex) is controlled by such regulatory T cells 11. These Ts cells, represented by the clone 178-4 Ts, restrict the antibody repertoire such that IgM is widely expressed. Only few dextran-specific IgG antibodies are found, which are predominantly of the J558 idiotype. Theisotype- and idiotype-distribution among hybridomas from euthymic BALB/c mice reflects this situation. The 178-4 Ts cells exclusively recognize a self peptide from the heavy-chain variable region (VH region) CDR3 (YCARDRYWYFDVW) of the J558 idiotype 12. They expand after immunization with Dex and can be found in a tight idiotypic connection with J558+ B cells 13 in the germinal centers of the spleen. It has to be pointed out that 178-4 Ts cells display no helper function on J558+ B cells, since Dex-specific IgG titers are by far lower in euthymic than in athymic mice. J558+ B cells are arrested in euthymic mice but can be activated upon adoptive transfer 11. The occupation of germinal centers by arrested J558+ B cells may leave only little space for the development of B cells carrying other idiotypes 14.

A lack of T cell control, e.g. in nude mice, results in a vigorous anti-Dex IgG response involving many different idiotypes. Furthermore, addition of 178-4 Ts cells to a culture of purified B cells from Dex-primed mice results in a dominance of B cells carrying the J558 idiotype 15. These T cells on the one hand suppress the IgG production and on the other hand provide a signal for the survival and clonal accumulation of J558-bearing B lymphocytes 15. We now demonstrate an autoimmune potential among the Dextran-specific IgG antibodies whose sequences have drifted away from the original J558 sequence, and which crossreact with the histone H3.

2 Results and discussion

2.1 Dex-specific IgG from nude mice deviates from the J558 sequence and crossreacts with autologous proteins

CD4+CD25+ Ts cells are important for maintaining peripheral tolerance. Depletion of CD25+ T cells, transfer of CD4+CD25 T cells to T cell deficient mice, or thymectomy before Ts cells have left the thymus leads to severe autoimmune processes 3. The idiotype-specific Ts cell clone 178-4 Ts restricts the idiotype distribution among Dex-specific IgG-producing B cells such that the J558 idiotype dominates 15. Dextran-specific IgG antibodies with other VH structures only occur when the T cell control fails, e.g. in nude mice (Table 1). This T-cell-mediated invariance of the antibody repertoire to the bacterial polysaccharide antigen Dex prevents the occurrence of autoantibodies, since Dex-specific IgG antibodies from hybridomas with non-J558 idiotypes can crossreact with BALB/c proteins (Table 1).

2.2 Phenotype of 178-4 Ts cells

The 178-4 Ts cells can be phenotypically characterized by flow cytometry as CD4+ CD25+ CD122 (Fig. 1), which is a typical phenotype of Ts cells. Furthermore, 178-4 Ts cells show a CD45RBlow phenotype, which is also characteristic for a number of Ts cells 16. The 178-4 Ts cells recognize a definite invariant peptide (amino acid positions 91–103) from the VH/D/JH joining region of the J558 idiotype 12 via their α βTCR. In this idiotypic T/B cell interaction, costimulation via CD28, which is expressed by 178-4 Ts cells, is required for T cell activation 14.

Table 1. Amino acid sequences of the VH regions and crossreactivity of different Dex-specific hybridomas from BALB/c and BALB/c nude micea)
  1. a) The original J558 sequence and deviations from this sequence are given. The peptide recognized by the J558-idiotype -specific 178-4 Ts cells is underlined. All mAb from euthymic BALB/c mice (3 representative hybridomas are depicted) were of the J558 idiotype, whereas the sequences of the VH regions of mAb from athymic BALB/c nu/nu mice (10 randomly chosen hybridomas aredepicted) diversify, especially within the CDR3 segment. Purified and biotinylated mAb were used in a protein blot of murine kidney proteins. None of the J558+ Dex-specific mAb, but 3 out of 10 mAb from nude mice with altered VH structures, crossreact with autologous proteins, despite having a comparable avidity to the original antigen, Dex. n.d., not determined.

original image
Figure 1.

 Phenotype of 178-4 Ts cells. Six-to-eight-day-old cultures of 178-4 Ts cells were stained with anti-CD28, anti-CD4, anti-CD45RB, anti-CD25 or anti-CD122.

2.3 Crossreactivity of Dex-specific mAb

Eight independently isolated Dex-specific IgG hybridomas from individual euthymic BALB/c mice (three are shown in Table 1) express VH structures that are relatively identical to the original J558 protein sequence 12. In IgG hybridomas from BALB/c nu/nu mice an obviously great variability is found, as demonstrated in 10 out of 60 randomly chosen hybridomas that were sequenced (Table 1).

Three out of the ten sequenced Dex-specific antibodies from athymic BALB/c nu/nu mice with altered VH regions crossreact with a 20 kDa BALB/c protein, whereas antibodies with the canonical J558 sequence, despite having a comparable avidity to Dex (Table 1), do not show any crossreactivity. The three mAb — N11H6, 7C6 and N10B10 — crossreact with a protein of same size and pI.

Since 3 out of 10 sequenced Dex-specific IgG mAb with altered VH sequences crossreact with a protein sized about 20 kDa, we isolated this protein from extracts of BALB/c kidney using crossreacting Dex-specific mAb, coupled to magnetic particles. Fig. 2 shows a Coomassie-stained protein blot of a 2D-electrophoresis of the protein, captured by the Dex-specific mAb 7C6 (see Table 1). This protein, with a size of about 20.7 kDa and an approximate pI of 7.5, does not bind to antibodies with the canonical J558 sequence (e.g. 7D7.5.4; Fig. 2), although they have the same avidity to the original antigen, Dex (Table 1).

N-terminal sequencing of the spot identified this protein as the murine histone H3 (Table 2). Fourteen out of the fifteen sequenced amino acid residues of the protein are most probably identical to the histone sequence. At positions 5 and 9, the sequence was not readable clearly without ambiguity; this may be due to the small amount of protein used for sequencing. Nevertheless, the protein captured by the Dex-specific antibodies 7C6, N11H6 and N10B10 could be clearly identified as histone H3.

As in the anti-Dex response in euthymic mice, the majority of the Dex-specific IgG from nude mice is of the IgG3 isotype 12. Addition of purified histones to immune sera results in a significant inhibition of specific binding of IgG3 antibodies to the antigen, Dex. Inhibition of Dex-binding by histones only occurs in immune sera of BALB/c nude mice, but not in immune sera of euthymic BALB/c mice or nude mice adoptively transferred with CD4+CD25+ T cells or 178-4 Ts cells (Fig. 3A).

In euthymic mice, low levels of Dex-specific IgG are found because 178-4 Ts analogous suppressor T cells reduce IgG production by silencing J558 Id+ Bγ-memory cells 12. Simultaneously, the idiotype distribution among the IgG3+ pool is directed by these suppressor cells such that the J558 idiotype dominates 15. The absence of Ts cells in nude mice leads to a vigorous Dex-specific IgG response characterized by a heterogeneous VH usage 14. A remarkable percentage of antibodies within this heterogeneous pool in nude mice bind to histones. In contrast, no crossreactivity to histones is found within the Dex-specific IgG of euthymic BALB/c mice, where the VH usage is homogenous, mainly restricted to the original J558 sequence by Ts cells (Fig. 3A). Because of the T cell mediated restriction of the VH usage to the original J558 sequence, i.v. injection of either the CD25+ T cell clone 178-4 Ts or purified autologous CD4+CD25+ T cells into nude mice before immunization leads to an impressive reduction of crossreactivity to histones among the IgG3 antibodies, comparable to the situation in euthymic BALB/c mice.

About 30% of culture supernatants of Dex-specific IgG-producing hybridomas from BALB/c nude mice crossreact to a remarkable degree with histones, as measured by histone-specific ELISA, whereas the histone-binding of culture supernatants from Dex-specific IgG-producing hybridomas of euthymic BALB/c mice is comparable to the medium control (Fig. 3B).

The presence of antinuclear antibodies (ANA) was clearly demonstrated in sera from Dex-immunized nude mice by indirect immunofluorescence (titer >1280; titer 640 shown in Fig. 4B). In contrast, sera from immunized euthymic BALB/c mice did not show any nuclear staining (titer 40; Fig. 4A). Adoptive transfer of purified CD4+CD25+ T cells protects nude mice from ANA during the course of immunization with Dex (Fig. 4C). Transfer of 178-4 Ts cells dramatically reduces the ANA-titer in nude mice. (titer ≥1280 reduced to titer 80). The homogenous fluorescence pattern in the nuclei of HEp-2 cells and the positive staining of chromosomes in mitotic cells by sera from nude mice (Fig. 4, right-hand panels) support our finding that antibodies in Dex-immunized nude mice exhibit crossreactivity to histones.

T cells analogous to 178-4 Ts cells are expressed in all euthymic BALB/c mice 13. The cognate T/B cell interaction on the one side silences Bγ memory cells 11 and on the other side retains the J558 VH CDR3 integrity of IgG antibodies by preventing the appearance of a high VH-region variability, which can result in potentially autoreactive IgG antibodies of a non-J558 idiotype.

Crossreactivity between peptides and carbohydrates 17, 18 is well documented. One approach using this phenomenon for vaccination is to link peptides that mimic polysaccharide structures to protein carriers. Immunization with these mimotope–carrier structures enhances immune responses against poorly immunogenic capsular polysaccharides 19. On the other hand, antigenic mimicry can break self tolerance by crossreaction of antibodies to bacterial polysaccharides with autologous structures. So, in the anti-Dex response a subtle balance — mediated by CD4+CD25+ Ts cells — exists, which ensures an antibacterial response by IgG antibodies of the canonical J558 idiotype and avoids autoimmunity by preventing the appearance of antibodies with other, potentially hazardous, VH structures. Self tolerance is often maintained by Ts cells 20, 21. Here, failure of Ts cell control, i.e. the lack of 178-4 Ts analogous cells in nude mice, results in the occurrence of polysaccharide-specific antibodies with enhanced VH variability, which can be autoreactive against histone H3 because of a kind of presumed antigenic mimicry between Dex and the histone H3.

Our data demonstrate a unique V/DH usage, controlled by T cells analogous to 178-4 Ts cells, and the appearance in athymic mice of unexpected heavy-chain structures, which can crossreact with the histone H3. This emphasizes the advantage of Ts cell mediated antibody-invariance in responses to antigens with highly conserved epitopes like bacterial polysaccharides, in contrast to responses to protein antigens where immune responses are usually associated with increased antibody diversity, resulting in affinity maturation.

Figure 2.

 Protein blot of kidney proteins captured by Dex-specific mAb. BALB/c kidney proteins, purified magnetically by Dynabeads conjugated to Dex-specific mAb 7D7.5.4 (J558 idiotype; left) or 7C6 (non-J558 idiotype; right), respectively, were transferred on a PVDF membrane and stained with Coomassie-blue after a 2D-electrophoresis. Only the non-J558-idiotype Dex-specific mAb 7C6, but not the J558+ mAb 7D7.5.4, crossreacts with a single 20 kDa protein.

Figure 3.

 Crossreactivity of Dex-specific IgG. Immune sera (A) and culture supernatants of Dex-specific IgG mAb (B) of athymic BALB/c nude (hollow bars), but not of euthymic BALB/c mice (black bars), crossreact with histones. Transfer of either purified CD4+ CD25+ T cells (hatched bars) or 178-4 Ts cells (crosshatched bars), transferred into BALB/c nude mice, prevents this crossreactivity. Preimmune sera and immune sera from Dex-immunized mice taken 12 days after each immunization and diluted 1:180 were monitored for Dex-specific IgG3 by ELISA. Inhibtion of Dex-binding by histones was tested by addition of 100 μl of histone solution (+) or control (BSA; –) per well. Culture supernatants of 50 Dex-specific IgG-producing hybridomas from athymic BALB/c nu/nu and 10 from euthymic BALB/c mice were monitored for crossreactivity with histones by ELISA (B). (C) Evaluation of the anti-histone ELISA. Plates were coated with ovalbumin (OVA), keyhole limpet hemocyanin (KLH), hen egg lysozyme (HEL), histones or Dex, respectively, and monitored for binding of the four Dex-specific IgG mAb 7D7.5.4 (hollow bars), 7C6 (hatched bars, rising to the right), N10B10 (hatched bars, rising to the left) and N11H6 (crosshatched bars). The mAb with the original J558 sequence (7D7.5.4) only binds to Dex, whereas the three mAb with altered CDR3 regions additionally crossreact with histones. All mAb were applied as nondiluted hybridoma supernatants.

Figure 4.

 Analysis of ANA in immune sera from Dex-primed mice. ANA were analyzed by indirect immunofluorescence, using HEp-2 cells and FITC-conjugated goat anti-mouse-IgG as detection Ab. Sera from nude mice (B) contain high titers of ANA (titer 640), whereas sera from euthymic BALB/c mice (A) do not show any nuclear staining (titer 40). Adoptive transfer of CD4+ CD25+ T cells (C) or idiotype-specific 178-4 Ts cells (D) before immunization inhibits or strongly decreases, respectively, the occurrence of ANA in nude mice.

Table 2. N-terminal amino acid-sequences of the precipitated 20.7 kDa protein and the mouse-histone H3 27a)</
  1. a); Since the amount of protein precipitated by the Dex-specific mAb 7C6 was at the detection limit for sequencing, the most likely sequence and possible alternatives are given.

original image

3 Materials and methods

3.1 Mice

BALB/c-AnNIcr and BALB/c-AnNIcr nu/nu mice were purchased from Charles River, Sulzfeld, Germany, or bred at the Institute's animal facilities. Animals were used for experiments at an age of 8–12 weeks.

3.2 Antigens and immunization

Dextran B1355S (Dex) (average molecular weight 4×107 Da) from Leuconostoc mesenteroides and containing 43% (1→3)-D-glucopyranosyl linkages was used as antigen. It was a kind gift from Dr M. E. Slodki (Northern Regional Research Laboratory U.S.D.A., Peoria, IL, USA). Mice were immunized by a single i.p. injection of 10 μg Dex in saline.

For monitoring crossreactivity of anti-Dex mAb, or immune sera, by ELISA, unfractionated whole histone 22 from calf thymus (Sigma, Deisenhofen, Germany) was used.

3.3 Cell culture procedures

Hybridoma formation was performed according to standard procedures 12, 23. Hybridomas were tested by ELISA for Dex-specific IgG Ab. Cells from positive wells were used to form monoclonal cultures by limiting-dilution. Hybridoma cells were cultured in DMEM supplemented with 10% FCS.

The T cell clone 178-4 Ts was maintained as previously described 12.

3.4 FACS analysis

Six-to-eight-day-old cultures of 178-4 Ts cells were harvested. Approximately 5×106 cells were stained with 2 μl of various antibodies: PE-conjugated hamster anti-mouse-CD28 (37.51; BD Pharmingen, Hamburg, Germany), Cy-Chrome® conjugated rat anti-mouse-CD4 (RN4–5; BD Pharmingen), FITC-conjugated rat anti-mouse-CD45RB (16A; BD Pharmingen), PE-conjugated rat anti-mouse-CD25 (PC61; BD Pharmingen), or FITC-conjugated rat anti-mouse-CD122 (5H4; BD Pharmingen). Stained cells were analyzed using a Becton Dickinson FACScan cytometer and LYSIS II software.

3.5 Purification and biotinylation of Dex-specific mAb

Culture supernatants of Dex-specific IgG hybridomas were harvested and IgG was purified using the protein-G sepharose technique 24. IgG mAb were eluted from the protein-G sepharose columns (Pharmacia, Uppsala, Sweden) using 100 mM glycine/HCl buffer at pH 2.8 and dialyzed. The concentration of antibodies was determined photometrically at 280 nm and diluted to 1 mg/ml in PBS.

Purified antibodies were biotinylated by incubating 1 ml of mAb with 15 μl of Biotin-7-NHS (D-Biotinoyl-ϵ-amminocarpronacid-N-hydroxysuccinimideester 25; Sigma) solution (2 mg/ml in dimethylsulfoxide) for 2 h at room temperature. Unbound biotin was removed by Sephadex-G-25 chromatography.

Purification and biotinylation were controlled by Dex-specific ELISA.

3.6 ELISA for Dex- and histone-specificity

Microtiter plates (96-well; Greiner, Nürtingen, Germany) were coated overnight with 100 μl Dex in PBS, pH 7.3 (25 μg/ml) per well. Plates were washed and saturated with 200 μl per well of 1% BSA (Sigma) in PBS for 1 h. The immune sera or affinity-purified mAb were serially diluted 1:3 in dilution buffer (PBS with 0.1% BSA / 0.05% Tween 20); immune sera were prediluted 1:20. Plates were incubated for 2 h at room temperature. When testing inhibition of Dex-binding by histones, 100 μl per well of histone solution (50 μg/ml in dilution buffer) (control: 100 μl BSA; 50 μg/ml in dilution buffer) was added. After washing, 100 μl per well of biotin-conjugated rat anti-mouse-IgG3 Ab (R40–82; BD Pharmingen), diluted 1:1000 in dilution buffer, was added as secondary Ab for 1 h. When testing biotinylation of mAb, no secondary antibody was used. After another washing step a volume of 100 μl per well of ExtrAvidin® peroxidase (Sigma) (1:1000 in dilution buffer) was added. The ELISA was developed with 1 mg/ml 1-2-phenylene-diamine (Sigma) in 0.2 M NaH2PO4, 0.1 M Na-citrate, pH 5.0 with 0.08% H2O2. The extinction was measured at 490 nm (MR4000, Dynatech, Bonn, Germany).

For monitoring crossreactivity of Dex-specific IgG mAb, microtiter plates were coated with 100 μl of purified histone (Sigma) in PBS, pH 7.3 (5 μg/ml) per well. The ELISA was processedas described above, using a biotinylated goat-anti-mouse-IgG (H+L; Dianova, Hamburg, Germany) as secondary antibody. The specificity of this ELISA was ensured by using a couple of charged and uncharged proteins like ovalbumin, keyhole limpet hemocyanin or hen egg lysozyme, respectively, for coating the plates (each protein was used at 5 μg/ml in PBS, 100 μl per well).

3.7 Coating of antibodies to magnetobeads

Dex-specific antibodies (15 μg) were incubated together with 1×108 Dynabeads M-280 streptavidin (deutsche DYNAL GmbH, Hamburg, Germany) according to the manufacturer's instructions. Beads were collected magnetically, the supernatant was removed and the beads were resuspended in PBS.

3.8 Magnetic protein-purification

Different organs of BALB/c mice were homogenized gently on ice using an Ultra Thurrax in buffer containing 50 mM Tris/HCl, 150 mM NaCl, 0.5% (w/v) triton X-100, 5 mM dithiothreitol and proteinase-inhibition-cocktail tablets (Boehringer Mannheim, Mannheim, Germany) according to the manufacturer's instructions. The suspension was stirred on ice for 30 min, centrifuged and filtered to remove insoluble parts.

Cell lysates (300 μl) were mixed with 1×107 Dynabeads–mAb conjugate at 4°C for 3 h under gently rotation, collected magnetically and washed three times in TBS-buffer. The pellet was resuspended in 50 μl of SDS-PAGE sample-loading buffer (62.5 mM Tris/HCl pH 6.8, 2% w/v SDS, 5% v/v β-mercaptoethanol and 25% v/v glycerol) and heated to 100°C for 3 min. The tube was placed in the magnet and the supernatant was carefully removed for loading onto polyacrylamide gels.

3.9 Gel electrophoresis and protein blots

Gel electrophoresis was performed using Bio-Rad's Protean IEF cell together with the ReadyPrep® 2-D Kit (Bio-Rad, Munich, Germany) with ReadyStrip IPG-strips in the first, and precast 12% SDS-PAGE polyacrylamide gels in the second dimension, according to the manufacturer's manual.

Molecular-mass estimates of bands were determined by comparing their electrophoretic mobilities with those of prestained molecular-mass marker proteins ranging from 7.2 to 237.0 kDa (Bio-Rad).

Proteins were transferred to Bio-Rad's Sequi-Blot PVDF membranes 26 using the Trans-Blot electrophoretic transfer cell (Bio-Rad) at 100 V and 350 mA for 1.5 h.

The PVDF membrane was stained with Coomassie-blue (Bio-Rad), spots were marked and the membrane was destained without using acetic acid for further use in Edman-sequencing.

3.10 Sequence analysis

Automated Edman degradation for the determination of the amino acid sequence was performed by SEQLAB Sequence Laboratories Göttingen GmbH (Göttingen, Germany). The sequences obtained were usedto identify the peptide by searching the Swiss-Prot protein database with the Blast algorithm.

3.11 Cell separation

Separation of CD4+CD25+ double-positive cells was performed in two steps. First CD4+ cells from 1×109 spleen cells from Dex-immunized mice were purified, using the mouse CD4+ T cell isolation kit, an LS column and a MidiMACS, (Miltenyi Biotec, Bergisch Gladbach, Germany), according to the manufacturer's instructions. This step was followed by flow cytometric separation. Cell suspensions were each stained with 20 μl of Cy-Chromen®-conjugated rat anti-mouse-CD4 and PE-conjugated rat anti-mouse CD25 and the CD4+CD25+ population (purity >95%), was positively sorted using a Becton Dickinson FACSvantage cytometer.

3.12 Adoptive transfer experiments

Either 2×106 purified CD4+CD25+ T cells or 1×107 178-4 Ts cells were injected i.v. into BALB/c nude mice one day before immunization with Dex. Twelve days after immunization, immune sera were analyzed for crossreactivity to histones and the content of ANA.

3.13 Immunofluorescence assay for ANA

ANA were demonstrated by indirect immunofluorescence. Slides coated with HEp-2 cells (Innogenetics, Heiden, Germany) were incubated with sera from BALB/c-AnNIcr and BALB/c-AnNIcr nu/nu mice atdilutions of 1:40 and twofold dilutions thereof for 30 min, washed and developed with a 1:100 dilution of FITC-conjugated goat anti-mouse-IgG (Fab-specific; Sigma), and then viewed with a fluorescence microscope (Leitz, Wetzlar, Germany). The titer (i.e. the reciprocal of the highest dilution of mice sera giving a positive nuclear staining) and the fluorescence pattern were recorded.


  1. 1


  2. 2


  3. 3


  4. 4