The Journal of Immunology, 2003, 170: 1158-1166.
Copyright © 2003 by The American Association of Immunologists
Comparative Effects of Human Ig
and Ig
in Inducing Autoreactive Antibodies Against B Cells in Mice 1
Jim J. C. Sheu*,
Tammy Cheng
,
Huan Y. Chen*,
Carmay Lim,
and
Tse-Wen Chang2,*,
Departments of
*
Life Science and
Chemistry, National Tsing Hua University, Hsinchu, Taiwan;
Institute of Biomedical Science, Academia Sinica, Taipei, Taiwan; and
Development Center for Biotechnology, Taipei, Taiwan
 |
Abstract
|
|---|
Human and mouse Ig
molecules share only 58% amino acid sequence identity in their extracellular regions. However, mice immunized with a recombinant Fc fusion protein containing the extracellular portion of human Ig produced significant amounts of IgG capable of binding to Ig on mouse B cells. The induced auto/cross-reactive Abs could down-regulate B cell levels and the consequent humoral immune responses against an irrelevant Ag in treated mice. Analogous immunization with an Fc fusion protein containing the extracellular portion of human Ig
gave a much weaker response to mouse Ig, although human and mouse Ig, like their Ig counterparts, share 56% sequence identity in their extracellular regions. Protein sequence analyses indicated that a potential immunogenic segment, located at the C-terminal loop of the extracellular domain, has an amino acid sequence that is identical between human and mouse Ig. A mAb A01, which could bind to both human and mouse Ig, was found to be specific to a peptide encompassing this immunogenic segment. These findings suggest that specific auto/cross-reactivity against self Ig can be induced by a molecular mimicry presented by a foreign Ig.
|
Introduction
|
|---|
The strength of an immune response is often related to the phylogenetic distance between the recipient and the immunizing Ag. Normally, the immune system does not produce autoreactive Abs and maintains tolerance toward self components. In patients with autoimmune diseases, autoreactive Abs or immune cells are generated. Among the various causes of autoimmune responses, certain microbial infection has been suggested (1, 2, 3). A plausible explanation is that some microbial proteins share similar antigenic epitopes with certain proteins in patients (known as molecular or antigenic mimicry), and as a result, the Abs and/or immune cells induced by the infectious microorganisms cross-react with autologous Ags (4, 5, 6, 7).
The effect of molecular mimicry in generating autoreactivity has also been found when a recipient is immunized with a homologous protein from a different species (8, 9, 10). These studies show that autoreactive B cells can present determinants of homologous foreign proteins to Th cells. The activated Th cells then release stimulatory signals triggering the autoreactive B cells to make Abs against epitopes shared by the immunizing Ag and the otherwise tolerated self protein. In some cases, the breaking down of B cell tolerance to self Ags can prime autoreactive T cells and elicit T cell-mediated autoimmune responses without the direct stimulation by the homologous foreign Ags (9, 11). Exploiting the immune cross-reactivity between homologous foreign and autologous proteins has provided the basis of new vaccine approaches (10, 12, 13).
Ig (also known as CD79a or mb-1) and Ig (CD79b or B29) form heterodimers through disulfide linkage and associate noncovalently with membrane-bound Igs (mIgs) as a part of the B cell Ag receptor (BCR)3 complex (14, 15). Ig and Ig, which both are expressed from early through late stages of B cell development and differentiation, play important roles in the surface expression of mIgs and signal transduction through the BCR (16, 17, 18). At the amino acid sequence level, human Ig (huIg) and mouse Ig (muIg) are highly homologous (>90%) in the intracellular and transmembrane regions, but are less so (58%) in the extracellular region (19, 20, 21). The pattern of homology between huIg and muIg is nearly the same as that for Ig.
In our earlier studies, we have demonstrated in mice the potential of developing an agent based on a fusion protein, composed of the exterior portions of self Ig and a foreign IgG.Fc, for inducing autoreactive Ab response against self Ig as a means to down-regulate B cells (22). In this study, we have investigated the molecular mimicry between huIg and muIg and explored its utility in inducing specific autoreactivity against self Ig. We also have found different effects of generating cross-reactivity between similar agents based on Ig and Ig, and analyzed the molecular basis for the difference.
|
Materials and Methods
|
|---|
Construction, expression, and purification of Fc fusion proteins
Human B cell RNA was extracted from Daudi cells (a human B cell line from American Type Culture Collection (ATCC), Manassas, VA) with TRIzol reagent (Life Technologies, Gaithersburg, MD), according to the procedure provided by the manufacturer. The cDNA was synthesized by Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI) using oligo(dT) as the reverse primer. A nest-PCR amplification was performed for the extracellular portion of huIg or huIg. The resultant PCR products were purified and cloned into the C terminus of huIg
1.Fc in a pcDNA3 vector (Invitrogen, San Diego, CA). The procedures for cell transfection, protein expression, and purification were as described in our previous study (22).
Immunization with BALB/c mice
Male BALB/c mice were purchased from National Laboratory Animal Breeding and Research Center (NLABRC, Taipei, Taiwan) and maintained until 6–8 wk of age for immunization. Mice were primed i.p. with huIg1.Fc-huIg or huIg1.Fc-huIg fusion protein in CFA (Sigma-Aldrich, St. Louis, MO) at the dosage of 50 µg/mouse. The treated mice were boost immunized with the recombinant proteins in IFA every 2 wk. Control groups were treated with either PBS, as a negative control, or equal amounts of huIgG.Fc, as a positive control in IFA. For B cell activity assay, OVA (Sigma-Aldrich) was given with IFA after 1 wk of the fourth immunization of above treatments at the dosage of 20 µg/mouse. The second and third OVA immunizations were boosted 1 wk after the fifth and sixth immunizations of above treatments, respectively. Antisera were prepared from treated mice 1 wk after each immunization.
Preparation of maltose-binding protein (MBP) fusion proteins
Gene segments of extracellular portions of huIg and huIg in Fc fusion protein expression vectors were subcloned into a pMALp2 vector (NEB, Beverly, MA). Mouse RNA was prepared from the splenocytes of BALB/c mice. Similar strategies as described above were adopted to clone muIg and muIg gene segments into the pMALp2 vector. The constructed expression vectors were introduced into Escherichia coli (BL21/pLys strain, from Culture Collection and Research Center, Hsinchu, Taiwan) and induced for protein expression at 0.1 mM isopropyl -D-thiogalactoside in Terrific Broth for 4 h at 20°C. After centrifugation, bacteria pellets were resuspended in PBS and sonicated for total 20 min (pulse 20 s, break 10 s) in ice. Cell debris were removed by centrifugation; soluble MBP fusion proteins in the supernatant were purified with amylose resin (NEB, Beverly, MA) by eluting with 10 mM maltose (Sigma-Aldrich).
ELISA analysis
For detecting the specific reactivity of antisera or the mAb A01, ELISA plates (Clontech, Palo Alto, CA) were coated overnight with various Ags (10 µg/ml, 100 µl/well), including huIgG, MBP, and MBP-fusion proteins (MBP-huIg, MBP-huIg, MBP-muIg, MBP-muIg), in 0.1 M sodium carbonate buffer (pH 9.6). For B cell activity assay, OVA was used as the solid-phase Ag in ELISA. Four peptides, listed in Table I, were synthesized and coated (1 µg/well) in ELISA plates for epitope identification with 50 mM phosphate buffer (pH 8.0) containing 0.5 M NaCl. After blocking with 5% nonfat milk, antisera or A01 at various dilutions were added to the plates and incubated for 2 h at 37°C. Specific binding was detected using HRP-conjugated goat anti-muIgG Abs (ICN, Aurora, OH) and tetramethylbenzidine substrate (Sigma-Aldrich) for color development. The reaction was stopped by adding 1 M H3PO4, and the absorption at OD 450 nm was determined by a Titertek Multiskan ELISA reader.
FACS analysis
A total of 1 x 106 Daudi or BALENLM-17 (Bal-17) cells, a mouse B cell line from J. T. Kung (Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan), were suspended in 100 µl 5% FBS in HBSS (Life Technologies) with 1 µl antisera from immunized mice. For A01, cells of both B cell lines were incubated with 100 µl cultured supernatants. After staining on ice for 30 min, cells were washed twice with HBSS and suspended in 100 µl 5% FBS in HBSS containing FITC-conjugated goat anti-muIgG Abs (BD Biosciences, San Diego, CA). For the analysis of fluorescent staining, cells were analyzed with a FACScan instrument (BD Biosciences) and the CellQuest software. To investigate the in vivo effects of huIg1.Fc-huIg on B cells, 
100 µl blood was collected from each immunized mouse 1 wk after every boost immunization, and mixed with 10 ml ice-cold RBC lysis buffer (150 mM NH4Cl, 15 mM KHCO3). Peripheral blood leukocytes were washed twice with RBC lysis buffer and double stained with FITC-conjugated rat anti-mouse 
-chain (for B cells) (a gift from T. K. Kung) plus PE-conjugated goat anti-mouse CD3 (for T cells) Abs (BD Biosciences). Cells were gated on the basis of forward and side scatter analyses as well as marker expression of lymphocyte.
Assays of cell death
Splenocytes were prepared from BALB/c mice and plated into wells of 24-well plates at a cell concentration of 5 x 105/well. Cells were incubated with polyclonal rabbit anti-mouse IgM (Jackson ImmunoResearch, West Grove, PA) at 10 µg/ml as positive controls or with antisera from huIg1.Fc-huIg-treated mice at the serum dilution of 1/50 for 24 h at 37°C. For the negative controls, splenocytes were cultured with pooled normal muIgG (Sigma-Aldrich) at 10 µg/ml. After the incubation, cells from each well were harvested and stained with FITC-conjugated goat anti-mouse CD19 Ab (BD Biosciences). CD19+ B cells in treated samples were gated for analysis by flow cytometry. The fractions of cell fragments and apoptotic bodies were revealed by side scatter channel (SSC) vs forward scatter channel (FSC) plot (23). For the experiment with Bal-17 cell line, the cells were incubated for 24 h with anti-IgM, anti-Ig antisera, or muIgG at the conditions as described above. Because phosphatidylserine expression on cell surface has been suggested as a good marker for apoptotic cells (24, 25), the treated Bal-17 cells were stained with 5 µl FITC-conjugated annexin V (BD Biosciences) and analyzed by flow cytometry.
Preparation of mouse monoclonal hybridoma
Mice, 1 wk after the fourth immunization with huIg1.Fc-huIg, were sacrificed by cervical dislocation, and splenocytes were collected from the spleen. After washing with DMEM (Life Technologies), the splenic cells were fused with NS0 cells (ATCC) at a ratio of 5:1 with 40% PEG4000 (Merck, Whitehouse Station, NJ). Hybridoma cells were selected with hypoxanthine/aminopterin/thymidine medium (Life Technologies) containing 10% FBS for 14 days, and supernatants from resultant clones were screened by ELISA using MBP-huIg and MBP-muIg proteins as solid-phase Ags. A specific clone, A01, identified to recognize both huIg and muIg, was used in the present study.
Secondary structure predictions of Ig and Ig
The Profile Network from HeiDelberg program (26, 27) based on neural networks was used to predict the secondary structure of huIg and huIg. The program gives the percentage of propensity of each residue to be a part of an -helix, a -strand, or a loop. We considered propensities greater than 50% as significant. For both molecules, the predicted helical regions have 
30% propensity, whereas the predicted -strand regions have >50% propensity; thus, huIg and huIg are predicted to be all molecules. Predicted strand regions with less than four contiguous residues were not considered as -strands. huIg was predicted to possess seven -strands: C4-K11 (1), A24-H30 (2), N37-R42 (3), E53-G56 (4), G63-Q68 (5), G75-Q83 (6), and T95-R100 (7), while huIg was predicted to have eight -strands: D8-N12 (1), V36-Y41 (2), S50-Q55 (3), Q63-G69 (4), L79-Q85 (5), G93-C101 (6), S105-Q109 (7), and T113-G119 (8).
|
Results
|
|---|
Comparative immunizations with huIg1.Fc-huIg and huIg1.Fc-huIg
As shown in Fig. 1A, antisera from BALB/c mice immunized with huIg1.Fc-huIg four times contained IgG that showed specific reactivity with MBP-huIg fusion protein, MBP-huIg, and muIg in MBP-muIg. In parallel experiments, antisera from huIg1.Fc-huIg-immunized mice showed specific reactivity to MBP-huIg fusion protein, MBP-huIg, but not to autologous Ig in MBP-muIg (Fig. 1B). Also, antisera from mice immunized with huIgG.Fc could only react with huIgG (Fig. 1C), and sera from mice treated with PBS showed weak or undetectable reactivity to all Ags used in the ELISA (Fig. 1D). All antisera or sera tested in Fig. 1, A–D, did not show significant reactivity against MBP protein. The data show that auto/cross-reactive Abs against autologous Ig were induced in the mice after immunization with an Fc fusion protein of huIg. In contrast, such cross-reactive Abs against autologous Ig could not be productively induced by immunization with a similar Fc fusion protein of huIg.
View larger version (19K):
[in this window]
[in a new window]
|
FIGURE 1. Specific reactivity with muIg and muIg of the antisera from mice immunized with huIg or huIg. The solid-phase Ags were: 1, MBP; 2, MBP-huIg; 3, MBP-muIg; 4, MBP-huIg; 5, MBP-muIg; and 6, huIgG (huIgG). Each of 4 groups of 15 mice was injected with A, huIg1.Fc-huIg fusion protein; B, huIg1.Fc-huIg fusion protein; C, huIgG.Fc (as a positive control); or D, PBS (as a negative control). The values (OD 450 nm absorption at 1/1000 dilution of antisera after the fourth immunization) were means ± SD, of results from 15 mice.
|
|
Reactivity of antisera with B cells
To characterize whether the antisera from mice immunized with huIg1.Fc-huIg or huIg1.Fc-huIg could recognize the native structure of Ig or Ig, cells of two B cell lines were incubated with the antisera in a routine staining procedure and examined by fluorescence flow cytometry. The B cell lines were Daudi, an IgM-expressing human B cell line, and BALENLM-17 (Bal-17), an IgM-expressing mouse B cell line. As shown in Fig. 2, antisera from huIg1.Fc-huIg-immunized mice could stain both Daudi and Bal-17 cells (A), while those from huIg1.Fc-huIg-immunized mice could bind to Daudi cells, but weakly to Bal-17 cells (B). These results suggest that the IgG induced by the Fc fusion protein of huIg could react with both huIg and muIg, while the IgG induced by a similar Fc fusion protein of huIg could react with huIg, but not with muIg. These findings are consistent with the ELISA results in Fig. 1.
Reduction of peripheral blood B cells in the immunized mice
Because the antisera from hu1.Fc-huIg-immunized mice could stain cells of Bal-17 B cell line, the induced auto/cross-reactive Abs might affect the B cell level in these mice. Peripheral blood was collected from treated mice 1 wk after each immunization, and equal aliquots were placed into tubes containing RBC lysis buffer, and the resultant blood leukocytes were subjected to flow cytometric analysis, in which the lymphocytes were double stained with anti--chain Ab and anti-CD3 Ab (Fig. 3A). After the fourth injection, B cells accounted for 14% among total peripheral blood leukocytes (1403 ± 485/104 total events) in PBS-treated control mice, while they accounted for 16% of total blood leukocytes (1607 ± 365/104 total events) in mice immunized with huIgG.Fc, a conventional foreign Ag. B cells in the peripheral blood leukocytes decreased upon sequential immunizations with hu1.Fc-huIg, 2.3% among total blood leukocytes (232 ± 92/104 total events), in mice after the fourth immunization with hu1.Fc-huIg, corresponding to decreases of 83 and 86%, respectively, of the levels of B cells in mice treated with PBS or huIgG.Fc (Fig. 3B).
View larger version (40K):
[in this window]
[in a new window]
|
FIGURE 3. Reduction of B cell level in mice making auto/cross-reactive anti-Ig Abs. Peripheral blood was collected from mice 1 wk after each immunization, and equal aliquots were subjected to flow cytometric analysis. B and T cells were distinguished by double staining with FITC-conjugated rat anti-mouse -chain (for B cells) and PE-conjugated goat anti-mouse CD3 (for T cells) Abs. A, The B cell levels in the mononuclear cells were shown in dot blots after the fourth immunization. B, The B cell numbers in total blood leukocytes after the fourth treatment with PBS and the fourth immunization with huIgG.Fc, and the second, third, and fourth immunizations with huIg1.Fc-huIg were shown in bars. The data were means ± SDs of results from 15 mice. *, p < 0.05, compared with huIgG.Fc control; **, p < 0.01, compared with huIgG.Fc control.
|
|
To address whether the anti-Ig autoantibody still attached to the surface Ig might influence the staining with an anti- Ab, a fluorescence flow cytometric experiment using cells of a human B cell line, Daudi, which express IgM on the surface, was performed. The cells were first stained with the antisera from mice immunized with huIg, coupled with FITC-labeled goat IgG anti-muIgG Ab. Those FITC-stained Daudi cells were then tested by incubation with PE-labeled rabbit IgG anti-human Ab. Clearly, all those stained with FITC could be doubly stained with PE (data not shown). The results indicated that the anti-Ig autoantibody still attached to the surface Ig did not influence the staining of the -chain in the BCR with an anti- Ab.
In vitro assay of cell death caused by antisera from the immunized mice
To investigate whether the reduction of B cells in hu1.Fc-huIg-treated mice was due to apoptosis as reported with B cell lines treated with cross-linking Abs against surface IgM (28, 29, 30), splenocytes and Bal-17 cells were incubated with anti-Ig antisera from treated mice for 24 h and analyzed for possible apoptotic results. In the in vitro assays, anti-muIgM Abs were used as positive controls and pooled normal muIgG were used as negative controls. The results showed that after incubation with the anti-Ig antisera, cell fragments or apoptotic bodies of B cells in the CD19+ subpopulation of splenocytes increased, as analyzed by SSC vs FSC plot using fluorescence flow cytometry (gated R1 regions in Fig. 4A). Among the Bal-17 cells incubated with anti-Ig antisera for 24 h, annexin V+ apoptotic cells increased, as compared with the cells incubated with muIgG (gated R2 regions in Fig. 4A). The results indicated that the extents of apoptosis in the splenocytes and Bal-17 cells induced by anti-Ig are 84 and 44%, respectively, of those induced by anti-IgM (Fig. 4B). Thus, the reduction of B cells in the peripheral blood of the hu1.Fc-huIg-immunized mice is partly due to cell death triggered by anti-Ig autoantibody cross-linking of BCR complexes.
View larger version (41K):
[in this window]
[in a new window]
|
FIGURE 4. Cell death induced by anti-Ig antisera from mice immunized with huIg1.Fc-huIg. In wells of 24-well culture plates, 5 x 105 splenocytes and Bal-17 cells were incubated with anti-Ig antisera (1/50), anti-IgM (10 µg/ml, as positive controls), or normal pooled muIgG (10 µg/ml, as negative controls) for 24 h at 37°C. The cells were then stained with FITC-conjugated anti-CD19 Ab (for slpenocytes) or FITC-conjugated annexin V (for Bal-17 cells) and analyzed by flow cytometry. A, Cell fragments and apoptotic bodies of CD19+ B cells in slpenocytes (gated R1 regions) were analyzed by SSC vs FSC plot, whereas apoptotic bodies in Bal-17 cells (annexin V+ cells, gated R2 regions) were analyzed by SSC vs FL1 plot. B, The numbers represent percentage of increases in apoptotic bodies, which are normalized to that caused by anti-IgM-positive control (100%).
|
|
Suppression of Ab response to OVA in the immunized mice
To study whether the reduction of B cell level affected the Ab response in the mice immunized with hu1.Fc-huIg, they were injected with an irrelevant Ag, OVA, after the fourth immunization with hu1.Fc-huIg. As shown in Fig. 5, the Ab response against OVA in hu1.Fc-huIg-immunized mice was reduced to 55% of control mice injected with PBS in an ELISA after the third immunization with OVA. In contrast, Ab response against OVA in huIgG.Fc-immunized mice was similar to that in control mice. These results indicate that eliciting auto/cross-reactive Abs against self Ig by immunizing with hu1.Fc-huIg reduced the B cell levels in total lymphocytes, which, in turn, led to the reduction in B cell-related immune responses.
Identification of a dual-specific mAb
A specific mAb, A01, with auto/cross-reactivity between huIg and muIg has been screened out from the huIg1.Fc-huIg-immunized mice. A01 could recognize both MBP-huIg and MBP-muIg molecules, but did not react with MBP or huIgG alone (Fig. 6A). A01 could also bind to the surface of Daudi cells and Bal-17 cells (Fig. 6B). These data show the auto/cross-reactivity of A01 to both huIg and muIg.
Predictions of possible antigenic sites on Ig and Ig
Because the induced anti-Ig Abs from huIg1.Fc-huIg-immunized mice recognize not only huIg, but also muIg (Figs. 1 and 2), the cross-reactive binding site on muIg should be similar to that on huIg. This could be the case if the residues comprising the Ab binding site on huIg and muIg are identical. Such a binding site on Ig should be solvent exposed to be recognized by the Ab. It should also be composed of at least 10 residues as suggested by x-ray structures of Ab-Ag complexes, showing an average of 12–15 residues of the Ag with an interface area between 1400 and 1700 Å2 (31, 32). In contrast to the anti-Ig Abs, the induced anti-Ig Abs from huIg1.Fc-huIg-immunized mice recognize only huIg (Figs. 1 and 2), indicating different epitope structures on muIg and huIg. Hence, the residues comprising the Ab binding sites on huIg and muIg would differ.
Sequence analyses, as described in Materials and Methods, predict huIg to be a classical member of the Ig-fold family (33, 34). The predicted -strands correspond to average negative antigenicity values (except 4) and/or low solvent-accessible surface propensities (Fig. 7A), and are thus unlikely antigenic-site candidates. Two loops in huIg, E84-G94 and Q101-R117, were predicted to be solvent exposed and antigenic, consisting of >10 residues. Because the C-terminal loop contains 14 residues of huIg (P104-R117) that are identical with those of muIg (P102-R115), it is likely to form an auto/cross-reactive binding site (Fig. 7B).
In analogy to huIg, huIg was predicted as an eight--stranded molecule with an Ig-fold. None of the -strands are contiguously exposed on the molecular surface, and are thus unlikely antigenic-site candidates (Fig. 8A). Two loops in huIg, P13-T35 and F121-D134, were predicted to be solvent exposed and antigenic. Within the first antigenic loop, P13-C18 and P25-F34 were identified as the immunodominant regions by antigenicity prediction. The P13-C18 region consists of less than 10 residues, and is thus unlikely to form an auto/cross-reactive binding site. The P25-F34 region contains four substitutions, three of which are conservative; however, the fourth substitution of a F34 in huIg to a S34 in muIg could affect Ab binding. Based on the sequence alignment, both of these two antigenic loops do not have a linear epitope containing 10 or more residues that are identical between huIg and muIg (Fig. 8B). Although the last loop region of huIg and muIg is identical except for two residues, the substitution of neutral A124 and Q128 in huIg to the charged D123 and R127 residues in muIg, respectively, would probably affect Ab binding.
Identification of the auto/cross-reactive antigenic sites on Ig molecules
To verify that the C-terminal loop of muIg or huIg is sufficient for Ab recognition, two peptide segments corresponding to a hydrophilic/antigenic sequence of huIg (huIg-1: R81-L97) and muIg (muIg-1: Q80-L95) were synthesized for the control experiments (Table I). Two other peptide segments corresponding to the C-terminal loop of huIg (huIg-2: V99-R117) and muIg (muIg-2: V97-R115), respectively, were also synthesized for the test. Antisera from huIgG.Fc-treated mice showed weak reactivity toward all four peptides (Fig. 9A). In contrast, antisera from huIg1.Fc-huIg-immunized mice could interact strongly with all peptides except muIg-1, indicating that the antigenic determinant of the C-terminal huIg region (huIg-2) suffices to induce auto/cross-reactivity between huIg and muIg against muIg-2 peptide (Fig. 9A).
In addition, A01 could recognize both C-terminal huIg-2 and muIg-2 peptides, but its specific reactivity toward the N-terminal huIg-1 and muIg-1 peptides was not statistically significant as compared with the irrelevant mAb (Fig. 9B). Blocking with MBP-huIg fusion protein could reduce the specific reactivity of A01 mAb or antisera from huIg1.Fc-huIg-immunized mice toward C-terminal huIg-2 and muIg-2 peptides (Fig. 9). Blocking with the N-terminal huIg-1 peptide did not affect the binding reactivity of anti-Ig antisera or A01 mAb toward huIg-2 and muIg-2 peptides (Fig. 9); blocking with huIg-2 peptide did not affect the binding reactivity of anti-Ig antisera toward huIg-1 peptide (Fig. 9A). These data show that the predicted Ig C-terminal loop could be the auto/cross-reactive anti-Ig binding site.
|
Discussion
|
|---|
The long-term goal of the present study is to understand how specific autoreactive Ab response against self Ig and Ig on B cells can be elicited effectively to down-regulate or purge B cells. For any of such processes to work, Th cells that assist B cells bearing receptors for respective self Ags in making Abs must be activated. In one such approach investigated in our earlier studies, we coupled the extracellular portion of autologous Ig to the Fc of a foreign IgG, viz, muIg-hu4.Fc (22). The foreign Fc provided the peptide segments for recognition by Th cells to activate autoreactive B cells.
In this study, we explored in a mouse model the utility of molecular mimicry between huIg and muIg in generating auto/cross-reactive Ab responses against self Ig. In this case, it was anticipated that certain peptide segments in the foreign Ig would provide antigenic epitopes for recognition by Th cells. At the same time, the sufficient conformational resemblance between certain parts of huIg and muIg would enable the induced Abs that react with the introduced foreign Ig to bind to the self Ig. The results in this study showed that the molecular mimicry between huIg and muIg enabled huIg to induce Abs effectively that cross-react with self Ig in mice. The results also showed that the induced auto/cross-reactive Abs can down-regulate B cell level and suppress B cell-related humoral immune responses against an irrelevant Ag.
Previous studies showed that proteins with overall 85–95% sequence identity to proteins from another species could provide sufficient antigenic specificity to induce auto/cross-reactive Abs (8, 9, 10). The results from this study show that a foreign Ig with moderate sequence homology to self (58%) can also break down the immunological tolerance and trigger auto/cross-reactivity in mice. In addition, different antigenic properties between Ig and Ig have been shown in this study, even though the sequence identity between huIg and muIg is similar to that between huIg and muIg. These results suggest that antigenic structure (and thus local sequence identity) may play a more important role than overall sequence homology in inducing cross-reactivity to self Ag.
A 14-aa peptide segment comprising part of the C-terminal loop of huIg was predicted and verified to be the antigenic site inducing the observed cross-reactivity in mice. The sequence of this segment, viz, PRPFLDMGEGTKNR, is identical between huIg and muIg. Interestingly, this sequence, excluding the last residue, is identical with the corresponding sequence in bovine, the only other species whose Ig sequence has been reported (Table II). Furthermore, this sequence appears to be unique to Ig, as no other significantly homologous sequence in the nonredundant SWISS-PROT protein sequence database (36) was found. Because Ig is expressed on the cell surface from an early stage of B cell development through the terminal differentiated stage of plasma cells (37, 38), this antigenic epitope may be exploited as an attractive target for B cell.
In our work, the induced anti-Ig autoantibodies in the sera of mice immunized with huIg could bind to cells of Bal-17 cell line from BALB/c mice (Fig. 2), as well as splenic B cells, which were also stained with an Ab against CD19 (unpublished observation). The sera from mice not immunized with huIg could not stain Bal-17 cells. These results clearly show that the putative target epitope on Ig is accessible to the anti-Ig autoantibody despite its proximity to the cell membrane. Recent studies have also shown that mAbs specific for the -chain of IgE.Fc receptors on mast cells and 
-chain of mIgE on B cells could recognize their respective epitopes proximal to the cell membrane (39, 40). In these two cases, the mAbs can bind with high affinity not only to recombinant target proteins, but also to the target cells.
Several studies have shown that Ig can be detected on the surface of neoplastic cells derived from B cells at various immature and mature stages (37, 41, 42). Thus, developing mAbs against Ig for potential use in targeting B cell tumors has been pursued (43). Furthermore, inducing active autoreactivity against disease-related autologous Ags using recombinant proteins capable of inducing cross-reactivity against self Ags has been suggested. Disis et al. (10) showed that human HER-2/neu protein immunization could circumvent tolerance to a self tumor Ag neu in treated rats. Apostolopoulos et al. (44, 45) also showed that peptide mimics of a tumor Ag could induce CTL to kill cancerous cells. Therapeutic vaccines for B cell lymphoma using the extracellular domain or a key antigenic segment of Ig, as identified in this study, would seem logical approaches.
In our previous studies of inducing autoreactivity against self Ig, the proportion of B cells in total lymphocytes dropped to 35% after the second immunization with muIg-hu4.Fc, but it did not decrease further with subsequent immunizations. Auto/cross-reactive Abs induced by hu1.Fc-huIg immunization produced more immunological effects on B cells than muIg-hu4.Fc, although these Abs presented similar binding intensity on B cell surface as shown by FACS. After multiple immunizations with hu1.Fc-huIg, the B cell levels were continuously reduced to 12% (Fig. 3B). The different regulatory effects after Ab binding may be due in part to the differential roles of Ig and Ig in B cell signaling. Several studies have shown that Ig and Ig play redundant regulatory roles, but interact with distinct cytoplasmic effectors during B cell development (46, 47, 48). Clark et al. (46) and Kim et al. (47) showed that Lyn and Fyn kinases, which play important roles in the first intracellular signal in B cells, associate preferentially and activate tyrosine kinases more readily with Ig than with Ig. Hence, Ig, which exhibits antigenic mimicry, seems to be a more suitable Ag than Ig in regulating B cell levels.
Clinical studies based on mAb therapeutics have shown that some mAbs specific for certain B cell-specific Ags, such as Rituximab (anti-CD20) (49), are efficacious in treating certain types of B cell lymphoma. Those mAbs cannot only recruit effector functions, but also mediate apoptotic mechanisms (49, 50, 51). The mAbs can cross-link target molecules on the cell surface and deliver transmembrane signals that control cell division (30, 50, 51). Glennie and colleagues (50, 51, 52) concluded that signaling activity on tumor cells appears to be more important than the recruitment of effectors. In our studies, apoptosis was found to occur after eliciting of anti-Ig autoreactive Abs (Fig. 4), suggesting that the induced Abs can cause the cross-linking of signal-transducing Ig.
There are possibly several advantages of using an immunogen to induce active autoreactive response against an autologous disease-associated component over a passively administered Ab. The ability to induce an active, specific immune response eliminates the need for passive infusions of large quantities of Abs. Therapeutic vaccines probably do not need to be administered repeatedly. The induced autoreactive Abs are fully immune tolerated, present in the body with long t1/2, and capable of recruiting effectors, such as complement and cytotoxic cells. The induced autoantibodies are polyclonal and may recognize several epitopes, conferring stronger effects on the targeted cells. The efficacy of using the cross-reactive foreign Ig or epitope-based vaccines for treating B cell lymphoma is now under investigation in murine models.
Other immune responses, such as cytokine secretion or activation of autoreactive T cells, may also be involved during hu1.Fc-huIg immunizations. Activation of autoreactive B or T cells is known to be a necessary, but not sufficient step in developing of strong autoimmune responses (53). Several factors, such as gene expression changes in cytokine profiles and mIg signaling, may participate in inducing a sufficient degree of clonal expansion (54, 55). It will be interesting to investigate the cytokine profile in those mice immunized with the cross-reactive huIg. The results should enable us to augment the effectiveness of the Ig immunogen in down-regulating or purging B cells with proper cytokines, such as IFN-, GM-CSF, or IL-2. In contrast, B cells specific for linear determinants distributed along the homologous foreign peptides have been found to provide the stimuli required for T cell to break tolerance to resembling self epitopes (9, 11, 56). In our ensuing studies, the existence of autoreactive T cell clones will be investigated by their ability to be activated by muIg peptides in the presence of APC or appropriate growth factors.
|
Acknowledgments
|
|---|
We thank Meek Huang, Maria Chen, and Steven Wu at National Tsing Hua University for technical assistance.
|
Footnotes
|
|---|
1 This study was supported by a grant to T.-W.C. (90-2815-c-007-068R-B) from the National Science Council, and by funds to C.L. from Institute of Biomedical Sciences, Academia Sinica. 
2 Address correspondence and reprint requests to Dr. Tse-Wen Chang, Development Center for Biotechnology, 81 Chang Hsing Street, Taipei 106, Taiwan. E-mail address: twchang{at}mail.dcb.org.tw
3 Abbreviations used in this paper: BCR, B cell Ag receptor; FSC, forward scatter channel; hu, human; MBP, maltose-binding protein; mIg, membrane-bound Ig; mu, mouse; SSC, side scatter channel.
Received for publication July 25, 2002.
Accepted for publication November 15, 2002.
|
References
|
|---|
- Beck, K., J. H. Larsen, J. M. Hansen, J. Nerup. 1974. Yersinia enterocolitica infection and thyroid disorder.. Lancet 2:951.
- Menser, M. A., J. M. Forrest, R. D. Bransby. 1978. Rubella infection and diabetes mellitus. Lancet 1:57.[Medline]
- Kurtzke, J. F.. 1993. Epidemiologic evidence for multiple sclerosis as an infection. Clin. Microbiol. Rev. 6:382.[Abstract/Free Full Text]
- Rowley, D., C. R. Jenkin. 1962. Antigenic cross-reaction between host and parasite as a possible cause of pathogenicity. Nature 193:151.[Medline]
- Damian, R. T.. 1964. Molecular mimicry: antigen sharing by parasite and host and its consequences. Am. Nat. 98:120.
- Oldstone, M. B. A.. 1990. Molecular mimicry and autoimmune disease. Cell 50:819.
- Theofilopoulos, A. N.. 1995. The basis of autoimmunity. I. Mechanisms of aberrant self-recognition. Immunol. Today 16:90.[Medline]
- Kazim, A. L., M. Z. Atassi. 1977. Antibodies against protein antigenic sites that are identical in the homologous protein of the immunized animal: autoreactivity in rabbits of antibodies to sperm-whale myoglobin. Biochim. Biophys. Acta 494:277.[Medline]
- Mamula, M. J., R. H. Lin, C. A. Janeway, J. A. Hardin. 1992. Breaking T cell tolerance with foreign and self co-immunogens: a study of autoimmune B and T cell epitopes of cytochrome c. J. Immunol. 149:789.[Abstract]
- Disis, M. L., F. M. Shiota, M. A. Cheever. 1998. Human HER-2/neu protein immunization circumvents tolerance to rat neu: a vaccine strategy for "self" tumor antigens. Immunology 93:192.[Medline]
- Lin, R. H., M. J. Mamula, J. A. Hardin, C. J. Janeway. 1991. Induction of auto-reactive B cells allows priming of auto-reactive T cells. J. Exp. Med. 173:1433.[Abstract/Free Full Text]
- Thurau, S. R., M. Diedrichs-Mohring, H. Fricke, S. Arbogast, G. Wildner. 1997. Molecular mimicry as a therapeutic approach for an autoimmune disease: oral treatment of uveitis patients with an MHC-peptide crossreactive with autoantigen-first results. Immunol. Lett. 57:193.[Medline]
- Apostolopoulos, V., M. S. Sandrin, I. F. C. McKenzie. 2000. Mimics and cross reactions of relevance to tumor immunotherapy. Vaccine 18:268.
- Reth, M.. 1992. Antigen receptors on B lymphocytes. Annu. Rev. Immunol. 10:97.[Medline]
- Venkitaraman, A. G., G. T. Willians, P. Dariavach, M. S. Neuberger. 1991. The B-cell antigen receptor of the five immunoglobulin classes. Nature 352:777.[Medline]
- Sanchez, M., Z. Misulovin, A. L. Burkhardt, S. Mahajan, T. Costa, R. Franke, J. B. Bolen. 1993. Signal transduction by immunoglobulin is mediated through Ig and Ig. J. Exp. Med. 178:1049.[Abstract/Free Full Text]
- Kurosaki, T.. 1997. Molecular mechanisms in B cell antigen receptor signaling. Curr. Opin. Immunol. 9:309.[Medline]
- Minegishi, Y., E. CoustanSmith, L. Rapalus, F. Ersoy, D. Campana, M. E. Conley. 1999. Mutations in Ig (CD79a) result in a complete block in B-cell development. J. Clin. Invest. 104:1115.[Medline]
- Yu, L. M., T. W. Chang. 1992. Human mb-1 gene: complete cDNA sequence and its expression in B cells bearing membrane Ig of various isotypes. J. Immunol. 148:633.[Abstract]
- Ha, H., H. Kubagawa, P. D. Burrows. 1992. Molecular cloning and expression pattern of a human gene homologous to the murine mb-1 gene. J. Immunol. 148:1526.[Abstract]
- Muller, B., L. Cooper, C. Terhorst. 1992. Cloning and sequencing of the cDNA encoding the human homologue of the murine immunoglobulin-associated protein B29. Eur. J. Immunol. 22:1621.[Medline]
- Sheu, J. J., J. Huang, T. W. Chang. 2002. Inducing specific reactivity against B cells in mice by immunizing with an Fc fusion protein containing self-Ig. Cancer Immunol. Immunother. 51:145.[Medline]
- Simak, J., K. Holada, J. G. Vostal. 2002. Release of annexin V-binding membrane microparticles from cultured human umbilical vein endothelial cells after treatment with camptothecin. BMC Cell Biol. 3:11.[Medline]
- Vermes, I., C. Haanen, H. Steffens-Nakken, C. Reutelingsperger. 1995. A novel assay for apoptosis: flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labelled annexin V. J. Immunol. Methods 184:39.[Medline]
- Koopman, G., C. P. Reutelingsperger, G. A. Kuijten, R. M. Keehnen, S. T. Pals, M. H. van Oers. 1994. Annexin V for flow cytometric detection of phosphatidylserine expression on B cells undergoing apoptosis. Blood 84:1415.[Abstract/Free Full Text]
- Rost, B., C. Sander. 1993. Prediction of protein secondary structure at better than 70% accuracy. J. Mol. Biol. 232:584.[Medline]
- Rost, B., C. Sander. 1994. Combining evolutionary information and neural networks to predict protein secondary structure. Proteins 19:55.[Medline]
- Carey, G. B., D. Donjerkovic, C. M. Mueller, S. Liu, J. A. Hinshaw, L. Tonnetti, W. Davidson, D. W. Scott. 2000. B-cell receptor and Fas-mediated signals for life and death. Immunol. Rev. 176:105.[Medline]
- Mayumi, M., Y. Ohshima, D. Hata, K. M. Kim, T. Heike, K. Katamura, K. Furusho. 1995. IgM-mediated B cell apoptosis. Crit. Rev. Immunol. 15:255.[Medline]
- Marches, R., E. Racila, T. F. Tucker, L. Picker, P. Mongini, R. Hsueh, E. S. Vitetta, R. H. Scheuermann, J. W. Uhr. 1995. Tumor dormancy and cell signalling. III. Role of hypercrosslinking of IgM and CD40 on the induction of cell cycle arrest and apoptosis in B lymphoma cells. Ther. Immunol. 2:125.[Medline]
- Abola, E. E., J. L. Sussman, J. Prilusky, N. O. Manning. 1997. Protein Data Bank archives of three-dimensional macromolecular structures. Methods Enzymol. 277:556.[Medline]
- Noskov, S. Y., C. Lim. 2001. Free energy decomposition of protein-protein interactions. Biophys. J. 81:737.[Medline]
- Halaby, D. M., A. Poupon, J. Mornon. 1999. The immunoglobulin fold family: sequence analysis and 3D structure comparisons. Protein Eng. 12:563.[Abstract/Free Full Text]
- Harpaz, Y., C. Chothia. 1994. Many of the immunoglobulin superfamily domains in cell adhesion molecules and surface receptors belong to a new structural set which is close to that containing variable domains. J. Mol. Biol. 238:528.[Medline]
- Jameson, B. A., H. Wolf. 1988. The antigenic index: a novel algorithm for predicting antigenic determinants. Comput. Appl. Biosci. 4:181.[Abstract/Free Full Text]
- Junker, V. L., R. Apweiler, A. Bairoch. 1999. Representation of functional information in the SWISS-PROT data bank. Bioinformatics 15:1066.[Abstract/Free Full Text]
- Mason, D. Y., J. L. Cordell, A. G. D. Tse, J. J. M. van Dongen, C. J. M. van Noesel, K. Micklem, K. A. F. Pulford, F. Valensi, W. M. Comans-Bitter, J. Borst, K. C. Gatter. 1991. The IgM-associated protein mb-1 as a marker of normal and neoplastic B cells. J. Immunol. 147:2474.
- Palacios, R., J. Samaridis. 1992. Fetal liver pro-B and pre-B lymphocyte clones: expression of lymphoid-specific genes, surface markers, growth requirements, colonization of the bone marrow, and generation of B lymphocytes in vivo and in vitro. Mol. Cell. Biol. 12:518.[Abstract/Free Full Text]
- Nechansky, A., M. W. Robertson, B. A. Albrecht, J. R. Apgar, F. Kricek. 2001. Inhibition of antigen-induced mediator release from IgE-sensitized cells by a monoclonal anti-FcRI -chain receptor antibody: implications for the involvement of the membrane-proximal -chain region in FcRI-mediated cell activation. J. Immunol. 166:5979.[Abstract/Free Full Text]
- Chen, H. Y., F. T. Liu, M. H. Hou, J. Huang, B. B. Sharma, T. W. Chang. 2002. Monoclonal antibodies against the CmX domain of human membrane-bound IgE and their potential use for targeting IgE-expressing B cells. Int. Arch. Allergy Immunol. 128:315.[Medline]
- Mason, D. Y., J. L. Cordell, J. B. Brown, M. Jones, K. Pulford, E. Jaffe, E. Ralfkiaer, F. Dallenbach, H. Stein, S. Pileri, K. C. Gatter. 1995. CD79a: a novel marker for B-cell neoplasms in routinely processed tissue samples. Blood 88:1453.
- Verschuren, M. C., W. M. Comans-Bitter, C. A. Kapteijn, D. Y. Mason, G. S. Brouns, J. Borst, H. G. Drexler, J. J. van Dongen. 1993. Transcription and protein expression of mb-1 and B29 genes in human hematopoietic malignancies and cell lines. Leukemia 7:1939.[Medline]
- Zhang, L., R. R. French, H. T. Chan, T. L. O’Keefe, M. S. Cragg, J. M. Power, M. J. Glennie. 1995. The development of anti-CD79 monoclonal antibodies for treatment of B cell neoplastic disease. Ther. Immunol. 2:191.[Medline]
- Apostolopoulos, V., J. Matsoukas, M. Plebanski, T. Mavromoustakos. 2002. Applications of peptide mimetics in cancer. Curr. Med. Chem. 9:411.[Medline]
- Apostolopoulos, V., S. A. Lofthouse, V. Popovski, G. Chevanayagam, M. S. Sandrim, I. F. C. McKenzie. 1998. Peptide mimics of a tumor antigen induce functional cytotoxic T cells. Nat. Biotechnol. 16:276.[Medline]
- Clark, M. R., K. S. Campbell, A. Dazlauskas, S. A. Johnson, M. Hertz, T. A. Potter, C. Pleiman, J. C. Cambier. 1992. The B cell antigen receptor complex: association of Ig and Ig with distinct cytoplasmic effectors. Science 258:123.[Abstract/Free Full Text]
- Kim, K. M., G. Alber, P. Weiser, M. Reth. 1993. Differential signaling through the Ig- and Ig- components of the B cell antigen receptor. Eur. J. Immunol. 132:125.
- Choquet, D., G. Ku, S. Cassard, B. Malissen, H. Korn, W. H. Fridman, C. Bonnerot. 1994. Different patterns of calcium signaling triggered through two components of the B lymphocyte antigen receptor. J. Biol. Chem. 269:6491.[Abstract/Free Full Text]
- Shan, D., J. A. Ledbetter, O. W. Press. 1998. Apoptosis of malignant human B cells by ligation of CD20 with monoclonal antibodies. Blood 91:1644.[Abstract/Free Full Text]
- Cragg, M. S., R. R. French, M. J. Glennie. 1999. Signaling antibodies in cancer therapy. Curr. Opin. Immunol. 11:541.[Medline]
- Glennie, M. J., P. W. M. Johnson. 2000. Clinical trials of antibody therapy. Immunol. Today 21:403.[Medline]
- Tutt, A. L., R. R. French, T. M. Illidge, J. Honeychurch, H. M. McBride, C. A. Penfold, D. T. Fearon, R. M. Parkhouse, G. G. Klaus, M. J. Glennie. 1998. Monoclonal antibody therapy of B cell lymphoma: signaling activity on tumor cells appears more important than recruitment of effectors. J. Immunol. 161:3176.[Abstract/Free Full Text]
- Verhasselt, V., M. Goldman. 2001. From autoimmune responses to autoimmune disease: what is needed?. J. Autoimmun. 16:327.[Medline]
- Ray, S. K., C. Putterman, B. Diamond. 1996. Pathogenic autoantibodies are routinely generated during the response to foreign antigen: a paradigm for autoimmune disease. Proc. Natl. Acad. Sci. USA 93:2019.[Abstract/Free Full Text]
- Santamaria, P.. 2001. Effector lymphocytes in autoimmunity. Curr. Opin. Immunol. 13:663.[Medline]
- Liang, B., M. J. Mamula. 2000. Molecular mimicry and the role of B lymphocytes in the processing of autoantigens. Cell. Mol. Life Sci. 57:561.[Medline]
Top
Abstract
Introduction
