Improved immunogenicity of a self tumor antigen by covalent linkage to CD40 ligand
Article first published online: 7 NOV 2003
Copyright © 2003 Wiley-Liss, Inc.
International Journal of Cancer
Volume 108, Issue 5, pages 696–703, 20 February 2004
How to Cite
Huang, H.-I., Wu, P.-Y., Teo, C.-Y., Chen, M.-N., Chen, Y.-C., Silin, D. and Tao, M.-H. (2004), Improved immunogenicity of a self tumor antigen by covalent linkage to CD40 ligand. Int. J. Cancer, 108: 696–703. doi: 10.1002/ijc.11612
- Issue published online: 19 DEC 2003
- Article first published online: 7 NOV 2003
- Manuscript Accepted: 3 SEP 2003
- Manuscript Revised: 18 JUL 2003
- Manuscript Received: 28 APR 2003
- National Science Council. Grant Number: NSC90-2323-B-001-003
- National Health Research Institutes, Taiwan. Grant Number: DOH90A1-PPLABAD01
- CD40 ligand;
- tumor vaccine;
- B-cell lymphoma;
The interaction between the CD40 ligand (CD40L) and CD40 on antigen-presenting cells (APCs) is critical in promoting humoral and cellular immune responses. Agonistic anti-CD40 monoclonal antibody and soluble CD40L can act as powerful adjuvants to promote vaccination, but usually require repeated high-dose treatment. In this study, we demonstrate that the adjuvant effect of CD40L can be greatly improved by directly linking the antigen to CD40L. We constructed a fusion protein (Id-CD40L) consisting of the extracellular domain of CD40L and the idiotype (Id) protein, a weakly immunogenic tumor-specific antigen derived from the murine 38C13 B-cell lymphoma. The soluble Id-CD40L fusion protein retained CD40 binding activity and stimulated CD80 and CD86 upregulation and interleukin (IL)-12 production by macrophages. Immunization of mice with Id-CD40L without adjuvants resulted in high titers of anti-Id Abs dominated by the IgG1 isotype and protected the mice from subsequent lethal tumor challenge. In a dose-response study, we demonstrated that Id-CD40L elicited anti-Id antibody (Ab) responses in all immunized animals, even at a dose as low as 0.5 μg. Immunization with free Id and an IgG-CD40L fusion protein, which was identical in structure to Id-CD40L but lost the Id determinant, resulted in significant lower anti-Id responses, indicating that physical linkage between the tumor antigen and CD40L was required for the optimal immune response. These results demonstrate that fusing CD40L to a candidate antigen can greatly improve the adjuvant activity of CD40L. This approach may be useful in developing vaccines for a variety of malignant and infectious diseases. © 2003 Wiley-Liss, Inc.
CD40 ligand (CD40L), a 35 kDa type 2 membrane protein belonging to the tumor necrosis factor (TNF) family, is mainly expressed on activated CD4+ T cells and a subset of activated CD8+ T cells.1 Its receptor, CD40, belongs to the TNF receptor family and is expressed on various antigen-presenting cell (APC) populations (B cells, dendritic cells and monocytes/macrophages).1 The CD40L/CD40 costimulatory pathway plays a crucial role in regulating immune responses. Ligation of CD40 on B cells promotes B-cell proliferation, differentiation and activation, as well as rescuing B cells from apoptotic death.1, 2 Binding of CD40L to CD40 on monocytes/macrophages and dendritic cells results in the activation of these APCs by increasing the production of costimulatory molecules (CD58, CD80 and CD86), proinflammatory cytokines [interleukin (IL)-12 and TNF-α] and chemokines [IL-8 and macrophage inflammatory protein (MIP)-1α]3 and therefore plays a critical role in generating cytotoxic T lymphocytes (CTLs) and T helper (Th)1-type immune responses.4, 5, 6, 7, 8 These characteristics make CD40-stimulating agents [agonistic anti-CD40 monoclonal antibody (mAb) and soluble CD40L trimer] powerful vaccine adjuvants. CD40 ligation circumvents the requirement for Th cells in conditioning APCs both in vitro and in vivo.4, 5, 6 Triggering of CD40 in vivo using agonistic anti-CD40 mAbs considerably enhances the efficacy of peptide-based vaccines in inducing primary and memory CTLs.9, 10, 11, 12, 13 CD40 stimulation also helps promote antibody (Ab) responses to Th cell-dependent antigens.14, 15
While CD40 activation has shown promise in potentiating immune responses in a number of therapeutic settings, the regimens used usually require repeated injection with high doses of anti-CD40 mAb or CD40L trimer, which produce certain undesirable side effects. Injection of lethally irradiated mice with anti-CD40 mAb induces an acute fatal reaction, mediated, in part, by interferon (IFN)-γ.16 Injection of anti-CD40 mAb also blocks the generation of long-term B-cell memory in a thymus-dependent immune response.17 In transgenic animal studies, long-term expression of the CD40L transgene induces the development of autoantibodies,18, 19, 20 possibly by blocking apoptosis of self-reactive B cells through CD40 activation. These results demonstrate that the conditions for CD40 activation need to be optimized to circumvent these potential side effects while retaining the therapeutic benefit.
The idiotype (Id) protein is a unique tumor-specific antigen expressed on malignant B cells. Like many other tumor-specific or tumor-associated antigens, immunization with the Id protein alone elicits weak, or no, immunity. Several approaches have been used to increase its immunogenicity; these include the coupling of Id to strongly immunogenic carrier proteins, such as keyhole limpet hemocyanin or fragment C of tetanus toxin, or the genetic fusion of Id with cytokines or chemokines.21, 22, 23, 24, 25, 26, 27, 28 Since CD40 is mainly expressed on APCs and serves as a central regulator of immune responses, we reasoned that coupling Id to CD40L might target the tumor antigen more efficiently to APCs via the CD40L/CD40 interaction, and that this interaction might in turn activate APCs, resulting in the induction of a stronger immune response. To test this hypothesis, we constructed an Id-CD40L fusion protein containing Id and the extracellular domain of mouse CD40L and demonstrated that Id-CD40L retained CD40 binding activity and stimulated potent antitumor immunity in vivo, whereas coinjection of the free Id and a soluble IgG-CD40L fusion protein (TBX-CD40L), which was identical in structure to Id-CD40L but lost the Id determinant, was much less effective, even at high IgG-CD40L doses. These results demonstrate that fusing CD40L to a candidate antigen can greatly improve the adjuvant activity of CD40L, thus possibly overcoming the potential side effects associated with high-dose CD40L treatment.
MATERIAL AND METHODS
38C13 murine B-cell lymphoma is a carcinogen [7,12-dimethylbenz(a)anthracene]-induced tumor originally produced in a T-cell-depleted C3H/eB mouse.29 The transfectoma cell lines producing Id or Id-murine granulocyte-macrophage colony-stimulating factor (Id-GM-CSF) have been reported previously.24 FO (CRL-1646; American Type Culture Collection, Manassas, VA) is a murine plasmacytoma cell line negative for immunoglobulin production. BALB/3T3 (CCL-163; American Type Culture Collection) is a murine fibroblast cell line.
Production of Id-CD40L fusion protein
Murine CD40L cDNA was obtained by RT-PCR amplification of RNA derived from mouse (C57BL/6) splenocytes stimulated for 4 hr with lipopolysaccharide (Gibco, Grand Island, NY). To construct the Id-CD40L fusion protein, the CD40L gene was engineered by overlap PCR as previously described to replace its signal sequence with a sequence encoding the (Gly4Ser)3 peptide linker and to introduce a stop codon upstream of the transmembrane domain.30 The forward primer 5′-GGAGGCGGGGGCTCGCATAGAAGATTGGATAAGG-3′ and reverse primer 5′-GTTTAGCGGCCGCTCAGAGTTTGAGTAAGCCAA-3′ were used in the first-round PCR on the mouse CD40L gene. The resulting PCR fragment was used as a template to perform a second-round PCR using the same reverse primer and a second forward primer, 5′-CTTATCGATGGAGGCGGGGGCTCGGGAGGCGGGGGC TCGGGAGGCGGGGGCTCG-3′. The 5′ end of the first forward primer and the 3′ end of the second forward primer have a 15-nucleotide overlap. The final PCR product was ligated into the pCR-Blunt vector (Invitrogen, Carlsbad, CA) for sequencing. The pCR-Blunt plasmid was digested at the ClaI and NotI sites, which were included in the PCR primers, and the resulting fragment was inserted into the immunoglobulin heavy-chain plasmid, p3079,24 at the end of the CH3 exon. The resulting plasmid, p472, contained the coding region for the VH region of 38C13 Id joined to those for the complete human γ1 constant region, a (Gly4Ser)3 linker, and the extracellular domain of murine CD40L. p472 and either the 38C13 Id light-chain vector, p3077,24 or an irrelevant light-chain vector were cotransfected into FO cells to make transfectomas producing, respectively, Id-CD40L or a chain-shuffle IgG-CD40L fusion protein, TBX-CD40L. Transfectoma clones were expanded for large-scale production in Dulbecco's modified Eagle medium containing 1% low IgG fetal calf serum (HyClone Laboratories, Logan, UT) and the fusion proteins purified by protein A chromatography as previously described.31 Sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) was used to analyze the size and assembly pattern of the CD40L fusion protein.
Transfection of 3T3 cells
Murine CD40 and CD40L cDNAs were obtained by RT-PCR from lipopolysaccharide-stimulated mouse splenocytes as described above and were separately inserted into the eukaryotic expression vector, pcDNA3 (Invitrogen), under the transcriptional control of a cytomegalovirus promoter. The resulting plasmids containing the CD40 or CD40L genes were designated p517 and p518, respectively. Mouse 3T3 fibroblasts were transfected with p517 or p518 using liposomes (lipofectamine 2000; Invitrogen) according to the manufacturer's protocol. Stable clones were selected by resistance to G418 (0.5 mg/ml) and screened for expression of CD40 or CD40L by immunostaining with biotin-conjugated rat antimouse CD40 mAb (3/23; PharMingen, San Diego, CA) or hamster antimouse CD40L mAb (MR1; PharMingen). Bound biotinylated Abs were detected using FITC-conjugated avidin (Cappel/ICN, Costa Mesa, CA) and analyzed by FACSCallibur (Becton Dickinson, Mountain View, CA). Cell lines expressing high levels of CD40 (3T3/CD40) or CD40L (3T3/CD40L) were isolated by sorting on a FACStarPlus (Becton Dickinson).
Interaction of Id-CD40L with CD40
A total of 5 × 105 3T3/CD40 cells or the parental 3T3 cells were reacted at 4°C for 30 min with 10 μg/ml of Id-CD40L or Id proteins, then bound proteins were detected using FITC-labeled goat antihuman κ Ab (Cappel/ICN), followed by fluorescent-activated cell sorting (FACS) analysis.
Macrophage activation assay
Peritoneal macrophages were prepared after intraperitoneal (i.p.) injection of C3H/HeN mice with 1 ml of 2.9% thioglycollate (DIFCO, Detroit, MI), as previously described.32 To activate macrophages, cells (1 × 106 per well) in 24-well plates were treated for 20–24 hr with 10 μg/ml of hamster anti-CD40 mAb (HM40-3; PharMingen), Id, or Id-CD40L. In some wells, 0.5 μg of Fc Block (2.4G2, anti-Fcγ receptor II/III; PharMingen) was included in the culture medium to prevent Id-containing proteins from binding to macrophages via the Fcγ receptor. IL-12 levels in the culture supernatants were measured using a sandwich ELISA kit (Mouse IL-12 p70 DuoSet; R&D Systems, Minneapolis, MN). CD80 or CD86 expression was measured by incubating the macrophages with biotin-conjugated hamster antimouse CD80 (16-10A1; PharMingen) or rat antimouse CD86 (GL1; PharMingen), respectively, followed by PE-labeled avidin (Cappel/ICN) and FACSCallibur analysis. FITC-conjugated anti-Mac1 Ab (M1/70; PharMingen) was used to gate Mac1+ macrophages for analysis.
The mice were bled from the tail and the serum samples assayed for anti-Id levels using ELISA plates coated with 38C13 Id, as previously described.30 Briefly, 50 μl of serial dilutions (starting at 1:50) of the test samples was added to the wells and the plates were incubated overnight at 4°C, then bound anti-Id Ab was detected using horseradish peroxidase-conjugated goat antimouse IgG Fc (Cappel) and quantified by reference to a mixture of purified anti-38C13 Id mAbs, containing the IgG1, IgG2a and IgG2b isotypes,33 in a 2:1:1 ratio. To measure IgG1 and IgG2a anti-Id isotypes, biotin-conjugated rat antimouse IgG1 (PharMingen) or antimouse IgG2a (PharMingen) was used as the detector, followed by avidin-horseradish peroxidase (PharMingen). Endpoint titers were defined as the highest serum dilution resulting in an absorbance twice that produced using nonimmune serum with a cutoff value of 0.12. Samples below the limit of detection were assigned a value of 50, since the first dilution of the test sample was 1:50.
FACS analysis of immune sera for anti-CD40L Ab
To determine whether the immune sera contained anti-CD40L Ab, 3T3 or 3T3/CD40L cells were reacted with serum samples (undiluted) from the Id-CD40L-immunized group, then with FITC-conjugated goat antimouse IgG Ab (PharMingen), followed by FACS analysis.
Groups of female C3H/HeN mice (6–8 weeks old), obtained from the National Laboratory Animal Breeding and Research Center (Taipei, Taiwan), received 2 i.p. injections at 2-week intervals of an appropriate amount of Id or Id-containing proteins in 200 μl of PBS. Blood samples were taken 2 weeks after each injection.
Two weeks after the second injection, mice were injected i.p. or subcutaneous (s.c.) with 1 × 103 38C13 cells, a dose previously shown to kill all unprotected mice. For s.c. challenge, tumor growth was measured 2–3 times per week, and the mean volume in cubic millimeters (mm3) was approximated by using the ellipsoidal formula: length (mm) × width (mm) × height (mm) × 0.52 (derived from π/6). The mean volume and SD of each group were calculated. The mice were killed when the tumor size was > 3,000 mm.3 For i.p. challenge, the survival of challenged mice was followed. Results were analyzed for significance using Student's t-test.
Construction and characterization of Id-CD40L fusion protein
The production of the chimeric Id protein, consisting of the 38C13 V region joined to the human IgG1 κ constant region, has been previously reported.24 To construct the Id-CD40L fusion protein, a PCR fragment encoding a peptide linker and the extracellular domain of murine CD40L was ligated to the 3′ end of the CH3 exon of the human heavy chain gene. This modified heavy-chain vector was cotransfected together with the Id light-chain vector to generate Id-CD40L. As shown in Figure 1(a), the Id-CD40L fusion protein consists of the V region of 38C13 Id and the human IgG1, κ constant region joined to the amino terminal end of the CD40L molecule by a flexible peptide linker (Gly4Ser)3. SDS-PAGE under nonreducing conditions showed that the heavy and light chains of Id-CD40L were correctly assembled, and that the fusion protein migrated more slowly than the Id protein (Fig. 1b, lanes 1 and 2). Under reducing conditions, the light chains of both the Id and Id-CD40L migrated at an apparent molecular weight of 28 kDa (Fig. 1b, lanes 3 and 4). However, the heavy chain of Id-CD40L migrated at an apparent molecular weight of 85 kDa, whereas the heavy chain of the Id protein migrated at 60 kDa, indicating the presence of CD40L molecules in the fusion protein. Indeed, the immunoblotting analysis showed that anti-CD40L mAb interacted with Id-CD40L, but not with Id (data not shown).
The ability of Id-CD40L to bind to CD40 was examined using flow cytometry. Mouse 3T3 fibroblasts were stably transfected with the mouse CD40 gene to generate a CD40-expressing cell line (3T3/CD40). As shown in Figure 1(c), Id-CD40L bound to 3T3/CD40 cells, but not to untransfected 3T3 cells, while the Id protein did not bind to either cell line. This shows that the Id-CD40L fusion protein retained CD40 binding activity.
Id-CD40L retains CD40L biologic function
To determine whether Id-CD40L was capable of activating APCs and inducing IL-12 production, peritoneal macrophages were incubated with either Id or Id-CD40L; cells treated with anti-CD40 mAb served as positive controls. As shown in Figure 2(a), macrophages treated with Id-CD40L produced a substantial amount of IL-12 (493 ± 21 pg/ml), approximately half that produced after stimulation by anti-CD40 mAb (920 ± 55 pg/ml), whereas macrophages cultured with Id alone did not produce detectable IL-12. We also demonstrated that macrophage activation by Id-CD40L did not occur by interaction with the Fcγ receptor, since addition of Fcγ receptor-blocking mAb to the culture medium did not affect the amount of IL-12 released by Id-CD40L-treated cells. We then examined CD80 and CD86 expression by macrophages using direct mAb staining and flow cytometric analysis after 48-hr culture with the various proteins. As shown in Figure 2(b), culture with either Id-CD40L or anti-CD40 mAb resulted in increased macrophage expression of CD80 and CD86, while the Id protein had much less of an effect.
Immune responses elicited by Id-CD40L
We then tested the in vivo immunogenicity of Id-CD40L. Groups of C3H/HeN mice (n = 10) were immunized i.p. with 10 μg of Id, Id-CD40L, or Id-GM-CSF in PBS, then boosted 2 weeks later with the same amount of antigen; Id-GM-CSF was previously shown to be the most potent immunogen among several Id-cytokine fusion proteins tested.25 As shown in Figure 3(a), after the first injection, all mice immunized with Id-CD40L produced anti-Id Abs at concentrations ranging from 3.7 μg/ml to 59.4 μg/ml, while most animals (7 out of 10) immunized with Id did not produce detectable anti-Id Abs. Booster immunization had little effect on the Id-immunized group, with 70% of the animals remaining seronegative, but increased the mean anti-Id concentration of the Id-CD40L-injected group from 24.7 ± 20.8 μg/ml to 89.0 ± 59.7 μg/ml. When compared to immunization with the highly effective Id-GM-CSF, Id-CD40L injection resulted in a lower titer of anti-Id Ab after primary immunization, but a similar anti-Id titer after booster immunization. We also measured anti-Id isotypes in the same serum samples. Immunization with Id-CD40L, like Id-GM-CSF, resulted in mainly IgG1 anti-Id Ab, with only low titers of IgG2a Ab being present (Fig. 3b). As expected, most animals in the Id-immunized group were negative for both IgG1 and IgG2a anti-Id Ab.
To characterize the effect of vaccine dose on the immune response, groups of mice (n = 5) were given 2 injections of 50, 10, 2, or 0.5 μg of Id-CD40L, then anti-Id Ab levels were assayed 2 weeks after the last injection. Mice immunized with 50 or 10 μg of Id served as negative controls. As shown in Figure 4, a clear dose-response relationship was seen between the dose of Id-CD40L used and the levels of anti-Id Ab generated: mice immunized with 50, 10, 2 and 0.5 μg of Id-CD40L producing, respectively, had 160.8 ± 62.1, 123.8 ± 57, 29.6 ± 15.7 and 12.0 ± 3.3 μg/ml of anti-Id Ab. In addition, all animals in the Id-CD40L-immunized groups, including those receiving the lowest dose of 0.5 μg immunogen, produced anti-Id Ab. In contrast, mice immunized with Id alone produced significantly less anti-Id Ab (7.2 ± 5.5 μg/ml and 3.2 ± 2.7 μg/ml for the 50 μg and 10 μg dose groups, respectively), and 20–40% of animals in the Id-immunized groups remained seronegative.
To determine whether the physical linkage between Id and CD40L was important for the adjuvant effect of CD40L, we constructed a chain-shuffle IgG-CD40L fusion protein, designated TBX-CD40L, which was identical in structure to Id-CD40L, except that its VL domain was derived from an irrelevant mouse Igκ. TBX-CD40L lost the Id determinant and did not induce anti-Id responses in vivo (column E, Fig. 5) but exhibited CD40-binding activity similar to Id-CD40L (data not shown). When mice were immunized with a mixture of 10 μg of Id and 10 μg of TBX-CD40L, their anti-Id titers (7.3 ± 6.7 μg/ml) were not significantly higher than those of mice immunized with 10 μg of Id alone (2.8 ± 2.1 μg/ml; p > 0.05; Fig. 5). Immunization with 50 μg of TBX-CD40L and 10 μg of Id increased the anti-Id titer to 20.2 ± 3.1 μg/ml, which was higher than that induced by 10 μg of Id alone (p < 0.05), but still significantly lower than that induced by 10 μg of Id-CD40L (74.4 ± 15.7 μg/ml; p < 0.01). These results demonstrate that CD40L must be physically linked to the antigen in order to achieve a maximal immune response.
Since Id-CD40L was such a powerful adjuvant, we also examined whether autoantibody to murine CD40L, a part of the immunogen, was induced in mice given 2 injections of Id-CD40L fusion protein. We generated a 3T3 fibroblast cell line stably expressing mouse CD40L molecule, which was recognized by a CD40L-specific mAb (Fig. 6, panel i), but not by sera from Id-CD40L-immunized animals (Fig. 6, panel ii), indicating that immunization with Id-CD40L did not elicit an Ab response to the CD40L part of the fusion protein.
Tumor protection elicited by Id-CD40L immunization
We then performed tumor protection experiments to determine whether immunization with the Id-CD40L fusion protein could elicit protective immunity against subsequent tumor challenge. In the first series of experiments involving s.c. tumor challenge, groups of 5 C3H/HeN mice were immunized twice i.p. with 10 μg of Id, Id-CD40L, or Id-GM-CSF, then challenged s.c. with a lethal dose of 38C13 tumor cells 2 weeks after the second immunization. As shown in Figure 7(a), all control animals injected with PBS or with Id developed rapidly growing s.c. tumors by day 15 and all died within 30 days. In contrast, 2 out of 5 (40%) mice in each of the Id-CD40L- and Id-GM-CSF-injected groups did not form tumors within 60 days, and tumor growth suppression was also observed in the 3 tumor-bearing animals in each group. By day 21, the mean tumor volume in the tumor-bearing mice in the Id-CD40L- and Id-GM-CSF-immunized groups was 659 ± 331 and 1,251 ± 671 mm,3 respectively, compared to values of 2,642 ± 801 in the Id-immunized group and 2,727 ± 611 mm3 in the PBS control group. When, in a second series of experiment (Fig. 7b), mice were immunized as described above but challenged i.p. with 38C13 tumor cells, all animals (10/10) in the PBS control group died within 30 days of tumor challenge, whereas Id-CD40L-immunized mice showed significant protection, with 60% (6/10 mice) of the animals surviving longer than 60 days (p < 0.005 vs. the PBS control group), the same percentage protection achieved using Id-GM-CSF. Immunization with Id resulted in little protection, with only 1 out of 10 mice surviving tumor challenge (p > 0.3 vs. the PBS control group).
Triggering of CD40 in vivo by agonistic anti-CD40 mAb can considerably enhance the efficacy of soluble protein- or peptide-based vaccines.5, 9, 13 However, in these previous studies, administration of high doses (100–400 μg) of CD40-stimulating agents was usually required to achieve maximal immune responses. We here report that fusion of CD40L to a self tumor antigen greatly increased the adjuvant activity of CD40L. Immunization with the Id-CD40L fusion protein resulted in substantially stronger anti-Id Ab responses at much lower antigen dose compared to vaccination with a simple mixture of the Id and free CD40L.
Id-CD40L contains 2 CD40L molecules covalently linked to the immunoglobulin heavy chain of the Id protein through a flexible peptide linker. Soluble Id-CD40L specifically bound to CD40 on the CD40 gene-transfected 3T3 cells (Fig. 1c) and stimulated macrophages to produce the proinflammatory cytokine IL-12 and upregulate CD80 and CD86 expression (Fig. 2). Chen et al.34 have also demonstrated that pulsing of bone marrow-derived DCs with Id-CD40L enhances their expression of several costimulatory molecules, class II MHC, and IL-12. However, in a comparative study, we showed that soluble Id-CD40L was significantly less effective than membrane-associated CD40L in triggering IL-12 production by macrophages (data not shown). CD40L has a tertiary structure similar to that of the TNF trimer,35 and according to a model predicted for the TNF-β/TNFR interaction, each CD40L trimer is capable of binding 3 CD40 molecules.36 The less potent stimulatory activity of Id-CD40L might be due to the presence of only 2 CD40 molecules in the fusion protein. This assumption is supported by the results of a previous study comparing the biologic activity of soluble monomeric, dimeric and trimeric CD40L, in which the dimeric form was less potent than the trimeric form in triggering biologic responses and monomeric CD40L was inactive.37 According to this hypothesis, a trimeric Id-CD40L, which can be produced by including an isoleucine zipper sequence in the construct,37 would be expected to be an even stronger immunogen than the Id-CD40L protein containing only 2 CD40L molecules used in this study. This is currently under investigation in our laboratory.
In this study, we demonstrated that Id-CD40L was a potent vaccine that induced anti-Id Ab titers comparable to those produced by immunization with the highly effective Id-GM-CSF vaccine (Fig. 3a). The anti-Id Ab isotypes induced by Id-CD40L and Id-GM-CSF were similar, both dominated by the IgG1 isotype, with only low titers of IgG2a (Fig. 3b), which is consistent with the induction of Th2-type immune response. This is a somewhat surprising result since activation of APCs through interaction of CD40 and CD40L favors Th1-type immune response.7, 8 The most likely explanation for the failure of Id-CD40L to induce Th1 immunity may be related to its suboptimal IL-12-stimulating activity, as discussed in the previous section. It has been known that IL-12 is a critical cytokine to regulate Th1 response and increased production of IgG2a Ab.38 We also showed that vaccination with CD40L-conjugated antigen was far more efficient in inducting immune responses than vaccination with an antigen and a CD40-stimulating agent separately. Mice immunized with 10 μg of Id-CD40L produced 10- and 3-fold more anti-Id Abs than animals immunized with a simple mixture of 10 μg of Id plus either 10 or 50 μg, respectively, of TBX-CD40L (Fig. 5), a fusion protein with a structure identical to Id-CD40L except that its VL domain is derived from an irrelevant mouse Igκ and has thus lost the Id determinant.
In a dose-response study, we showed that Id-CD40L, even at a dose as low as 0.5 μg, elicited anti-Id Ab responses in all immunized animals (Fig. 4), while 20–40% of animals remained seronegative after immunization with up to 50 μg of Id. Our approach of conjugating CD40L with the antigen greatly reduces the amount of CD40-stimulating agent needed in the vaccine regimen and may thus avoid certain forms of toxicity induced by long-term and systemic CD40 activation.16, 18, 19, 20 Another advantage of using CD40L-antigen fusion protein is to avoid possible suppression of T cell and Ab responses.17, 39 Treatment of tumor-bearing mice with agonistic anti-CD40 Ab was reported to accelerate the deletion of tumor antigen-specific T cells, but this effect was avoided by combined treatment with anti-CD40 Ab and tumor antigen.39
The precise mechanisms by which Id-CD40L exerts its adjuvant effect are not clear, but may involve the specific targeting of the antigen to APCs by interaction with CD40 and/or APCs activation. The immunogenicity of a protein antigen can be increased by directing immunogens to APCs through class II MHC, the Fcγ receptor, or surface immunoglobulin.40, 41, 42, 43 We and others have previously shown that the Id protein of B-cell lymphoma can be rendered immunogenic and elicit protective immunity by coupling it to a variety of cytokines (GM-CSF, IL-2, IL-4 and IL-1β peptide), chemokines (MCP-3, IP-10 and MIP-3a), antimicrobial peptide β-defensins, CTLA-4, or fragment C of tetanus toxin.23, 24, 25, 26, 27, 28, 30 The promotion of immunogenicity by tetanus toxin probably involves the classic carrier effect, providing cognate T cell help for the Id response. In contrast, the adjuvant activity of cytokines, chemokines, defensins and CTLA-4 is dependent on their biologic activity, since mutant proteins that have lost the ability to bind to and/or activate APCs fail to stimulate antitumor immunity.27, 28, 30 In the present study, we showed that the vaccine efficacy of Id-CD40L depended on the covalent linkage of Id and CD40L, immunization with a simple mixture of Id and an irrelevant CD40L fusion protein resulting in only minimal anti-Id responses (Fig. 5). These results suggest that direct targeting of the Id protein to APCs via the CD40L/CD40 interaction plays a critical role in promoting anti-Id responses. Since Id-CD40L treatment led to activation of macrophages and dendritic cells (Fig. 2 and Chen et al.34), it is conceivable that the stimulatory effect of Id-CD40L on APCs also plays a role in enhancing the anti-Id responses.
Treatment of tumor-bearing mice with agonistic anti-CD40 mAb is effective in eradicating CD40-positive B-cell tumors of murine or human origin.44, 45, 46 In these tumors, CD40 activation directly inhibits tumor cell proliferation, increases apoptosis and favors Ab-dependent cell-mediated cytotoxicity.47 Moreover, as a result of upregulation of MHC class II molecules and CD80/CD86 and increased IL-12 secretion, a rapid CTL response is induced, even in the absence of Th cells.44 The efficacy of CD40-stimulating treatment has also been demonstrated in CD40-expressing human breast and ovarian carcinomas, the effect being mainly mediated by direct induction of tumor cell apoptosis.48, 49 Moreover, anti-CD40 mAb has recently been demonstrated to have substantial antitumor and antimetastatic effects on CD40-negative tumors by direct or indirect activation of CTLs and natural killer cells.50, 51 Although the 38C13 B-cell lymphoma used in the present study expresses significant level of CD40 (data not shown), we believe that the contribution of direct CD40 triggering to the antitumor activity of Id-CD40L is minimal for the following reasons. First, treatment with Id-CD40L at concentrations up to 50 μg/ml did not inhibit 38C13 tumor cell growth in vitro (data not shown); this lack of an in vitro suppressive effect of Id-CD40L on CD40-positive neoplastic B cells can be attributed to its less potent dimeric structure. Second, and more important, we were unable to find any circulating Id-CD40L protein (<10 ng/ml) in Id-CD40L-immunized mice at the time of 38C13 tumor challenge. Instead, high levels of anti-Id Abs were present in the sera. We believe that this humoral response plays a major role in the antitumor activity elicited by Id-CD40L immunization, since, in the 38C13 model, anti-Id Ab alone has been shown to provide tumor protection.52, 53 Since the Id-CD40L immunization induced mostly IgG1 isotype (Fig. 3b), it is conceivable that the IgG1 anti-Id Abs contribute to most of the protection seen in our study. In contrast, the contribution of CTLs to antitumor activity against 38C13 is minimal, since we were unable to detect Id-specific CTLs in Id-CD40L-immunized animals (data not shown).
In summary, we have demonstrated that fusion of CD40L to a weak tumor antigen, Id of B-cell lymphoma, dramatically increases its immunogenicity and substantially promotes specific Ab responses and protective immunity. Compared to immunization with a simple mixture of Id and CD40L, the Id-CD40L fusion protein induces a much greater immune response at a much lower dose of vaccine. The general approach of fusing CD40L to a candidate antigen may be applicable to the development of vaccines for use in a variety of malignant and infectious diseases.
The authors thank Dr. Steve Roffler (Institute of Biomedical Medicine, Academia Sinica, Taipei, Taiwan) for helpful discussions.
- 32Macrophages and monocytes. In: ColiganJE, KruisbeekAM, MarguliesDH, ShevachEM, StroberW, eds. Current protocols in immunology. New York: John Wiley and Sons, 1997. 14.1. 1–9., .
- 39CD40 stimulation accelerates deletion of tumor-specific CD8+ T cells in the absence of tumor-antigen vaccination. Proc Natl Acad Sci USA 2001; 28: 28., , , , , .
- 52Idiotype vaccination against murine B cell lymphoma: humoral and cellular requirements for the full expression of antitumor immunity. J Immunol 1990; 145: 102936., , , , .